Datasets:

acmc

/

beamit-annotated_full_texts_dataset

like 0

Effects of oral alkali drug therapy on clinical outcomes in pre-dialysis chronic kidney disease patients: a systematic review and meta-analysis.

35176947

Metabolic acidosis accelerates the progression of chronic kidney disease (CKD) and increases the mortality rate. Whether oral alkali drug therapy benefits pre-dialysis CKD patients is controversial. We performed a meta-analysis of the effects of oral alkali drug therapy on major clinical outcomes in pre-dialysis CKD patients.

BACKGROUND

We systematically searched MEDLINE using the Ovid, EMBASE, and Cochrane Library databases without language restriction. We included all eligible clinical studies that involved pre-dialysis CKD adults and compared those who received oral alkali drug therapy with controls.

METHODS

A total of 18 eligible studies, including 14 randomized controlled trials and 4 cohort studies reported in 19 publications with 3695 participants, were included. Oral alkali drug therapy led to a 55% reduction in renal failure events (relative risk [RR]: 0.45; 95% confidence interval [CI]: 0.25-0.82), a rate of decline in the estimated glomerular filtration rate (eGFR) of 2.59 mL/min/1.73 m2 per year (95% CI, 0.88-4.31). There was no significant effect on decline in eGFR events (RR: 0.34; 95% CI: 0.09-1.23), proteinuria (standardized mean difference: -0.32; 95% CI: -1.08 to 0.43), all-cause mortality events (RR: 0.90; 95% CI: 0.40-2.02) and cardiovascular (CV) events (RR: 1.03; 95% CI: 0.32-3.37) compared with the control groups.

RESULTS

Based on the available and low-to-moderate certainty evidence, oral alkali drug therapy might potentially reduce the risk of kidney failure events, but no benefit in reducing all-cause mortality events, CV events, decline in eGFR and porteninuria.

CONCLUSION

8865123

Introduction

Metabolic acidosis (MA), a common complication of chronic kidney disease (CKD) caused by failure to balance the daily acid load, causes kidney damage, leading to protein-energy consumption, chronic inflammation, endocrine disorders, and the aggravation of metabolic osteopathy [1]. MA is associated with adverse outcomes in CKD patients, including the progression of CKD, all-cause mortality, and cardiovascular (CV) events [2,3]. Oral bicarbonate supplementation is commonly used to correct MA and improve the prognosis of CKD patients. According to the 2020 KDIGO guidelines for glomerulonephritis, MA should be treated with supplementation with oral sodium bicarbonate if the serum bicarbonate level is < 22 mmol/L [4]. However, it remains unclear whether treatment of MA based on oral alkali supplementation would translate into improved clinical outcomes, including delaying the progression of MA and decreasing the risks of all-cause mortality and CV events. In the UBI study, treatment of MA with sodium bicarbonate in patients with stage 3–5 CKD was safe and reduced the risks of CKD progression and all-cause mortality [5]. However, other studies did not confirm the benefits of alkali supplementation in terms of delaying CKD progression and improving survival. Randomized controlled trials (RCTs) showed no significant effects of oral alkali supplementation on renal outcomes and mortality [6,7]. In a cohort study of 386 CKD patients, compared with those who did not receive bicarbonate supplementation, the risk of ischemic heart disease was significantly lower in patients who received bicarbonate supplementation [8]. In a multi-center RCT, no differences were found in CV events between the sodium bicarbonate group and the placebo group [6]. Therefore, the association between oral alkali drug supplementation and clinical outcomes in pre-dialysis CKD patients is unclear. In this systematic review, we summarized all available clinical study data to evaluate the benefits of oral alkali drug therapy regarding kidney outcomes, all-cause mortality, and CV events in pre-dialysis CKD patients.

Results

Overview of included trials The literature search yielded 9284 potentially relevant records, of which the full texts of 185 publications were reviewed (Figure 1). After screening and eligibility assessment, 14 RCTs [5–7,20–32] and 4 cohort studies [8,29,33,34] reported in 19 publications with 3695 individuals were included in this systematic review and meta-analysis. Baseline and key characteristics of the enrolled studies are presented in Supplementary Table S1 and Table S2. The median follow-up time was 19.5 months. Individuals were enrolled at an average age of 55.78 years, and male participants accounted for 59.55% of the total. The average eGFR of participants was 31.51 mL/min/1.73 m2. A total of 2 oral alkali drug therapies were studied, including those featuring sodium bicarbonate in 17 studies, veverimer in 1 study,. PRISMA flow chart for the included studies. The Jadad score for each included RCT is presented in Supplementary Table S1. Eleven trials had a Jadad score of 3–5, and the others scored less than 3. Of all RCTs, 78.57% were associated with a low risk of bias arising from the randomization process, and all studies had a low risk of bias due to deviations from intended interventions, due to missing outcome data, associated with measurements of outcomes, and associated with selection of the reported results. In terms of overall bias, 78.57% of the research trials were assessed as at low risk, and 21.43% as subject to some concerns (Supplementary Table S3). As shown in Supplementary Tables S1 and S4, all included cohort studies were considered of high quality (with scores of 7★–8★) according to the NOS checklist. The literature search yielded 9284 potentially relevant records, of which the full texts of 185 publications were reviewed (Figure 1). After screening and eligibility assessment, 14 RCTs [5–7,20–32] and 4 cohort studies [8,29,33,34] reported in 19 publications with 3695 individuals were included in this systematic review and meta-analysis. Baseline and key characteristics of the enrolled studies are presented in Supplementary Table S1 and Table S2. The median follow-up time was 19.5 months. Individuals were enrolled at an average age of 55.78 years, and male participants accounted for 59.55% of the total. The average eGFR of participants was 31.51 mL/min/1.73 m2. A total of 2 oral alkali drug therapies were studied, including those featuring sodium bicarbonate in 17 studies, veverimer in 1 study,. PRISMA flow chart for the included studies. The Jadad score for each included RCT is presented in Supplementary Table S1. Eleven trials had a Jadad score of 3–5, and the others scored less than 3. Of all RCTs, 78.57% were associated with a low risk of bias arising from the randomization process, and all studies had a low risk of bias due to deviations from intended interventions, due to missing outcome data, associated with measurements of outcomes, and associated with selection of the reported results. In terms of overall bias, 78.57% of the research trials were assessed as at low risk, and 21.43% as subject to some concerns (Supplementary Table S3). As shown in Supplementary Tables S1 and S4, all included cohort studies were considered of high quality (with scores of 7★–8★) according to the NOS checklist. Effects of oral alkali drug therapy on renal outcomes Nine RCTs with 1833 participants reported 250 renal failure events. Compared with the control group, oral alkali drug therapy was associated with a 55% reduction in the risk of renal failure events (RR: 0.45; 95% CI: 0.25–0.82), with significant heterogeneity across studies (I2 = 67.8%, p = 0.005; Figure 2). No significant heterogeneity was observed in any subgroup analysis (Table 1). Forest plot for renal failure events and decline in eGFR events. Renal failure was defined as a more than 50% decline in eGFR from baseline during follow-up, doubling of serum creatinine or ESRD. CI: confidence interval; RR: relative risk. Subgroup analysis of renal failure events. Note. a p value calculated by χ2 statistics was shown. CI: confidence interval; n: number of patients; RCT: randomized parallel-group controlled trial; RR: relative risk. Data regarding the effects of oral alkali drug therapy on decline in eGFR events were available from three RCTs that included 404 individuals and 123 events. Overall, there was no significant effect of oral alkali drug therapy on decline in eGFR events (RR: 0.34; 95% CI: 0.09–1.23) compared with the control group. Moderate heterogeneity across these trials (I2 =54.1%, p = 0.113; Figure 2) was found. Thirtine RCTs and three cohort studies with 2746 participants provided data on differences in the rate of change in eGFR. Compared with the control group, oral alkali drug therapy slowed the rate of eGFR decline by 2.59 mL/min/1.73 m2 per year (95% CI: 0.88–4.31), with significant heterogeneity observed (I2 = 97.6%, p < 0.001; Figure 3). Subgroup analyses showed that effect sizes were greater in studies that enrolled patients baseline serum bicarbonate < 20.95 mmol/L, baseline eGFR 30-59 mL/min/1.73 m2, age < 55 years (p < 0.001; Supplementary Table S5). Forest plot for rate of change in estimated glomerular filtration rate (eGFR). CI: confidence interval; MD: mean difference; SD: standard deviation. Data on the effects of oral alkali drug therapy on proteinuria or albuminuria were available in only five studies (four RCTs and one cohort study) with 591 participants, and no significant effect was found (SMD: −0.32; 95% CI: −1.08 to 0.43). I2 statistics (88.2%, p < 0.001; Figure 4) indicated significant heterogeneity across studies. Subgroup analyses did not reveal heterogeneity regarding pre-specified characteristics (Supplementary Table S5). Forest plot for the change in proteinuria or albuminuria. CI: confidence interval; SD: standard deviation; SMD: standard mean difference. Nine RCTs with 1833 participants reported 250 renal failure events. Compared with the control group, oral alkali drug therapy was associated with a 55% reduction in the risk of renal failure events (RR: 0.45; 95% CI: 0.25–0.82), with significant heterogeneity across studies (I2 = 67.8%, p = 0.005; Figure 2). No significant heterogeneity was observed in any subgroup analysis (Table 1). Forest plot for renal failure events and decline in eGFR events. Renal failure was defined as a more than 50% decline in eGFR from baseline during follow-up, doubling of serum creatinine or ESRD. CI: confidence interval; RR: relative risk. Subgroup analysis of renal failure events. Note. a p value calculated by χ2 statistics was shown. CI: confidence interval; n: number of patients; RCT: randomized parallel-group controlled trial; RR: relative risk. Data regarding the effects of oral alkali drug therapy on decline in eGFR events were available from three RCTs that included 404 individuals and 123 events. Overall, there was no significant effect of oral alkali drug therapy on decline in eGFR events (RR: 0.34; 95% CI: 0.09–1.23) compared with the control group. Moderate heterogeneity across these trials (I2 =54.1%, p = 0.113; Figure 2) was found. Thirtine RCTs and three cohort studies with 2746 participants provided data on differences in the rate of change in eGFR. Compared with the control group, oral alkali drug therapy slowed the rate of eGFR decline by 2.59 mL/min/1.73 m2 per year (95% CI: 0.88–4.31), with significant heterogeneity observed (I2 = 97.6%, p < 0.001; Figure 3). Subgroup analyses showed that effect sizes were greater in studies that enrolled patients baseline serum bicarbonate < 20.95 mmol/L, baseline eGFR 30-59 mL/min/1.73 m2, age < 55 years (p < 0.001; Supplementary Table S5). Forest plot for rate of change in estimated glomerular filtration rate (eGFR). CI: confidence interval; MD: mean difference; SD: standard deviation. Data on the effects of oral alkali drug therapy on proteinuria or albuminuria were available in only five studies (four RCTs and one cohort study) with 591 participants, and no significant effect was found (SMD: −0.32; 95% CI: −1.08 to 0.43). I2 statistics (88.2%, p < 0.001; Figure 4) indicated significant heterogeneity across studies. Subgroup analyses did not reveal heterogeneity regarding pre-specified characteristics (Supplementary Table S5). Forest plot for the change in proteinuria or albuminuria. CI: confidence interval; SD: standard deviation; SMD: standard mean difference. Effects of oral alkali drug therapy on all-cause mortality and CV events Seven RCTs involving 1709 individuals reported 127 all-cause mortality events. There was no significant effect of oral alkali drug therapy on the risk of all-cause mortality compared with the control groups (RR: 0.90; 95% CI: 0.40–2.02). Significant heterogeneity was noted across the included trials (I2 = 54.7%, p = 0.05; Figure 5). No significant heterogeneity was found for all-cause mortality in the subgroup analyses (Supplementary Table S5). Forest plot for all-cause mortality and cardiovascular events. Cardiovascular events were defined as a composite, including fatal or non-fatal myocardial infarction, fatal or non-fatal stroke, coronary artery revascularization, cardiovascular disease and cardiovascular death. CI: confidence interval; N: number of trials; RR: relative risk. Data for CV events were available from three RCTs and one cohort studie that included 1098 participants and 160 events. There was no significant difference in the risk of CV events between treatment and control groups (RR: 1.03; 95% CI: 0.32–3.37), with significant heterogeneity observed among trials (I2 = 65.1%, p = 0.057; Figure 5). Subgroup analyses revealed no heterogeneity for CV events (Supplementary Table S5). Seven RCTs involving 1709 individuals reported 127 all-cause mortality events. There was no significant effect of oral alkali drug therapy on the risk of all-cause mortality compared with the control groups (RR: 0.90; 95% CI: 0.40–2.02). Significant heterogeneity was noted across the included trials (I2 = 54.7%, p = 0.05; Figure 5). No significant heterogeneity was found for all-cause mortality in the subgroup analyses (Supplementary Table S5). Forest plot for all-cause mortality and cardiovascular events. Cardiovascular events were defined as a composite, including fatal or non-fatal myocardial infarction, fatal or non-fatal stroke, coronary artery revascularization, cardiovascular disease and cardiovascular death. CI: confidence interval; N: number of trials; RR: relative risk. Data for CV events were available from three RCTs and one cohort studie that included 1098 participants and 160 events. There was no significant difference in the risk of CV events between treatment and control groups (RR: 1.03; 95% CI: 0.32–3.37), with significant heterogeneity observed among trials (I2 = 65.1%, p = 0.057; Figure 5). Subgroup analyses revealed no heterogeneity for CV events (Supplementary Table S5). Sensitivity analysis The results did not change after the exclusion of studies with a follow-up duration of < 12 months, with a sample size < 50, or assessed as low quality, or when different random-effects estimination methods were used (Supplementary Table S6). The results did not change after the exclusion of studies with a follow-up duration of < 12 months, with a sample size < 50, or assessed as low quality, or when different random-effects estimination methods were used (Supplementary Table S6).

Introduction: Metabolic acidosis (MA), a common complication of chronic kidney disease (CKD) caused by failure to balance the daily acid load, causes kidney damage, leading to protein-energy consumption, chronic inflammation, endocrine disorders, and the aggravation of metabolic osteopathy [1]. MA is associated with adverse outcomes in CKD patients, including the progression of CKD, all-cause mortality, and cardiovascular (CV) events [2,3]. Oral bicarbonate supplementation is commonly used to correct MA and improve the prognosis of CKD patients. According to the 2020 KDIGO guidelines for glomerulonephritis, MA should be treated with supplementation with oral sodium bicarbonate if the serum bicarbonate level is < 22 mmol/L [4]. However, it remains unclear whether treatment of MA based on oral alkali supplementation would translate into improved clinical outcomes, including delaying the progression of MA and decreasing the risks of all-cause mortality and CV events. In the UBI study, treatment of MA with sodium bicarbonate in patients with stage 3–5 CKD was safe and reduced the risks of CKD progression and all-cause mortality [5]. However, other studies did not confirm the benefits of alkali supplementation in terms of delaying CKD progression and improving survival. Randomized controlled trials (RCTs) showed no significant effects of oral alkali supplementation on renal outcomes and mortality [6,7]. In a cohort study of 386 CKD patients, compared with those who did not receive bicarbonate supplementation, the risk of ischemic heart disease was significantly lower in patients who received bicarbonate supplementation [8]. In a multi-center RCT, no differences were found in CV events between the sodium bicarbonate group and the placebo group [6]. Therefore, the association between oral alkali drug supplementation and clinical outcomes in pre-dialysis CKD patients is unclear. In this systematic review, we summarized all available clinical study data to evaluate the benefits of oral alkali drug therapy regarding kidney outcomes, all-cause mortality, and CV events in pre-dialysis CKD patients. Materials and methods: Data sources and search strategy We performed this systematic review according to a pre-specified protocol [9] registered in the International Prospective Register of Systematic Reviews (CRD42018111030), and the reporting was in line with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A comprehensive search was conducted using the following databases: MEDLINE by Ovid (1946 to February 2020), EMBASE (1966 to February 2020), and Cochrane Central Register of Controlled Trials (no date restriction), with relevant keywords and medical subject headings that included various spellings of ‘CKD’, ‘RCT’, ‘Cohort Studies’, and ‘Oral Alkali Therapy’ (the terms ‘Sodium Bicarbonate’, ‘Alkali’, (see item S1). Studies were considered without any language restriction. To ensure a comprehensive literature search, we also screened reference lists from included articles. The ClinicalTrials.gov website was searched for ongoing but unpublished trials in this field. We performed this systematic review according to a pre-specified protocol [9] registered in the International Prospective Register of Systematic Reviews (CRD42018111030), and the reporting was in line with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A comprehensive search was conducted using the following databases: MEDLINE by Ovid (1946 to February 2020), EMBASE (1966 to February 2020), and Cochrane Central Register of Controlled Trials (no date restriction), with relevant keywords and medical subject headings that included various spellings of ‘CKD’, ‘RCT’, ‘Cohort Studies’, and ‘Oral Alkali Therapy’ (the terms ‘Sodium Bicarbonate’, ‘Alkali’, (see item S1). Studies were considered without any language restriction. To ensure a comprehensive literature search, we also screened reference lists from included articles. The ClinicalTrials.gov website was searched for ongoing but unpublished trials in this field. Study selection and outcome estimation We included data from RCTs and cohort studies in which oral alkali drug therapy was provided to adults with pre-dialysis CKD (participants who were pregnant, had malignancies or acute illnesses, or had a follow-up time of less than 3 months were excluded) and comparisons were made with subjects receiving the usual therapy. Pre-defined outcomes that contained analyzable data were extracted as follows. A renal failure event was defined as a more than 50% decline in estimated glomerular filtration rate (eGFR) from baseline during follow-up, doubling of serum creatinine, or progression to end-stage renal disease (ESRD) [10]. Decline in eGFR was defined as a decrease of eGFR >3 mL/min/1.73 m2 per year [11]. The rate of change in eGFR per year and changes in urinary protein or urinary albumin during follow-up, including urinary protein excretion, urinary albumin excretion, and the urinary albumin/creatinine ratio, were recorded. Additionally, the incidences of all-cause mortality events and CV events, defined as a composite, including fatal or non-fatal myocardial infarction, fatal or non-fatal stroke, coronary artery revascularization, CV disease, and CV death, were recorded. We included data from RCTs and cohort studies in which oral alkali drug therapy was provided to adults with pre-dialysis CKD (participants who were pregnant, had malignancies or acute illnesses, or had a follow-up time of less than 3 months were excluded) and comparisons were made with subjects receiving the usual therapy. Pre-defined outcomes that contained analyzable data were extracted as follows. A renal failure event was defined as a more than 50% decline in estimated glomerular filtration rate (eGFR) from baseline during follow-up, doubling of serum creatinine, or progression to end-stage renal disease (ESRD) [10]. Decline in eGFR was defined as a decrease of eGFR >3 mL/min/1.73 m2 per year [11]. The rate of change in eGFR per year and changes in urinary protein or urinary albumin during follow-up, including urinary protein excretion, urinary albumin excretion, and the urinary albumin/creatinine ratio, were recorded. Additionally, the incidences of all-cause mortality events and CV events, defined as a composite, including fatal or non-fatal myocardial infarction, fatal or non-fatal stroke, coronary artery revascularization, CV disease, and CV death, were recorded. Data extraction and quality assessment Two independent reviewers (H.S. and X.S.) extracted data and assessed their quality according to the pre-specified protocol. Disagreements were resolved by a third reviewer (L.W.). Data from all eligible studies were extracted into a spreadsheet. The data sought included the characteristics of the studies (study type, randomization method, follow-up time, withdrawals/dropouts), baseline patient characteristics (age, sex, baseline eGFR), intake of alkali drug supplementation, and outcome events. We used the Cochrane Collaboration risk-of-bias tool [12] to assess all potential sources of bias for the included RCTs. Trials were assessed as being at low or high risk of bias or subject to other risks or some concerns, and the overall risk of bias generally corresponded to the worst risk of bias in any of the domains. However, if a study was judged to be subject to some concerns about the risk of bias for multiple domains, it might be judged as being at high risk of bias overall. In addition, the quality of the RCTs was assessed using the Jadad scale [13]. We used the Newcastle-Ottawa Scale (NOS) to assess the quality of cohort studies in terms of selection of cohorts, comparability of cohorts, and assessments of outcomes [14]. Two independent reviewers (H.S. and X.S.) extracted data and assessed their quality according to the pre-specified protocol. Disagreements were resolved by a third reviewer (L.W.). Data from all eligible studies were extracted into a spreadsheet. The data sought included the characteristics of the studies (study type, randomization method, follow-up time, withdrawals/dropouts), baseline patient characteristics (age, sex, baseline eGFR), intake of alkali drug supplementation, and outcome events. We used the Cochrane Collaboration risk-of-bias tool [12] to assess all potential sources of bias for the included RCTs. Trials were assessed as being at low or high risk of bias or subject to other risks or some concerns, and the overall risk of bias generally corresponded to the worst risk of bias in any of the domains. However, if a study was judged to be subject to some concerns about the risk of bias for multiple domains, it might be judged as being at high risk of bias overall. In addition, the quality of the RCTs was assessed using the Jadad scale [13]. We used the Newcastle-Ottawa Scale (NOS) to assess the quality of cohort studies in terms of selection of cohorts, comparability of cohorts, and assessments of outcomes [14]. Data synthesis and statistical analyses When dichotomous outcome data from individual studies were analyzed, relative risks (RRs) and their 95% confidence intervals (CIs) were calculated. If the RR for an individual study was unavailable in the original article, the RR and 95% CI were calculated from event numbers extracted from each study before data pooling. In calculating the RR values, we used the total number of patients randomized in each group as the denominator. Continuous outcome data from individual trials were analyzed using differences in means (MDs) with 95% CIs to pool eGFR data, whereas the standardized mean differences (SMDs) with 95% CIs were used to pool proteinuria or albuminuria data. When continuous outcome data were analyzed, the difference in the mean change between values at baseline and the end of treatment was used. If data on changes between baseline and end-of-treatment values were not available in the studies, we calculated them using correlations estimated from other included studies that had a similar follow-up period and reported their results in considerable detail according to the imputed formulation and its related interpretations in the Cochrane Handbook [15]. Because of the poor stability of the Der Simonian-Laird procedure for small numbers of studies, we used the empirical Bayes procedure to estimate all outcomes [16,17]. We also used the Der Simonian-Laird random effects model and restricted maximum likelihood approach to assess summary effects as part of sensitivity analyses [18,19]. Considering the inevitable heterogeneity among studies, subgroup and sensitivity analyses were performed. Subgroup analyses were performed based on a pre-specified protocol according to the study type, baseline serum bicarbonate, baseline eGFR, mean age, follow-up time, and sample size. In addition, we performed sensitivity analyses using different random-effects estimation methods, excluding studies with a sample size <50, those with a follow-up of < 12 months, and studies of low quality (Jadad score <3, NOS score <5★). Heterogeneity among studies was evaluated using the I2 or τ2 statistic. Stata version 15.0 (StataCorp LP, College Station, TX) was used for statistical analysis, and a two-sided p-value < 0.05 was considered indicative of significance. When dichotomous outcome data from individual studies were analyzed, relative risks (RRs) and their 95% confidence intervals (CIs) were calculated. If the RR for an individual study was unavailable in the original article, the RR and 95% CI were calculated from event numbers extracted from each study before data pooling. In calculating the RR values, we used the total number of patients randomized in each group as the denominator. Continuous outcome data from individual trials were analyzed using differences in means (MDs) with 95% CIs to pool eGFR data, whereas the standardized mean differences (SMDs) with 95% CIs were used to pool proteinuria or albuminuria data. When continuous outcome data were analyzed, the difference in the mean change between values at baseline and the end of treatment was used. If data on changes between baseline and end-of-treatment values were not available in the studies, we calculated them using correlations estimated from other included studies that had a similar follow-up period and reported their results in considerable detail according to the imputed formulation and its related interpretations in the Cochrane Handbook [15]. Because of the poor stability of the Der Simonian-Laird procedure for small numbers of studies, we used the empirical Bayes procedure to estimate all outcomes [16,17]. We also used the Der Simonian-Laird random effects model and restricted maximum likelihood approach to assess summary effects as part of sensitivity analyses [18,19]. Considering the inevitable heterogeneity among studies, subgroup and sensitivity analyses were performed. Subgroup analyses were performed based on a pre-specified protocol according to the study type, baseline serum bicarbonate, baseline eGFR, mean age, follow-up time, and sample size. In addition, we performed sensitivity analyses using different random-effects estimation methods, excluding studies with a sample size <50, those with a follow-up of < 12 months, and studies of low quality (Jadad score <3, NOS score <5★). Heterogeneity among studies was evaluated using the I2 or τ2 statistic. Stata version 15.0 (StataCorp LP, College Station, TX) was used for statistical analysis, and a two-sided p-value < 0.05 was considered indicative of significance. Data sources and search strategy: We performed this systematic review according to a pre-specified protocol [9] registered in the International Prospective Register of Systematic Reviews (CRD42018111030), and the reporting was in line with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A comprehensive search was conducted using the following databases: MEDLINE by Ovid (1946 to February 2020), EMBASE (1966 to February 2020), and Cochrane Central Register of Controlled Trials (no date restriction), with relevant keywords and medical subject headings that included various spellings of ‘CKD’, ‘RCT’, ‘Cohort Studies’, and ‘Oral Alkali Therapy’ (the terms ‘Sodium Bicarbonate’, ‘Alkali’, (see item S1). Studies were considered without any language restriction. To ensure a comprehensive literature search, we also screened reference lists from included articles. The ClinicalTrials.gov website was searched for ongoing but unpublished trials in this field. Study selection and outcome estimation: We included data from RCTs and cohort studies in which oral alkali drug therapy was provided to adults with pre-dialysis CKD (participants who were pregnant, had malignancies or acute illnesses, or had a follow-up time of less than 3 months were excluded) and comparisons were made with subjects receiving the usual therapy. Pre-defined outcomes that contained analyzable data were extracted as follows. A renal failure event was defined as a more than 50% decline in estimated glomerular filtration rate (eGFR) from baseline during follow-up, doubling of serum creatinine, or progression to end-stage renal disease (ESRD) [10]. Decline in eGFR was defined as a decrease of eGFR >3 mL/min/1.73 m2 per year [11]. The rate of change in eGFR per year and changes in urinary protein or urinary albumin during follow-up, including urinary protein excretion, urinary albumin excretion, and the urinary albumin/creatinine ratio, were recorded. Additionally, the incidences of all-cause mortality events and CV events, defined as a composite, including fatal or non-fatal myocardial infarction, fatal or non-fatal stroke, coronary artery revascularization, CV disease, and CV death, were recorded. Data extraction and quality assessment: Two independent reviewers (H.S. and X.S.) extracted data and assessed their quality according to the pre-specified protocol. Disagreements were resolved by a third reviewer (L.W.). Data from all eligible studies were extracted into a spreadsheet. The data sought included the characteristics of the studies (study type, randomization method, follow-up time, withdrawals/dropouts), baseline patient characteristics (age, sex, baseline eGFR), intake of alkali drug supplementation, and outcome events. We used the Cochrane Collaboration risk-of-bias tool [12] to assess all potential sources of bias for the included RCTs. Trials were assessed as being at low or high risk of bias or subject to other risks or some concerns, and the overall risk of bias generally corresponded to the worst risk of bias in any of the domains. However, if a study was judged to be subject to some concerns about the risk of bias for multiple domains, it might be judged as being at high risk of bias overall. In addition, the quality of the RCTs was assessed using the Jadad scale [13]. We used the Newcastle-Ottawa Scale (NOS) to assess the quality of cohort studies in terms of selection of cohorts, comparability of cohorts, and assessments of outcomes [14]. Data synthesis and statistical analyses: When dichotomous outcome data from individual studies were analyzed, relative risks (RRs) and their 95% confidence intervals (CIs) were calculated. If the RR for an individual study was unavailable in the original article, the RR and 95% CI were calculated from event numbers extracted from each study before data pooling. In calculating the RR values, we used the total number of patients randomized in each group as the denominator. Continuous outcome data from individual trials were analyzed using differences in means (MDs) with 95% CIs to pool eGFR data, whereas the standardized mean differences (SMDs) with 95% CIs were used to pool proteinuria or albuminuria data. When continuous outcome data were analyzed, the difference in the mean change between values at baseline and the end of treatment was used. If data on changes between baseline and end-of-treatment values were not available in the studies, we calculated them using correlations estimated from other included studies that had a similar follow-up period and reported their results in considerable detail according to the imputed formulation and its related interpretations in the Cochrane Handbook [15]. Because of the poor stability of the Der Simonian-Laird procedure for small numbers of studies, we used the empirical Bayes procedure to estimate all outcomes [16,17]. We also used the Der Simonian-Laird random effects model and restricted maximum likelihood approach to assess summary effects as part of sensitivity analyses [18,19]. Considering the inevitable heterogeneity among studies, subgroup and sensitivity analyses were performed. Subgroup analyses were performed based on a pre-specified protocol according to the study type, baseline serum bicarbonate, baseline eGFR, mean age, follow-up time, and sample size. In addition, we performed sensitivity analyses using different random-effects estimation methods, excluding studies with a sample size <50, those with a follow-up of < 12 months, and studies of low quality (Jadad score <3, NOS score <5★). Heterogeneity among studies was evaluated using the I2 or τ2 statistic. Stata version 15.0 (StataCorp LP, College Station, TX) was used for statistical analysis, and a two-sided p-value < 0.05 was considered indicative of significance. Results: Overview of included trials The literature search yielded 9284 potentially relevant records, of which the full texts of 185 publications were reviewed (Figure 1). After screening and eligibility assessment, 14 RCTs [5–7,20–32] and 4 cohort studies [8,29,33,34] reported in 19 publications with 3695 individuals were included in this systematic review and meta-analysis. Baseline and key characteristics of the enrolled studies are presented in Supplementary Table S1 and Table S2. The median follow-up time was 19.5 months. Individuals were enrolled at an average age of 55.78 years, and male participants accounted for 59.55% of the total. The average eGFR of participants was 31.51 mL/min/1.73 m2. A total of 2 oral alkali drug therapies were studied, including those featuring sodium bicarbonate in 17 studies, veverimer in 1 study,. PRISMA flow chart for the included studies. The Jadad score for each included RCT is presented in Supplementary Table S1. Eleven trials had a Jadad score of 3–5, and the others scored less than 3. Of all RCTs, 78.57% were associated with a low risk of bias arising from the randomization process, and all studies had a low risk of bias due to deviations from intended interventions, due to missing outcome data, associated with measurements of outcomes, and associated with selection of the reported results. In terms of overall bias, 78.57% of the research trials were assessed as at low risk, and 21.43% as subject to some concerns (Supplementary Table S3). As shown in Supplementary Tables S1 and S4, all included cohort studies were considered of high quality (with scores of 7★–8★) according to the NOS checklist. The literature search yielded 9284 potentially relevant records, of which the full texts of 185 publications were reviewed (Figure 1). After screening and eligibility assessment, 14 RCTs [5–7,20–32] and 4 cohort studies [8,29,33,34] reported in 19 publications with 3695 individuals were included in this systematic review and meta-analysis. Baseline and key characteristics of the enrolled studies are presented in Supplementary Table S1 and Table S2. The median follow-up time was 19.5 months. Individuals were enrolled at an average age of 55.78 years, and male participants accounted for 59.55% of the total. The average eGFR of participants was 31.51 mL/min/1.73 m2. A total of 2 oral alkali drug therapies were studied, including those featuring sodium bicarbonate in 17 studies, veverimer in 1 study,. PRISMA flow chart for the included studies. The Jadad score for each included RCT is presented in Supplementary Table S1. Eleven trials had a Jadad score of 3–5, and the others scored less than 3. Of all RCTs, 78.57% were associated with a low risk of bias arising from the randomization process, and all studies had a low risk of bias due to deviations from intended interventions, due to missing outcome data, associated with measurements of outcomes, and associated with selection of the reported results. In terms of overall bias, 78.57% of the research trials were assessed as at low risk, and 21.43% as subject to some concerns (Supplementary Table S3). As shown in Supplementary Tables S1 and S4, all included cohort studies were considered of high quality (with scores of 7★–8★) according to the NOS checklist. Effects of oral alkali drug therapy on renal outcomes Nine RCTs with 1833 participants reported 250 renal failure events. Compared with the control group, oral alkali drug therapy was associated with a 55% reduction in the risk of renal failure events (RR: 0.45; 95% CI: 0.25–0.82), with significant heterogeneity across studies (I2 = 67.8%, p = 0.005; Figure 2). No significant heterogeneity was observed in any subgroup analysis (Table 1). Forest plot for renal failure events and decline in eGFR events. Renal failure was defined as a more than 50% decline in eGFR from baseline during follow-up, doubling of serum creatinine or ESRD. CI: confidence interval; RR: relative risk. Subgroup analysis of renal failure events. Note. a p value calculated by χ2 statistics was shown. CI: confidence interval; n: number of patients; RCT: randomized parallel-group controlled trial; RR: relative risk. Data regarding the effects of oral alkali drug therapy on decline in eGFR events were available from three RCTs that included 404 individuals and 123 events. Overall, there was no significant effect of oral alkali drug therapy on decline in eGFR events (RR: 0.34; 95% CI: 0.09–1.23) compared with the control group. Moderate heterogeneity across these trials (I2 =54.1%, p = 0.113; Figure 2) was found. Thirtine RCTs and three cohort studies with 2746 participants provided data on differences in the rate of change in eGFR. Compared with the control group, oral alkali drug therapy slowed the rate of eGFR decline by 2.59 mL/min/1.73 m2 per year (95% CI: 0.88–4.31), with significant heterogeneity observed (I2 = 97.6%, p < 0.001; Figure 3). Subgroup analyses showed that effect sizes were greater in studies that enrolled patients baseline serum bicarbonate < 20.95 mmol/L, baseline eGFR 30-59 mL/min/1.73 m2, age < 55 years (p < 0.001; Supplementary Table S5). Forest plot for rate of change in estimated glomerular filtration rate (eGFR). CI: confidence interval; MD: mean difference; SD: standard deviation. Data on the effects of oral alkali drug therapy on proteinuria or albuminuria were available in only five studies (four RCTs and one cohort study) with 591 participants, and no significant effect was found (SMD: −0.32; 95% CI: −1.08 to 0.43). I2 statistics (88.2%, p < 0.001; Figure 4) indicated significant heterogeneity across studies. Subgroup analyses did not reveal heterogeneity regarding pre-specified characteristics (Supplementary Table S5). Forest plot for the change in proteinuria or albuminuria. CI: confidence interval; SD: standard deviation; SMD: standard mean difference. Nine RCTs with 1833 participants reported 250 renal failure events. Compared with the control group, oral alkali drug therapy was associated with a 55% reduction in the risk of renal failure events (RR: 0.45; 95% CI: 0.25–0.82), with significant heterogeneity across studies (I2 = 67.8%, p = 0.005; Figure 2). No significant heterogeneity was observed in any subgroup analysis (Table 1). Forest plot for renal failure events and decline in eGFR events. Renal failure was defined as a more than 50% decline in eGFR from baseline during follow-up, doubling of serum creatinine or ESRD. CI: confidence interval; RR: relative risk. Subgroup analysis of renal failure events. Note. a p value calculated by χ2 statistics was shown. CI: confidence interval; n: number of patients; RCT: randomized parallel-group controlled trial; RR: relative risk. Data regarding the effects of oral alkali drug therapy on decline in eGFR events were available from three RCTs that included 404 individuals and 123 events. Overall, there was no significant effect of oral alkali drug therapy on decline in eGFR events (RR: 0.34; 95% CI: 0.09–1.23) compared with the control group. Moderate heterogeneity across these trials (I2 =54.1%, p = 0.113; Figure 2) was found. Thirtine RCTs and three cohort studies with 2746 participants provided data on differences in the rate of change in eGFR. Compared with the control group, oral alkali drug therapy slowed the rate of eGFR decline by 2.59 mL/min/1.73 m2 per year (95% CI: 0.88–4.31), with significant heterogeneity observed (I2 = 97.6%, p < 0.001; Figure 3). Subgroup analyses showed that effect sizes were greater in studies that enrolled patients baseline serum bicarbonate < 20.95 mmol/L, baseline eGFR 30-59 mL/min/1.73 m2, age < 55 years (p < 0.001; Supplementary Table S5). Forest plot for rate of change in estimated glomerular filtration rate (eGFR). CI: confidence interval; MD: mean difference; SD: standard deviation. Data on the effects of oral alkali drug therapy on proteinuria or albuminuria were available in only five studies (four RCTs and one cohort study) with 591 participants, and no significant effect was found (SMD: −0.32; 95% CI: −1.08 to 0.43). I2 statistics (88.2%, p < 0.001; Figure 4) indicated significant heterogeneity across studies. Subgroup analyses did not reveal heterogeneity regarding pre-specified characteristics (Supplementary Table S5). Forest plot for the change in proteinuria or albuminuria. CI: confidence interval; SD: standard deviation; SMD: standard mean difference. Effects of oral alkali drug therapy on all-cause mortality and CV events Seven RCTs involving 1709 individuals reported 127 all-cause mortality events. There was no significant effect of oral alkali drug therapy on the risk of all-cause mortality compared with the control groups (RR: 0.90; 95% CI: 0.40–2.02). Significant heterogeneity was noted across the included trials (I2 = 54.7%, p = 0.05; Figure 5). No significant heterogeneity was found for all-cause mortality in the subgroup analyses (Supplementary Table S5). Forest plot for all-cause mortality and cardiovascular events. Cardiovascular events were defined as a composite, including fatal or non-fatal myocardial infarction, fatal or non-fatal stroke, coronary artery revascularization, cardiovascular disease and cardiovascular death. CI: confidence interval; N: number of trials; RR: relative risk. Data for CV events were available from three RCTs and one cohort studie that included 1098 participants and 160 events. There was no significant difference in the risk of CV events between treatment and control groups (RR: 1.03; 95% CI: 0.32–3.37), with significant heterogeneity observed among trials (I2 = 65.1%, p = 0.057; Figure 5). Subgroup analyses revealed no heterogeneity for CV events (Supplementary Table S5). Seven RCTs involving 1709 individuals reported 127 all-cause mortality events. There was no significant effect of oral alkali drug therapy on the risk of all-cause mortality compared with the control groups (RR: 0.90; 95% CI: 0.40–2.02). Significant heterogeneity was noted across the included trials (I2 = 54.7%, p = 0.05; Figure 5). No significant heterogeneity was found for all-cause mortality in the subgroup analyses (Supplementary Table S5). Forest plot for all-cause mortality and cardiovascular events. Cardiovascular events were defined as a composite, including fatal or non-fatal myocardial infarction, fatal or non-fatal stroke, coronary artery revascularization, cardiovascular disease and cardiovascular death. CI: confidence interval; N: number of trials; RR: relative risk. Data for CV events were available from three RCTs and one cohort studie that included 1098 participants and 160 events. There was no significant difference in the risk of CV events between treatment and control groups (RR: 1.03; 95% CI: 0.32–3.37), with significant heterogeneity observed among trials (I2 = 65.1%, p = 0.057; Figure 5). Subgroup analyses revealed no heterogeneity for CV events (Supplementary Table S5). Sensitivity analysis The results did not change after the exclusion of studies with a follow-up duration of < 12 months, with a sample size < 50, or assessed as low quality, or when different random-effects estimination methods were used (Supplementary Table S6). The results did not change after the exclusion of studies with a follow-up duration of < 12 months, with a sample size < 50, or assessed as low quality, or when different random-effects estimination methods were used (Supplementary Table S6). Overview of included trials: The literature search yielded 9284 potentially relevant records, of which the full texts of 185 publications were reviewed (Figure 1). After screening and eligibility assessment, 14 RCTs [5–7,20–32] and 4 cohort studies [8,29,33,34] reported in 19 publications with 3695 individuals were included in this systematic review and meta-analysis. Baseline and key characteristics of the enrolled studies are presented in Supplementary Table S1 and Table S2. The median follow-up time was 19.5 months. Individuals were enrolled at an average age of 55.78 years, and male participants accounted for 59.55% of the total. The average eGFR of participants was 31.51 mL/min/1.73 m2. A total of 2 oral alkali drug therapies were studied, including those featuring sodium bicarbonate in 17 studies, veverimer in 1 study,. PRISMA flow chart for the included studies. The Jadad score for each included RCT is presented in Supplementary Table S1. Eleven trials had a Jadad score of 3–5, and the others scored less than 3. Of all RCTs, 78.57% were associated with a low risk of bias arising from the randomization process, and all studies had a low risk of bias due to deviations from intended interventions, due to missing outcome data, associated with measurements of outcomes, and associated with selection of the reported results. In terms of overall bias, 78.57% of the research trials were assessed as at low risk, and 21.43% as subject to some concerns (Supplementary Table S3). As shown in Supplementary Tables S1 and S4, all included cohort studies were considered of high quality (with scores of 7★–8★) according to the NOS checklist. Effects of oral alkali drug therapy on renal outcomes: Nine RCTs with 1833 participants reported 250 renal failure events. Compared with the control group, oral alkali drug therapy was associated with a 55% reduction in the risk of renal failure events (RR: 0.45; 95% CI: 0.25–0.82), with significant heterogeneity across studies (I2 = 67.8%, p = 0.005; Figure 2). No significant heterogeneity was observed in any subgroup analysis (Table 1). Forest plot for renal failure events and decline in eGFR events. Renal failure was defined as a more than 50% decline in eGFR from baseline during follow-up, doubling of serum creatinine or ESRD. CI: confidence interval; RR: relative risk. Subgroup analysis of renal failure events. Note. a p value calculated by χ2 statistics was shown. CI: confidence interval; n: number of patients; RCT: randomized parallel-group controlled trial; RR: relative risk. Data regarding the effects of oral alkali drug therapy on decline in eGFR events were available from three RCTs that included 404 individuals and 123 events. Overall, there was no significant effect of oral alkali drug therapy on decline in eGFR events (RR: 0.34; 95% CI: 0.09–1.23) compared with the control group. Moderate heterogeneity across these trials (I2 =54.1%, p = 0.113; Figure 2) was found. Thirtine RCTs and three cohort studies with 2746 participants provided data on differences in the rate of change in eGFR. Compared with the control group, oral alkali drug therapy slowed the rate of eGFR decline by 2.59 mL/min/1.73 m2 per year (95% CI: 0.88–4.31), with significant heterogeneity observed (I2 = 97.6%, p < 0.001; Figure 3). Subgroup analyses showed that effect sizes were greater in studies that enrolled patients baseline serum bicarbonate < 20.95 mmol/L, baseline eGFR 30-59 mL/min/1.73 m2, age < 55 years (p < 0.001; Supplementary Table S5). Forest plot for rate of change in estimated glomerular filtration rate (eGFR). CI: confidence interval; MD: mean difference; SD: standard deviation. Data on the effects of oral alkali drug therapy on proteinuria or albuminuria were available in only five studies (four RCTs and one cohort study) with 591 participants, and no significant effect was found (SMD: −0.32; 95% CI: −1.08 to 0.43). I2 statistics (88.2%, p < 0.001; Figure 4) indicated significant heterogeneity across studies. Subgroup analyses did not reveal heterogeneity regarding pre-specified characteristics (Supplementary Table S5). Forest plot for the change in proteinuria or albuminuria. CI: confidence interval; SD: standard deviation; SMD: standard mean difference. Effects of oral alkali drug therapy on all-cause mortality and CV events: Seven RCTs involving 1709 individuals reported 127 all-cause mortality events. There was no significant effect of oral alkali drug therapy on the risk of all-cause mortality compared with the control groups (RR: 0.90; 95% CI: 0.40–2.02). Significant heterogeneity was noted across the included trials (I2 = 54.7%, p = 0.05; Figure 5). No significant heterogeneity was found for all-cause mortality in the subgroup analyses (Supplementary Table S5). Forest plot for all-cause mortality and cardiovascular events. Cardiovascular events were defined as a composite, including fatal or non-fatal myocardial infarction, fatal or non-fatal stroke, coronary artery revascularization, cardiovascular disease and cardiovascular death. CI: confidence interval; N: number of trials; RR: relative risk. Data for CV events were available from three RCTs and one cohort studie that included 1098 participants and 160 events. There was no significant difference in the risk of CV events between treatment and control groups (RR: 1.03; 95% CI: 0.32–3.37), with significant heterogeneity observed among trials (I2 = 65.1%, p = 0.057; Figure 5). Subgroup analyses revealed no heterogeneity for CV events (Supplementary Table S5). Sensitivity analysis: The results did not change after the exclusion of studies with a follow-up duration of < 12 months, with a sample size < 50, or assessed as low quality, or when different random-effects estimination methods were used (Supplementary Table S6). Discussion: MA, a common complication of CKD, is associated with CKD progression and higher mortality. The benefits of oral alkali drug supplementation for renal outcome events, all-cause mortality, and CV events in pre-dialysis CKD patients are controversial. This meta-analysis of 18 studies including 3695 participants suggests that oral alkali drug therapy produces a 55% reduction in renal failure events. No significant effects were observed for decline in eGFR, proteinuria, the risk of all-cause mortality events and CV events. The results were broadly consistent across major subgroups, as demonstrated by the sensitivity analyses. Of note, the existence of significant heterogeneity may limit the interpretation and clinical application of these results. Heterogeneity in the study included different CKD stages of patients, different baseline serum bicarbonate levels across studies, considerable variation in follow-up time, and different strategies to correct MA. Although we preformed subgroup analyses, which remains a concern for meaningful interpretation of the results. MA is associated with the progression of CKD [35,36]. However, there are sparse data on the effects of oral alkali drug supplementation on renal function in CKD patients, with inconsistent effects reported to date. In accordance with our study, a 2012 systematic review suggested that oral alkali therapy could slow the decline of the eGFR in patients with MA [37]. A 2019 systematic review indicated that oral alkali supplementation was associated with an improvement in eGFR and a reduction in the risk of progression to ESRD [38]. However, only two studies evaluated the effect of oral alkali therapy on the incidence of ESRD. Compared with previous meta-analyses [37,38], our study included many new studies on the effects of oral alkali therapy on changes in eGFR and ESRD events. Our summary data showed smaller reduction in kidney function decline, although with significant heterogeneity between the included studies. These data should be interpreted with caution. . . Additional well- designed trials are needed to explore the effect of treatment of MA on the risk of kidney disease progression with these different types of interventions. In the Chronic Renal Insufficiency Cohort study [2], the role of serum bicarbonate level as a risk factor for renal outcomes (ESRD or 50% reduction in eGFR) was evaluated in 3939 individuals with stage 2–4 CKD. After adjusting for covariates, the risk of developing a renal endpoint was 3% lower per 1 mmol/L increase in serum bicarbonate level [2]. A retrospective study from the National Health and Nutrition Examination Survey III involving 1486 CKD patients with a median 14.2 years of follow-up demonstrated that a higher dietary acid load was independently associated with an increased risk of ESRD, and this association was more pronounced in individuals with advanced CKD than in those with mild or moderate CKD [39]. Several studies included in this meta-analysis reported that oral alkali supplementation can delay the progression of CKD in patients with MA [28,29,32,35,37]. However, Mahajan [21] found that oral alkali supplementation delayed the progression of CKD in stage 2 CKD patients without MA. Wesson and Simoni [40] demonstrated that oral alkali dietary supplementation prevented eGFR decline in the two-thirds nephrectomy rat model (a model of early-stage CKD that does not include MA) compared with control rats over a 24-week period. This implies that mechanisms other than the correction of MA are involved in the renoprotective effect of oral alkali supplementation and raises important questions regarding the potential use of oral alkali supplementation in other conditions. Several potential mechanisms may be involved, including decreasing interstitial ammonium levels and reducing complement activation; the correction of interstitial acidosis and decreasing the local production of endothelin-1 and angiotensin II; decreasing tubular H+ secretion, which can limit tubular cast formation; activation of the cholinergic anti-inflammatory pathway and decreasing renal inflammation; or the correction of MA leading to enhanced blood glucose control [41]. The mechanisms underlying the renoprotective effect of oral alkali therapy need to be further explored. No high-quality study has assessed the effects of oral alkali therapy on CKD progression in patients with and without MA. Therefore, well-designed studies are needed. Both the BiCARB Study Group [7] and Raphael KL [28] studies failed to find a benefit of oral alkali therapy in terms of preventing an eGFR decline in CKD. In both studies, the mean age of participants was around 72.5 years, which may explain why some patients were not responsive to oral alkali therapy. The eGFR typically declines with age. It is possible that sodium bicarbonate is less effective in older patients with CKD compared to younger patients [7,28], which was consistent with the results of subgroup analyses. Further well-designed studies are needed to explore this. In our study, there was no compelling evidence that oral alkali drug therapy was associated with a lower incidence of all-cause mortality events and CV events. The scarcity of data on all-cause mortality events (seven studies with 127 events) and CV events (four studies with 160 events) available for the meta-analysis might have introduced a risk of false-negative results because of low statistical power. Among 740 individuals with 3 years of follow-up enrolled in the UBI study, the correction of MA reduced the risk of all-cause mortality in stage 3–5 CKD (fully adjusted hazard ratio: 0.36; 95% CI: 0.18–0.74) [5]. The Large Chronic Renal Insufficiency Cohort study showed that maintenance of a serum bicarbonate level >26 mmol/L was associated with increased risks of congestive heart failure events and mortality [42]. Numerous trials showed a U-shaped relationship between serum bicarbonate and mortality in patients with CKD [43,44]. The overall effects of alkali supplementation on all-cause mortality and CV events are uncertain. Chronic alkaline therapy for renoprotection may impact vascular calcification. In animal studies, alkali supplementation worsened arterial calcification [45,46]. Mixing sodium bicarbonate and calcium results in an insoluble precipitate, calcium carbonate (CaCO3). Supplementation with sodium bicarbonate increased the levels of serum phosphorous and FGF-23, risk factors for CV events and mortality [47,48] . In addition, a high sodium retention level is a cause for concern. There are 0.0123 mmol of sodium in every 1 mg of sodium bicarbonate, and high sodium can cause hypertension, a fluid overload, and an increased risk of heart failure in CKD patients [2]. Therefore, salt restrictions should be stricter in patients taking oral sodium bicarbonate. The optimal dosage of supplementary oral alkali drugs that provides renal and cardiovascular protection and minimizes side effects is uncertain. It is important to determine the optimal serum bicarbonate level and safe dose of oral alkali drugs according to CKD stage as well as monitor the serum bicarbonate level when an oral alkali supplement is given to pre-dialysis CKD patients. This study has several potential limitations. First, as a result of different abilities to regulate the acid-base balance according to CKD stage, different baseline serum bicarbonate levels across studies, the inclusion of some patients without MA at baseline, and different target serum bicarbonate levels across studies, the data do not provide insight into the safe and upper dosage limits of oral alkali supplementation or the optimal serum bicarbonate level for patients with CKD. Second, findings related to proteinuria, decline in eGFR, and all-cause mortality and CV events were based on limited studies, restricting the reliability of the results related to these outcomes. Third, the small sample size in some studies, as well as the existence of statistical heterogeneity and clinical heterogeneity, limited the reliability of our conclusions. Fourth, veverimer, a novel drug correcting MA, corrected MA by selectively binding and removing hydrochloric acid from the gastrointestinal tract, resulting in increased serum bicarbonate concentrations. However, there is only one RCT currently comparing veverimer and placebo. Futher studies including comparison between veverimer and sodium bicarbonate supplementation would be helpful to determine if there are any significant differences between the two strategies. Finally, none of the studies included patients with uncontrolled hypertension or obvious chronic heart failure. Whether oral alkali drug therapy is safe in these patients is unclear and needs to be further explored. In summary, based on the available and low-to-moderate certainty evidence, oral alkali drug therapy might potentially reduce the risk of kidney failure events, but no benefits in reducing all-cause mortality events, CV events, decline in eGFR and porteninuria. Notably, due to significant heterogeneity among studies the findings are not the final word. Further studies are needed to confirm these results for patients with CKD. Supplementary Material: Click here for additional data file.

Background: Metabolic acidosis accelerates the progression of chronic kidney disease (CKD) and increases the mortality rate. Whether oral alkali drug therapy benefits pre-dialysis CKD patients is controversial. We performed a meta-analysis of the effects of oral alkali drug therapy on major clinical outcomes in pre-dialysis CKD patients. Methods: We systematically searched MEDLINE using the Ovid, EMBASE, and Cochrane Library databases without language restriction. We included all eligible clinical studies that involved pre-dialysis CKD adults and compared those who received oral alkali drug therapy with controls. Results: A total of 18 eligible studies, including 14 randomized controlled trials and 4 cohort studies reported in 19 publications with 3695 participants, were included. Oral alkali drug therapy led to a 55% reduction in renal failure events (relative risk [RR]: 0.45; 95% confidence interval [CI]: 0.25-0.82), a rate of decline in the estimated glomerular filtration rate (eGFR) of 2.59 mL/min/1.73 m2 per year (95% CI, 0.88-4.31). There was no significant effect on decline in eGFR events (RR: 0.34; 95% CI: 0.09-1.23), proteinuria (standardized mean difference: -0.32; 95% CI: -1.08 to 0.43), all-cause mortality events (RR: 0.90; 95% CI: 0.40-2.02) and cardiovascular (CV) events (RR: 1.03; 95% CI: 0.32-3.37) compared with the control groups. Conclusions: Based on the available and low-to-moderate certainty evidence, oral alkali drug therapy might potentially reduce the risk of kidney failure events, but no benefit in reducing all-cause mortality events, CV events, decline in eGFR and porteninuria.

[CONTENT] Oral alkali drug therapy | pre-dialysis chronic kidney disease | meta-analysis | renal outcomes | all-cause mortality | cardiovascular events [SUMMARY]

[CONTENT] Oral alkali drug therapy | pre-dialysis chronic kidney disease | meta-analysis | renal outcomes | all-cause mortality | cardiovascular events [SUMMARY]

[CONTENT] Oral alkali drug therapy | pre-dialysis chronic kidney disease | meta-analysis | renal outcomes | all-cause mortality | cardiovascular events [SUMMARY]

[CONTENT] Acidosis | Administration, Oral | Adult | Alkalies | Cause of Death | Disease Progression | Glomerular Filtration Rate | Humans | Proteinuria | Randomized Controlled Trials as Topic | Renal Dialysis | Renal Insufficiency, Chronic [SUMMARY]

[CONTENT] Acidosis | Administration, Oral | Adult | Alkalies | Cause of Death | Disease Progression | Glomerular Filtration Rate | Humans | Proteinuria | Randomized Controlled Trials as Topic | Renal Dialysis | Renal Insufficiency, Chronic [SUMMARY]

[CONTENT] Acidosis | Administration, Oral | Adult | Alkalies | Cause of Death | Disease Progression | Glomerular Filtration Rate | Humans | Proteinuria | Randomized Controlled Trials as Topic | Renal Dialysis | Renal Insufficiency, Chronic [SUMMARY]

[CONTENT] oral bicarbonate supplementation | bicarbonate supplementation risk | effects alkali supplementation | supplementation renal outcomes | alkali supplementation renal [SUMMARY]

[CONTENT] oral bicarbonate supplementation | bicarbonate supplementation risk | effects alkali supplementation | supplementation renal outcomes | alkali supplementation renal [SUMMARY]

[CONTENT] oral bicarbonate supplementation | bicarbonate supplementation risk | effects alkali supplementation | supplementation renal outcomes | alkali supplementation renal [SUMMARY]

[CONTENT] studies | events | alkali | data | oral | oral alkali | egfr | risk | included | therapy [SUMMARY]

[CONTENT] studies | events | alkali | data | oral | oral alkali | egfr | risk | included | therapy [SUMMARY]

[CONTENT] studies | events | alkali | data | oral | oral alkali | egfr | risk | included | therapy [SUMMARY]

[CONTENT] ckd | ma | supplementation | ckd patients | patients | bicarbonate | mortality | progression | oral | clinical [SUMMARY]

[CONTENT] events | significant | ci | table | heterogeneity | supplementary | supplementary table | significant heterogeneity | figure | studies [SUMMARY]

[CONTENT] studies | events | data | risk | egfr | ckd | alkali | bias | oral | significant [SUMMARY]

[CONTENT] ||| ||| [SUMMARY]

[CONTENT] 18 | 14 | 4 | 19 | 3695 ||| 55% | 0.45 | 95% | CI | 0.25-0.82 | 2.59 mL | 95% | CI | 0.88 ||| 0.34 | 95% | CI | 0.09-1.23 | proteinuria ( | 95% | CI | -1.08 | 0.43 | 0.90 | 95% | CI | 0.40 | CV | 1.03 | 95% | CI | 0.32-3.37 [SUMMARY]

[CONTENT] ||| ||| ||| MEDLINE | Ovid | EMBASE | Cochrane Library ||| CKD ||| ||| 18 | 14 | 4 | 19 | 3695 ||| 55% | 0.45 | 95% | CI | 0.25-0.82 | 2.59 mL | 95% | CI | 0.88 ||| 0.34 | 95% | CI | 0.09-1.23 | proteinuria ( | 95% | CI | -1.08 | 0.43 | 0.90 | 95% | CI | 0.40 | CV | 1.03 | 95% | CI | 0.32-3.37 ||| CV [SUMMARY]

Survey about Intention to Engage in Specific Disaster Activities among Disaster Medical Assistance Team Members.

34658320

Different disaster activities should be performed smoothly. In relation to this, human resources for disaster activities must be secured. To achieve a stable supply of human resources, it is essential to improve the intentions of individuals responding to each type of disaster. However, the current intention of Disaster Medical Assistance Team (DMAT) members has not yet been assessed.

INTRODUCTION

An anonymous web questionnaire survey was conducted. Japanese DMAT members in the nuclear disaster-affected area (Group A; n = 79) and the non-affected area (Group N; n = 99) were included in the analysis. The outcome was the answer to the following question: "Will you actively engage in activities during natural, human-made, and chemical (C), biological (B), radiological/nuclear (R/N), and explosive (E) (CBRNE) disasters?" Then, questionnaire responses were compared according to disaster type.

METHODS

The intention to engage in C (50), B (47), R/N (58), and E (52) disasters was significantly lower than that in natural (82) and human-made (82) disasters (P <.001). The intention to engage in CBRNE disasters among younger participants (age ≤39 years) was significantly higher in Group A than in Group N. By contrast, the intention to engage in R/N disasters alone among older participants (age ≥40 years) was higher in Group A than in Group N. However, there was no difference between the two groups in terms of intention to engage in C, B, and E disasters. Moreover, the intention to engage in all disasters between younger and older participants in Group A did not differ. In Group N, older participants had a significantly higher intention to engage in B and R/N disasters.

RESULTS

Experience with a specific type of calamity at a young age may improve intention to engage in not only disasters encountered, but also other types. In addition, the intention to engage in CBRNE disasters improved with age in the non-experienced population. To respond smoothly to specific disasters in the future, measures must be taken to improve the intention to engage in CBRNE disasters among DMAT members.

CONCLUSION

8607140

Introduction

In patients with critical conditions, the initial response of the rapid response team or medical emergency team is the most important factor correlated with prognosis.1–3 In recent years, people have sustained injuries caused by different types of disasters, which can be classified as natural (ie, earthquakes), human-made (eg, transport accidents), and specific (ie, coronavirus disease 2019 [COVID-19] and chemical terrorism). Hence, the need to manage these disasters is increasing.4 Among them, chemical (C), biological (B), radiological (R), nuclear (N), and explosive (E) (CBRNE) disasters are considered specific. In such disasters, a rapid and smooth response is required to save the lives of patients. However, in the Fukushima Daiichi Nuclear Power Plant (FDNPP; Ōkuma, Fukushima, Japan) accident (2011), one of the most well-known radiological/nuclear (R/N) disasters, it was challenging to smoothly run disaster response activities at all times.5 Therefore, when providing medical treatment in areas with various hazards, measures should be taken in advance to facilitate disaster activities. With consideration of factors that can prevent a smooth response to CBRNE disasters, the lack of human resources is a major concern. Disaster responders have a low intention to engage in specific disaster activities. Some surveys have shown that even individuals who are willing to respond to natural hazards avoid involvement in nuclear disasters or those involving communicable diseases due to anxiety and lack of knowledge.6–8 Hence, this is a major cause for the lack of human resources and is associated with difficulties in facilitating CBRNE disaster activities. A previous study revealed that factors such as self-confidence, incentives, and family understanding affect the intention of firefighters to engage in nuclear disaster activities.9 However, the current intention of all medical responders to participate in CBRNE disaster activities has not been fully elucidated. In Japan, the Great Hanshin earthquake of 1995 has led to the development of a disaster medical system. Moreover, the Disaster Medical Assistance Team (DMAT), which responds to various disasters, has been established. The DMAT comprises physicians, nurses, and logisticians, as defined in the Basic Disaster Management Plan based on the Japan’s Disaster Countermeasures Basic Act.10,11 Japanese DMAT members can decide whether or not to respond when they receive dispatch requests. On the other hand, to date, the team plays a major role in large-scale disasters in Japan, and its members are the most important disaster medical responders in Japan. Therefore, a survey about the intention of DMAT members to engage in short-term CBRNE disaster activities must be conducted to facilitate a smooth response. This study aimed to conduct a web questionnaire survey among DMAT members from two different areas (one with nuclear disaster experience and the other without). To smoothen each specific disaster response, the current intentions of DMAT members to engage in CBRNE disaster activities were examined. Moreover, future measures that can improve such intentions were evaluated.

Methods

This was a cross-sectional study. An anonymous web questionnaire survey was conducted from October 2020 through November 2020. The website URL of the questionnaire was distributed by research members via e-mail to DMAT members in the two different areas. That is, one was a nuclear disaster-affected area (Group A) and the other was a non-affected area (Group N). In total, 204 participants from both areas responded. However, only 178 provided complete responses (effective response rate: total 87.3%; Group A 84.9%; Group N 89.2%). These data were then included in the analysis (Figure 1). The sample size was estimated using the pwr.anova.test function of R 4.0.3 software (R Foundation for Statistical Computing; Vienna, Austria). The following three parameters were included: group size (k = 4), medium effect (f = 0.25), and power (0.8). The estimated sample size was 45 per group; therefore, the total size of the response group was determined to be 180. The questionnaire was used to collect information such as sex, age, occupation, family status, and experience in disaster activities. To validate the intention to engage in disaster activities (natural, human-made, CBRNE disasters), the following question was created: “Will you actively engage in response activities during a natural, human-made, or CBRNE disaster?” The participants were required to answer using the Engagement Intent Score (EIS), which indicates their agreement to the abovementioned question (0%-100%). Participants with an EIS of <50% were instructed to provide a free answer as to why they did not wish to engage. Figure 1.Flow Chart Showing the Selection of Participants. Note: The web URL of the questionnaire was sent via e-mail; addresses were in the two DMAT mailing lists. That is, one list was for a nuclear disaster-affected area and the other was for a non-affected area. In total, 204 members answered the questionnaire. After excluding 26 incomplete response data, 178 participants were finally included in the analysis.Abbreviation: DMAT, Disaster Medical Assistance Team. Flow Chart Showing the Selection of Participants. Note: The web URL of the questionnaire was sent via e-mail; addresses were in the two DMAT mailing lists. That is, one list was for a nuclear disaster-affected area and the other was for a non-affected area. In total, 204 members answered the questionnaire. After excluding 26 incomplete response data, 178 participants were finally included in the analysis. Abbreviation: DMAT, Disaster Medical Assistance Team. The participants were divided into four groups according to age and area: younger Group A (≤39 years old; nuclear disaster-affected area), older Group A (≥40 years old; nuclear disaster-affected area), younger Group N (≤39 years old; non-affected area), and older Group N (≥40 years old; non-affected area; Figure 2). The reference age was set at 39 years as the mean age of DMAT members was 38.8 years, and a specific medical checkup is available for those aged >40 years in Japan.12,13 The background characteristics were compared between the four groups using the chi-squared test. Each EIS was presented as the mean and standard deviation (SD). The EIS between each disaster was compared with the analysis of variance and the Tukey-Kramer test for multiple comparisons. Sub-analysis was performed for male participants only. Further analyses were conducted according to age and area. The EIS was compared between the younger and older and nuclear disaster-affected and non-affected groups using the student’s t-test. All statistical analyses were performed using JMP 14 (SAS Institute Inc.; Cary, North Carolina USA), and a P value of .05 was considered statistically significant. Figure 2.Diagrammatic Image Representation of a Stratified Comparison. Note: The horizontal axis indicates age and the vertical axis represents disaster experience. Each comparison in Figures 3A, 3B, 4A, and 4B is depicted with a black bidirectional arrow. Abbreviations: Group A, nuclear disaster-affected area; Group N, non-affected area. Diagrammatic Image Representation of a Stratified Comparison. Note: The horizontal axis indicates age and the vertical axis represents disaster experience. Each comparison in Figures 3A, 3B, 4A, and 4B is depicted with a black bidirectional arrow. Abbreviations: Group A, nuclear disaster-affected area; Group N, non-affected area. Ethics Committee Approval The ethics committee of Fukushima Medical University approved the study protocol (Fukushima, Japan; approval number: 2020-130). The ethics committee of Fukushima Medical University approved the study protocol (Fukushima, Japan; approval number: 2020-130).

Results

The characteristics of younger participants in Group A (n = 28), younger participants in Group N (n = 56), older participants in Group A (n = 51), and older participants in Group N (n = 43) are depicted in Table 1. There were differences in terms of background characteristics between the four groups in terms of occupation and experience in natural disaster activities (Table 1). According to the primary outcome, the mean EIS for each type of disaster were as follows: natural, 82.2 (SD = 20.3); human-made, 81.7 (SD = 23.2); C, 50.0 (SD = 34.9); B, 47.4 (SD = 35.3); R/N, 57.6 (SD = 35.5); and E, 52.4 (SD = 36.1). After multiple comparisons, the proportion of natural and human-made disasters was significantly higher than that of C, B, R/N, and E disasters (all P values <.001). Furthermore, R/N disasters had a higher EIS than B disasters (P <.05; Table 2). In addition, a sub-analysis of only the male participants showed the same results (Supplemental Table 1; available online only). Table 1.Characteristics of the ParticipantsYounger Group A(n = 28)Younger Group N(n = 56)Older Group A(n = 51)Older Group N(n = 43) P ValueSex, n (%)  Female6 (21.4)17 (30.4)18 (35.3)9 (20.9).368  Male22 (78.6)39 (69.6)33 (64.7)34 (79.1)Age (years), n (%)  20-296 (21.4)8 (14.3)−−.408a   30-3922 (78.6)48 (85.7)−−  40-49−−33 (64.7)34 (79.1).125b   Over 50−−18 (35.3)9 (20.9)Occupation, n (%)  Physician3 (10.7)8 (14.3)17 (33.3)13 (30.2).048  Nurse7 (25.0)27 (48.2)20 (39.3)15 (34.9)  Administrative Staff7 (25.0)9 (16.1)7 (13.7)7 (16.3)  Others11 (39.3)12 (21.4)7 (13.7)8 (18.6)Family, n (%)  Without6 (21.4)16 (28.6)11 (21.6)10 (23.3).822  With22 (78.6)40 (71.4)40 (78.4)33 (76.7)Experience in Natural Disaster Activities, n (%)  No12 (42.9)24 (42.9)9 (17.6)10 (23.3).012  Yes16 (57.1)32 (57.1)42 (82.4)33 (76.7)Experience in CBRNE Disaster Activities, n (%)  No27 (96.4)53 (94.6)45 (88.2)39 (90.7).495  Yes1 (3.6)3 (5.4)6 (11.8)4 (9.3)Abbreviations: CBRNE, chemical, biological, radiological, nuclear, and explosive; Group A, nuclear disaster-affected area; Group N, non-affected area.a Comparison between younger Group A and younger Group N.b Comparison between older Group A and older Group N. Characteristics of the Participants Abbreviations: CBRNE, chemical, biological, radiological, nuclear, and explosive; Group A, nuclear disaster-affected area; Group N, non-affected area. Comparison between younger Group A and younger Group N. Comparison between older Group A and older Group N. Table 2.Multiple Comparison of EIS between the Six Types of DisastersNo.Disaster TypeMean (SD) EIS95% CI P Valuea vs. 2vs. 3vs. 4vs. 5vs. 61Natural82.3 (SD = 20.3)79.2-85.31.00<.01<.01<.01<.012Human-Made81.7 (SD = 23.2)78.3-85.1—<.01<.01<.01<.013Chemical50.0 (SD = 34.9)44.8-55.1——.97.21.984Biological47.4 (SD = 35.3)42.2-52.6———.03.675Radiological/Nuclear57.6 (SD = 35.5)52.3-62.8————.636Explosive52.4 (SD = 36.1)47.0-57.7—————Abbreviation: EIS, enga_gement intent score.a P values <.05 were considered statistically significant. Multiple Comparison of EIS between the Six Types of Disasters Abbreviation: EIS, enga_gement intent score. P values <.05 were considered statistically significant. Based on the intention to engage in various types of disasters, the EIS for all CBRNE disasters among younger participants was significantly higher in younger Group A than in younger Group N (C: 60.9 [SD = 30.8] versus 37.4 [SD = 32.3], P <.01; B: 55.3 [SD = 31.9] versus 35.5 [SD = 32.6], P <.01; R/N: 63.0 [SD = 31.9] versus 41.3 [SD = 35.0], P <.01; E: 61.1 [SD = 32.4] versus 44.6 [SD = 35.2], P <.05; Figure 3A and Table 3). Meanwhile, the EIS for R/N disasters alone among older participants was significantly higher in older Group A than in older Group N (72.1 [SD = 31.2] versus 58.0 [SD = 35.3]; P <.05), but those for other disasters (natural, human-made, C, B, and E) did not significantly differ between the two groups (Figure 3B and Table 3). According to age, there was no difference in the intention to engage in all types of disasters between younger and older participants in Group A (Figure 4A and Table 4). However, older participants in Group N had a significantly higher EIS for B (35.5 [SD = 32.6] versus 49.4 [SD = 35.7]; P <.05) and R/N (41.3 [SD = 35.0] versus 58.0 [SD = 35.3]; P <.05) disasters than younger participants in Group N (Figure 4B and Table 4). Figure 3.Comparison of Engagement Intent Score According to the Type of Disaster in Each Age Group. A. There was no significant difference between younger participants in Group A and Group N in terms of intention to engage in natural and human-made disasters. Group A had a significantly higher intention to engage in all CBRNE disaster activities. B. The score for radiological/nuclear disaster alone was significantly higher among older participants in Group A than in Group N. However, the results for other disasters, except radiological/nuclear ones, did not significantly differ between the two groups.Abbreviations: CBRNE, chemical, biological, radiological, nuclear, and explosive; Group A, nuclear disaster-affected area; Group N, non-affected area.* P <.05;** P <.01. Comparison of Engagement Intent Score According to the Type of Disaster in Each Age Group. A. There was no significant difference between younger participants in Group A and Group N in terms of intention to engage in natural and human-made disasters. Group A had a significantly higher intention to engage in all CBRNE disaster activities. B. The score for radiological/nuclear disaster alone was significantly higher among older participants in Group A than in Group N. However, the results for other disasters, except radiological/nuclear ones, did not significantly differ between the two groups. Abbreviations: CBRNE, chemical, biological, radiological, nuclear, and explosive; Group A, nuclear disaster-affected area; Group N, non-affected area. * P <.05; ** P <.01. Table 3.Comparison of EIS for Each Type of Disaster among the Same Age GroupDisaster TypeYounger Group A(n = 28)Younger Group N(n = 56) P Valuea Older Group A(n = 51)Older Group N(n = 43) P Valuea Natural78.0 (SD = 25.9)80.1 (SD = 20.9).6885.9 (SD = 17.8)83.4 (SD = 18.1).50Human-Made78.0 (SD = 27.1)82.9 (SD = 21.2).3681.2 (SD = 25.6)83.1 (SD = 20.2).70Chemical60.9 (SD = 30.8)37.4 (SD = 32.3)<.0157.6 (SD = 35.7)50.3 (SD = 35.6).33Biological55.3 (SD = 31.9)35.5 (SD = 32.6)<.0154.5 (SD = 37.3)49.4 (SD = 35.7).50Radiological/Nuclear63.0 (SD = 31.9)41.3 (SD = 35.0)<.0172.1 (SD = 31.2)58.0 (SD = 35.3).04Explosive61.1 (SD = 32.4)44.6 (SD = 35.2).0456.8 (SD = 36.4)51.6 (SD = 38.3).51Abbreviation: EIS, engagement intent score.a P values <.05 were considered statistically significant. Comparison of EIS for Each Type of Disaster among the Same Age Group Abbreviation: EIS, engagement intent score. P values <.05 were considered statistically significant. Figure 4.Comparison of Engagement Intent Score in Terms of the Type of Disasters in Each Group. A. There was no difference in the intention to engage in all types of disasters between younger and older participants in Group A. B. Older participants in Group N had a significantly higher intention to engage in biological and radiological/nuclear disaster activities. The same trend was observed for chemical disasters. However, the results did not significantly differ.Abbreviations: Group A, nuclear disaster-affected area; Group N, non-affected area.*P <.05. Comparison of Engagement Intent Score in Terms of the Type of Disasters in Each Group. A. There was no difference in the intention to engage in all types of disasters between younger and older participants in Group A. B. Older participants in Group N had a significantly higher intention to engage in biological and radiological/nuclear disaster activities. The same trend was observed for chemical disasters. However, the results did not significantly differ. Abbreviations: Group A, nuclear disaster-affected area; Group N, non-affected area. *P <.05. Table 4.Comparison of EIS for Each Type of Disaster among the Same AreaDisaster TypeYounger Group A(n = 28)Older Group A(n = 51) P Valuea Younger Group N(n = 56)Older Group N(n = 43) P Valuea Natural78.0 (SD = 25.9)85.9 (SD = 17.8).1180.1 (SD = 20.9)83.4 (SD = 18.1).42Human-Made78.0 (SD = 27.1)81.2 (SD = 25.6).6182.9 (SD = 21.2)83.1 (SD = 20.2).97Chemical60.9 (SD = 30.8)57.6 (SD = 35.7).6837.4 (SD = 35.5)50.3 (SD = 35.6).06Biological55.3 (SD = 31.9)54.5 (SD = 37.3).9335.5 (SD = 32.6)49.4 (SD = 35.7).05Radiological/Nuclear63.0 (SD = 31.9)72.1 (SD = 31.2).2241.3 (SD = 35.0)58.0 (SD = 35.3).02Explosive61.1 (SD = 32.4)56.8 (SD = 36.4).6044.6 (SD = 35.2)51.6 (SD = 38.3).34Abbreviation: EIS, engagement intent score.a P values <.05 were considered statistically significant. Comparison of EIS for Each Type of Disaster among the Same Area Abbreviation: EIS, engagement intent score. P values <.05 were considered statistically significant. According to the free answers, the DMAT members were not willing to engage in natural and human-made disasters mainly due to safety and the thought of leaving their family members behind in case of injuries or death. Nevertheless, the main reasons why DMAT members were not willing to engage in CBRN disaster activities were lack of knowledge and skills along with the anxiety and fear attributable to the fact that the R or C agents cannot be visualized with the naked eye. Meanwhile, the reasons for not engaging in E disasters slightly differed. That is, several participants answered lack of security due to the risk of second or third explosion.

Conclusion

Japanese DMAT members had a low intention to engage in CBRNE disaster activities compared with natural and human-made disaster activities. To respond smoothly to specific disasters, measures to efficiently improve the intention to engage in CBRNE disaster activities are required.

Introduction: In patients with critical conditions, the initial response of the rapid response team or medical emergency team is the most important factor correlated with prognosis.1–3 In recent years, people have sustained injuries caused by different types of disasters, which can be classified as natural (ie, earthquakes), human-made (eg, transport accidents), and specific (ie, coronavirus disease 2019 [COVID-19] and chemical terrorism). Hence, the need to manage these disasters is increasing.4 Among them, chemical (C), biological (B), radiological (R), nuclear (N), and explosive (E) (CBRNE) disasters are considered specific. In such disasters, a rapid and smooth response is required to save the lives of patients. However, in the Fukushima Daiichi Nuclear Power Plant (FDNPP; Ōkuma, Fukushima, Japan) accident (2011), one of the most well-known radiological/nuclear (R/N) disasters, it was challenging to smoothly run disaster response activities at all times.5 Therefore, when providing medical treatment in areas with various hazards, measures should be taken in advance to facilitate disaster activities. With consideration of factors that can prevent a smooth response to CBRNE disasters, the lack of human resources is a major concern. Disaster responders have a low intention to engage in specific disaster activities. Some surveys have shown that even individuals who are willing to respond to natural hazards avoid involvement in nuclear disasters or those involving communicable diseases due to anxiety and lack of knowledge.6–8 Hence, this is a major cause for the lack of human resources and is associated with difficulties in facilitating CBRNE disaster activities. A previous study revealed that factors such as self-confidence, incentives, and family understanding affect the intention of firefighters to engage in nuclear disaster activities.9 However, the current intention of all medical responders to participate in CBRNE disaster activities has not been fully elucidated. In Japan, the Great Hanshin earthquake of 1995 has led to the development of a disaster medical system. Moreover, the Disaster Medical Assistance Team (DMAT), which responds to various disasters, has been established. The DMAT comprises physicians, nurses, and logisticians, as defined in the Basic Disaster Management Plan based on the Japan’s Disaster Countermeasures Basic Act.10,11 Japanese DMAT members can decide whether or not to respond when they receive dispatch requests. On the other hand, to date, the team plays a major role in large-scale disasters in Japan, and its members are the most important disaster medical responders in Japan. Therefore, a survey about the intention of DMAT members to engage in short-term CBRNE disaster activities must be conducted to facilitate a smooth response. This study aimed to conduct a web questionnaire survey among DMAT members from two different areas (one with nuclear disaster experience and the other without). To smoothen each specific disaster response, the current intentions of DMAT members to engage in CBRNE disaster activities were examined. Moreover, future measures that can improve such intentions were evaluated. Methods: This was a cross-sectional study. An anonymous web questionnaire survey was conducted from October 2020 through November 2020. The website URL of the questionnaire was distributed by research members via e-mail to DMAT members in the two different areas. That is, one was a nuclear disaster-affected area (Group A) and the other was a non-affected area (Group N). In total, 204 participants from both areas responded. However, only 178 provided complete responses (effective response rate: total 87.3%; Group A 84.9%; Group N 89.2%). These data were then included in the analysis (Figure 1). The sample size was estimated using the pwr.anova.test function of R 4.0.3 software (R Foundation for Statistical Computing; Vienna, Austria). The following three parameters were included: group size (k = 4), medium effect (f = 0.25), and power (0.8). The estimated sample size was 45 per group; therefore, the total size of the response group was determined to be 180. The questionnaire was used to collect information such as sex, age, occupation, family status, and experience in disaster activities. To validate the intention to engage in disaster activities (natural, human-made, CBRNE disasters), the following question was created: “Will you actively engage in response activities during a natural, human-made, or CBRNE disaster?” The participants were required to answer using the Engagement Intent Score (EIS), which indicates their agreement to the abovementioned question (0%-100%). Participants with an EIS of <50% were instructed to provide a free answer as to why they did not wish to engage. Figure 1.Flow Chart Showing the Selection of Participants. Note: The web URL of the questionnaire was sent via e-mail; addresses were in the two DMAT mailing lists. That is, one list was for a nuclear disaster-affected area and the other was for a non-affected area. In total, 204 members answered the questionnaire. After excluding 26 incomplete response data, 178 participants were finally included in the analysis.Abbreviation: DMAT, Disaster Medical Assistance Team. Flow Chart Showing the Selection of Participants. Note: The web URL of the questionnaire was sent via e-mail; addresses were in the two DMAT mailing lists. That is, one list was for a nuclear disaster-affected area and the other was for a non-affected area. In total, 204 members answered the questionnaire. After excluding 26 incomplete response data, 178 participants were finally included in the analysis. Abbreviation: DMAT, Disaster Medical Assistance Team. The participants were divided into four groups according to age and area: younger Group A (≤39 years old; nuclear disaster-affected area), older Group A (≥40 years old; nuclear disaster-affected area), younger Group N (≤39 years old; non-affected area), and older Group N (≥40 years old; non-affected area; Figure 2). The reference age was set at 39 years as the mean age of DMAT members was 38.8 years, and a specific medical checkup is available for those aged >40 years in Japan.12,13 The background characteristics were compared between the four groups using the chi-squared test. Each EIS was presented as the mean and standard deviation (SD). The EIS between each disaster was compared with the analysis of variance and the Tukey-Kramer test for multiple comparisons. Sub-analysis was performed for male participants only. Further analyses were conducted according to age and area. The EIS was compared between the younger and older and nuclear disaster-affected and non-affected groups using the student’s t-test. All statistical analyses were performed using JMP 14 (SAS Institute Inc.; Cary, North Carolina USA), and a P value of .05 was considered statistically significant. Figure 2.Diagrammatic Image Representation of a Stratified Comparison. Note: The horizontal axis indicates age and the vertical axis represents disaster experience. Each comparison in Figures 3A, 3B, 4A, and 4B is depicted with a black bidirectional arrow. Abbreviations: Group A, nuclear disaster-affected area; Group N, non-affected area. Diagrammatic Image Representation of a Stratified Comparison. Note: The horizontal axis indicates age and the vertical axis represents disaster experience. Each comparison in Figures 3A, 3B, 4A, and 4B is depicted with a black bidirectional arrow. Abbreviations: Group A, nuclear disaster-affected area; Group N, non-affected area. Ethics Committee Approval The ethics committee of Fukushima Medical University approved the study protocol (Fukushima, Japan; approval number: 2020-130). The ethics committee of Fukushima Medical University approved the study protocol (Fukushima, Japan; approval number: 2020-130). Ethics Committee Approval: The ethics committee of Fukushima Medical University approved the study protocol (Fukushima, Japan; approval number: 2020-130). Results: The characteristics of younger participants in Group A (n = 28), younger participants in Group N (n = 56), older participants in Group A (n = 51), and older participants in Group N (n = 43) are depicted in Table 1. There were differences in terms of background characteristics between the four groups in terms of occupation and experience in natural disaster activities (Table 1). According to the primary outcome, the mean EIS for each type of disaster were as follows: natural, 82.2 (SD = 20.3); human-made, 81.7 (SD = 23.2); C, 50.0 (SD = 34.9); B, 47.4 (SD = 35.3); R/N, 57.6 (SD = 35.5); and E, 52.4 (SD = 36.1). After multiple comparisons, the proportion of natural and human-made disasters was significantly higher than that of C, B, R/N, and E disasters (all P values <.001). Furthermore, R/N disasters had a higher EIS than B disasters (P <.05; Table 2). In addition, a sub-analysis of only the male participants showed the same results (Supplemental Table 1; available online only). Table 1.Characteristics of the ParticipantsYounger Group A(n = 28)Younger Group N(n = 56)Older Group A(n = 51)Older Group N(n = 43) P ValueSex, n (%)  Female6 (21.4)17 (30.4)18 (35.3)9 (20.9).368  Male22 (78.6)39 (69.6)33 (64.7)34 (79.1)Age (years), n (%)  20-296 (21.4)8 (14.3)−−.408a   30-3922 (78.6)48 (85.7)−−  40-49−−33 (64.7)34 (79.1).125b   Over 50−−18 (35.3)9 (20.9)Occupation, n (%)  Physician3 (10.7)8 (14.3)17 (33.3)13 (30.2).048  Nurse7 (25.0)27 (48.2)20 (39.3)15 (34.9)  Administrative Staff7 (25.0)9 (16.1)7 (13.7)7 (16.3)  Others11 (39.3)12 (21.4)7 (13.7)8 (18.6)Family, n (%)  Without6 (21.4)16 (28.6)11 (21.6)10 (23.3).822  With22 (78.6)40 (71.4)40 (78.4)33 (76.7)Experience in Natural Disaster Activities, n (%)  No12 (42.9)24 (42.9)9 (17.6)10 (23.3).012  Yes16 (57.1)32 (57.1)42 (82.4)33 (76.7)Experience in CBRNE Disaster Activities, n (%)  No27 (96.4)53 (94.6)45 (88.2)39 (90.7).495  Yes1 (3.6)3 (5.4)6 (11.8)4 (9.3)Abbreviations: CBRNE, chemical, biological, radiological, nuclear, and explosive; Group A, nuclear disaster-affected area; Group N, non-affected area.a Comparison between younger Group A and younger Group N.b Comparison between older Group A and older Group N. Characteristics of the Participants Abbreviations: CBRNE, chemical, biological, radiological, nuclear, and explosive; Group A, nuclear disaster-affected area; Group N, non-affected area. Comparison between younger Group A and younger Group N. Comparison between older Group A and older Group N. Table 2.Multiple Comparison of EIS between the Six Types of DisastersNo.Disaster TypeMean (SD) EIS95% CI P Valuea vs. 2vs. 3vs. 4vs. 5vs. 61Natural82.3 (SD = 20.3)79.2-85.31.00<.01<.01<.01<.012Human-Made81.7 (SD = 23.2)78.3-85.1—<.01<.01<.01<.013Chemical50.0 (SD = 34.9)44.8-55.1——.97.21.984Biological47.4 (SD = 35.3)42.2-52.6———.03.675Radiological/Nuclear57.6 (SD = 35.5)52.3-62.8————.636Explosive52.4 (SD = 36.1)47.0-57.7—————Abbreviation: EIS, enga_gement intent score.a P values <.05 were considered statistically significant. Multiple Comparison of EIS between the Six Types of Disasters Abbreviation: EIS, enga_gement intent score. P values <.05 were considered statistically significant. Based on the intention to engage in various types of disasters, the EIS for all CBRNE disasters among younger participants was significantly higher in younger Group A than in younger Group N (C: 60.9 [SD = 30.8] versus 37.4 [SD = 32.3], P <.01; B: 55.3 [SD = 31.9] versus 35.5 [SD = 32.6], P <.01; R/N: 63.0 [SD = 31.9] versus 41.3 [SD = 35.0], P <.01; E: 61.1 [SD = 32.4] versus 44.6 [SD = 35.2], P <.05; Figure 3A and Table 3). Meanwhile, the EIS for R/N disasters alone among older participants was significantly higher in older Group A than in older Group N (72.1 [SD = 31.2] versus 58.0 [SD = 35.3]; P <.05), but those for other disasters (natural, human-made, C, B, and E) did not significantly differ between the two groups (Figure 3B and Table 3). According to age, there was no difference in the intention to engage in all types of disasters between younger and older participants in Group A (Figure 4A and Table 4). However, older participants in Group N had a significantly higher EIS for B (35.5 [SD = 32.6] versus 49.4 [SD = 35.7]; P <.05) and R/N (41.3 [SD = 35.0] versus 58.0 [SD = 35.3]; P <.05) disasters than younger participants in Group N (Figure 4B and Table 4). Figure 3.Comparison of Engagement Intent Score According to the Type of Disaster in Each Age Group. A. There was no significant difference between younger participants in Group A and Group N in terms of intention to engage in natural and human-made disasters. Group A had a significantly higher intention to engage in all CBRNE disaster activities. B. The score for radiological/nuclear disaster alone was significantly higher among older participants in Group A than in Group N. However, the results for other disasters, except radiological/nuclear ones, did not significantly differ between the two groups.Abbreviations: CBRNE, chemical, biological, radiological, nuclear, and explosive; Group A, nuclear disaster-affected area; Group N, non-affected area.* P <.05;** P <.01. Comparison of Engagement Intent Score According to the Type of Disaster in Each Age Group. A. There was no significant difference between younger participants in Group A and Group N in terms of intention to engage in natural and human-made disasters. Group A had a significantly higher intention to engage in all CBRNE disaster activities. B. The score for radiological/nuclear disaster alone was significantly higher among older participants in Group A than in Group N. However, the results for other disasters, except radiological/nuclear ones, did not significantly differ between the two groups. Abbreviations: CBRNE, chemical, biological, radiological, nuclear, and explosive; Group A, nuclear disaster-affected area; Group N, non-affected area. * P <.05; ** P <.01. Table 3.Comparison of EIS for Each Type of Disaster among the Same Age GroupDisaster TypeYounger Group A(n = 28)Younger Group N(n = 56) P Valuea Older Group A(n = 51)Older Group N(n = 43) P Valuea Natural78.0 (SD = 25.9)80.1 (SD = 20.9).6885.9 (SD = 17.8)83.4 (SD = 18.1).50Human-Made78.0 (SD = 27.1)82.9 (SD = 21.2).3681.2 (SD = 25.6)83.1 (SD = 20.2).70Chemical60.9 (SD = 30.8)37.4 (SD = 32.3)<.0157.6 (SD = 35.7)50.3 (SD = 35.6).33Biological55.3 (SD = 31.9)35.5 (SD = 32.6)<.0154.5 (SD = 37.3)49.4 (SD = 35.7).50Radiological/Nuclear63.0 (SD = 31.9)41.3 (SD = 35.0)<.0172.1 (SD = 31.2)58.0 (SD = 35.3).04Explosive61.1 (SD = 32.4)44.6 (SD = 35.2).0456.8 (SD = 36.4)51.6 (SD = 38.3).51Abbreviation: EIS, engagement intent score.a P values <.05 were considered statistically significant. Comparison of EIS for Each Type of Disaster among the Same Age Group Abbreviation: EIS, engagement intent score. P values <.05 were considered statistically significant. Figure 4.Comparison of Engagement Intent Score in Terms of the Type of Disasters in Each Group. A. There was no difference in the intention to engage in all types of disasters between younger and older participants in Group A. B. Older participants in Group N had a significantly higher intention to engage in biological and radiological/nuclear disaster activities. The same trend was observed for chemical disasters. However, the results did not significantly differ.Abbreviations: Group A, nuclear disaster-affected area; Group N, non-affected area.*P <.05. Comparison of Engagement Intent Score in Terms of the Type of Disasters in Each Group. A. There was no difference in the intention to engage in all types of disasters between younger and older participants in Group A. B. Older participants in Group N had a significantly higher intention to engage in biological and radiological/nuclear disaster activities. The same trend was observed for chemical disasters. However, the results did not significantly differ. Abbreviations: Group A, nuclear disaster-affected area; Group N, non-affected area. *P <.05. Table 4.Comparison of EIS for Each Type of Disaster among the Same AreaDisaster TypeYounger Group A(n = 28)Older Group A(n = 51) P Valuea Younger Group N(n = 56)Older Group N(n = 43) P Valuea Natural78.0 (SD = 25.9)85.9 (SD = 17.8).1180.1 (SD = 20.9)83.4 (SD = 18.1).42Human-Made78.0 (SD = 27.1)81.2 (SD = 25.6).6182.9 (SD = 21.2)83.1 (SD = 20.2).97Chemical60.9 (SD = 30.8)57.6 (SD = 35.7).6837.4 (SD = 35.5)50.3 (SD = 35.6).06Biological55.3 (SD = 31.9)54.5 (SD = 37.3).9335.5 (SD = 32.6)49.4 (SD = 35.7).05Radiological/Nuclear63.0 (SD = 31.9)72.1 (SD = 31.2).2241.3 (SD = 35.0)58.0 (SD = 35.3).02Explosive61.1 (SD = 32.4)56.8 (SD = 36.4).6044.6 (SD = 35.2)51.6 (SD = 38.3).34Abbreviation: EIS, engagement intent score.a P values <.05 were considered statistically significant. Comparison of EIS for Each Type of Disaster among the Same Area Abbreviation: EIS, engagement intent score. P values <.05 were considered statistically significant. According to the free answers, the DMAT members were not willing to engage in natural and human-made disasters mainly due to safety and the thought of leaving their family members behind in case of injuries or death. Nevertheless, the main reasons why DMAT members were not willing to engage in CBRN disaster activities were lack of knowledge and skills along with the anxiety and fear attributable to the fact that the R or C agents cannot be visualized with the naked eye. Meanwhile, the reasons for not engaging in E disasters slightly differed. That is, several participants answered lack of security due to the risk of second or third explosion. Discussion: This study aimed to assess and elucidate the response provided by DMAT members to CBRNE disasters; this is considering the fact that an organized response is imperative in disaster management. To achieve the goal, this study initially focused on securing human resources. As a first step toward this goal, this study investigated the DMAT’s intention to engage in various types of disaster activities. Some questionnaire surveys asked about the intention to engage in disaster activities. However, most studies were based on a two-to-ten scale of responses to the questionnaire. To the best of the authors’ knowledge, this is the first survey about the intention of DMAT members to participate in a specific disaster activity, and the intention was scored on a scale of 100%. Using a continuous scale, the intention to engage in disaster activities could be accurately evaluated. In previous studies, approximately 40% of medical personnel or firefighter-training students in a non-affected area6,7 and 56.3% in a nuclear disaster-affected area9 were willing to engage in R/N disaster activities. Based on these reports, disaster responders in a nuclear disaster-affected area had a higher intention to engage in R/N disasters, and this finding is understandable. With actual experience in helping individuals affected by a disaster, people can accept things as their own, and their interest in disaster activities increases. A greater interest leads to intention to take action.14 In this study, both age groups who experienced nuclear disasters had a higher EIS for R/N disasters, and the result was comparable to that of previous reports about firefighters in nuclear disaster areas. This finding indicated that experiences with disasters might have a positive influence on intention among not only firefighters, but also DMAT members. As shown in Table 2, the EIS of CBRNE disasters was significantly lower than that of natural or human-made disasters. Moreover, the SD of the EIS in each CBRNE disaster was higher than that of natural or human-made disasters. This result indicates that the intention to engage in CBRNE disasters varies according to the member’s experience, knowledge, or skill; however, difference in the intention to engage in natural and human-made disasters appears to be less. According to the free answer, there was a trend for the reasons for not willing to engage in disaster activities. For natural and human-made disasters, the main reasons were safety and leaving family members behind in case of critical injuries or death. However, the main reasons for not willing to engage in CBRN disasters were lack of knowledge and skills, anxiety, and fear since these disasters cannot be prevented. The reasons for not willing to engage in E disasters slightly differed, and the most common reason was the lack of assurance regarding safety from the effects of second or third impact. Thus, DMAT members must have appropriate knowledge and skills to help improve their intention to engage in CBRNE disaster activities. Biological disasters had the lowest EIS. This result may be attributed to the recent COVID-19 pandemic. That is, it is similar to a previous survey, which reported that there is a trend toward lower intention to engage in situations involving communicable diseases compared with R/N hazards.7 Japan experienced large C, R/N, or E disasters, such as the Tokyo subway sarin attack (1995), FDNPP accident (2011), and atomic bomb explosion during World War II (1945).5,15,16 By contrast, within the previous decades, there was no significant B disaster, except the COVID-19 pandemic. Moreover, lack of knowledge to disaster-related matters may result in an increase in the DMAT members’ perception of risks, which may consequently lead to avoidance in engaging to disaster activities.17 Based on these aspects, the fact of unknown might lead to a lack of disaster response image for this study population, and this might affect the result of low EIS in B disasters. In the younger group (age: ≤39 years), the EIS was significantly higher in Group A than in Group N for all CBRNE disasters. It is easy to imagine that those who have experienced R/N disasters (Group A) have a higher EIS for R/N disasters. However, the result showed that they also had a higher EIS for C, B, and E disasters. In other words, the results indicated that the experience with R/N disasters had a positive impact on the intention to engage in other CBRNE disasters. Thus, an experience with one specific disaster may help develop adaptability or gain confidence to face other specific disasters in general, thereby improving one’s intention to engage to such disasters. In this study, the reason for the abovementioned phenomenon has not been validated. However, authors have considered the high-risk perception of young-aged Japanese population. A previous study has revealed that there is a relationship between behavior and risk perception.18 The latter is strongly influenced by not only experience, but also culture or nationality, and Japanese are known to have a higher-risk perception than other nationalities.19,20 Moreover, age affects risk perception.19 Individuals may perceive CBRNE disasters as similar based on experiences with a certain type of CBRNE disaster at a young age. Moreover, they can lower their risk perception for other CBRNE disasters. By contrast, the EIS for R/N disasters alone in older participants was significantly higher in Group A than in Group N, and there was no difference in the EIS for other CBRNE disasters between the two older groups. In the sub-analysis comparing the EIS of younger participants in Group A and older participants in Group N, there was no significant difference according to all types of disasters. Group A experienced R/N disasters at a young age, indicating that they acquired knowledge and experience at an early stage, which could have developed over a long period of time. As shown in Table 1, the background characteristics significantly differed in terms of occupation and natural disaster activity experience in each group. This study compared the EIS between the two groups according to each factor because it is impossible to exclude background characteristics if a simultaneous comparison of the four groups is performed. Younger participants in Group N had a low EIS for CBRNE disasters. Thus, EIS increases with age and experience. Assessing what they have gained through aging or experience will help to identify the important points for interventions to elevate the EIS of individuals who engage in disaster activities in the future. There are several possible factors with consideration of age and experience. These include acquisition of knowledge and skills or interest. However, further research on specific measures should be conducted. By efficiently and accurately addressing these factors, the number of people who will engage in disaster activities will increase. Then, this will lead to a stable supply of human resources in the future. Limitations: The website URL of the questionnaire was sent to the registered e-mail addresses in the mailing list. However, whether all DMAT members have received the e-mail was not validated. Moreover, some people might have registered more than one e-mail address. Meanwhile, others might not have received the e-mail due to changes in address. Therefore, the actual collection rate was not verified. This survey was conducted on DMAT members from two areas only. Thus, DMAT members from other areas or individuals with other occupations that might involve engagement in various disasters were not included. Hence, further surveys must be conducted. Conclusion: Japanese DMAT members had a low intention to engage in CBRNE disaster activities compared with natural and human-made disaster activities. To respond smoothly to specific disasters, measures to efficiently improve the intention to engage in CBRNE disaster activities are required.

Background: Different disaster activities should be performed smoothly. In relation to this, human resources for disaster activities must be secured. To achieve a stable supply of human resources, it is essential to improve the intentions of individuals responding to each type of disaster. However, the current intention of Disaster Medical Assistance Team (DMAT) members has not yet been assessed. Methods: An anonymous web questionnaire survey was conducted. Japanese DMAT members in the nuclear disaster-affected area (Group A; n = 79) and the non-affected area (Group N; n = 99) were included in the analysis. The outcome was the answer to the following question: "Will you actively engage in activities during natural, human-made, and chemical (C), biological (B), radiological/nuclear (R/N), and explosive (E) (CBRNE) disasters?" Then, questionnaire responses were compared according to disaster type. Results: The intention to engage in C (50), B (47), R/N (58), and E (52) disasters was significantly lower than that in natural (82) and human-made (82) disasters (P <.001). The intention to engage in CBRNE disasters among younger participants (age ≤39 years) was significantly higher in Group A than in Group N. By contrast, the intention to engage in R/N disasters alone among older participants (age ≥40 years) was higher in Group A than in Group N. However, there was no difference between the two groups in terms of intention to engage in C, B, and E disasters. Moreover, the intention to engage in all disasters between younger and older participants in Group A did not differ. In Group N, older participants had a significantly higher intention to engage in B and R/N disasters. Conclusions: Experience with a specific type of calamity at a young age may improve intention to engage in not only disasters encountered, but also other types. In addition, the intention to engage in CBRNE disasters improved with age in the non-experienced population. To respond smoothly to specific disasters in the future, measures must be taken to improve the intention to engage in CBRNE disasters among DMAT members.

Introduction: In patients with critical conditions, the initial response of the rapid response team or medical emergency team is the most important factor correlated with prognosis.1–3 In recent years, people have sustained injuries caused by different types of disasters, which can be classified as natural (ie, earthquakes), human-made (eg, transport accidents), and specific (ie, coronavirus disease 2019 [COVID-19] and chemical terrorism). Hence, the need to manage these disasters is increasing.4 Among them, chemical (C), biological (B), radiological (R), nuclear (N), and explosive (E) (CBRNE) disasters are considered specific. In such disasters, a rapid and smooth response is required to save the lives of patients. However, in the Fukushima Daiichi Nuclear Power Plant (FDNPP; Ōkuma, Fukushima, Japan) accident (2011), one of the most well-known radiological/nuclear (R/N) disasters, it was challenging to smoothly run disaster response activities at all times.5 Therefore, when providing medical treatment in areas with various hazards, measures should be taken in advance to facilitate disaster activities. With consideration of factors that can prevent a smooth response to CBRNE disasters, the lack of human resources is a major concern. Disaster responders have a low intention to engage in specific disaster activities. Some surveys have shown that even individuals who are willing to respond to natural hazards avoid involvement in nuclear disasters or those involving communicable diseases due to anxiety and lack of knowledge.6–8 Hence, this is a major cause for the lack of human resources and is associated with difficulties in facilitating CBRNE disaster activities. A previous study revealed that factors such as self-confidence, incentives, and family understanding affect the intention of firefighters to engage in nuclear disaster activities.9 However, the current intention of all medical responders to participate in CBRNE disaster activities has not been fully elucidated. In Japan, the Great Hanshin earthquake of 1995 has led to the development of a disaster medical system. Moreover, the Disaster Medical Assistance Team (DMAT), which responds to various disasters, has been established. The DMAT comprises physicians, nurses, and logisticians, as defined in the Basic Disaster Management Plan based on the Japan’s Disaster Countermeasures Basic Act.10,11 Japanese DMAT members can decide whether or not to respond when they receive dispatch requests. On the other hand, to date, the team plays a major role in large-scale disasters in Japan, and its members are the most important disaster medical responders in Japan. Therefore, a survey about the intention of DMAT members to engage in short-term CBRNE disaster activities must be conducted to facilitate a smooth response. This study aimed to conduct a web questionnaire survey among DMAT members from two different areas (one with nuclear disaster experience and the other without). To smoothen each specific disaster response, the current intentions of DMAT members to engage in CBRNE disaster activities were examined. Moreover, future measures that can improve such intentions were evaluated. Conclusion: Japanese DMAT members had a low intention to engage in CBRNE disaster activities compared with natural and human-made disaster activities. To respond smoothly to specific disasters, measures to efficiently improve the intention to engage in CBRNE disaster activities are required.

Background: Different disaster activities should be performed smoothly. In relation to this, human resources for disaster activities must be secured. To achieve a stable supply of human resources, it is essential to improve the intentions of individuals responding to each type of disaster. However, the current intention of Disaster Medical Assistance Team (DMAT) members has not yet been assessed. Methods: An anonymous web questionnaire survey was conducted. Japanese DMAT members in the nuclear disaster-affected area (Group A; n = 79) and the non-affected area (Group N; n = 99) were included in the analysis. The outcome was the answer to the following question: "Will you actively engage in activities during natural, human-made, and chemical (C), biological (B), radiological/nuclear (R/N), and explosive (E) (CBRNE) disasters?" Then, questionnaire responses were compared according to disaster type. Results: The intention to engage in C (50), B (47), R/N (58), and E (52) disasters was significantly lower than that in natural (82) and human-made (82) disasters (P <.001). The intention to engage in CBRNE disasters among younger participants (age ≤39 years) was significantly higher in Group A than in Group N. By contrast, the intention to engage in R/N disasters alone among older participants (age ≥40 years) was higher in Group A than in Group N. However, there was no difference between the two groups in terms of intention to engage in C, B, and E disasters. Moreover, the intention to engage in all disasters between younger and older participants in Group A did not differ. In Group N, older participants had a significantly higher intention to engage in B and R/N disasters. Conclusions: Experience with a specific type of calamity at a young age may improve intention to engage in not only disasters encountered, but also other types. In addition, the intention to engage in CBRNE disasters improved with age in the non-experienced population. To respond smoothly to specific disasters in the future, measures must be taken to improve the intention to engage in CBRNE disasters among DMAT members.

[CONTENT] disaster | emergency responders | hazard | human resources | intention [SUMMARY]

[CONTENT] disaster | emergency responders | hazard | human resources | intention [SUMMARY]

[CONTENT] disaster | emergency responders | hazard | human resources | intention [SUMMARY]

[CONTENT] disaster | emergency responders | hazard | human resources | intention [SUMMARY]

[CONTENT] disaster | emergency responders | hazard | human resources | intention [SUMMARY]

[CONTENT] disaster | emergency responders | hazard | human resources | intention [SUMMARY]

[CONTENT] Adult | Disaster Planning | Disasters | Humans | Intention | Medical Assistance | Surveys and Questionnaires | Workforce [SUMMARY]

[CONTENT] Adult | Disaster Planning | Disasters | Humans | Intention | Medical Assistance | Surveys and Questionnaires | Workforce [SUMMARY]

[CONTENT] Adult | Disaster Planning | Disasters | Humans | Intention | Medical Assistance | Surveys and Questionnaires | Workforce [SUMMARY]

[CONTENT] Adult | Disaster Planning | Disasters | Humans | Intention | Medical Assistance | Surveys and Questionnaires | Workforce [SUMMARY]

[CONTENT] Adult | Disaster Planning | Disasters | Humans | Intention | Medical Assistance | Surveys and Questionnaires | Workforce [SUMMARY]

[CONTENT] Adult | Disaster Planning | Disasters | Humans | Intention | Medical Assistance | Surveys and Questionnaires | Workforce [SUMMARY]

[CONTENT] nuclear disaster activities | firefighters nuclear disaster | disaster medical | radiological nuclear disasters | important disaster medical [SUMMARY]

[CONTENT] nuclear disaster activities | firefighters nuclear disaster | disaster medical | radiological nuclear disasters | important disaster medical [SUMMARY]

[CONTENT] nuclear disaster activities | firefighters nuclear disaster | disaster medical | radiological nuclear disasters | important disaster medical [SUMMARY]

[CONTENT] nuclear disaster activities | firefighters nuclear disaster | disaster medical | radiological nuclear disasters | important disaster medical [SUMMARY]

[CONTENT] nuclear disaster activities | firefighters nuclear disaster | disaster medical | radiological nuclear disasters | important disaster medical [SUMMARY]

[CONTENT] nuclear disaster activities | firefighters nuclear disaster | disaster medical | radiological nuclear disasters | important disaster medical [SUMMARY]

[CONTENT] group | disaster | sd | disasters | eis | engage | nuclear | participants | affected | activities [SUMMARY]

[CONTENT] group | disaster | sd | disasters | eis | engage | nuclear | participants | affected | activities [SUMMARY]

[CONTENT] group | disaster | sd | disasters | eis | engage | nuclear | participants | affected | activities [SUMMARY]

[CONTENT] group | disaster | sd | disasters | eis | engage | nuclear | participants | affected | activities [SUMMARY]

[CONTENT] group | disaster | sd | disasters | eis | engage | nuclear | participants | affected | activities [SUMMARY]

[CONTENT] group | disaster | sd | disasters | eis | engage | nuclear | participants | affected | activities [SUMMARY]

[CONTENT] disaster | response | activities | disasters | disaster activities | nuclear | medical | team | cbrne | japan [SUMMARY]

[CONTENT] area | affected | group | affected area | disaster | participants | non affected area | non | disaster affected | non affected [SUMMARY]

[CONTENT] sd | group | 35 | sd 35 | older | participants | disaster | younger | 05 | participants group [SUMMARY]

[CONTENT] activities | disaster activities | disaster | intention engage cbrne disaster | intention engage cbrne | cbrne disaster activities | engage cbrne disaster activities | engage cbrne | engage cbrne disaster | engage [SUMMARY]

[CONTENT] disaster | group | disasters | activities | sd | disaster activities | engage | cbrne | intention | dmat [SUMMARY]

[CONTENT] disaster | group | disasters | activities | sd | disaster activities | engage | cbrne | intention | dmat [SUMMARY]

[CONTENT] ||| ||| ||| Disaster Medical Assistance Team (DMAT [SUMMARY]

[CONTENT] ||| Japanese | DMAT | Group A | 79 | Group N | 99 ||| ||| [SUMMARY]

[CONTENT] 50 | R/N | 58 | 52 | 82 | 82 ||| CBRNE | age ≤39 years | Group A | Group N. By | Group A ||| Group N. However | two ||| Group A ||| Group N [SUMMARY]

[CONTENT] ||| CBRNE ||| CBRNE | DMAT [SUMMARY]

[CONTENT] ||| ||| ||| Disaster Medical Assistance Team (DMAT ||| ||| Japanese | DMAT | Group A | 79 | Group N | 99 ||| ||| ||| ||| 50 | R/N | 58 | 52 | 82 | 82 ||| CBRNE | age ≤39 years | Group A | Group N. By | Group A ||| Group N. However | two ||| Group A ||| Group N ||| ||| CBRNE ||| CBRNE | DMAT [SUMMARY]

[CONTENT] ||| ||| ||| Disaster Medical Assistance Team (DMAT ||| ||| Japanese | DMAT | Group A | 79 | Group N | 99 ||| ||| ||| ||| 50 | R/N | 58 | 52 | 82 | 82 ||| CBRNE | age ≤39 years | Group A | Group N. By | Group A ||| Group N. However | two ||| Group A ||| Group N ||| ||| CBRNE ||| CBRNE | DMAT [SUMMARY]

Comparison of 4 Methods for Dynamization of Locking Plates: Differences in the Amount and Type of Fracture Motion.

28657927

Decreasing the stiffness of locked plating constructs can promote natural fracture healing by controlled dynamization of the fracture. This biomechanical study compared the effect of 4 different stiffness reduction methods on interfragmentary motion by measuring axial motion and shear motion at the fracture site.

BACKGROUND

Distal femur locking plates were applied to bridge a metadiaphyseal fracture in femur surrogates. A locked construct with a short-bridge span served as the nondynamized control group (LOCKED). Four different methods for stiffness reduction were evaluated: replacing diaphyseal locking screws with nonlocked screws (NONLOCKED); bridge dynamization (BRIDGE) with 2 empty screw holes proximal to the fracture; screw dynamization with far cortical locking (FCL) screws; and plate dynamization with active locking plates (ACTIVE). Construct stiffness, axial motion, and shear motion at the fracture site were measured to characterize each dynamization methods.

METHODS

Compared with LOCKED control constructs, NONLOCKED constructs had a similar stiffness (P = 0.08), axial motion (P = 0.07), and shear motion (P = 0.97). BRIDGE constructs reduced stiffness by 45% compared with LOCKED constructs (P < 0.001), but interfragmentary motion was dominated by shear. Compared with LOCKED constructs, FCL and ACTIVE constructs reduced stiffness by 62% (P < 0.001) and 75% (P < 0.001), respectively, and significantly increased axial motion, but not shear motion.

RESULTS

In a surrogate model of a distal femur fracture, replacing locked with nonlocked diaphyseal screws does not significantly decrease construct stiffness and does not enhance interfragmentary motion. A longer bridge span primarily increases shear motion, not axial motion. The use of FCL screws or active plating delivers axial dynamization without introducing shear motion.

CONCLUSIONS

5603978

INTRODUCTION

Angle-stable locked plating has become the standard treatment for most difficult fractures of the distal femur. Despite excellent early results, there is growing concern surrounding the relatively high nonunion rates after locked plate fixation of distal femur fractures. Most recent studies quote nonunion or fixation failure rates after locked plating of distal femur fractures of 10%–23%.1–6 There is abundant evidence that deficient fracture motion caused by overly stiff locking plates can suppress natural fracture healing, contributing to delayed unions, nonunions, and fixation failure.3,7–9 Conversely, research over the past 50 years has consistently demonstrated that controlled axial dynamization can improve the speed and strength of fracture healing by dynamically stimulating natural bone healing through callus formation.7,10–15 Two primary mechanical conditions critical for natural fracture healing have been identified: Callus formation is promoted by axial interfragmentary motion greater than 0.2 mm16,17; and fracture healing is inhibited when interfragmentary motion is dominated by shear displacement.18 Strategies aimed at altering the mechanical environment created by locked plating constructs and at promoting fracture healing by spontaneous callus formation were proposed as early as 2003.19 Four principal methods are currently promoted to reduce the stiffness of locked bridge plating constructs: diaphyseal fixation with nonlocking screws rather than locking screws; increasing the length of the bridge spanning the fracture zone19; screw dynamization with far cortical locking (FCL) screws8; and plate dynamization with active plates that have elastically suspended locking holes.20 It is not clear which of these 4 constructs provide the best mechanical environment to achieve the goal of early fracture dynamization to promote healing while minimizing any detrimental effects from motion. This study measured construct stiffness as well as axial and shear motion at the fracture site to assess the efficacy by which each strategy can satisfy the basic conditions for mechanical stimulation of fracture healing. Specifically, this study tested the hypotheses that the 4 dynamization strategies will differ in their efficacy to decrease construct stiffness, to increase interfragmentary axial motion, and to prevent excessive shear motion.

RESULTS

Construct Stiffness There was no significant difference between the stiffness of LOCKED control constructs (2998 ± 361 N/mm) and NONLOCKED constructs (2549 ± 355 N/mm, P = 0.08) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A994, summarizing construct stiffness and interfragmentary motion). However, compared with the LOCKED control group, BRIDGE group constructs had a 45% lower stiffness (P < 0.001), FCL group constructs had a 62% lower stiffness (P < 0.001), and ACTIVE group constructs had a 75% lower stiffness (P < 0.001) (Fig. 3). Construct stiffness achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. There was no significant difference between the stiffness of LOCKED control constructs (2998 ± 361 N/mm) and NONLOCKED constructs (2549 ± 355 N/mm, P = 0.08) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A994, summarizing construct stiffness and interfragmentary motion). However, compared with the LOCKED control group, BRIDGE group constructs had a 45% lower stiffness (P < 0.001), FCL group constructs had a 62% lower stiffness (P < 0.001), and ACTIVE group constructs had a 75% lower stiffness (P < 0.001) (Fig. 3). Construct stiffness achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. Axial Interfragmentary Motion In each group, near cortex motion dA, NC was smaller than far cortex motion dA, FC (Fig. 4). Near cortex motion in response to one body weight loading (700 N) was the same for LOCKED control constructs and NONLOCKED constructs (0.10 ± 0.02 mm, P = 0.85) and remained below the 0.2-mm axial motion threshold required for callus stimulation. Compared with the LOCKED control group, dA, NC was 2 times greater in the BRIDGE group (P < 0.001), over 4 times greater in the FCL group (P < 0.001), and over 7 times greater in the ACTIVE group (P < 0.001). Similarly, far cortex motion was not significantly different between the LOCKED control group (0.37 ± 0.04 mm) and the NONLOCKED group (0.46 ± 0.08 mm) (P = 0.07). However, compared with the LOCKED control group, dA, FC was 73% greater in the BRIDGE group (P < 0.001), 105% greater in the FCL group (P < 0.001), and 303% greater in the ACTIVE group (P < 0.001). Axial motion at the near and far cortex, achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. In each group, near cortex motion dA, NC was smaller than far cortex motion dA, FC (Fig. 4). Near cortex motion in response to one body weight loading (700 N) was the same for LOCKED control constructs and NONLOCKED constructs (0.10 ± 0.02 mm, P = 0.85) and remained below the 0.2-mm axial motion threshold required for callus stimulation. Compared with the LOCKED control group, dA, NC was 2 times greater in the BRIDGE group (P < 0.001), over 4 times greater in the FCL group (P < 0.001), and over 7 times greater in the ACTIVE group (P < 0.001). Similarly, far cortex motion was not significantly different between the LOCKED control group (0.37 ± 0.04 mm) and the NONLOCKED group (0.46 ± 0.08 mm) (P = 0.07). However, compared with the LOCKED control group, dA, FC was 73% greater in the BRIDGE group (P < 0.001), 105% greater in the FCL group (P < 0.001), and 303% greater in the ACTIVE group (P < 0.001). Axial motion at the near and far cortex, achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. Shear Motion Shear motion dS remained below 0.2 mm in all groups except in the BRIDGE group, which exhibited on average 0.96 ± 0.14 mm shear motion in response to one body weight loading (Fig. 5A). In BRIDGE constructs, this magnitude of shear motion was 50% greater than the corresponding far cortex motion and over 3 times greater than the near cortex motion. Using image analysis, shear-dominant motion in BRIDGE constructs was attributed to rotation of the femoral diaphysis around the proximal plate segment because of plate bending under axial loading, which caused the proximal osteotomy surface to be translated toward the locking plate (Fig. 5B). Resulting transverse shear, resulting from the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. Shear motion dS remained below 0.2 mm in all groups except in the BRIDGE group, which exhibited on average 0.96 ± 0.14 mm shear motion in response to one body weight loading (Fig. 5A). In BRIDGE constructs, this magnitude of shear motion was 50% greater than the corresponding far cortex motion and over 3 times greater than the near cortex motion. Using image analysis, shear-dominant motion in BRIDGE constructs was attributed to rotation of the femoral diaphysis around the proximal plate segment because of plate bending under axial loading, which caused the proximal osteotomy surface to be translated toward the locking plate (Fig. 5B). Resulting transverse shear, resulting from the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article.

INTRODUCTION: Angle-stable locked plating has become the standard treatment for most difficult fractures of the distal femur. Despite excellent early results, there is growing concern surrounding the relatively high nonunion rates after locked plate fixation of distal femur fractures. Most recent studies quote nonunion or fixation failure rates after locked plating of distal femur fractures of 10%–23%.1–6 There is abundant evidence that deficient fracture motion caused by overly stiff locking plates can suppress natural fracture healing, contributing to delayed unions, nonunions, and fixation failure.3,7–9 Conversely, research over the past 50 years has consistently demonstrated that controlled axial dynamization can improve the speed and strength of fracture healing by dynamically stimulating natural bone healing through callus formation.7,10–15 Two primary mechanical conditions critical for natural fracture healing have been identified: Callus formation is promoted by axial interfragmentary motion greater than 0.2 mm16,17; and fracture healing is inhibited when interfragmentary motion is dominated by shear displacement.18 Strategies aimed at altering the mechanical environment created by locked plating constructs and at promoting fracture healing by spontaneous callus formation were proposed as early as 2003.19 Four principal methods are currently promoted to reduce the stiffness of locked bridge plating constructs: diaphyseal fixation with nonlocking screws rather than locking screws; increasing the length of the bridge spanning the fracture zone19; screw dynamization with far cortical locking (FCL) screws8; and plate dynamization with active plates that have elastically suspended locking holes.20 It is not clear which of these 4 constructs provide the best mechanical environment to achieve the goal of early fracture dynamization to promote healing while minimizing any detrimental effects from motion. This study measured construct stiffness as well as axial and shear motion at the fracture site to assess the efficacy by which each strategy can satisfy the basic conditions for mechanical stimulation of fracture healing. Specifically, this study tested the hypotheses that the 4 dynamization strategies will differ in their efficacy to decrease construct stiffness, to increase interfragmentary axial motion, and to prevent excessive shear motion. MATERIAL AND METHODS: In a biomechanical bench-top study, periarticular locking plates were applied to bridge a metadiaphyseal fracture in femur surrogates. Construct stiffness was assessed under quasi-physiologic loading by measuring the resulting axial and shear motion at the fracture site. In the LOCKED control group, the periarticular plate was applied to the diaphysis with bicortical locking screws. The first locking screw was placed adjacent to the fracture to achieve a short-bridge span. Subsequently, 4 different strategies to decrease construct stiffness and to dynamize the fracture site were evaluated (Fig. 1): replacing diaphyseal locking screws with nonlocking screws (NONLOCKED group); bridge dynamization (BRIDGE group) by increasing the bridging span with locked screws in the diaphysis19; screw dynamization with FCL screws (FCL group)8; and plate dynamization with active locking plates (ACTIVE group).10,20 Construct stiffness was characterized by measuring the construct deformation in response to quasi-physiologic loading. Dynamization of the fracture site was characterized by measuring the interfragmentary motion in axial and shear direction. Finally, the stiffness and interfragmentary motion results of the 4 strategies for stiffness reduction were compared with the LOCKED control group to determine their effectiveness in dynamizing the fracture site. Strategies to dynamize a locked plating construct (LOCKED) for distal femur fractures. Editor's Note: A color image accompanies the online version of this article. Specimens Plating constructs were evaluated in fourth generation femur surrogate specimens made of fiber-reinforced epoxy composite (#3406, large size; Sawbones, Vashon, WA) to minimize interspecimen variability. An unstable distal femur fracture (AO/Orthopaedic Trauma Association 33-A3) was modeled by introducing a 10-mm gap osteotomy located 60 mm proximal to the intercondylar notch.21,22 This gap osteotomy simulated the biomechanical constraints of a comminuted fracture that relies on full-load transfer through the bridge plating construct because of a lack of bony apposition at the fracture site. In the LOCKED control group, this gap osteotomy was stabilized with a 286-mm long distal femur plate (ZPLP; Zimmer, Warsaw, IN) made of stainless steel. The plate had 13 holes for diaphyseal fixation, 7 of which were locking holes (Fig. 2A). The diaphyseal plate segment was applied using 3 evenly spaced 4.5-mm locking screws placed in the first, fourth, and seventh locking hole from the fracture site, resulting in a short-bridge span of 25 mm. A plate elevation of 1 mm over the proximal diaphysis was achieved with temporary spacers to simulate biological fixation with preservation of periosteal perfusion.23 The distal plate segment was applied to the metaphysis using six 5.5-mm cannulated locking screws in accordance with the manufacturer's technique guide. All screws were tightened to 4 Nm. A, Distal femur plate with alternating locked and nonlocked holes; (B) 3 distinct screws used with standard femur plate; (C and D) active plate with screw holes located in elastically suspended sliding elements. Editor's Note: A color image accompanies the online version of this article. Subsequently, 4 additional construct configurations were assembled by changing one variable of the LOCKED control group constructs at a time: For constructs of the NONLOCKED group, nonlocking screws were used in place of locking screws for diaphyseal fixation, using the first, fourth, and sixth nonlocking hole from the fracture site (Fig. 2B). Because of the alternating locking/nonlocking screw hole configuration of this plate, the nonlocking construct had an intermediate bridge span of 40 mm. BRIDGE group constructs used a longer bridge span (87 mm) than LOCKED control group constructs (25 mm) by placing diaphyseal locking screws in the third, fifth, and seventh locking hole from the fracture site. FCL group constructs replaced the 3 diaphyseal locking screws of LOCKED control group constructs with 3 FCL screws (4.5 mm MotionLoc; Zimmer) made of stainless steel. FCL screws rigidly lock into the plate and the far cortex, but they are not rigidly constrained in the near cortex underlying the plate. The elastic shaft of FCL screws can flex within the near cortex motion envelope to generate symmetric interfragmentary motion.8 ACTIVE group constructs had a screw configuration identical to that of the LOCKED control group, but used an active locking plate. Screw holes of active locking plates are integrated into individual sliding elements that are elastically suspended in a silicone envelope inside lateral plate pockets (Fig. 2C). Lateral pockets are arranged in an alternating pattern from both plate sides, resulting in a staggered locking hole configuration. The pocket geometry combined with the silicone suspension allows controlled axial translation, which enables up to 1.5 mm of axial motion across a fracture while providing stable fixation in response to bending and torsional loading.24 The silicone suspension consisted of long-term implantable medical-grade silicone elastomer. The active locking plate was made of stainless steel and was geometrically equivalent to the standard locking plate of the LOCKED control group (Fig. 2D). Five specimens of each of the 5 constructs were tested for reproducibility requiring a total of 25 construct tests. Plating constructs were evaluated in fourth generation femur surrogate specimens made of fiber-reinforced epoxy composite (#3406, large size; Sawbones, Vashon, WA) to minimize interspecimen variability. An unstable distal femur fracture (AO/Orthopaedic Trauma Association 33-A3) was modeled by introducing a 10-mm gap osteotomy located 60 mm proximal to the intercondylar notch.21,22 This gap osteotomy simulated the biomechanical constraints of a comminuted fracture that relies on full-load transfer through the bridge plating construct because of a lack of bony apposition at the fracture site. In the LOCKED control group, this gap osteotomy was stabilized with a 286-mm long distal femur plate (ZPLP; Zimmer, Warsaw, IN) made of stainless steel. The plate had 13 holes for diaphyseal fixation, 7 of which were locking holes (Fig. 2A). The diaphyseal plate segment was applied using 3 evenly spaced 4.5-mm locking screws placed in the first, fourth, and seventh locking hole from the fracture site, resulting in a short-bridge span of 25 mm. A plate elevation of 1 mm over the proximal diaphysis was achieved with temporary spacers to simulate biological fixation with preservation of periosteal perfusion.23 The distal plate segment was applied to the metaphysis using six 5.5-mm cannulated locking screws in accordance with the manufacturer's technique guide. All screws were tightened to 4 Nm. A, Distal femur plate with alternating locked and nonlocked holes; (B) 3 distinct screws used with standard femur plate; (C and D) active plate with screw holes located in elastically suspended sliding elements. Editor's Note: A color image accompanies the online version of this article. Subsequently, 4 additional construct configurations were assembled by changing one variable of the LOCKED control group constructs at a time: For constructs of the NONLOCKED group, nonlocking screws were used in place of locking screws for diaphyseal fixation, using the first, fourth, and sixth nonlocking hole from the fracture site (Fig. 2B). Because of the alternating locking/nonlocking screw hole configuration of this plate, the nonlocking construct had an intermediate bridge span of 40 mm. BRIDGE group constructs used a longer bridge span (87 mm) than LOCKED control group constructs (25 mm) by placing diaphyseal locking screws in the third, fifth, and seventh locking hole from the fracture site. FCL group constructs replaced the 3 diaphyseal locking screws of LOCKED control group constructs with 3 FCL screws (4.5 mm MotionLoc; Zimmer) made of stainless steel. FCL screws rigidly lock into the plate and the far cortex, but they are not rigidly constrained in the near cortex underlying the plate. The elastic shaft of FCL screws can flex within the near cortex motion envelope to generate symmetric interfragmentary motion.8 ACTIVE group constructs had a screw configuration identical to that of the LOCKED control group, but used an active locking plate. Screw holes of active locking plates are integrated into individual sliding elements that are elastically suspended in a silicone envelope inside lateral plate pockets (Fig. 2C). Lateral pockets are arranged in an alternating pattern from both plate sides, resulting in a staggered locking hole configuration. The pocket geometry combined with the silicone suspension allows controlled axial translation, which enables up to 1.5 mm of axial motion across a fracture while providing stable fixation in response to bending and torsional loading.24 The silicone suspension consisted of long-term implantable medical-grade silicone elastomer. The active locking plate was made of stainless steel and was geometrically equivalent to the standard locking plate of the LOCKED control group (Fig. 2D). Five specimens of each of the 5 constructs were tested for reproducibility requiring a total of 25 construct tests. Loading For stiffness assessment, constructs were tested under quasi-physiological loading in a material test system according to an established loading protocol (see Figure, Supplemental Digital Content 1, http://links.lww.com/BOT/A993, describing specimen loading and outcome assessment).21,25 The femoral condyles were embedded in a mounting fixture using bone cement and were rigidly connected to the base of the test system (8874, Instron, Canton, MA). The metaphyseal plate segment was coated with soft clay to prevent nonphysiologic plate constraints. The femoral head was placed in a spherical recess of a polymer block that was attached to the test system actuator. This enabled axial load application while allowing unconstrained rotation of the femoral head. Load was induced along the mechanical axis of the femur, with the load vector intersecting the femoral head and the epicondylar center. Each construct was loaded in 50-N increments up to 700 N, corresponding to approximately one body weight. For stiffness assessment, constructs were tested under quasi-physiological loading in a material test system according to an established loading protocol (see Figure, Supplemental Digital Content 1, http://links.lww.com/BOT/A993, describing specimen loading and outcome assessment).21,25 The femoral condyles were embedded in a mounting fixture using bone cement and were rigidly connected to the base of the test system (8874, Instron, Canton, MA). The metaphyseal plate segment was coated with soft clay to prevent nonphysiologic plate constraints. The femoral head was placed in a spherical recess of a polymer block that was attached to the test system actuator. This enabled axial load application while allowing unconstrained rotation of the femoral head. Load was induced along the mechanical axis of the femur, with the load vector intersecting the femoral head and the epicondylar center. Each construct was loaded in 50-N increments up to 700 N, corresponding to approximately one body weight. Outcome Assessment Constructs were characterized by determining their construct stiffness and interfragmentary motion using noncontact optical photogrammetry. For this purpose, an array of 4 active luminescent markers consisting of miniature light emitting diodes were glued to the osteotomy surfaces. An 18 megapixel digital camera (Canon EOS T6) captured the marker locations with a resolution of 0.01 mm after each incremental loading step. ImageJ quantitative image analysis software developed by the National Institute of Health (www.imagej.net) was used to extract marker displacement and to calculate the average axial motion dA and shear motion dS between osteotomy surfaces in response to incremental load steps. Because plate bending induces different amounts of axial motion at the near cortex and far cortex,26 axial motion dA was extracted individually for the near cortex (dA, NC) from markers 1 and 3, and for the far cortex (dA, FC) from markers 2 and 4, as depicted in Supplemental Digital Content 1 (see Figure, http://links.lww.com/BOT/A993). Construct stiffness SC was calculated by dividing the applied axial load by the axial motion dA at the midpoint between the near and far cortex, with dA = (dA, FC + dA, NC)/2. Constructs were characterized by determining their construct stiffness and interfragmentary motion using noncontact optical photogrammetry. For this purpose, an array of 4 active luminescent markers consisting of miniature light emitting diodes were glued to the osteotomy surfaces. An 18 megapixel digital camera (Canon EOS T6) captured the marker locations with a resolution of 0.01 mm after each incremental loading step. ImageJ quantitative image analysis software developed by the National Institute of Health (www.imagej.net) was used to extract marker displacement and to calculate the average axial motion dA and shear motion dS between osteotomy surfaces in response to incremental load steps. Because plate bending induces different amounts of axial motion at the near cortex and far cortex,26 axial motion dA was extracted individually for the near cortex (dA, NC) from markers 1 and 3, and for the far cortex (dA, FC) from markers 2 and 4, as depicted in Supplemental Digital Content 1 (see Figure, http://links.lww.com/BOT/A993). Construct stiffness SC was calculated by dividing the applied axial load by the axial motion dA at the midpoint between the near and far cortex, with dA = (dA, FC + dA, NC)/2. Statistical Analysis All results are reported as their mean and SD. Construct stiffness SC and interfragmentary motion parameters dS, dA, FC, and dA, FC of the 4 experimental groups was statistically compared with the LOCKED control group results using one-way analysis of variance testing including a post hoc Turkey honest significant difference (HSD) to identify significant differences. Each outcome parameter was analyzed individually, and a level of significance of α = 0.05 was used to detect significant differences. All results are reported as their mean and SD. Construct stiffness SC and interfragmentary motion parameters dS, dA, FC, and dA, FC of the 4 experimental groups was statistically compared with the LOCKED control group results using one-way analysis of variance testing including a post hoc Turkey honest significant difference (HSD) to identify significant differences. Each outcome parameter was analyzed individually, and a level of significance of α = 0.05 was used to detect significant differences. Specimens: Plating constructs were evaluated in fourth generation femur surrogate specimens made of fiber-reinforced epoxy composite (#3406, large size; Sawbones, Vashon, WA) to minimize interspecimen variability. An unstable distal femur fracture (AO/Orthopaedic Trauma Association 33-A3) was modeled by introducing a 10-mm gap osteotomy located 60 mm proximal to the intercondylar notch.21,22 This gap osteotomy simulated the biomechanical constraints of a comminuted fracture that relies on full-load transfer through the bridge plating construct because of a lack of bony apposition at the fracture site. In the LOCKED control group, this gap osteotomy was stabilized with a 286-mm long distal femur plate (ZPLP; Zimmer, Warsaw, IN) made of stainless steel. The plate had 13 holes for diaphyseal fixation, 7 of which were locking holes (Fig. 2A). The diaphyseal plate segment was applied using 3 evenly spaced 4.5-mm locking screws placed in the first, fourth, and seventh locking hole from the fracture site, resulting in a short-bridge span of 25 mm. A plate elevation of 1 mm over the proximal diaphysis was achieved with temporary spacers to simulate biological fixation with preservation of periosteal perfusion.23 The distal plate segment was applied to the metaphysis using six 5.5-mm cannulated locking screws in accordance with the manufacturer's technique guide. All screws were tightened to 4 Nm. A, Distal femur plate with alternating locked and nonlocked holes; (B) 3 distinct screws used with standard femur plate; (C and D) active plate with screw holes located in elastically suspended sliding elements. Editor's Note: A color image accompanies the online version of this article. Subsequently, 4 additional construct configurations were assembled by changing one variable of the LOCKED control group constructs at a time: For constructs of the NONLOCKED group, nonlocking screws were used in place of locking screws for diaphyseal fixation, using the first, fourth, and sixth nonlocking hole from the fracture site (Fig. 2B). Because of the alternating locking/nonlocking screw hole configuration of this plate, the nonlocking construct had an intermediate bridge span of 40 mm. BRIDGE group constructs used a longer bridge span (87 mm) than LOCKED control group constructs (25 mm) by placing diaphyseal locking screws in the third, fifth, and seventh locking hole from the fracture site. FCL group constructs replaced the 3 diaphyseal locking screws of LOCKED control group constructs with 3 FCL screws (4.5 mm MotionLoc; Zimmer) made of stainless steel. FCL screws rigidly lock into the plate and the far cortex, but they are not rigidly constrained in the near cortex underlying the plate. The elastic shaft of FCL screws can flex within the near cortex motion envelope to generate symmetric interfragmentary motion.8 ACTIVE group constructs had a screw configuration identical to that of the LOCKED control group, but used an active locking plate. Screw holes of active locking plates are integrated into individual sliding elements that are elastically suspended in a silicone envelope inside lateral plate pockets (Fig. 2C). Lateral pockets are arranged in an alternating pattern from both plate sides, resulting in a staggered locking hole configuration. The pocket geometry combined with the silicone suspension allows controlled axial translation, which enables up to 1.5 mm of axial motion across a fracture while providing stable fixation in response to bending and torsional loading.24 The silicone suspension consisted of long-term implantable medical-grade silicone elastomer. The active locking plate was made of stainless steel and was geometrically equivalent to the standard locking plate of the LOCKED control group (Fig. 2D). Five specimens of each of the 5 constructs were tested for reproducibility requiring a total of 25 construct tests. Loading: For stiffness assessment, constructs were tested under quasi-physiological loading in a material test system according to an established loading protocol (see Figure, Supplemental Digital Content 1, http://links.lww.com/BOT/A993, describing specimen loading and outcome assessment).21,25 The femoral condyles were embedded in a mounting fixture using bone cement and were rigidly connected to the base of the test system (8874, Instron, Canton, MA). The metaphyseal plate segment was coated with soft clay to prevent nonphysiologic plate constraints. The femoral head was placed in a spherical recess of a polymer block that was attached to the test system actuator. This enabled axial load application while allowing unconstrained rotation of the femoral head. Load was induced along the mechanical axis of the femur, with the load vector intersecting the femoral head and the epicondylar center. Each construct was loaded in 50-N increments up to 700 N, corresponding to approximately one body weight. Outcome Assessment: Constructs were characterized by determining their construct stiffness and interfragmentary motion using noncontact optical photogrammetry. For this purpose, an array of 4 active luminescent markers consisting of miniature light emitting diodes were glued to the osteotomy surfaces. An 18 megapixel digital camera (Canon EOS T6) captured the marker locations with a resolution of 0.01 mm after each incremental loading step. ImageJ quantitative image analysis software developed by the National Institute of Health (www.imagej.net) was used to extract marker displacement and to calculate the average axial motion dA and shear motion dS between osteotomy surfaces in response to incremental load steps. Because plate bending induces different amounts of axial motion at the near cortex and far cortex,26 axial motion dA was extracted individually for the near cortex (dA, NC) from markers 1 and 3, and for the far cortex (dA, FC) from markers 2 and 4, as depicted in Supplemental Digital Content 1 (see Figure, http://links.lww.com/BOT/A993). Construct stiffness SC was calculated by dividing the applied axial load by the axial motion dA at the midpoint between the near and far cortex, with dA = (dA, FC + dA, NC)/2. Statistical Analysis: All results are reported as their mean and SD. Construct stiffness SC and interfragmentary motion parameters dS, dA, FC, and dA, FC of the 4 experimental groups was statistically compared with the LOCKED control group results using one-way analysis of variance testing including a post hoc Turkey honest significant difference (HSD) to identify significant differences. Each outcome parameter was analyzed individually, and a level of significance of α = 0.05 was used to detect significant differences. RESULTS: Construct Stiffness There was no significant difference between the stiffness of LOCKED control constructs (2998 ± 361 N/mm) and NONLOCKED constructs (2549 ± 355 N/mm, P = 0.08) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A994, summarizing construct stiffness and interfragmentary motion). However, compared with the LOCKED control group, BRIDGE group constructs had a 45% lower stiffness (P < 0.001), FCL group constructs had a 62% lower stiffness (P < 0.001), and ACTIVE group constructs had a 75% lower stiffness (P < 0.001) (Fig. 3). Construct stiffness achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. There was no significant difference between the stiffness of LOCKED control constructs (2998 ± 361 N/mm) and NONLOCKED constructs (2549 ± 355 N/mm, P = 0.08) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A994, summarizing construct stiffness and interfragmentary motion). However, compared with the LOCKED control group, BRIDGE group constructs had a 45% lower stiffness (P < 0.001), FCL group constructs had a 62% lower stiffness (P < 0.001), and ACTIVE group constructs had a 75% lower stiffness (P < 0.001) (Fig. 3). Construct stiffness achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. Axial Interfragmentary Motion In each group, near cortex motion dA, NC was smaller than far cortex motion dA, FC (Fig. 4). Near cortex motion in response to one body weight loading (700 N) was the same for LOCKED control constructs and NONLOCKED constructs (0.10 ± 0.02 mm, P = 0.85) and remained below the 0.2-mm axial motion threshold required for callus stimulation. Compared with the LOCKED control group, dA, NC was 2 times greater in the BRIDGE group (P < 0.001), over 4 times greater in the FCL group (P < 0.001), and over 7 times greater in the ACTIVE group (P < 0.001). Similarly, far cortex motion was not significantly different between the LOCKED control group (0.37 ± 0.04 mm) and the NONLOCKED group (0.46 ± 0.08 mm) (P = 0.07). However, compared with the LOCKED control group, dA, FC was 73% greater in the BRIDGE group (P < 0.001), 105% greater in the FCL group (P < 0.001), and 303% greater in the ACTIVE group (P < 0.001). Axial motion at the near and far cortex, achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. In each group, near cortex motion dA, NC was smaller than far cortex motion dA, FC (Fig. 4). Near cortex motion in response to one body weight loading (700 N) was the same for LOCKED control constructs and NONLOCKED constructs (0.10 ± 0.02 mm, P = 0.85) and remained below the 0.2-mm axial motion threshold required for callus stimulation. Compared with the LOCKED control group, dA, NC was 2 times greater in the BRIDGE group (P < 0.001), over 4 times greater in the FCL group (P < 0.001), and over 7 times greater in the ACTIVE group (P < 0.001). Similarly, far cortex motion was not significantly different between the LOCKED control group (0.37 ± 0.04 mm) and the NONLOCKED group (0.46 ± 0.08 mm) (P = 0.07). However, compared with the LOCKED control group, dA, FC was 73% greater in the BRIDGE group (P < 0.001), 105% greater in the FCL group (P < 0.001), and 303% greater in the ACTIVE group (P < 0.001). Axial motion at the near and far cortex, achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. Shear Motion Shear motion dS remained below 0.2 mm in all groups except in the BRIDGE group, which exhibited on average 0.96 ± 0.14 mm shear motion in response to one body weight loading (Fig. 5A). In BRIDGE constructs, this magnitude of shear motion was 50% greater than the corresponding far cortex motion and over 3 times greater than the near cortex motion. Using image analysis, shear-dominant motion in BRIDGE constructs was attributed to rotation of the femoral diaphysis around the proximal plate segment because of plate bending under axial loading, which caused the proximal osteotomy surface to be translated toward the locking plate (Fig. 5B). Resulting transverse shear, resulting from the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. Shear motion dS remained below 0.2 mm in all groups except in the BRIDGE group, which exhibited on average 0.96 ± 0.14 mm shear motion in response to one body weight loading (Fig. 5A). In BRIDGE constructs, this magnitude of shear motion was 50% greater than the corresponding far cortex motion and over 3 times greater than the near cortex motion. Using image analysis, shear-dominant motion in BRIDGE constructs was attributed to rotation of the femoral diaphysis around the proximal plate segment because of plate bending under axial loading, which caused the proximal osteotomy surface to be translated toward the locking plate (Fig. 5B). Resulting transverse shear, resulting from the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. Construct Stiffness: There was no significant difference between the stiffness of LOCKED control constructs (2998 ± 361 N/mm) and NONLOCKED constructs (2549 ± 355 N/mm, P = 0.08) (see Table, Supplemental Digital Content 2, http://links.lww.com/BOT/A994, summarizing construct stiffness and interfragmentary motion). However, compared with the LOCKED control group, BRIDGE group constructs had a 45% lower stiffness (P < 0.001), FCL group constructs had a 62% lower stiffness (P < 0.001), and ACTIVE group constructs had a 75% lower stiffness (P < 0.001) (Fig. 3). Construct stiffness achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. Axial Interfragmentary Motion: In each group, near cortex motion dA, NC was smaller than far cortex motion dA, FC (Fig. 4). Near cortex motion in response to one body weight loading (700 N) was the same for LOCKED control constructs and NONLOCKED constructs (0.10 ± 0.02 mm, P = 0.85) and remained below the 0.2-mm axial motion threshold required for callus stimulation. Compared with the LOCKED control group, dA, NC was 2 times greater in the BRIDGE group (P < 0.001), over 4 times greater in the FCL group (P < 0.001), and over 7 times greater in the ACTIVE group (P < 0.001). Similarly, far cortex motion was not significantly different between the LOCKED control group (0.37 ± 0.04 mm) and the NONLOCKED group (0.46 ± 0.08 mm) (P = 0.07). However, compared with the LOCKED control group, dA, FC was 73% greater in the BRIDGE group (P < 0.001), 105% greater in the FCL group (P < 0.001), and 303% greater in the ACTIVE group (P < 0.001). Axial motion at the near and far cortex, achieved with the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. Shear Motion: Shear motion dS remained below 0.2 mm in all groups except in the BRIDGE group, which exhibited on average 0.96 ± 0.14 mm shear motion in response to one body weight loading (Fig. 5A). In BRIDGE constructs, this magnitude of shear motion was 50% greater than the corresponding far cortex motion and over 3 times greater than the near cortex motion. Using image analysis, shear-dominant motion in BRIDGE constructs was attributed to rotation of the femoral diaphysis around the proximal plate segment because of plate bending under axial loading, which caused the proximal osteotomy surface to be translated toward the locking plate (Fig. 5B). Resulting transverse shear, resulting from the 4 strategies for plate dynamization, relative to the LOCKED control construct. Editor's Note: A color image accompanies the online version of this article. DISCUSSION: Results confirmed the study hypothesis by demonstrating that the 4 dynamization strategies yielded not only different amounts of construct stiffness and interfragmentary motion, but also different types of interfragmentary motion. LOCKED group results confirmed that a locking plate with a short-bridge span results in asymmetric interfragmentary motion that is deficient for callus formation.3,13,16,26 Although one body weight loading may be excessive for early postoperative loading, it resulted only in 0.1-mm motion at the near cortex. This motion remained below the 0.2-mm motion threshold that has been established as the lower boundary for fracture motion required to promote callus formation.17 This result is supported by several in vivo and clinical studies that demonstrate suppression of callus formation and healing at the near cortex adjacent to a locking plate.3,7,9,20 The far cortex motion was greater than measured at the near cortex, likely secondary to plate bending. Clinically, the increased far cortex motion may allow callus to form, but the repetitive bending may also play a role in eventual fatigue failure of the plate before fracture healing occurs. NONLOCKED constructs represent an intuitive response to the stiffness concern associated with locked plating by reverting to nonlocking screws in the diaphysis. However, substituting nonlocking diaphyseal fixation had no significant effect on construct stiffness or interfragmentary motion. This may be explained by the rigid compression of the plate onto the bone surface, which is required to retain stable fixation. Compressing the plate to the bone prevents any motion at the plate–bone interface, which is a prerequisite to induce symmetric interfragmentary motion.26 In contrast to locked plating constructs, the stiffness of a nonlocked construct will gradually decay as a result of dynamic loading.27 Although this can lead to increased fracture motion over time, the resulting uncontrolled motion is not a reliable strategy for dynamization. In addition, the natural fracture healing process responds with much more robust callus formation when exposed to early motion relative to delayed motion.28 BRIDGE constructs resembled the earliest and most widely proposed strategy to dynamize locked plating constructs. In a biomechanical study, Stoffel et al19 reported in 2003 that axial stiffness of locked plating constructs was mainly influenced by their bridge span. They recommended that one or 2 holes should be omitted on each side of the fracture to allow callus formation. They found that omitting 2 holes made the construct almost twice as flexible, but also 42% less strong. This study found a 45% stiffness reduction by omitting 2 screw holes. However, the greater flexibility of the longer bridge span increased motion primarily at the far cortex, whereas near cortex motion remained deficient. In addition, the longer bridge span induced up to 3 times more shear motion than axial motion. Although shear motion does not necessarily inhibit healing,29,30 several studies have shown that excessive or predominant shear motion will significantly delay healing.18,31 A recent study on the effect of bridge span on fracture motion also confirmed a disproportionate increase in shear motion.32 By analyzing 66 distal femur fractures stabilized with locking plates, they furthermore established a direct association between shear-dominated fracture motion and callus inhibition. Their findings, combined with the results of this study question the technique of increasing the bridge span to dynamize a locked construct, because this may weaken the construct and may cause asymmetric axial motion and excessive shear motion that inhibits fracture healing. FCL and ACTIVE group constructs reduced stiffness compared with the LOCKED control group by 62%–75% to 1130 and 759 N/mm, respectively. Nevertheless, their stiffness remained substantially higher than the stiffness range of 50–400 N/mm reported for external fixators33,34 and Ilizarov frames.35 The fact that external fixators and Ilizarov frames are established clinical tools that promote fracture healing by callus formation suggests that the stiffness reduction of FCL and ACTIVE constructs is rather conservative and does not introduce excessive dynamization. FCL and ACTIVE constructs enhanced interfragmentary motion at the near and far cortex well above the 0.2-mm threshold needed to stimulate callus formation. The highest axial motion of 1.1 mm was observed at the far cortex of ACTIVE constructs, and remained at the lower limit of the 1–4 mm motion range reported for functional bracing.36 Most importantly, FCL and ACTIVE constructs delivered dynamization that was dominated by axial motion, not shear motion. These constructs allow a screw to be placed close to the fracture site without affecting stiffness, and therefore, limit the amount of shear motion possible. The controlled axial dynamization provided by FCL and ACTIVE constructs delivers faster and stronger healing. In an ovine fracture healing study, FCL constructs induced consistent and circumferential callus bridging and yielded 157% stronger healing compared with standard locked plating.7 Clinically, a prospective study of 31 consecutive distal femur fractures stabilized with FCL constructs reported no implant or fixation failure, an average time to union of 16 weeks and a nonunion rate of 3%.37 Similar to FCL constructs, ACTIVE plating induced 6 times more callus at 3 weeks postsurgery, and yielded 4 times stronger healing compared with rigid locked plating in an ovine fracture healing study.20 These in vivo and clinical studies of FCL and ACTIVE plating constructs demonstrated that controlled axial dynamization reliably promoted natural fracture healing. Results of this study are limited by the use of femur surrogates. Validated surrogates were used to extract relative differences between constructs under highly reproducible test conditions.38 Because the surrogates represented a strong, nonosteoporotic femur, results may not be extrapolated to fracture fixation in the osteopenic femur. Moreover, this study only investigated construct stability in terms of stiffness and related interfragmentary motion, without loading constructs to failure to determine their strength. The strength of the tested constructs has been evaluated in previous studies, showing that increasing the bridge span will decrease construct strength,19 whereas the strength of FCL and active plating constructs is comparable with that of standard locked plating constructs.8,24 Testing was furthermore limited to static loading and did not investigate gradual loosening or fatigue of constructs under dynamic loading. Moreover, this study has been limited to a principal loading mode that combines axial compression and bending but not torsion. Only plates made of stainless steel were tested, which are approximately twice as stiff as geometrically equivalent plates made of Titanium alloy. Although results of this study raise concerns on the negative effect of shear-dominated interfragmentary motion on fracture healing, this concern should be formally investigated in a future in vivo study. Most importantly, emerging implant technologies that can provide controlled dynamization will require more clinical studies to document their effect on fracture healing, and to better define the range of interfragmentary motion that will promote healing of different fracture patterns at specific fracture locations. In conclusion, results of this study indicate that intuitive technical tricks, such as reverting to nonlocking screws or using long plates to maximize the bridge span may not reliably achieve relative stability and adequate interfragmentary motion for promoting natural fracture healing. Conversely, engineered implant solutions in the form of FCL screws or active plates can reliably dynamize a locked plating construct to stimulate fracture healing. As such, results should encourage implant manufacturers to provide engineered solutions that reliably promote rather than potentially hinder fracture healing to avoid the need for and uncertainty of technical tricks intended to optimize construct stability. Supplementary Material:

Background: Decreasing the stiffness of locked plating constructs can promote natural fracture healing by controlled dynamization of the fracture. This biomechanical study compared the effect of 4 different stiffness reduction methods on interfragmentary motion by measuring axial motion and shear motion at the fracture site. Methods: Distal femur locking plates were applied to bridge a metadiaphyseal fracture in femur surrogates. A locked construct with a short-bridge span served as the nondynamized control group (LOCKED). Four different methods for stiffness reduction were evaluated: replacing diaphyseal locking screws with nonlocked screws (NONLOCKED); bridge dynamization (BRIDGE) with 2 empty screw holes proximal to the fracture; screw dynamization with far cortical locking (FCL) screws; and plate dynamization with active locking plates (ACTIVE). Construct stiffness, axial motion, and shear motion at the fracture site were measured to characterize each dynamization methods. Results: Compared with LOCKED control constructs, NONLOCKED constructs had a similar stiffness (P = 0.08), axial motion (P = 0.07), and shear motion (P = 0.97). BRIDGE constructs reduced stiffness by 45% compared with LOCKED constructs (P < 0.001), but interfragmentary motion was dominated by shear. Compared with LOCKED constructs, FCL and ACTIVE constructs reduced stiffness by 62% (P < 0.001) and 75% (P < 0.001), respectively, and significantly increased axial motion, but not shear motion. Conclusions: In a surrogate model of a distal femur fracture, replacing locked with nonlocked diaphyseal screws does not significantly decrease construct stiffness and does not enhance interfragmentary motion. A longer bridge span primarily increases shear motion, not axial motion. The use of FCL screws or active plating delivers axial dynamization without introducing shear motion.

[CONTENT] plate fixation | locked plating | dynamization | interfragmentary motion | stiffness | active plating | femur [SUMMARY]

[CONTENT] plate fixation | locked plating | dynamization | interfragmentary motion | stiffness | active plating | femur [SUMMARY]

[CONTENT] plate fixation | locked plating | dynamization | interfragmentary motion | stiffness | active plating | femur [SUMMARY]

[CONTENT] Biomechanical Phenomena | Bone Plates | Bone Screws | Diaphyses | Equipment Design | Femoral Fractures | Fracture Fixation, Internal | Humans | Models, Anatomic | Shear Strength [SUMMARY]

[CONTENT] Biomechanical Phenomena | Bone Plates | Bone Screws | Diaphyses | Equipment Design | Femoral Fractures | Fracture Fixation, Internal | Humans | Models, Anatomic | Shear Strength [SUMMARY]

[CONTENT] Biomechanical Phenomena | Bone Plates | Bone Screws | Diaphyses | Equipment Design | Femoral Fractures | Fracture Fixation, Internal | Humans | Models, Anatomic | Shear Strength [SUMMARY]

[CONTENT] locked distal femur | fracture healing inhibited | fracture fixation osteopenic | fracture motion callus | plating distal femur [SUMMARY]

[CONTENT] locked distal femur | fracture healing inhibited | fracture fixation osteopenic | fracture motion callus | plating distal femur [SUMMARY]

[CONTENT] locked distal femur | fracture healing inhibited | fracture fixation osteopenic | fracture motion callus | plating distal femur [SUMMARY]

[CONTENT] motion | group | plate | constructs | locked | mm | fracture | construct | locking | cortex [SUMMARY]

[CONTENT] motion | group | plate | constructs | locked | mm | fracture | construct | locking | cortex [SUMMARY]

[CONTENT] motion | group | plate | constructs | locked | mm | fracture | construct | locking | cortex [SUMMARY]

[CONTENT] healing | fracture | fracture healing | mechanical | motion | plating | fixation | dynamization | formation | natural [SUMMARY]

[CONTENT] group | 001 | motion | group 001 | greater | locked control | control | locked | mm | constructs [SUMMARY]

[CONTENT] motion | group | constructs | locked | plate | mm | fracture | da | cortex | stiffness [SUMMARY]

[CONTENT] ||| 4 [SUMMARY]

[CONTENT] NONLOCKED | 0.08 | 0.07 | 0.97 ||| 45% ||| FCL | 62% | 75% [SUMMARY]

[CONTENT] ||| 4 ||| ||| ||| Four | 2 ||| ||| NONLOCKED | 0.08 | 0.07 | 0.97 ||| 45% ||| FCL | 62% | 75% ||| ||| ||| FCL [SUMMARY]

Efficacy and safety of endoscopic submucosal dissection for gastric tube cancer: A multicenter retrospective study.

33776371

Recent improvements in the prognosis of patients with esophageal cancer have led to the increased occurrence of gastric tube cancer (GTC) in the reconstructed gastric tube. However, there are few reports on the treatment results of endoscopic submucosal dissection (ESD) for GTC.

BACKGROUND

We retrospectively investigated 48 GTC lesions in 38 consecutive patients with GTC in the reconstructed gastric tube after esophagectomy who had undergone ESD between January 2005 and December 2019 at 8 institutions participating in the Okayama Gut Study group. The clinical indications of ESD for early gastric cancer were similarly applied for GTC after esophagectomy. ESD specimens were evaluated in 2-mm slices according to the Japanese Classification of Gastric Carcinoma with curability assessments divided into curative and non-curative resection based on the Gastric Cancer Treatment Guidelines. Patient characteristics, treatment results, clinical course, and treatment outcomes were analyzed.

METHODS

The median age of patients was 71.5 years (range, 57-84years), and there were 34 men and 4 women. The median observation period after ESD was 884 d (range, 8-4040 d). The median procedure time was 81 min (range, 29-334 min), the en bloc resection rate was 91.7% (44/48), and the curative resection rate was 79% (38/48). Complications during ESD were seen in 4% (2/48) of case, and those after ESD were seen in 10% (5/48) of case. The survival rate at 5 years was 59.5%. During the observation period after ESD, 10 patients died of other diseases. Although there were differences in the procedure time between institutions, a multivariate analysis showed that tumor size was the only factor associated with prolonged procedure time.

RESULTS

ESD for GTC after esophagectomy was shown to be safe and effective.

CONCLUSION

7985736

INTRODUCTION

Recently, the survival of patients with esophageal cancer after esophagectomy has improved[1-5]. However, the risk of a subsequent occurrence of primary cancer is high in these patients. The most frequent cancer that overlaps with esophageal cancer is head and neck cancer, while the second most common is gastric cancer, including gastric tube cancer (GTC)[6-9]. Therefore, the improved prognosis of esophageal cancer patients has led to an increase in the occurrence of GTC in the reconstructed gastric tube. For the treatment of GTC after esophagectomy, total gastric tube resection (TGTR) or partial gastric tube resection (PGTR) has been proposed. However, surgical resection for GTC, being a secondary operation following esophagectomy, may lead to high mortality and morbidity[10,11]. On the other hand, in recent years, endoscopic therapy for early gastric cancer (EGC) has developed and become widespread[12]. Endoscopic submucosal dissection (ESD) enables the treatment of large lesions with a higher rate of en bloc resection that cannot be achieved by using conventional endoscopic mucosal resection. In addition, ESD is less invasive than surgery. For this reason, ESD has become widely used as a standard treatment for EGC, and ESD is often performed for GTC. However, ESD for GTC after esophagectomy is a technically difficult procedure compared with that for an unresected stomach because of the limited working space, unusual fluid-pooling area, food residue, bile reflux, fibrosis, and staples under the suture line[13]. Therefore, a high degree of skill is required for ESD of GTC. There are few reports about ESD for GTC after esophagectomy, and most are case reports and case series of a small number of patients[13-17]. A study by Nonaka et al[18] reported the effectiveness and safety of ESD for GTC in a high-volume national center, which had largest number of cases but was nonetheless a single-center study. Therefore, the aim of this study was to evaluate the efficacy and safety of ESD for GTC after esophagectomy in a multicenter context.

MATERIALS AND METHODS

Patients We retrospectively investigated patients with GTC in the reconstructed gastric tube after esophagectomy for esophageal squamous cell carcinoma who had undergone ESD between January 2005 and December 2019 at 8 institutions participating in the Okayama Gut Study group (O-GUTS). All of the participating institutions in O-GUTS, except Okayama University Hospital (OUH), were considered core hospitals in each area. During the study period, 48 GTC lesions in 38 consecutive patients were treated. The clinical indications of ESD for EGC were based on the Gastric Cancer Treatment Guidelines[19]. These indications were similarly applied for GTC after esophagectomy with gastric tube reconstruction. Study measurements were as follows: patient characteristics, endoscopic findings, treatment results, adverse events, histopathological results, and clinical courses. In addition, we defined OUH as a high-volume center and compared the patients’ background and clinical outcomes between OUH and other facilities. The institutional review board of each hospital approved this study, and informed consent was obtained from all patients. We retrospectively investigated patients with GTC in the reconstructed gastric tube after esophagectomy for esophageal squamous cell carcinoma who had undergone ESD between January 2005 and December 2019 at 8 institutions participating in the Okayama Gut Study group (O-GUTS). All of the participating institutions in O-GUTS, except Okayama University Hospital (OUH), were considered core hospitals in each area. During the study period, 48 GTC lesions in 38 consecutive patients were treated. The clinical indications of ESD for EGC were based on the Gastric Cancer Treatment Guidelines[19]. These indications were similarly applied for GTC after esophagectomy with gastric tube reconstruction. Study measurements were as follows: patient characteristics, endoscopic findings, treatment results, adverse events, histopathological results, and clinical courses. In addition, we defined OUH as a high-volume center and compared the patients’ background and clinical outcomes between OUH and other facilities. The institutional review board of each hospital approved this study, and informed consent was obtained from all patients. Endoscopic procedures All endoscopic procedures were performed by experts in ESD who had experience with more than 500 clinical cases. There were no restrictions on the scopes and devices used by each endoscopist for ESD. The scopes used were GIF-Q260J or GIF-H260 (Olympus, Tokyo, Japan), and the devices were an insulation-tipped diathermic knife (IT Knife), IT Knife 2, IT Knife nano, or Dual Knife J (Olympus, Tokyo, Japan). Other devices, such as an argon plasma coagulation probe (ERBE, Tubingen, Germany) for marking dots or a needle knife (ZEON MEDICAL, Tokyo, Japan) for the initial incision, were occasionally used. First, marking dots for the incision lines were placed around the lesion. Next, fructose-added glycerol (Glyceol; TAIYO Pharma CO, Tokyo, Japan) with a minute amount of indigo carmine dye was injected into the submucosal layer. In some cases, 0.4% sodium hyaluronate (MucoUp; Boston Scientific, Tokyo, Japan) was used. After submucosal injection, a precut was made with the Dual Knife J or needle knife, followed by a circumferential mucosal incision around the lesion using the dots as a landmark and submucosal dissection with the IT Knife, IT Knife 2, IT Knife nano, or Dual Knife J. The resected specimens were evaluated pathologically. All endoscopic procedures were performed by experts in ESD who had experience with more than 500 clinical cases. There were no restrictions on the scopes and devices used by each endoscopist for ESD. The scopes used were GIF-Q260J or GIF-H260 (Olympus, Tokyo, Japan), and the devices were an insulation-tipped diathermic knife (IT Knife), IT Knife 2, IT Knife nano, or Dual Knife J (Olympus, Tokyo, Japan). Other devices, such as an argon plasma coagulation probe (ERBE, Tubingen, Germany) for marking dots or a needle knife (ZEON MEDICAL, Tokyo, Japan) for the initial incision, were occasionally used. First, marking dots for the incision lines were placed around the lesion. Next, fructose-added glycerol (Glyceol; TAIYO Pharma CO, Tokyo, Japan) with a minute amount of indigo carmine dye was injected into the submucosal layer. In some cases, 0.4% sodium hyaluronate (MucoUp; Boston Scientific, Tokyo, Japan) was used. After submucosal injection, a precut was made with the Dual Knife J or needle knife, followed by a circumferential mucosal incision around the lesion using the dots as a landmark and submucosal dissection with the IT Knife, IT Knife 2, IT Knife nano, or Dual Knife J. The resected specimens were evaluated pathologically. Histopathological assessment of curability ESD specimens were evaluated in 2-mm slices according to the Japanese Classification of Gastric Carcinoma with curability assessments divided into curative and non-curative resection based on the Gastric Cancer Treatment Guidelines[20]. R0 resection indicated that the lesion was resected en bloc with both the horizontal and vertical margins tumor-free histopathologically, but did not include findings regarding lymphovascular infiltration, the type of adenocarcinoma, or an assessment of the depth of invasion for curability. A curative resection was divided into eCura A and eCura B. A non-curative resection was defined as not meeting the criteria of curative resection and was further separated into 2 groups, eCura C-1 and eCura C-2, based on histopathological results per the Gastric Cancer Treatment Guidelines[19]. ESD specimens were evaluated in 2-mm slices according to the Japanese Classification of Gastric Carcinoma with curability assessments divided into curative and non-curative resection based on the Gastric Cancer Treatment Guidelines[20]. R0 resection indicated that the lesion was resected en bloc with both the horizontal and vertical margins tumor-free histopathologically, but did not include findings regarding lymphovascular infiltration, the type of adenocarcinoma, or an assessment of the depth of invasion for curability. A curative resection was divided into eCura A and eCura B. A non-curative resection was defined as not meeting the criteria of curative resection and was further separated into 2 groups, eCura C-1 and eCura C-2, based on histopathological results per the Gastric Cancer Treatment Guidelines[19]. Statistical analysis Continuous and categorical variables are expressed as median (range) and n (%), respectively. Overall survival was calculated according to the Kaplan-Meier method. Differences in the clinical outcomes of ESD for GTC between institutions were evaluated using the Mann-Whitney U test for continuous data and the Chi-squared test for categorical variables. The risk factors for long procedure time were evaluated using logistic regression analysis. All statistical analyses were performed using the statistical analysis software JMP Pro, version 15 (SAS Institute Inc., Cary, NC, United States). P values < 0.05 were considered statistically significant. Continuous and categorical variables are expressed as median (range) and n (%), respectively. Overall survival was calculated according to the Kaplan-Meier method. Differences in the clinical outcomes of ESD for GTC between institutions were evaluated using the Mann-Whitney U test for continuous data and the Chi-squared test for categorical variables. The risk factors for long procedure time were evaluated using logistic regression analysis. All statistical analyses were performed using the statistical analysis software JMP Pro, version 15 (SAS Institute Inc., Cary, NC, United States). P values < 0.05 were considered statistically significant.

CONCLUSION

As some patients were observed for only a short time, the assessment of long-term prognosis after ESD for GTC was insufficient. Further accumulation and follow-up of cases of GTC are necessary in the future.

INTRODUCTION: Recently, the survival of patients with esophageal cancer after esophagectomy has improved[1-5]. However, the risk of a subsequent occurrence of primary cancer is high in these patients. The most frequent cancer that overlaps with esophageal cancer is head and neck cancer, while the second most common is gastric cancer, including gastric tube cancer (GTC)[6-9]. Therefore, the improved prognosis of esophageal cancer patients has led to an increase in the occurrence of GTC in the reconstructed gastric tube. For the treatment of GTC after esophagectomy, total gastric tube resection (TGTR) or partial gastric tube resection (PGTR) has been proposed. However, surgical resection for GTC, being a secondary operation following esophagectomy, may lead to high mortality and morbidity[10,11]. On the other hand, in recent years, endoscopic therapy for early gastric cancer (EGC) has developed and become widespread[12]. Endoscopic submucosal dissection (ESD) enables the treatment of large lesions with a higher rate of en bloc resection that cannot be achieved by using conventional endoscopic mucosal resection. In addition, ESD is less invasive than surgery. For this reason, ESD has become widely used as a standard treatment for EGC, and ESD is often performed for GTC. However, ESD for GTC after esophagectomy is a technically difficult procedure compared with that for an unresected stomach because of the limited working space, unusual fluid-pooling area, food residue, bile reflux, fibrosis, and staples under the suture line[13]. Therefore, a high degree of skill is required for ESD of GTC. There are few reports about ESD for GTC after esophagectomy, and most are case reports and case series of a small number of patients[13-17]. A study by Nonaka et al[18] reported the effectiveness and safety of ESD for GTC in a high-volume national center, which had largest number of cases but was nonetheless a single-center study. Therefore, the aim of this study was to evaluate the efficacy and safety of ESD for GTC after esophagectomy in a multicenter context. MATERIALS AND METHODS: Patients We retrospectively investigated patients with GTC in the reconstructed gastric tube after esophagectomy for esophageal squamous cell carcinoma who had undergone ESD between January 2005 and December 2019 at 8 institutions participating in the Okayama Gut Study group (O-GUTS). All of the participating institutions in O-GUTS, except Okayama University Hospital (OUH), were considered core hospitals in each area. During the study period, 48 GTC lesions in 38 consecutive patients were treated. The clinical indications of ESD for EGC were based on the Gastric Cancer Treatment Guidelines[19]. These indications were similarly applied for GTC after esophagectomy with gastric tube reconstruction. Study measurements were as follows: patient characteristics, endoscopic findings, treatment results, adverse events, histopathological results, and clinical courses. In addition, we defined OUH as a high-volume center and compared the patients’ background and clinical outcomes between OUH and other facilities. The institutional review board of each hospital approved this study, and informed consent was obtained from all patients. We retrospectively investigated patients with GTC in the reconstructed gastric tube after esophagectomy for esophageal squamous cell carcinoma who had undergone ESD between January 2005 and December 2019 at 8 institutions participating in the Okayama Gut Study group (O-GUTS). All of the participating institutions in O-GUTS, except Okayama University Hospital (OUH), were considered core hospitals in each area. During the study period, 48 GTC lesions in 38 consecutive patients were treated. The clinical indications of ESD for EGC were based on the Gastric Cancer Treatment Guidelines[19]. These indications were similarly applied for GTC after esophagectomy with gastric tube reconstruction. Study measurements were as follows: patient characteristics, endoscopic findings, treatment results, adverse events, histopathological results, and clinical courses. In addition, we defined OUH as a high-volume center and compared the patients’ background and clinical outcomes between OUH and other facilities. The institutional review board of each hospital approved this study, and informed consent was obtained from all patients. Endoscopic procedures All endoscopic procedures were performed by experts in ESD who had experience with more than 500 clinical cases. There were no restrictions on the scopes and devices used by each endoscopist for ESD. The scopes used were GIF-Q260J or GIF-H260 (Olympus, Tokyo, Japan), and the devices were an insulation-tipped diathermic knife (IT Knife), IT Knife 2, IT Knife nano, or Dual Knife J (Olympus, Tokyo, Japan). Other devices, such as an argon plasma coagulation probe (ERBE, Tubingen, Germany) for marking dots or a needle knife (ZEON MEDICAL, Tokyo, Japan) for the initial incision, were occasionally used. First, marking dots for the incision lines were placed around the lesion. Next, fructose-added glycerol (Glyceol; TAIYO Pharma CO, Tokyo, Japan) with a minute amount of indigo carmine dye was injected into the submucosal layer. In some cases, 0.4% sodium hyaluronate (MucoUp; Boston Scientific, Tokyo, Japan) was used. After submucosal injection, a precut was made with the Dual Knife J or needle knife, followed by a circumferential mucosal incision around the lesion using the dots as a landmark and submucosal dissection with the IT Knife, IT Knife 2, IT Knife nano, or Dual Knife J. The resected specimens were evaluated pathologically. All endoscopic procedures were performed by experts in ESD who had experience with more than 500 clinical cases. There were no restrictions on the scopes and devices used by each endoscopist for ESD. The scopes used were GIF-Q260J or GIF-H260 (Olympus, Tokyo, Japan), and the devices were an insulation-tipped diathermic knife (IT Knife), IT Knife 2, IT Knife nano, or Dual Knife J (Olympus, Tokyo, Japan). Other devices, such as an argon plasma coagulation probe (ERBE, Tubingen, Germany) for marking dots or a needle knife (ZEON MEDICAL, Tokyo, Japan) for the initial incision, were occasionally used. First, marking dots for the incision lines were placed around the lesion. Next, fructose-added glycerol (Glyceol; TAIYO Pharma CO, Tokyo, Japan) with a minute amount of indigo carmine dye was injected into the submucosal layer. In some cases, 0.4% sodium hyaluronate (MucoUp; Boston Scientific, Tokyo, Japan) was used. After submucosal injection, a precut was made with the Dual Knife J or needle knife, followed by a circumferential mucosal incision around the lesion using the dots as a landmark and submucosal dissection with the IT Knife, IT Knife 2, IT Knife nano, or Dual Knife J. The resected specimens were evaluated pathologically. Histopathological assessment of curability ESD specimens were evaluated in 2-mm slices according to the Japanese Classification of Gastric Carcinoma with curability assessments divided into curative and non-curative resection based on the Gastric Cancer Treatment Guidelines[20]. R0 resection indicated that the lesion was resected en bloc with both the horizontal and vertical margins tumor-free histopathologically, but did not include findings regarding lymphovascular infiltration, the type of adenocarcinoma, or an assessment of the depth of invasion for curability. A curative resection was divided into eCura A and eCura B. A non-curative resection was defined as not meeting the criteria of curative resection and was further separated into 2 groups, eCura C-1 and eCura C-2, based on histopathological results per the Gastric Cancer Treatment Guidelines[19]. ESD specimens were evaluated in 2-mm slices according to the Japanese Classification of Gastric Carcinoma with curability assessments divided into curative and non-curative resection based on the Gastric Cancer Treatment Guidelines[20]. R0 resection indicated that the lesion was resected en bloc with both the horizontal and vertical margins tumor-free histopathologically, but did not include findings regarding lymphovascular infiltration, the type of adenocarcinoma, or an assessment of the depth of invasion for curability. A curative resection was divided into eCura A and eCura B. A non-curative resection was defined as not meeting the criteria of curative resection and was further separated into 2 groups, eCura C-1 and eCura C-2, based on histopathological results per the Gastric Cancer Treatment Guidelines[19]. Statistical analysis Continuous and categorical variables are expressed as median (range) and n (%), respectively. Overall survival was calculated according to the Kaplan-Meier method. Differences in the clinical outcomes of ESD for GTC between institutions were evaluated using the Mann-Whitney U test for continuous data and the Chi-squared test for categorical variables. The risk factors for long procedure time were evaluated using logistic regression analysis. All statistical analyses were performed using the statistical analysis software JMP Pro, version 15 (SAS Institute Inc., Cary, NC, United States). P values < 0.05 were considered statistically significant. Continuous and categorical variables are expressed as median (range) and n (%), respectively. Overall survival was calculated according to the Kaplan-Meier method. Differences in the clinical outcomes of ESD for GTC between institutions were evaluated using the Mann-Whitney U test for continuous data and the Chi-squared test for categorical variables. The risk factors for long procedure time were evaluated using logistic regression analysis. All statistical analyses were performed using the statistical analysis software JMP Pro, version 15 (SAS Institute Inc., Cary, NC, United States). P values < 0.05 were considered statistically significant. Patients: We retrospectively investigated patients with GTC in the reconstructed gastric tube after esophagectomy for esophageal squamous cell carcinoma who had undergone ESD between January 2005 and December 2019 at 8 institutions participating in the Okayama Gut Study group (O-GUTS). All of the participating institutions in O-GUTS, except Okayama University Hospital (OUH), were considered core hospitals in each area. During the study period, 48 GTC lesions in 38 consecutive patients were treated. The clinical indications of ESD for EGC were based on the Gastric Cancer Treatment Guidelines[19]. These indications were similarly applied for GTC after esophagectomy with gastric tube reconstruction. Study measurements were as follows: patient characteristics, endoscopic findings, treatment results, adverse events, histopathological results, and clinical courses. In addition, we defined OUH as a high-volume center and compared the patients’ background and clinical outcomes between OUH and other facilities. The institutional review board of each hospital approved this study, and informed consent was obtained from all patients. Endoscopic procedures: All endoscopic procedures were performed by experts in ESD who had experience with more than 500 clinical cases. There were no restrictions on the scopes and devices used by each endoscopist for ESD. The scopes used were GIF-Q260J or GIF-H260 (Olympus, Tokyo, Japan), and the devices were an insulation-tipped diathermic knife (IT Knife), IT Knife 2, IT Knife nano, or Dual Knife J (Olympus, Tokyo, Japan). Other devices, such as an argon plasma coagulation probe (ERBE, Tubingen, Germany) for marking dots or a needle knife (ZEON MEDICAL, Tokyo, Japan) for the initial incision, were occasionally used. First, marking dots for the incision lines were placed around the lesion. Next, fructose-added glycerol (Glyceol; TAIYO Pharma CO, Tokyo, Japan) with a minute amount of indigo carmine dye was injected into the submucosal layer. In some cases, 0.4% sodium hyaluronate (MucoUp; Boston Scientific, Tokyo, Japan) was used. After submucosal injection, a precut was made with the Dual Knife J or needle knife, followed by a circumferential mucosal incision around the lesion using the dots as a landmark and submucosal dissection with the IT Knife, IT Knife 2, IT Knife nano, or Dual Knife J. The resected specimens were evaluated pathologically. Histopathological assessment of curability: ESD specimens were evaluated in 2-mm slices according to the Japanese Classification of Gastric Carcinoma with curability assessments divided into curative and non-curative resection based on the Gastric Cancer Treatment Guidelines[20]. R0 resection indicated that the lesion was resected en bloc with both the horizontal and vertical margins tumor-free histopathologically, but did not include findings regarding lymphovascular infiltration, the type of adenocarcinoma, or an assessment of the depth of invasion for curability. A curative resection was divided into eCura A and eCura B. A non-curative resection was defined as not meeting the criteria of curative resection and was further separated into 2 groups, eCura C-1 and eCura C-2, based on histopathological results per the Gastric Cancer Treatment Guidelines[19]. Statistical analysis: Continuous and categorical variables are expressed as median (range) and n (%), respectively. Overall survival was calculated according to the Kaplan-Meier method. Differences in the clinical outcomes of ESD for GTC between institutions were evaluated using the Mann-Whitney U test for continuous data and the Chi-squared test for categorical variables. The risk factors for long procedure time were evaluated using logistic regression analysis. All statistical analyses were performed using the statistical analysis software JMP Pro, version 15 (SAS Institute Inc., Cary, NC, United States). P values < 0.05 were considered statistically significant. RESULTS: Patients’ characteristics and endoscopic findings A total of 38 consecutive patients with 48 GTC lesions were treated with ESD between January 2005 and December 2019 (Table 1). The median age of these patients was 71.5 years (range, 57-84 years), and they included 34 men and 4 women. The median period from esophagectomy to the treatment of GTC was 2106 d (range, 38-9523 d). This included patients who had a diagnosis of EGC before surgery for esophageal cancer and had undergone ESD after esophagectomy (5 patients). The reconstruction routes were antethoracic, retrosternal, and posterior mediastinal in 7, 11, and 20 patients, respectively. The location of the GTC lesion was upper, middle, and low in 2, 18, and 28 patients, respectively. Regarding the macroscopic type, there were 21 lesions of 0-IIa, 22 lesions of 0-IIc, 2 lesions of 0-IIb, 1 lesion of 0-III, and 2 combined lesions. The median observation period after ESD was 884 d (range, 8-4040 d). Patients’ characteristics and endoscopic findings ESD: Endoscopic submucosal dissection; GTC: Gastric tube cancer; U: Upper; M: Medium; L: Lower. A total of 38 consecutive patients with 48 GTC lesions were treated with ESD between January 2005 and December 2019 (Table 1). The median age of these patients was 71.5 years (range, 57-84 years), and they included 34 men and 4 women. The median period from esophagectomy to the treatment of GTC was 2106 d (range, 38-9523 d). This included patients who had a diagnosis of EGC before surgery for esophageal cancer and had undergone ESD after esophagectomy (5 patients). The reconstruction routes were antethoracic, retrosternal, and posterior mediastinal in 7, 11, and 20 patients, respectively. The location of the GTC lesion was upper, middle, and low in 2, 18, and 28 patients, respectively. Regarding the macroscopic type, there were 21 lesions of 0-IIa, 22 lesions of 0-IIc, 2 lesions of 0-IIb, 1 lesion of 0-III, and 2 combined lesions. The median observation period after ESD was 884 d (range, 8-4040 d). Patients’ characteristics and endoscopic findings ESD: Endoscopic submucosal dissection; GTC: Gastric tube cancer; U: Upper; M: Medium; L: Lower. Treatment results of ESD and histopathological findings Treatment results of ESD for GTC after esophagectomy and pathological findings are shown in Table 2. The median procedure time was 81 min (range, 29-334 min). En bloc resection was performed in 44 of 48 lesions (91.7%). The median tumor size of the resected specimen was 15 mm (range, 4-60 mm). Among the 48 lesions, 43 were differentiated (90%) and 5 were undifferentiated (10%). Regarding the tumor depth, 40 lesions were intramucosal carcinoma (M, 84%), 4 were submucosal superficial carcinoma (SM1, 8%), and 4 were submucosal deep invasive carcinoma (SM2 or deeper, 8%). Ulcerative findings were seen in 6 lesions (13%). Lymphatic infiltration was seen in 3 lesions (6%), and vascular infiltration was seen in 1 lesion (2%). According to the Japanese Gastric Cancer Treatment Guidelines, 38 lesions (79%) achieved curative resection (eCura A) and 10 lesions (21%) were classified as non-curative resection. The reasons for non-curative resection were as follows: 3 lesions were horizontal margin positive (HM1) or cutting into the lesion (eCura C-1), 2 were undifferentiated and showed SM invasion, 2 showed lymphatic infiltration, 2 showed SM invasion with ulcerative findings, and 1 was undifferentiated and showed SM invasion and lymphatic and vascular infiltration (eCura C-2). Treatment results of endoscopic submucosal dissection for gastric tube cancer and histopathological findings ESD: Endoscopic submucosal dissection; M: Intramucosal; SM1: Submucosal superficial; SM2: Submucosal deep invasive. Treatment results of ESD for GTC after esophagectomy and pathological findings are shown in Table 2. The median procedure time was 81 min (range, 29-334 min). En bloc resection was performed in 44 of 48 lesions (91.7%). The median tumor size of the resected specimen was 15 mm (range, 4-60 mm). Among the 48 lesions, 43 were differentiated (90%) and 5 were undifferentiated (10%). Regarding the tumor depth, 40 lesions were intramucosal carcinoma (M, 84%), 4 were submucosal superficial carcinoma (SM1, 8%), and 4 were submucosal deep invasive carcinoma (SM2 or deeper, 8%). Ulcerative findings were seen in 6 lesions (13%). Lymphatic infiltration was seen in 3 lesions (6%), and vascular infiltration was seen in 1 lesion (2%). According to the Japanese Gastric Cancer Treatment Guidelines, 38 lesions (79%) achieved curative resection (eCura A) and 10 lesions (21%) were classified as non-curative resection. The reasons for non-curative resection were as follows: 3 lesions were horizontal margin positive (HM1) or cutting into the lesion (eCura C-1), 2 were undifferentiated and showed SM invasion, 2 showed lymphatic infiltration, 2 showed SM invasion with ulcerative findings, and 1 was undifferentiated and showed SM invasion and lymphatic and vascular infiltration (eCura C-2). Treatment results of endoscopic submucosal dissection for gastric tube cancer and histopathological findings ESD: Endoscopic submucosal dissection; M: Intramucosal; SM1: Submucosal superficial; SM2: Submucosal deep invasive. Adverse events Complications during ESD were seen in 2 cases (4%), with 1 case of perforation, and 1 case of bleeding. Complications after ESD were seen in 5 cases (10%), with 2 cases of bleeding, 1 case of subcutaneous abscess, 1 case of liver failure, and 1 case of respiratory failure (Table 2). It was the same patient who had perforation during ESD and who formed subcutaneous abscess after ESD (Figure 1). In this case, perforation during ESD was sealed immediately with endoclips. Nevertheless, 20 d after ESD, the patient was admitted to the hospital with redness of the skin in the precordial area and excretion of pus from the skin. Computed tomography showed formation of a subcutaneous abscess around the gastric tube of the antethoracic reconstruction route. The patient was treated conservatively with antibiotics and percutaneous drainage and was discharged on the 16th day after the start of re-admission. A case of subcutaneous abscess formation after perforation during endoscopic submucosal dissection for gastric tube cancer. A: Gastric tube cancer located at the anterior wall of gastric body; B: Marking dots were placed around the lesion, and endoscopic submucosal dissection (ESD) was performed as usual; C: Perforation occurred during ESD; D: Perforation was sealed immediately with 4 endoclips; E: Redness of the skin in the precordial area, 20 d after ESD; F: Computed tomography performed 20 d after ESD. A subcutaneous abscess (yellow arrow) had formed around the gastric tube of the antethoracic reconstruction route (orange arrow). Complications during ESD were seen in 2 cases (4%), with 1 case of perforation, and 1 case of bleeding. Complications after ESD were seen in 5 cases (10%), with 2 cases of bleeding, 1 case of subcutaneous abscess, 1 case of liver failure, and 1 case of respiratory failure (Table 2). It was the same patient who had perforation during ESD and who formed subcutaneous abscess after ESD (Figure 1). In this case, perforation during ESD was sealed immediately with endoclips. Nevertheless, 20 d after ESD, the patient was admitted to the hospital with redness of the skin in the precordial area and excretion of pus from the skin. Computed tomography showed formation of a subcutaneous abscess around the gastric tube of the antethoracic reconstruction route. The patient was treated conservatively with antibiotics and percutaneous drainage and was discharged on the 16th day after the start of re-admission. A case of subcutaneous abscess formation after perforation during endoscopic submucosal dissection for gastric tube cancer. A: Gastric tube cancer located at the anterior wall of gastric body; B: Marking dots were placed around the lesion, and endoscopic submucosal dissection (ESD) was performed as usual; C: Perforation occurred during ESD; D: Perforation was sealed immediately with 4 endoclips; E: Redness of the skin in the precordial area, 20 d after ESD; F: Computed tomography performed 20 d after ESD. A subcutaneous abscess (yellow arrow) had formed around the gastric tube of the antethoracic reconstruction route (orange arrow). Patients’ clinical courses Of the 38 cases, 2 had local recurrence and 3 had metachronous recurrence. In the 2 cases with local recurrence, 1 received additional surgery and the other received additional ESD. In the 3 cases with metachronous recurrence, 1 received additional surgery and the others received additional ESD. The patients’ overall survival curve is shown in Figure 2. The survival rate at 5 years was 59.5%. During the observation period after ESD, no patient died of GTC. However, 10 patients died of other diseases, including pneumonia, which was the most common and occurred in 4 patients, heart failure and hepatocellular carcinoma in 1 patient each, and other unknown diseases. Overall survival curve after endoscopic submucosal dissection for gastric tube cancer. The survival rate at 5 year was 59.5%. ESD: Endoscopic submucosal dissection. Of the 38 cases, 2 had local recurrence and 3 had metachronous recurrence. In the 2 cases with local recurrence, 1 received additional surgery and the other received additional ESD. In the 3 cases with metachronous recurrence, 1 received additional surgery and the others received additional ESD. The patients’ overall survival curve is shown in Figure 2. The survival rate at 5 years was 59.5%. During the observation period after ESD, no patient died of GTC. However, 10 patients died of other diseases, including pneumonia, which was the most common and occurred in 4 patients, heart failure and hepatocellular carcinoma in 1 patient each, and other unknown diseases. Overall survival curve after endoscopic submucosal dissection for gastric tube cancer. The survival rate at 5 year was 59.5%. ESD: Endoscopic submucosal dissection. Comparison of clinical outcomes A comparison of the patients’ background and clinical outcomes between OUH and other hospitals is shown in Table 3. In terms of the patients’ backgrounds, the posterior mediastinal route was used as a reconstruction route in more cases at other hospitals. Treatment results were generally similar in both groups; however, procedure time was significantly longer at other hospitals. Comparison of clinical outcomes between Okayama University Hospital and other hospitals OUH: Okayama University Hospital; ESD: Endoscopic submucosal dissection; M: Intramucosal; SM: Submucosal. Since there were differences in procedure time between institutions, we divided patients into two groups, a short procedure time group (< 90 min) and a long procedure time group (≥ 90 min), and examined the factors affecting the procedure time. In univariate analysis (Table 4), the treatment institution and tumor size showed significant differences between the two groups. However, in multivariate analysis (Table 5), tumor size was the only factor associated with a long procedure time. Comparison of short (< 90 min) and long (≥ 90 min) procedure time groups U: Upper; M: Medium; L: Lower; M: Intramucosal; SM: Submucosal. Multivariate analysis about risk factors for a long procedure time of endoscopic submucosal dissection for gastric tube cancer U: Upper; M: Medium; CI: Confidence interval. A comparison of the patients’ background and clinical outcomes between OUH and other hospitals is shown in Table 3. In terms of the patients’ backgrounds, the posterior mediastinal route was used as a reconstruction route in more cases at other hospitals. Treatment results were generally similar in both groups; however, procedure time was significantly longer at other hospitals. Comparison of clinical outcomes between Okayama University Hospital and other hospitals OUH: Okayama University Hospital; ESD: Endoscopic submucosal dissection; M: Intramucosal; SM: Submucosal. Since there were differences in procedure time between institutions, we divided patients into two groups, a short procedure time group (< 90 min) and a long procedure time group (≥ 90 min), and examined the factors affecting the procedure time. In univariate analysis (Table 4), the treatment institution and tumor size showed significant differences between the two groups. However, in multivariate analysis (Table 5), tumor size was the only factor associated with a long procedure time. Comparison of short (< 90 min) and long (≥ 90 min) procedure time groups U: Upper; M: Medium; L: Lower; M: Intramucosal; SM: Submucosal. Multivariate analysis about risk factors for a long procedure time of endoscopic submucosal dissection for gastric tube cancer U: Upper; M: Medium; CI: Confidence interval. Patients’ characteristics and endoscopic findings: A total of 38 consecutive patients with 48 GTC lesions were treated with ESD between January 2005 and December 2019 (Table 1). The median age of these patients was 71.5 years (range, 57-84 years), and they included 34 men and 4 women. The median period from esophagectomy to the treatment of GTC was 2106 d (range, 38-9523 d). This included patients who had a diagnosis of EGC before surgery for esophageal cancer and had undergone ESD after esophagectomy (5 patients). The reconstruction routes were antethoracic, retrosternal, and posterior mediastinal in 7, 11, and 20 patients, respectively. The location of the GTC lesion was upper, middle, and low in 2, 18, and 28 patients, respectively. Regarding the macroscopic type, there were 21 lesions of 0-IIa, 22 lesions of 0-IIc, 2 lesions of 0-IIb, 1 lesion of 0-III, and 2 combined lesions. The median observation period after ESD was 884 d (range, 8-4040 d). Patients’ characteristics and endoscopic findings ESD: Endoscopic submucosal dissection; GTC: Gastric tube cancer; U: Upper; M: Medium; L: Lower. Treatment results of ESD and histopathological findings: Treatment results of ESD for GTC after esophagectomy and pathological findings are shown in Table 2. The median procedure time was 81 min (range, 29-334 min). En bloc resection was performed in 44 of 48 lesions (91.7%). The median tumor size of the resected specimen was 15 mm (range, 4-60 mm). Among the 48 lesions, 43 were differentiated (90%) and 5 were undifferentiated (10%). Regarding the tumor depth, 40 lesions were intramucosal carcinoma (M, 84%), 4 were submucosal superficial carcinoma (SM1, 8%), and 4 were submucosal deep invasive carcinoma (SM2 or deeper, 8%). Ulcerative findings were seen in 6 lesions (13%). Lymphatic infiltration was seen in 3 lesions (6%), and vascular infiltration was seen in 1 lesion (2%). According to the Japanese Gastric Cancer Treatment Guidelines, 38 lesions (79%) achieved curative resection (eCura A) and 10 lesions (21%) were classified as non-curative resection. The reasons for non-curative resection were as follows: 3 lesions were horizontal margin positive (HM1) or cutting into the lesion (eCura C-1), 2 were undifferentiated and showed SM invasion, 2 showed lymphatic infiltration, 2 showed SM invasion with ulcerative findings, and 1 was undifferentiated and showed SM invasion and lymphatic and vascular infiltration (eCura C-2). Treatment results of endoscopic submucosal dissection for gastric tube cancer and histopathological findings ESD: Endoscopic submucosal dissection; M: Intramucosal; SM1: Submucosal superficial; SM2: Submucosal deep invasive. Adverse events: Complications during ESD were seen in 2 cases (4%), with 1 case of perforation, and 1 case of bleeding. Complications after ESD were seen in 5 cases (10%), with 2 cases of bleeding, 1 case of subcutaneous abscess, 1 case of liver failure, and 1 case of respiratory failure (Table 2). It was the same patient who had perforation during ESD and who formed subcutaneous abscess after ESD (Figure 1). In this case, perforation during ESD was sealed immediately with endoclips. Nevertheless, 20 d after ESD, the patient was admitted to the hospital with redness of the skin in the precordial area and excretion of pus from the skin. Computed tomography showed formation of a subcutaneous abscess around the gastric tube of the antethoracic reconstruction route. The patient was treated conservatively with antibiotics and percutaneous drainage and was discharged on the 16th day after the start of re-admission. A case of subcutaneous abscess formation after perforation during endoscopic submucosal dissection for gastric tube cancer. A: Gastric tube cancer located at the anterior wall of gastric body; B: Marking dots were placed around the lesion, and endoscopic submucosal dissection (ESD) was performed as usual; C: Perforation occurred during ESD; D: Perforation was sealed immediately with 4 endoclips; E: Redness of the skin in the precordial area, 20 d after ESD; F: Computed tomography performed 20 d after ESD. A subcutaneous abscess (yellow arrow) had formed around the gastric tube of the antethoracic reconstruction route (orange arrow). Patients’ clinical courses: Of the 38 cases, 2 had local recurrence and 3 had metachronous recurrence. In the 2 cases with local recurrence, 1 received additional surgery and the other received additional ESD. In the 3 cases with metachronous recurrence, 1 received additional surgery and the others received additional ESD. The patients’ overall survival curve is shown in Figure 2. The survival rate at 5 years was 59.5%. During the observation period after ESD, no patient died of GTC. However, 10 patients died of other diseases, including pneumonia, which was the most common and occurred in 4 patients, heart failure and hepatocellular carcinoma in 1 patient each, and other unknown diseases. Overall survival curve after endoscopic submucosal dissection for gastric tube cancer. The survival rate at 5 year was 59.5%. ESD: Endoscopic submucosal dissection. Comparison of clinical outcomes: A comparison of the patients’ background and clinical outcomes between OUH and other hospitals is shown in Table 3. In terms of the patients’ backgrounds, the posterior mediastinal route was used as a reconstruction route in more cases at other hospitals. Treatment results were generally similar in both groups; however, procedure time was significantly longer at other hospitals. Comparison of clinical outcomes between Okayama University Hospital and other hospitals OUH: Okayama University Hospital; ESD: Endoscopic submucosal dissection; M: Intramucosal; SM: Submucosal. Since there were differences in procedure time between institutions, we divided patients into two groups, a short procedure time group (< 90 min) and a long procedure time group (≥ 90 min), and examined the factors affecting the procedure time. In univariate analysis (Table 4), the treatment institution and tumor size showed significant differences between the two groups. However, in multivariate analysis (Table 5), tumor size was the only factor associated with a long procedure time. Comparison of short (< 90 min) and long (≥ 90 min) procedure time groups U: Upper; M: Medium; L: Lower; M: Intramucosal; SM: Submucosal. Multivariate analysis about risk factors for a long procedure time of endoscopic submucosal dissection for gastric tube cancer U: Upper; M: Medium; CI: Confidence interval. DISCUSSION: This study was the first multicenter study on ESD for GTC in the reconstructed gastric tube after esophagectomy, and it included the second largest number of patients. According to a systematic review of GTC after esophagectomy, there are two surgical options for the treatment of GTC: PGTR or TGTR plus lymphadenectomy with colon or jejunal reconstruction[21]. However, surgical treatment for GTC is highly invasive and carries a certain degree of risk. Sugiura et al[10] reported that 5 of 7 cases of TGTR had surgical complications (leakage) and 2 died. In addition, 1 of 3 cases of PTGR had fatal complications. Akita et al[11] reported that 1 of 5 cases of TGTR died of postoperative complications. On the other hand, in previous studies on ESD for GTC, the proportions of R0 resection and curative resection were 87.5%-92% and 65%-85%, respectively, and complications were seen in 12.5%-18% of patients[13-18]. In the present study, the proportions of R0 resection and curative resection were 91.7% and 79%, respectively, and complications were seen in 10%. Overall, the treatment results of ESD for GTC in this study were similar to those of previous studies. In a previous study on gastric ESD of the unresected stomach, the proportions of R0 resection, curative resections, and complications were 92%-94.9%, 80.4%-94.7%, and 5.9%-6.3%, respectively[22,23]. Furthermore, in gastric ESD of the remnant stomach after gastrectomy, the proportions of R0 resection, curative resection, and complications were 84.7%-85%, 70.9%-78%, and 2.8%-21.1%, respectively[24,25]. ESD for GTC was considered a minimally invasive, effective, and relatively safe treatment. There are some points of note in GTC. First, detection of early GTC requires long-term regular endoscopic surveillance after esophagectomy. GTC is often found long after esophagectomy; in some cases, GTC is detected after more than 10 years and the risk of metachronous GTC is high[15-18]. Second, GTC may be difficult to diagnose. The reasons are as follows: Food residue and bile reflux are often seen in the gastric tube, and the lumen of the gastric tube is long and narrow and can constrain endoscopic observation[13]. Therefore, it is necessary to pay attention to these points during endoscopy for patients with gastric tube reconstruction after esophagectomy. Third, when performing ESD for GTC, it is necessary to pay attention to complications specific to GTC. For example, in our study, a subcutaneous abscess formed after treatment in a case with perforation during ESD for GTC in the antethoracic reconstruction route. This case was cured by conservative treatment with antibiotics and percutaneous drainage. Moreover, Miyagi et al[26] reported that post-treatment precordial skin burns occurred in 5 of 8 patients with GTC in the antethoracic reconstruction route. In this report, all burns were diagnosed as first-degree burns based on the clinical classification of burn depth, developed on postoperative day 1-2, and took 4-7 d to heal. In this study, since approximately half of the patients were treated at OUH, we defined it as a high-volume center and compared clinical outcomes with those of other institutions. As a result, there were no significant differences in the clinical outcomes of ESD between institutions. In addition, lesion size was the only factor related to long procedure time in multivariate analysis. We believe these results were attributable to the fact that all of the participating institutions specialized in gastrointestinal diseases with more than 500 cases of ESD for EGC. Moreover, ESD for GTC may have been performed by leading specialists given the relative rarity of GTC. For these reasons, ESD for GTC seems safe if performed by specialists with sufficient ESD experience. Previously, not a few patients had complications or died of other diseases during the course after esophagectomy[27,28]. However, due to the widespread use of minimally invasive esophagectomy, such as thoracoscopic and laparoscopic surgery, the incidence of postoperative complications, including respiratory complications, has decreased and the general condition of patients after esophagectomy has improved in recent years[29-32]. With continued improvements in the prognosis of esophageal cancer, the number of cases of GTC after esophagectomy will likely increase in the future and the demand for ESD for GTC is expected to increase further. There were several limitations to this study. First, as this was a retrospective study, the ESD indications and devices used for treatment were not standardized. However, treatment was performed according to the typical standards. Second, since data on Helicobacter pylori infection status were missing in some patients, the association between GTC and Helicobacter pylori could not be evaluated. Third, as some patients were observed for only a short time, the assessment of long-term prognosis after ESD for GTC was insufficient. Further follow-up studies are needed in the future. CONCLUSION: In conclusion, ESD for GTC after esophagectomy is a safe and effective treatment that can be performed without significant variability in treatment results at any specialized institution where standard gastric ESD can be performed with sufficient expertise. Further accumulation and follow-up of cases of GTC are necessary in the future.

Background: Recent improvements in the prognosis of patients with esophageal cancer have led to the increased occurrence of gastric tube cancer (GTC) in the reconstructed gastric tube. However, there are few reports on the treatment results of endoscopic submucosal dissection (ESD) for GTC. Methods: We retrospectively investigated 48 GTC lesions in 38 consecutive patients with GTC in the reconstructed gastric tube after esophagectomy who had undergone ESD between January 2005 and December 2019 at 8 institutions participating in the Okayama Gut Study group. The clinical indications of ESD for early gastric cancer were similarly applied for GTC after esophagectomy. ESD specimens were evaluated in 2-mm slices according to the Japanese Classification of Gastric Carcinoma with curability assessments divided into curative and non-curative resection based on the Gastric Cancer Treatment Guidelines. Patient characteristics, treatment results, clinical course, and treatment outcomes were analyzed. Results: The median age of patients was 71.5 years (range, 57-84years), and there were 34 men and 4 women. The median observation period after ESD was 884 d (range, 8-4040 d). The median procedure time was 81 min (range, 29-334 min), the en bloc resection rate was 91.7% (44/48), and the curative resection rate was 79% (38/48). Complications during ESD were seen in 4% (2/48) of case, and those after ESD were seen in 10% (5/48) of case. The survival rate at 5 years was 59.5%. During the observation period after ESD, 10 patients died of other diseases. Although there were differences in the procedure time between institutions, a multivariate analysis showed that tumor size was the only factor associated with prolonged procedure time. Conclusions: ESD for GTC after esophagectomy was shown to be safe and effective.

INTRODUCTION: Recently, the survival of patients with esophageal cancer after esophagectomy has improved[1-5]. However, the risk of a subsequent occurrence of primary cancer is high in these patients. The most frequent cancer that overlaps with esophageal cancer is head and neck cancer, while the second most common is gastric cancer, including gastric tube cancer (GTC)[6-9]. Therefore, the improved prognosis of esophageal cancer patients has led to an increase in the occurrence of GTC in the reconstructed gastric tube. For the treatment of GTC after esophagectomy, total gastric tube resection (TGTR) or partial gastric tube resection (PGTR) has been proposed. However, surgical resection for GTC, being a secondary operation following esophagectomy, may lead to high mortality and morbidity[10,11]. On the other hand, in recent years, endoscopic therapy for early gastric cancer (EGC) has developed and become widespread[12]. Endoscopic submucosal dissection (ESD) enables the treatment of large lesions with a higher rate of en bloc resection that cannot be achieved by using conventional endoscopic mucosal resection. In addition, ESD is less invasive than surgery. For this reason, ESD has become widely used as a standard treatment for EGC, and ESD is often performed for GTC. However, ESD for GTC after esophagectomy is a technically difficult procedure compared with that for an unresected stomach because of the limited working space, unusual fluid-pooling area, food residue, bile reflux, fibrosis, and staples under the suture line[13]. Therefore, a high degree of skill is required for ESD of GTC. There are few reports about ESD for GTC after esophagectomy, and most are case reports and case series of a small number of patients[13-17]. A study by Nonaka et al[18] reported the effectiveness and safety of ESD for GTC in a high-volume national center, which had largest number of cases but was nonetheless a single-center study. Therefore, the aim of this study was to evaluate the efficacy and safety of ESD for GTC after esophagectomy in a multicenter context. CONCLUSION: As some patients were observed for only a short time, the assessment of long-term prognosis after ESD for GTC was insufficient. Further accumulation and follow-up of cases of GTC are necessary in the future.

Background: Recent improvements in the prognosis of patients with esophageal cancer have led to the increased occurrence of gastric tube cancer (GTC) in the reconstructed gastric tube. However, there are few reports on the treatment results of endoscopic submucosal dissection (ESD) for GTC. Methods: We retrospectively investigated 48 GTC lesions in 38 consecutive patients with GTC in the reconstructed gastric tube after esophagectomy who had undergone ESD between January 2005 and December 2019 at 8 institutions participating in the Okayama Gut Study group. The clinical indications of ESD for early gastric cancer were similarly applied for GTC after esophagectomy. ESD specimens were evaluated in 2-mm slices according to the Japanese Classification of Gastric Carcinoma with curability assessments divided into curative and non-curative resection based on the Gastric Cancer Treatment Guidelines. Patient characteristics, treatment results, clinical course, and treatment outcomes were analyzed. Results: The median age of patients was 71.5 years (range, 57-84years), and there were 34 men and 4 women. The median observation period after ESD was 884 d (range, 8-4040 d). The median procedure time was 81 min (range, 29-334 min), the en bloc resection rate was 91.7% (44/48), and the curative resection rate was 79% (38/48). Complications during ESD were seen in 4% (2/48) of case, and those after ESD were seen in 10% (5/48) of case. The survival rate at 5 years was 59.5%. During the observation period after ESD, 10 patients died of other diseases. Although there were differences in the procedure time between institutions, a multivariate analysis showed that tumor size was the only factor associated with prolonged procedure time. Conclusions: ESD for GTC after esophagectomy was shown to be safe and effective.

[CONTENT] Endoscopic submucosal dissection | Gastric tube | Gastric cancer | Eso-phagectomy | Multicenter study | Retrospective study [SUMMARY]

[CONTENT] Endoscopic submucosal dissection | Gastric tube | Gastric cancer | Eso-phagectomy | Multicenter study | Retrospective study [SUMMARY]

[CONTENT] Endoscopic submucosal dissection | Gastric tube | Gastric cancer | Eso-phagectomy | Multicenter study | Retrospective study [SUMMARY]

[CONTENT] Endoscopic submucosal dissection | Gastric tube | Gastric cancer | Eso-phagectomy | Multicenter study | Retrospective study [SUMMARY]

[CONTENT] Endoscopic submucosal dissection | Gastric tube | Gastric cancer | Eso-phagectomy | Multicenter study | Retrospective study [SUMMARY]

[CONTENT] Aged | Aged, 80 and over | Dissection | Endoscopic Mucosal Resection | Female | Gastric Mucosa | Humans | Male | Middle Aged | Retrospective Studies | Stomach Neoplasms | Treatment Outcome [SUMMARY]

[CONTENT] Aged | Aged, 80 and over | Dissection | Endoscopic Mucosal Resection | Female | Gastric Mucosa | Humans | Male | Middle Aged | Retrospective Studies | Stomach Neoplasms | Treatment Outcome [SUMMARY]

[CONTENT] Aged | Aged, 80 and over | Dissection | Endoscopic Mucosal Resection | Female | Gastric Mucosa | Humans | Male | Middle Aged | Retrospective Studies | Stomach Neoplasms | Treatment Outcome [SUMMARY]

[CONTENT] Aged | Aged, 80 and over | Dissection | Endoscopic Mucosal Resection | Female | Gastric Mucosa | Humans | Male | Middle Aged | Retrospective Studies | Stomach Neoplasms | Treatment Outcome [SUMMARY]

[CONTENT] Aged | Aged, 80 and over | Dissection | Endoscopic Mucosal Resection | Female | Gastric Mucosa | Humans | Male | Middle Aged | Retrospective Studies | Stomach Neoplasms | Treatment Outcome [SUMMARY]

[CONTENT] gtc esophagectomy total | gtc esophagectomy pathological | review gtc esophagectomy | gastric tube esophagectomy | gtc esophagectomy gastric [SUMMARY]

[CONTENT] gtc esophagectomy total | gtc esophagectomy pathological | review gtc esophagectomy | gastric tube esophagectomy | gtc esophagectomy gastric [SUMMARY]

[CONTENT] gtc esophagectomy total | gtc esophagectomy pathological | review gtc esophagectomy | gastric tube esophagectomy | gtc esophagectomy gastric [SUMMARY]

[CONTENT] gtc esophagectomy total | gtc esophagectomy pathological | review gtc esophagectomy | gastric tube esophagectomy | gtc esophagectomy gastric [SUMMARY]

[CONTENT] gtc esophagectomy total | gtc esophagectomy pathological | review gtc esophagectomy | gastric tube esophagectomy | gtc esophagectomy gastric [SUMMARY]

[CONTENT] esd | patients | gtc | gastric | submucosal | treatment | endoscopic | cancer | lesions | resection [SUMMARY]

[CONTENT] esd | patients | gtc | gastric | submucosal | treatment | endoscopic | cancer | lesions | resection [SUMMARY]

[CONTENT] esd | patients | gtc | gastric | submucosal | treatment | endoscopic | cancer | lesions | resection [SUMMARY]

[CONTENT] esd | patients | gtc | gastric | submucosal | treatment | endoscopic | cancer | lesions | resection [SUMMARY]

[CONTENT] esd | patients | gtc | gastric | submucosal | treatment | endoscopic | cancer | lesions | resection [SUMMARY]

[CONTENT] gtc | cancer | esophagectomy | esd | resection | high | esd gtc | gastric | gtc esophagectomy | study [SUMMARY]

[CONTENT] knife | knife knife | tokyo japan | japan | tokyo | curative | resection | study | ecura | clinical [SUMMARY]

[CONTENT] significant variability treatment | accumulation follow cases gtc | expertise accumulation follow cases | gtc esophagectomy safe effective | results specialized institution standard | results specialized institution | results specialized | accumulation follow | accumulation follow cases | treatment performed significant [SUMMARY]

[CONTENT] esd | patients | gtc | knife | gastric | resection | lesions | submucosal | treatment | cancer [SUMMARY]

[CONTENT] esd | patients | gtc | knife | gastric | resection | lesions | submucosal | treatment | cancer [SUMMARY]

[CONTENT] GTC ||| ESD | GTC [SUMMARY]

[CONTENT] 48 | GTC | 38 | GTC | ESD | between January 2005 | December 2019 | 8 ||| ESD | GTC ||| ESD | 2-mm | the Japanese Classification of Gastric Carcinoma | the Gastric Cancer Treatment Guidelines ||| [SUMMARY]

[CONTENT] ESD | GTC [SUMMARY]

[CONTENT] GTC ||| ESD | GTC ||| 48 | GTC | 38 | GTC | ESD | between January 2005 | December 2019 | 8 ||| ESD | GTC ||| ESD | 2-mm | the Japanese Classification of Gastric Carcinoma | the Gastric Cancer Treatment Guidelines ||| ||| ||| 71.5 years | 57-84years | 34 | 4 ||| ESD | 884 | 8-4040 ||| 81 | 29 | 91.7% | 44/48 | 79% | 38/48 ||| ESD | 4% | 2/48 | ESD | 10% | 5/48 ||| 5 years | 59.5% ||| ESD | 10 ||| ||| GTC [SUMMARY]

[CONTENT] GTC ||| ESD | GTC ||| 48 | GTC | 38 | GTC | ESD | between January 2005 | December 2019 | 8 ||| ESD | GTC ||| ESD | 2-mm | the Japanese Classification of Gastric Carcinoma | the Gastric Cancer Treatment Guidelines ||| ||| ||| 71.5 years | 57-84years | 34 | 4 ||| ESD | 884 | 8-4040 ||| 81 | 29 | 91.7% | 44/48 | 79% | 38/48 ||| ESD | 4% | 2/48 | ESD | 10% | 5/48 ||| 5 years | 59.5% ||| ESD | 10 ||| ||| GTC [SUMMARY]

OATP1B1 and tumour OATP1B3 modulate exposure, toxicity, and survival after irinotecan-based chemotherapy.

25611302

Treatment of advanced and metastatic colorectal cancer with irinotecan is hampered by severe toxicities. The active metabolite of irinotecan, SN-38, is a known substrate of drug-metabolising enzymes, including UGT1A1, as well as OATP and ABC drug transporters.

BACKGROUND

Blood samples (n=127) and tumour tissue (n=30) were obtained from advanced cancer patients treated with irinotecan-based regimens for pharmacogenetic and drug level analysis and transporter expression. Clinical variables, toxicity, and outcomes data were collected.

METHODS

SLCO1B1 521C was significantly associated with increased SN-38 exposure (P<0.001), which was additive with UGT1A1*28. ABCC5 (rs562) carriers had significantly reduced SN-38 glucuronide and APC metabolite levels. Reduced risk of neutropenia and diarrhoea was associated with ABCC2-24C/T (odds ratio (OR)=0.22, 0.06-0.85) and CES1 (rs2244613; OR=0.29, 0.09-0.89), respectively. Progression-free survival (PFS) was significantly longer in SLCO1B1 388G/G patients and reduced in ABCC2-24T/T and UGT1A1*28 carriers. Notably, higher OATP1B3 tumour expression was associated with reduced PFS.

RESULTS

Clarifying the association of host genetic variation in OATP and ABC transporters to SN-38 exposure, toxicity and PFS provides rationale for personalising irinotecan-based chemotherapy. Our findings suggest that OATP polymorphisms and expression in tumour tissue may serve as important new biomarkers.

CONCLUSIONS

4453959

Results

Study population Patient and tumour characteristics (n=127) are described in Table 1. The treatment profile of patients is presented in Table 2. Approximately half (55%) of the study population had not received previous chemotherapy and the majority of patients (82%) were treated with bevacizumab (BEV)-FOLFIRI or FOLFIRI chemotherapy regimens as first-line therapy (Table 2). Patient and tumour characteristics (n=127) are described in Table 1. The treatment profile of patients is presented in Table 2. Approximately half (55%) of the study population had not received previous chemotherapy and the majority of patients (82%) were treated with bevacizumab (BEV)-FOLFIRI or FOLFIRI chemotherapy regimens as first-line therapy (Table 2). OATP1B1 and ABCC5 are important determinants of SN-38 and SN-38G levels The primary objective was to determine covariates associated with interindividual variability of plasma concentrations of CPT-11 and metabolites. Plasma concentrations of CPT-11, SN-38, SN-38G, and APC were measured from blood samples collected immediately following the end of CPT-11 infusion. Multiple linear regression was performed on dose-normalised drug levels (natural log transformed) adjusting for age at time of treatment initiation, sex and treatment regimen in the final models. ABCB1 (c.3435 C>T) was significantly associated with CPT-11 exposure as homozygous variant (T/T) carriers had lower levels compared with wild-type patients (P<0.05; Figure 1A and Supplementary Table 1). This model had one significant adjusting covariate (FOLFIRI treatment regimen, P <0.05). Several genotypes were significantly associated with variation in SN-38 levels as part of a model that explained approximately 27% of the variation in exposure (Table 3). SLCO1B1 521C allele carriers had significantly increased systemic exposure to SN-38 (P<0.001, Figure 1B). As expected, heterozygous (*1/*28) and homozygous variant (*28/*28) UGT1A1 carriers had increased SN-38 plasma levels (Figure 1C). Interestingly, a significant increase in SN-38 level was observed with an increasing number of combined SLCO1B1 521C and UGT1A1*28 variant alleles, suggesting an additive effect of polymorphisms in these two genes (Figure 1D). A corresponding decrease in the SN-38G/SN-38 ratio was observed with an increasing number of combined variant alleles (Figure 1E). Together, this suggests that patients heterozygous for both SLCO1B1 521C and UGT1A1*28 may have an equivalent risk of increased SN-38 exposure compared with patients homozygous for either polymorphism, respectively. Carriers of a rare SNP in CES1 (rs71647871, allele frequency, 0.024) had significantly decreased SN-38 levels (P<0.05, Table 3). Interestingly, SN-38G levels were significantly affected by ABCC5 genotype. Patients harbouring the ABCC5 rs562 C allele had reduced SN-38G plasma exposure compared with wild-type patients (P <0.001, Figure 1F, Supplementary Table 2). In addition, patients expressing CYP3A5 also had significantly reduced SN-38G plasma levels (P <0.05). FOLFIRI and FOLFIRINOX treatment regimens were significantly associated with SN-38G (P<0.001 and P <0.05, respectively). The model for the inactive metabolite APC explained approximately 24% of the variability in plasma exposure (Supplementary Table 3). SLCO1B3 g.699 and ABCC2 c.1249G>A were positively associated with APC levels (P<0.01). Decreased plasma levels were observed in patients carrying the ABCC5 rs562 C allele compared with wild-type patients (P<0.001). Significant adjusting covariates in this model included sex (P<0.05) and treatment regimen (P<0.01). The primary objective was to determine covariates associated with interindividual variability of plasma concentrations of CPT-11 and metabolites. Plasma concentrations of CPT-11, SN-38, SN-38G, and APC were measured from blood samples collected immediately following the end of CPT-11 infusion. Multiple linear regression was performed on dose-normalised drug levels (natural log transformed) adjusting for age at time of treatment initiation, sex and treatment regimen in the final models. ABCB1 (c.3435 C>T) was significantly associated with CPT-11 exposure as homozygous variant (T/T) carriers had lower levels compared with wild-type patients (P<0.05; Figure 1A and Supplementary Table 1). This model had one significant adjusting covariate (FOLFIRI treatment regimen, P <0.05). Several genotypes were significantly associated with variation in SN-38 levels as part of a model that explained approximately 27% of the variation in exposure (Table 3). SLCO1B1 521C allele carriers had significantly increased systemic exposure to SN-38 (P<0.001, Figure 1B). As expected, heterozygous (*1/*28) and homozygous variant (*28/*28) UGT1A1 carriers had increased SN-38 plasma levels (Figure 1C). Interestingly, a significant increase in SN-38 level was observed with an increasing number of combined SLCO1B1 521C and UGT1A1*28 variant alleles, suggesting an additive effect of polymorphisms in these two genes (Figure 1D). A corresponding decrease in the SN-38G/SN-38 ratio was observed with an increasing number of combined variant alleles (Figure 1E). Together, this suggests that patients heterozygous for both SLCO1B1 521C and UGT1A1*28 may have an equivalent risk of increased SN-38 exposure compared with patients homozygous for either polymorphism, respectively. Carriers of a rare SNP in CES1 (rs71647871, allele frequency, 0.024) had significantly decreased SN-38 levels (P<0.05, Table 3). Interestingly, SN-38G levels were significantly affected by ABCC5 genotype. Patients harbouring the ABCC5 rs562 C allele had reduced SN-38G plasma exposure compared with wild-type patients (P <0.001, Figure 1F, Supplementary Table 2). In addition, patients expressing CYP3A5 also had significantly reduced SN-38G plasma levels (P <0.05). FOLFIRI and FOLFIRINOX treatment regimens were significantly associated with SN-38G (P<0.001 and P <0.05, respectively). The model for the inactive metabolite APC explained approximately 24% of the variability in plasma exposure (Supplementary Table 3). SLCO1B3 g.699 and ABCC2 c.1249G>A were positively associated with APC levels (P<0.01). Decreased plasma levels were observed in patients carrying the ABCC5 rs562 C allele compared with wild-type patients (P<0.001). Significant adjusting covariates in this model included sex (P<0.05) and treatment regimen (P<0.01). Drug transporters predict CPT-11-related toxicities CPT-11-related adverse reactions and frequencies of grade severity are described in Supplementary Table 4. Multinomial logistic regression analysis was performed to determine association of genotypes to adverse reactions comparing no event vs low-grade vs high-grade toxicity (Table 4). ABCC2 -24C/T carriers had significantly lower risk of grade 3/4 neutropenia compared with wild-type patients (odds ratio (OR)=0.22, 95% CI=0.06–0.85). UGT1A1*28 carriers were at increased risk for severe neutropenia (grade 3/4) compared with wild-type patients following binary logistic regression analysing low- vs high-grade events (OR=3.67; 95% CI=1.19–11.33; Table 4). In addition to their increased SN-38 exposure, patients with two or more combined SLCO1B1 521C and UGT1A1*28 variant alleles were also at significantly increased risk of severe neutropenia (grade 3/4) compared with patients wild-type for both alleles (OR=4.154, 95% CI=1.06–16.36). Non-haematological toxicity was associated with drug transporters and metabolising enzymes (Table 4). Carriers of a common SNP in CES1 (rs2244613) had significantly lower risk of higher-grade diarrhoea (OR=0.29, 95% CI=0.09–0.89). Patients with ABCB1 3435 C/T and T/T genotypes were much more likely to experience higher-grade nausea/vomiting (OR=9.06, 95% CI=1.03–79.41 and OR=10.52, 95% CI=1.10–100.2, respectively). Higher-grade oral mucositis was observed in patients expressing CYP3A5 (OR=8.10, 95% CI=1.57–41.90), whereas those heterozygous for SLCO1B1 388A/G had a significantly lower risk (OR=0.19, 95% CI=0.05–0.72). No significant adjusting covariates were found in the binary logistic or multinomial logistic regression models. CPT-11-related adverse reactions and frequencies of grade severity are described in Supplementary Table 4. Multinomial logistic regression analysis was performed to determine association of genotypes to adverse reactions comparing no event vs low-grade vs high-grade toxicity (Table 4). ABCC2 -24C/T carriers had significantly lower risk of grade 3/4 neutropenia compared with wild-type patients (odds ratio (OR)=0.22, 95% CI=0.06–0.85). UGT1A1*28 carriers were at increased risk for severe neutropenia (grade 3/4) compared with wild-type patients following binary logistic regression analysing low- vs high-grade events (OR=3.67; 95% CI=1.19–11.33; Table 4). In addition to their increased SN-38 exposure, patients with two or more combined SLCO1B1 521C and UGT1A1*28 variant alleles were also at significantly increased risk of severe neutropenia (grade 3/4) compared with patients wild-type for both alleles (OR=4.154, 95% CI=1.06–16.36). Non-haematological toxicity was associated with drug transporters and metabolising enzymes (Table 4). Carriers of a common SNP in CES1 (rs2244613) had significantly lower risk of higher-grade diarrhoea (OR=0.29, 95% CI=0.09–0.89). Patients with ABCB1 3435 C/T and T/T genotypes were much more likely to experience higher-grade nausea/vomiting (OR=9.06, 95% CI=1.03–79.41 and OR=10.52, 95% CI=1.10–100.2, respectively). Higher-grade oral mucositis was observed in patients expressing CYP3A5 (OR=8.10, 95% CI=1.57–41.90), whereas those heterozygous for SLCO1B1 388A/G had a significantly lower risk (OR=0.19, 95% CI=0.05–0.72). No significant adjusting covariates were found in the binary logistic or multinomial logistic regression models. Biomarkers of PFS Approximately 73% of patients were considered responders having radiographic evidence of no change or tumour shrinkage during CPT-11-based therapy. At the time of analysis (censor date: 13 August 2013), disease progression was evident in the majority of patients with a median PFS time of 10.5 months (range, 0.2–43 months; Table 2). Cox-regression analysis for PFS was performed on patients (excluding pancreatic cancer patients) treated with BEV-FOLFIRI or FOLFIRI regimens only (n=103, Table 5). Patients homozygous for SLCO1B1 388G/G alleles had a significantly increased PFS compared with wild-type patients (HR=1.60, 95% CI=1.04–2.46). Patients carrying two ABCC2 c.-24T alleles had decreased PFS (HR=0.62, 95% CI=0.40–0.95). Reduced PFS was also observed for UGT1A1*28 (*1/*28, HR=0.69, 95% CI=0.52–0.92; *28/*28, HR=0.60, 95% CI=0.38–0.97). ABCC5 was associated with PFS in the univariate analysis (P=0.05), but did not remain significant following adjustment in the final model. No significant adjusting covariates were found in the multivariate analysis. Approximately 73% of patients were considered responders having radiographic evidence of no change or tumour shrinkage during CPT-11-based therapy. At the time of analysis (censor date: 13 August 2013), disease progression was evident in the majority of patients with a median PFS time of 10.5 months (range, 0.2–43 months; Table 2). Cox-regression analysis for PFS was performed on patients (excluding pancreatic cancer patients) treated with BEV-FOLFIRI or FOLFIRI regimens only (n=103, Table 5). Patients homozygous for SLCO1B1 388G/G alleles had a significantly increased PFS compared with wild-type patients (HR=1.60, 95% CI=1.04–2.46). Patients carrying two ABCC2 c.-24T alleles had decreased PFS (HR=0.62, 95% CI=0.40–0.95). Reduced PFS was also observed for UGT1A1*28 (*1/*28, HR=0.69, 95% CI=0.52–0.92; *28/*28, HR=0.60, 95% CI=0.38–0.97). ABCC5 was associated with PFS in the univariate analysis (P=0.05), but did not remain significant following adjustment in the final model. No significant adjusting covariates were found in the multivariate analysis. Tumour OATP1B3 expression suggests poorer clinical outcomes The role of OATP1B3 expression, known to transport SN-38, in response to CPT-11-based chemotherapy is unknown. Recently, OATP1B3 expression within colon tumours has been shown to be a cancer-specific isoform that may be expressed primarily as an intracellular protein calling into question the functional relevance of this transporter within the tumour (Thakkar et al, 2013). Paired normal and tumour samples were available for a subset of patients treated with CPT-11-based chemotherapy regimens (n=30: BEV-FOLFIRI, n=22; FOLFIRI, n=5; FOLFIRINOX, n=1; BEV-irinotecan, n=1; cetuximab-irinotecan, n=1). Staining for OATP1B3 was performed and intensity of expression was scored. Tumour tissue had a significantly higher OATP1B3 score compared with paired normal tissue (P<0.05, Figure 2A and C). Progression-free survival was significantly reduced in patients with high (score, 2 or 3) OATP1B3 expression compared with patients with low (score, 0 or 1) expression (Figure 2B), suggesting that OATP1B3 expression may correlate with poorer clinical response to CPT-11 therapy. The role of OATP1B3 expression, known to transport SN-38, in response to CPT-11-based chemotherapy is unknown. Recently, OATP1B3 expression within colon tumours has been shown to be a cancer-specific isoform that may be expressed primarily as an intracellular protein calling into question the functional relevance of this transporter within the tumour (Thakkar et al, 2013). Paired normal and tumour samples were available for a subset of patients treated with CPT-11-based chemotherapy regimens (n=30: BEV-FOLFIRI, n=22; FOLFIRI, n=5; FOLFIRINOX, n=1; BEV-irinotecan, n=1; cetuximab-irinotecan, n=1). Staining for OATP1B3 was performed and intensity of expression was scored. Tumour tissue had a significantly higher OATP1B3 score compared with paired normal tissue (P<0.05, Figure 2A and C). Progression-free survival was significantly reduced in patients with high (score, 2 or 3) OATP1B3 expression compared with patients with low (score, 0 or 1) expression (Figure 2B), suggesting that OATP1B3 expression may correlate with poorer clinical response to CPT-11 therapy.

Materials and Methods: Study population Metastatic colorectal cancer and advanced or metastatic pancreatic cancer patients (n=127) being treated with CPT-11-based chemotherapy regimens were prospectively recruited between January 2010 and November 2012 from the London Regional Cancer Program, London Health Sciences Centre, London, Ontario, Canada. The majority of patients were prescribed CPT-11 at 180 mg m−2 biweekly in combination with 5-fluorouracil and leucovorin (FOLFIRI regimen) with or without bevacuzimab. Patients were included if they were aged 18 years or above with an ECOG (Eastern Cooperative Oncology Group) performance status ⩽2. Exclusion criteria included: >35 μmol l−1 total bilirubin, >3X upper normal limit AST or ALT without liver metastases or >5X with liver metastases, known hypersensitivity to CPT-11, known history of Gilbert's syndrome, and concurrent use of ketoconazole. All participants provided written informed consent. The study was approved by the Research Ethics Board at Western University. Metastatic colorectal cancer and advanced or metastatic pancreatic cancer patients (n=127) being treated with CPT-11-based chemotherapy regimens were prospectively recruited between January 2010 and November 2012 from the London Regional Cancer Program, London Health Sciences Centre, London, Ontario, Canada. The majority of patients were prescribed CPT-11 at 180 mg m−2 biweekly in combination with 5-fluorouracil and leucovorin (FOLFIRI regimen) with or without bevacuzimab. Patients were included if they were aged 18 years or above with an ECOG (Eastern Cooperative Oncology Group) performance status ⩽2. Exclusion criteria included: >35 μmol l−1 total bilirubin, >3X upper normal limit AST or ALT without liver metastases or >5X with liver metastases, known hypersensitivity to CPT-11, known history of Gilbert's syndrome, and concurrent use of ketoconazole. All participants provided written informed consent. The study was approved by the Research Ethics Board at Western University. Chart review Paper and electronic record chart review for each consented patient was conducted by a single reviewer. Age recorded was the age at initiation of CPT-11-based chemotherapy. A cycle of CPT-11 was defined as a single administration of CPT-11 alone or in combination, irrespective of the chemotherapy regimen used. Treatment-related toxicities were recorded for the cycle where the blood sample was obtained in addition to any toxicities occurring throughout the duration of treatment with CPT-11. The blood sample cycle was considered to extend to the subsequent measurement of basic laboratory values and/or clinical assessment. Toxicities during this period were not considered to be associated with the particular treatment if they were documented more than 3 weeks after sample blood drawn. NCI-CTCAE version 3.0 (Bethesda, MD, USA) was used to grade toxicities. Toxicity grade was determined using subjective measures (described in clinical notes) when exact grading was not documented. Responders were defined as having a stable or reduced tumour size on the first CT scan following CPT-11-based chemotherapy initiation. The interval between CT scans was at the discretion of the treating oncologist. Progression was defined as the date the CT or MRI reported an increased tumour size. Progression-free survival was considered to be the time from initiation of CPT-11-based chemotherapy to the date of progression, death, last contact, or censor date (13 August 2013), whichever occurred first. Paper and electronic record chart review for each consented patient was conducted by a single reviewer. Age recorded was the age at initiation of CPT-11-based chemotherapy. A cycle of CPT-11 was defined as a single administration of CPT-11 alone or in combination, irrespective of the chemotherapy regimen used. Treatment-related toxicities were recorded for the cycle where the blood sample was obtained in addition to any toxicities occurring throughout the duration of treatment with CPT-11. The blood sample cycle was considered to extend to the subsequent measurement of basic laboratory values and/or clinical assessment. Toxicities during this period were not considered to be associated with the particular treatment if they were documented more than 3 weeks after sample blood drawn. NCI-CTCAE version 3.0 (Bethesda, MD, USA) was used to grade toxicities. Toxicity grade was determined using subjective measures (described in clinical notes) when exact grading was not documented. Responders were defined as having a stable or reduced tumour size on the first CT scan following CPT-11-based chemotherapy initiation. The interval between CT scans was at the discretion of the treating oncologist. Progression was defined as the date the CT or MRI reported an increased tumour size. Progression-free survival was considered to be the time from initiation of CPT-11-based chemotherapy to the date of progression, death, last contact, or censor date (13 August 2013), whichever occurred first. Genotyping DNA was extracted from whole blood using the Gentra Purgene Blood Kit (Qiagen, Toronto, Ontario, Canada). The following TaqMan allelic discrimination assays (Applied Biosystems, Carlsbad, CA, USA) were used for genotyping: ABCB1 (c.3435C>T, rs1045642), ABCG2 (c.421C>A, rs2231142; c.34G>A, rs2231137), ABCC2 (c.-24C>T, rs717620; c.1249G>A, rs2273697), ABCC5 (T>C, rs562), SLCO1B1*1b (c.388A>G, rs2306283), SLCO1B1*5 (c.521T>C, rs4149056), SLCO1B3 (g.699G>A, rs7311358), UGT1A1*28 (TA(6/7), rs8175347), CES1 (g.14506G>A, rs71647871; g.27467A>C, rs2244613), CYP3A4*22 (intron 6 C>T, rs35599367), and CYP3A5*3 (g.6986A>G, rs776746). Hardy–Weinburg equilibrium was assessed for all genotypes using the χ2 goodness-of-fit test. DNA was extracted from whole blood using the Gentra Purgene Blood Kit (Qiagen, Toronto, Ontario, Canada). The following TaqMan allelic discrimination assays (Applied Biosystems, Carlsbad, CA, USA) were used for genotyping: ABCB1 (c.3435C>T, rs1045642), ABCG2 (c.421C>A, rs2231142; c.34G>A, rs2231137), ABCC2 (c.-24C>T, rs717620; c.1249G>A, rs2273697), ABCC5 (T>C, rs562), SLCO1B1*1b (c.388A>G, rs2306283), SLCO1B1*5 (c.521T>C, rs4149056), SLCO1B3 (g.699G>A, rs7311358), UGT1A1*28 (TA(6/7), rs8175347), CES1 (g.14506G>A, rs71647871; g.27467A>C, rs2244613), CYP3A4*22 (intron 6 C>T, rs35599367), and CYP3A5*3 (g.6986A>G, rs776746). Hardy–Weinburg equilibrium was assessed for all genotypes using the χ2 goodness-of-fit test. Drug levels Blood samples were obtained by venipuncture of the opposite arm immediately following the end of the CPT-11 infusion (90 min) during any cycle of CPT-11-based chemotherapy. Plasma was collected and stored at −80 °C until analysis. Plasma concentrations of CPT-11 and metabolites SN-38, SN-38G, and APC (7-ethyl-10-[4-N-(5-aminopentanoic acid)-1-piperidino] carbonyloxycamptothecin) were measured by liquid chromatography-tandem mass spectrometry. Standards were obtained from Sigma-Aldrich (Oakville, Ontario, Canada; CPT-11, SN-38) and Toronto Research Chemicals (Toronto, Ontario, Canada; SN-38-glucuronide, APC). Plasma samples (100 μl) were precipitated upon addition of 3 volumes of acetonitrile and 0.1% formic acid (FA) spiked with 15 μl internal standard (camptothecin, 500 ng ml−1, Toronto Research Chemicals). Samples were vortexed, centrifuged, and diluted with H2O (0.1% FA). Analytes were injected (30 μl) into the liquid chromatograph (Agilent 1200) and separated on a reverse-phase column (Hypersil Gold, 50 × 5 mm, 5 μM particle size) over 6 min using gradient elution with H2O (0.1% FA) and acetonitrile (0.1% FA) (10–90%). Standard curves and quality controls (co-efficient of variation (%), high (13.5), med (2.6), low (3.7)) were prepared in drug-free plasma. The mass spectrometer (Thermo TSQ Vantage, Burlington, ON, Canada) with heated electrospray ionisation source was set in positive mode for detection of CPT-11, SN-38, SN-38G, APC, and camptothecin with transitions 587→124 m/z, 393→349.3 m/z, 569→393.3 m/z, 691→227 m/z, 349→305 m/z, respectively. Blood samples were obtained by venipuncture of the opposite arm immediately following the end of the CPT-11 infusion (90 min) during any cycle of CPT-11-based chemotherapy. Plasma was collected and stored at −80 °C until analysis. Plasma concentrations of CPT-11 and metabolites SN-38, SN-38G, and APC (7-ethyl-10-[4-N-(5-aminopentanoic acid)-1-piperidino] carbonyloxycamptothecin) were measured by liquid chromatography-tandem mass spectrometry. Standards were obtained from Sigma-Aldrich (Oakville, Ontario, Canada; CPT-11, SN-38) and Toronto Research Chemicals (Toronto, Ontario, Canada; SN-38-glucuronide, APC). Plasma samples (100 μl) were precipitated upon addition of 3 volumes of acetonitrile and 0.1% formic acid (FA) spiked with 15 μl internal standard (camptothecin, 500 ng ml−1, Toronto Research Chemicals). Samples were vortexed, centrifuged, and diluted with H2O (0.1% FA). Analytes were injected (30 μl) into the liquid chromatograph (Agilent 1200) and separated on a reverse-phase column (Hypersil Gold, 50 × 5 mm, 5 μM particle size) over 6 min using gradient elution with H2O (0.1% FA) and acetonitrile (0.1% FA) (10–90%). Standard curves and quality controls (co-efficient of variation (%), high (13.5), med (2.6), low (3.7)) were prepared in drug-free plasma. The mass spectrometer (Thermo TSQ Vantage, Burlington, ON, Canada) with heated electrospray ionisation source was set in positive mode for detection of CPT-11, SN-38, SN-38G, APC, and camptothecin with transitions 587→124 m/z, 393→349.3 m/z, 569→393.3 m/z, 691→227 m/z, 349→305 m/z, respectively. Immunohistochemistry Archived normal and tumour tissue biopsy slides were obtained from a subset of study participants (n=30) following approval by the Tissue and Archive Committee (Department of Pathology, London Health Sciences Centre). Antigen retrieval was performed with citrate buffer and slides were subsequently incubated with pre-immune serum or anti-OATP1B3 polyclonal antibody (1 : 200) followed by an avidin-biotin immunoperoxidase assay and developed using an AEC (3-Amino-9-ethylcarbazole) staining kit (Sigma-Aldrich) using a modified protocol (Lee et al, 2008). Nuclei were counterstained using Mayer's haematoxylin (Sigma-Aldrich). Scoring for OATP1B3 expression in normal, normal adjacent, and tumour tissue was performed independently by one pathologist (JP). Staining intensity for OATP1B3 was defined and evaluated using the following semi-quantitative previously published method: (0) no staining, (1) weakly positive, (2) moderately positive, and (3) strongly positive (Lee et al, 2008). Archived normal and tumour tissue biopsy slides were obtained from a subset of study participants (n=30) following approval by the Tissue and Archive Committee (Department of Pathology, London Health Sciences Centre). Antigen retrieval was performed with citrate buffer and slides were subsequently incubated with pre-immune serum or anti-OATP1B3 polyclonal antibody (1 : 200) followed by an avidin-biotin immunoperoxidase assay and developed using an AEC (3-Amino-9-ethylcarbazole) staining kit (Sigma-Aldrich) using a modified protocol (Lee et al, 2008). Nuclei were counterstained using Mayer's haematoxylin (Sigma-Aldrich). Scoring for OATP1B3 expression in normal, normal adjacent, and tumour tissue was performed independently by one pathologist (JP). Staining intensity for OATP1B3 was defined and evaluated using the following semi-quantitative previously published method: (0) no staining, (1) weakly positive, (2) moderately positive, and (3) strongly positive (Lee et al, 2008). Statistics The primary objective was to determine covariates associated with interindividual variability of CPT-11 and metabolite plasma concentration. All statistical analysis was performed in GraphPad Prism and the statistical software R. One-way analysis of variance with Bonferroni correction and Student's t-tests was used to compare drug levels between genotypic groups. Multiple linear regression analysis was performed to determine significant covariates on the interindividual variability of dose-normalised CPT-11 and metabolite plasma concentrations (natural log-transformed). Covariates considered included: age, sex, treatment regimen, ABCB1, ABCG2, ABCC2, ABCC5, SLCO1B1, SLCO1B3, UGT1A1, CES1, CYP3A4, and CYP3A5 genotype. Covariates were assessed individually and were considered for the final model at a significance level of P<0.2. Covariates meeting these criteria were entered into a multiple linear regression model adjusting for age, sex, and treatment regimen and remained in the final model if P<0.1. Secondary outcomes included assessing covariates associated with toxicity and PFS. Multinomial logistic regression analysis was used to determine association of genotype with toxicity events after adjustment for sex, age, and treatment regimen. Toxicity categories used in the regression analysis were: neutropenia (no event vs low (grade 1 or 2) vs high (grade 3 or 4)) and diarrhoea, nausea/vomiting and oral mucositis (no event vs low (grade 1) vs high (grade 2 or 3)). Univariate analysis was performed for each covariate and only significant genotypes were included in the final model with the exception of adjustment covariates as indicated above. Logistic regression analysis was also performed to determine covariates associated with neutropenia (low (grade 0, 1 or 2) vs high (grade 3 or 4)). Cox regression analysis was used to determine association of covariates with PFS in patients treated with BEV-FOLFIRI or FOLFIRI regimens (excluding patients with pancreatic cancer). Univariate analysis was performed and covariates with a cut value of P<0.2 were included in the final multivariate analysis with the exception of adjusting covariates, age at enrolment, sex and treatment regimen. Kruskal–Wallis with Dunn's multiple comparative test and Wilcoxon signed rank test was used to compare OATP1B3 pathology scores. Mann–Whitney U-test was used to compare OATP1B3 pathology scores and PFS in patients treated with CPT-11-based chemotherapy regimens scored for OATP1B3 tumour expression (0–1 vs 2–3). The primary objective was to determine covariates associated with interindividual variability of CPT-11 and metabolite plasma concentration. All statistical analysis was performed in GraphPad Prism and the statistical software R. One-way analysis of variance with Bonferroni correction and Student's t-tests was used to compare drug levels between genotypic groups. Multiple linear regression analysis was performed to determine significant covariates on the interindividual variability of dose-normalised CPT-11 and metabolite plasma concentrations (natural log-transformed). Covariates considered included: age, sex, treatment regimen, ABCB1, ABCG2, ABCC2, ABCC5, SLCO1B1, SLCO1B3, UGT1A1, CES1, CYP3A4, and CYP3A5 genotype. Covariates were assessed individually and were considered for the final model at a significance level of P<0.2. Covariates meeting these criteria were entered into a multiple linear regression model adjusting for age, sex, and treatment regimen and remained in the final model if P<0.1. Secondary outcomes included assessing covariates associated with toxicity and PFS. Multinomial logistic regression analysis was used to determine association of genotype with toxicity events after adjustment for sex, age, and treatment regimen. Toxicity categories used in the regression analysis were: neutropenia (no event vs low (grade 1 or 2) vs high (grade 3 or 4)) and diarrhoea, nausea/vomiting and oral mucositis (no event vs low (grade 1) vs high (grade 2 or 3)). Univariate analysis was performed for each covariate and only significant genotypes were included in the final model with the exception of adjustment covariates as indicated above. Logistic regression analysis was also performed to determine covariates associated with neutropenia (low (grade 0, 1 or 2) vs high (grade 3 or 4)). Cox regression analysis was used to determine association of covariates with PFS in patients treated with BEV-FOLFIRI or FOLFIRI regimens (excluding patients with pancreatic cancer). Univariate analysis was performed and covariates with a cut value of P<0.2 were included in the final multivariate analysis with the exception of adjusting covariates, age at enrolment, sex and treatment regimen. Kruskal–Wallis with Dunn's multiple comparative test and Wilcoxon signed rank test was used to compare OATP1B3 pathology scores. Mann–Whitney U-test was used to compare OATP1B3 pathology scores and PFS in patients treated with CPT-11-based chemotherapy regimens scored for OATP1B3 tumour expression (0–1 vs 2–3). Study population: Metastatic colorectal cancer and advanced or metastatic pancreatic cancer patients (n=127) being treated with CPT-11-based chemotherapy regimens were prospectively recruited between January 2010 and November 2012 from the London Regional Cancer Program, London Health Sciences Centre, London, Ontario, Canada. The majority of patients were prescribed CPT-11 at 180 mg m−2 biweekly in combination with 5-fluorouracil and leucovorin (FOLFIRI regimen) with or without bevacuzimab. Patients were included if they were aged 18 years or above with an ECOG (Eastern Cooperative Oncology Group) performance status ⩽2. Exclusion criteria included: >35 μmol l−1 total bilirubin, >3X upper normal limit AST or ALT without liver metastases or >5X with liver metastases, known hypersensitivity to CPT-11, known history of Gilbert's syndrome, and concurrent use of ketoconazole. All participants provided written informed consent. The study was approved by the Research Ethics Board at Western University. Chart review: Paper and electronic record chart review for each consented patient was conducted by a single reviewer. Age recorded was the age at initiation of CPT-11-based chemotherapy. A cycle of CPT-11 was defined as a single administration of CPT-11 alone or in combination, irrespective of the chemotherapy regimen used. Treatment-related toxicities were recorded for the cycle where the blood sample was obtained in addition to any toxicities occurring throughout the duration of treatment with CPT-11. The blood sample cycle was considered to extend to the subsequent measurement of basic laboratory values and/or clinical assessment. Toxicities during this period were not considered to be associated with the particular treatment if they were documented more than 3 weeks after sample blood drawn. NCI-CTCAE version 3.0 (Bethesda, MD, USA) was used to grade toxicities. Toxicity grade was determined using subjective measures (described in clinical notes) when exact grading was not documented. Responders were defined as having a stable or reduced tumour size on the first CT scan following CPT-11-based chemotherapy initiation. The interval between CT scans was at the discretion of the treating oncologist. Progression was defined as the date the CT or MRI reported an increased tumour size. Progression-free survival was considered to be the time from initiation of CPT-11-based chemotherapy to the date of progression, death, last contact, or censor date (13 August 2013), whichever occurred first. Genotyping: DNA was extracted from whole blood using the Gentra Purgene Blood Kit (Qiagen, Toronto, Ontario, Canada). The following TaqMan allelic discrimination assays (Applied Biosystems, Carlsbad, CA, USA) were used for genotyping: ABCB1 (c.3435C>T, rs1045642), ABCG2 (c.421C>A, rs2231142; c.34G>A, rs2231137), ABCC2 (c.-24C>T, rs717620; c.1249G>A, rs2273697), ABCC5 (T>C, rs562), SLCO1B1*1b (c.388A>G, rs2306283), SLCO1B1*5 (c.521T>C, rs4149056), SLCO1B3 (g.699G>A, rs7311358), UGT1A1*28 (TA(6/7), rs8175347), CES1 (g.14506G>A, rs71647871; g.27467A>C, rs2244613), CYP3A4*22 (intron 6 C>T, rs35599367), and CYP3A5*3 (g.6986A>G, rs776746). Hardy–Weinburg equilibrium was assessed for all genotypes using the χ2 goodness-of-fit test. Drug levels: Blood samples were obtained by venipuncture of the opposite arm immediately following the end of the CPT-11 infusion (90 min) during any cycle of CPT-11-based chemotherapy. Plasma was collected and stored at −80 °C until analysis. Plasma concentrations of CPT-11 and metabolites SN-38, SN-38G, and APC (7-ethyl-10-[4-N-(5-aminopentanoic acid)-1-piperidino] carbonyloxycamptothecin) were measured by liquid chromatography-tandem mass spectrometry. Standards were obtained from Sigma-Aldrich (Oakville, Ontario, Canada; CPT-11, SN-38) and Toronto Research Chemicals (Toronto, Ontario, Canada; SN-38-glucuronide, APC). Plasma samples (100 μl) were precipitated upon addition of 3 volumes of acetonitrile and 0.1% formic acid (FA) spiked with 15 μl internal standard (camptothecin, 500 ng ml−1, Toronto Research Chemicals). Samples were vortexed, centrifuged, and diluted with H2O (0.1% FA). Analytes were injected (30 μl) into the liquid chromatograph (Agilent 1200) and separated on a reverse-phase column (Hypersil Gold, 50 × 5 mm, 5 μM particle size) over 6 min using gradient elution with H2O (0.1% FA) and acetonitrile (0.1% FA) (10–90%). Standard curves and quality controls (co-efficient of variation (%), high (13.5), med (2.6), low (3.7)) were prepared in drug-free plasma. The mass spectrometer (Thermo TSQ Vantage, Burlington, ON, Canada) with heated electrospray ionisation source was set in positive mode for detection of CPT-11, SN-38, SN-38G, APC, and camptothecin with transitions 587→124 m/z, 393→349.3 m/z, 569→393.3 m/z, 691→227 m/z, 349→305 m/z, respectively. Immunohistochemistry: Archived normal and tumour tissue biopsy slides were obtained from a subset of study participants (n=30) following approval by the Tissue and Archive Committee (Department of Pathology, London Health Sciences Centre). Antigen retrieval was performed with citrate buffer and slides were subsequently incubated with pre-immune serum or anti-OATP1B3 polyclonal antibody (1 : 200) followed by an avidin-biotin immunoperoxidase assay and developed using an AEC (3-Amino-9-ethylcarbazole) staining kit (Sigma-Aldrich) using a modified protocol (Lee et al, 2008). Nuclei were counterstained using Mayer's haematoxylin (Sigma-Aldrich). Scoring for OATP1B3 expression in normal, normal adjacent, and tumour tissue was performed independently by one pathologist (JP). Staining intensity for OATP1B3 was defined and evaluated using the following semi-quantitative previously published method: (0) no staining, (1) weakly positive, (2) moderately positive, and (3) strongly positive (Lee et al, 2008). Statistics: The primary objective was to determine covariates associated with interindividual variability of CPT-11 and metabolite plasma concentration. All statistical analysis was performed in GraphPad Prism and the statistical software R. One-way analysis of variance with Bonferroni correction and Student's t-tests was used to compare drug levels between genotypic groups. Multiple linear regression analysis was performed to determine significant covariates on the interindividual variability of dose-normalised CPT-11 and metabolite plasma concentrations (natural log-transformed). Covariates considered included: age, sex, treatment regimen, ABCB1, ABCG2, ABCC2, ABCC5, SLCO1B1, SLCO1B3, UGT1A1, CES1, CYP3A4, and CYP3A5 genotype. Covariates were assessed individually and were considered for the final model at a significance level of P<0.2. Covariates meeting these criteria were entered into a multiple linear regression model adjusting for age, sex, and treatment regimen and remained in the final model if P<0.1. Secondary outcomes included assessing covariates associated with toxicity and PFS. Multinomial logistic regression analysis was used to determine association of genotype with toxicity events after adjustment for sex, age, and treatment regimen. Toxicity categories used in the regression analysis were: neutropenia (no event vs low (grade 1 or 2) vs high (grade 3 or 4)) and diarrhoea, nausea/vomiting and oral mucositis (no event vs low (grade 1) vs high (grade 2 or 3)). Univariate analysis was performed for each covariate and only significant genotypes were included in the final model with the exception of adjustment covariates as indicated above. Logistic regression analysis was also performed to determine covariates associated with neutropenia (low (grade 0, 1 or 2) vs high (grade 3 or 4)). Cox regression analysis was used to determine association of covariates with PFS in patients treated with BEV-FOLFIRI or FOLFIRI regimens (excluding patients with pancreatic cancer). Univariate analysis was performed and covariates with a cut value of P<0.2 were included in the final multivariate analysis with the exception of adjusting covariates, age at enrolment, sex and treatment regimen. Kruskal–Wallis with Dunn's multiple comparative test and Wilcoxon signed rank test was used to compare OATP1B3 pathology scores. Mann–Whitney U-test was used to compare OATP1B3 pathology scores and PFS in patients treated with CPT-11-based chemotherapy regimens scored for OATP1B3 tumour expression (0–1 vs 2–3). Results: Study population Patient and tumour characteristics (n=127) are described in Table 1. The treatment profile of patients is presented in Table 2. Approximately half (55%) of the study population had not received previous chemotherapy and the majority of patients (82%) were treated with bevacizumab (BEV)-FOLFIRI or FOLFIRI chemotherapy regimens as first-line therapy (Table 2). Patient and tumour characteristics (n=127) are described in Table 1. The treatment profile of patients is presented in Table 2. Approximately half (55%) of the study population had not received previous chemotherapy and the majority of patients (82%) were treated with bevacizumab (BEV)-FOLFIRI or FOLFIRI chemotherapy regimens as first-line therapy (Table 2). OATP1B1 and ABCC5 are important determinants of SN-38 and SN-38G levels The primary objective was to determine covariates associated with interindividual variability of plasma concentrations of CPT-11 and metabolites. Plasma concentrations of CPT-11, SN-38, SN-38G, and APC were measured from blood samples collected immediately following the end of CPT-11 infusion. Multiple linear regression was performed on dose-normalised drug levels (natural log transformed) adjusting for age at time of treatment initiation, sex and treatment regimen in the final models. ABCB1 (c.3435 C>T) was significantly associated with CPT-11 exposure as homozygous variant (T/T) carriers had lower levels compared with wild-type patients (P<0.05; Figure 1A and Supplementary Table 1). This model had one significant adjusting covariate (FOLFIRI treatment regimen, P <0.05). Several genotypes were significantly associated with variation in SN-38 levels as part of a model that explained approximately 27% of the variation in exposure (Table 3). SLCO1B1 521C allele carriers had significantly increased systemic exposure to SN-38 (P<0.001, Figure 1B). As expected, heterozygous (*1/*28) and homozygous variant (*28/*28) UGT1A1 carriers had increased SN-38 plasma levels (Figure 1C). Interestingly, a significant increase in SN-38 level was observed with an increasing number of combined SLCO1B1 521C and UGT1A1*28 variant alleles, suggesting an additive effect of polymorphisms in these two genes (Figure 1D). A corresponding decrease in the SN-38G/SN-38 ratio was observed with an increasing number of combined variant alleles (Figure 1E). Together, this suggests that patients heterozygous for both SLCO1B1 521C and UGT1A1*28 may have an equivalent risk of increased SN-38 exposure compared with patients homozygous for either polymorphism, respectively. Carriers of a rare SNP in CES1 (rs71647871, allele frequency, 0.024) had significantly decreased SN-38 levels (P<0.05, Table 3). Interestingly, SN-38G levels were significantly affected by ABCC5 genotype. Patients harbouring the ABCC5 rs562 C allele had reduced SN-38G plasma exposure compared with wild-type patients (P <0.001, Figure 1F, Supplementary Table 2). In addition, patients expressing CYP3A5 also had significantly reduced SN-38G plasma levels (P <0.05). FOLFIRI and FOLFIRINOX treatment regimens were significantly associated with SN-38G (P<0.001 and P <0.05, respectively). The model for the inactive metabolite APC explained approximately 24% of the variability in plasma exposure (Supplementary Table 3). SLCO1B3 g.699 and ABCC2 c.1249G>A were positively associated with APC levels (P<0.01). Decreased plasma levels were observed in patients carrying the ABCC5 rs562 C allele compared with wild-type patients (P<0.001). Significant adjusting covariates in this model included sex (P<0.05) and treatment regimen (P<0.01). The primary objective was to determine covariates associated with interindividual variability of plasma concentrations of CPT-11 and metabolites. Plasma concentrations of CPT-11, SN-38, SN-38G, and APC were measured from blood samples collected immediately following the end of CPT-11 infusion. Multiple linear regression was performed on dose-normalised drug levels (natural log transformed) adjusting for age at time of treatment initiation, sex and treatment regimen in the final models. ABCB1 (c.3435 C>T) was significantly associated with CPT-11 exposure as homozygous variant (T/T) carriers had lower levels compared with wild-type patients (P<0.05; Figure 1A and Supplementary Table 1). This model had one significant adjusting covariate (FOLFIRI treatment regimen, P <0.05). Several genotypes were significantly associated with variation in SN-38 levels as part of a model that explained approximately 27% of the variation in exposure (Table 3). SLCO1B1 521C allele carriers had significantly increased systemic exposure to SN-38 (P<0.001, Figure 1B). As expected, heterozygous (*1/*28) and homozygous variant (*28/*28) UGT1A1 carriers had increased SN-38 plasma levels (Figure 1C). Interestingly, a significant increase in SN-38 level was observed with an increasing number of combined SLCO1B1 521C and UGT1A1*28 variant alleles, suggesting an additive effect of polymorphisms in these two genes (Figure 1D). A corresponding decrease in the SN-38G/SN-38 ratio was observed with an increasing number of combined variant alleles (Figure 1E). Together, this suggests that patients heterozygous for both SLCO1B1 521C and UGT1A1*28 may have an equivalent risk of increased SN-38 exposure compared with patients homozygous for either polymorphism, respectively. Carriers of a rare SNP in CES1 (rs71647871, allele frequency, 0.024) had significantly decreased SN-38 levels (P<0.05, Table 3). Interestingly, SN-38G levels were significantly affected by ABCC5 genotype. Patients harbouring the ABCC5 rs562 C allele had reduced SN-38G plasma exposure compared with wild-type patients (P <0.001, Figure 1F, Supplementary Table 2). In addition, patients expressing CYP3A5 also had significantly reduced SN-38G plasma levels (P <0.05). FOLFIRI and FOLFIRINOX treatment regimens were significantly associated with SN-38G (P<0.001 and P <0.05, respectively). The model for the inactive metabolite APC explained approximately 24% of the variability in plasma exposure (Supplementary Table 3). SLCO1B3 g.699 and ABCC2 c.1249G>A were positively associated with APC levels (P<0.01). Decreased plasma levels were observed in patients carrying the ABCC5 rs562 C allele compared with wild-type patients (P<0.001). Significant adjusting covariates in this model included sex (P<0.05) and treatment regimen (P<0.01). Drug transporters predict CPT-11-related toxicities CPT-11-related adverse reactions and frequencies of grade severity are described in Supplementary Table 4. Multinomial logistic regression analysis was performed to determine association of genotypes to adverse reactions comparing no event vs low-grade vs high-grade toxicity (Table 4). ABCC2 -24C/T carriers had significantly lower risk of grade 3/4 neutropenia compared with wild-type patients (odds ratio (OR)=0.22, 95% CI=0.06–0.85). UGT1A1*28 carriers were at increased risk for severe neutropenia (grade 3/4) compared with wild-type patients following binary logistic regression analysing low- vs high-grade events (OR=3.67; 95% CI=1.19–11.33; Table 4). In addition to their increased SN-38 exposure, patients with two or more combined SLCO1B1 521C and UGT1A1*28 variant alleles were also at significantly increased risk of severe neutropenia (grade 3/4) compared with patients wild-type for both alleles (OR=4.154, 95% CI=1.06–16.36). Non-haematological toxicity was associated with drug transporters and metabolising enzymes (Table 4). Carriers of a common SNP in CES1 (rs2244613) had significantly lower risk of higher-grade diarrhoea (OR=0.29, 95% CI=0.09–0.89). Patients with ABCB1 3435 C/T and T/T genotypes were much more likely to experience higher-grade nausea/vomiting (OR=9.06, 95% CI=1.03–79.41 and OR=10.52, 95% CI=1.10–100.2, respectively). Higher-grade oral mucositis was observed in patients expressing CYP3A5 (OR=8.10, 95% CI=1.57–41.90), whereas those heterozygous for SLCO1B1 388A/G had a significantly lower risk (OR=0.19, 95% CI=0.05–0.72). No significant adjusting covariates were found in the binary logistic or multinomial logistic regression models. CPT-11-related adverse reactions and frequencies of grade severity are described in Supplementary Table 4. Multinomial logistic regression analysis was performed to determine association of genotypes to adverse reactions comparing no event vs low-grade vs high-grade toxicity (Table 4). ABCC2 -24C/T carriers had significantly lower risk of grade 3/4 neutropenia compared with wild-type patients (odds ratio (OR)=0.22, 95% CI=0.06–0.85). UGT1A1*28 carriers were at increased risk for severe neutropenia (grade 3/4) compared with wild-type patients following binary logistic regression analysing low- vs high-grade events (OR=3.67; 95% CI=1.19–11.33; Table 4). In addition to their increased SN-38 exposure, patients with two or more combined SLCO1B1 521C and UGT1A1*28 variant alleles were also at significantly increased risk of severe neutropenia (grade 3/4) compared with patients wild-type for both alleles (OR=4.154, 95% CI=1.06–16.36). Non-haematological toxicity was associated with drug transporters and metabolising enzymes (Table 4). Carriers of a common SNP in CES1 (rs2244613) had significantly lower risk of higher-grade diarrhoea (OR=0.29, 95% CI=0.09–0.89). Patients with ABCB1 3435 C/T and T/T genotypes were much more likely to experience higher-grade nausea/vomiting (OR=9.06, 95% CI=1.03–79.41 and OR=10.52, 95% CI=1.10–100.2, respectively). Higher-grade oral mucositis was observed in patients expressing CYP3A5 (OR=8.10, 95% CI=1.57–41.90), whereas those heterozygous for SLCO1B1 388A/G had a significantly lower risk (OR=0.19, 95% CI=0.05–0.72). No significant adjusting covariates were found in the binary logistic or multinomial logistic regression models. Biomarkers of PFS Approximately 73% of patients were considered responders having radiographic evidence of no change or tumour shrinkage during CPT-11-based therapy. At the time of analysis (censor date: 13 August 2013), disease progression was evident in the majority of patients with a median PFS time of 10.5 months (range, 0.2–43 months; Table 2). Cox-regression analysis for PFS was performed on patients (excluding pancreatic cancer patients) treated with BEV-FOLFIRI or FOLFIRI regimens only (n=103, Table 5). Patients homozygous for SLCO1B1 388G/G alleles had a significantly increased PFS compared with wild-type patients (HR=1.60, 95% CI=1.04–2.46). Patients carrying two ABCC2 c.-24T alleles had decreased PFS (HR=0.62, 95% CI=0.40–0.95). Reduced PFS was also observed for UGT1A1*28 (*1/*28, HR=0.69, 95% CI=0.52–0.92; *28/*28, HR=0.60, 95% CI=0.38–0.97). ABCC5 was associated with PFS in the univariate analysis (P=0.05), but did not remain significant following adjustment in the final model. No significant adjusting covariates were found in the multivariate analysis. Approximately 73% of patients were considered responders having radiographic evidence of no change or tumour shrinkage during CPT-11-based therapy. At the time of analysis (censor date: 13 August 2013), disease progression was evident in the majority of patients with a median PFS time of 10.5 months (range, 0.2–43 months; Table 2). Cox-regression analysis for PFS was performed on patients (excluding pancreatic cancer patients) treated with BEV-FOLFIRI or FOLFIRI regimens only (n=103, Table 5). Patients homozygous for SLCO1B1 388G/G alleles had a significantly increased PFS compared with wild-type patients (HR=1.60, 95% CI=1.04–2.46). Patients carrying two ABCC2 c.-24T alleles had decreased PFS (HR=0.62, 95% CI=0.40–0.95). Reduced PFS was also observed for UGT1A1*28 (*1/*28, HR=0.69, 95% CI=0.52–0.92; *28/*28, HR=0.60, 95% CI=0.38–0.97). ABCC5 was associated with PFS in the univariate analysis (P=0.05), but did not remain significant following adjustment in the final model. No significant adjusting covariates were found in the multivariate analysis. Tumour OATP1B3 expression suggests poorer clinical outcomes The role of OATP1B3 expression, known to transport SN-38, in response to CPT-11-based chemotherapy is unknown. Recently, OATP1B3 expression within colon tumours has been shown to be a cancer-specific isoform that may be expressed primarily as an intracellular protein calling into question the functional relevance of this transporter within the tumour (Thakkar et al, 2013). Paired normal and tumour samples were available for a subset of patients treated with CPT-11-based chemotherapy regimens (n=30: BEV-FOLFIRI, n=22; FOLFIRI, n=5; FOLFIRINOX, n=1; BEV-irinotecan, n=1; cetuximab-irinotecan, n=1). Staining for OATP1B3 was performed and intensity of expression was scored. Tumour tissue had a significantly higher OATP1B3 score compared with paired normal tissue (P<0.05, Figure 2A and C). Progression-free survival was significantly reduced in patients with high (score, 2 or 3) OATP1B3 expression compared with patients with low (score, 0 or 1) expression (Figure 2B), suggesting that OATP1B3 expression may correlate with poorer clinical response to CPT-11 therapy. The role of OATP1B3 expression, known to transport SN-38, in response to CPT-11-based chemotherapy is unknown. Recently, OATP1B3 expression within colon tumours has been shown to be a cancer-specific isoform that may be expressed primarily as an intracellular protein calling into question the functional relevance of this transporter within the tumour (Thakkar et al, 2013). Paired normal and tumour samples were available for a subset of patients treated with CPT-11-based chemotherapy regimens (n=30: BEV-FOLFIRI, n=22; FOLFIRI, n=5; FOLFIRINOX, n=1; BEV-irinotecan, n=1; cetuximab-irinotecan, n=1). Staining for OATP1B3 was performed and intensity of expression was scored. Tumour tissue had a significantly higher OATP1B3 score compared with paired normal tissue (P<0.05, Figure 2A and C). Progression-free survival was significantly reduced in patients with high (score, 2 or 3) OATP1B3 expression compared with patients with low (score, 0 or 1) expression (Figure 2B), suggesting that OATP1B3 expression may correlate with poorer clinical response to CPT-11 therapy. Study population: Patient and tumour characteristics (n=127) are described in Table 1. The treatment profile of patients is presented in Table 2. Approximately half (55%) of the study population had not received previous chemotherapy and the majority of patients (82%) were treated with bevacizumab (BEV)-FOLFIRI or FOLFIRI chemotherapy regimens as first-line therapy (Table 2). OATP1B1 and ABCC5 are important determinants of SN-38 and SN-38G levels: The primary objective was to determine covariates associated with interindividual variability of plasma concentrations of CPT-11 and metabolites. Plasma concentrations of CPT-11, SN-38, SN-38G, and APC were measured from blood samples collected immediately following the end of CPT-11 infusion. Multiple linear regression was performed on dose-normalised drug levels (natural log transformed) adjusting for age at time of treatment initiation, sex and treatment regimen in the final models. ABCB1 (c.3435 C>T) was significantly associated with CPT-11 exposure as homozygous variant (T/T) carriers had lower levels compared with wild-type patients (P<0.05; Figure 1A and Supplementary Table 1). This model had one significant adjusting covariate (FOLFIRI treatment regimen, P <0.05). Several genotypes were significantly associated with variation in SN-38 levels as part of a model that explained approximately 27% of the variation in exposure (Table 3). SLCO1B1 521C allele carriers had significantly increased systemic exposure to SN-38 (P<0.001, Figure 1B). As expected, heterozygous (*1/*28) and homozygous variant (*28/*28) UGT1A1 carriers had increased SN-38 plasma levels (Figure 1C). Interestingly, a significant increase in SN-38 level was observed with an increasing number of combined SLCO1B1 521C and UGT1A1*28 variant alleles, suggesting an additive effect of polymorphisms in these two genes (Figure 1D). A corresponding decrease in the SN-38G/SN-38 ratio was observed with an increasing number of combined variant alleles (Figure 1E). Together, this suggests that patients heterozygous for both SLCO1B1 521C and UGT1A1*28 may have an equivalent risk of increased SN-38 exposure compared with patients homozygous for either polymorphism, respectively. Carriers of a rare SNP in CES1 (rs71647871, allele frequency, 0.024) had significantly decreased SN-38 levels (P<0.05, Table 3). Interestingly, SN-38G levels were significantly affected by ABCC5 genotype. Patients harbouring the ABCC5 rs562 C allele had reduced SN-38G plasma exposure compared with wild-type patients (P <0.001, Figure 1F, Supplementary Table 2). In addition, patients expressing CYP3A5 also had significantly reduced SN-38G plasma levels (P <0.05). FOLFIRI and FOLFIRINOX treatment regimens were significantly associated with SN-38G (P<0.001 and P <0.05, respectively). The model for the inactive metabolite APC explained approximately 24% of the variability in plasma exposure (Supplementary Table 3). SLCO1B3 g.699 and ABCC2 c.1249G>A were positively associated with APC levels (P<0.01). Decreased plasma levels were observed in patients carrying the ABCC5 rs562 C allele compared with wild-type patients (P<0.001). Significant adjusting covariates in this model included sex (P<0.05) and treatment regimen (P<0.01). Drug transporters predict CPT-11-related toxicities: CPT-11-related adverse reactions and frequencies of grade severity are described in Supplementary Table 4. Multinomial logistic regression analysis was performed to determine association of genotypes to adverse reactions comparing no event vs low-grade vs high-grade toxicity (Table 4). ABCC2 -24C/T carriers had significantly lower risk of grade 3/4 neutropenia compared with wild-type patients (odds ratio (OR)=0.22, 95% CI=0.06–0.85). UGT1A1*28 carriers were at increased risk for severe neutropenia (grade 3/4) compared with wild-type patients following binary logistic regression analysing low- vs high-grade events (OR=3.67; 95% CI=1.19–11.33; Table 4). In addition to their increased SN-38 exposure, patients with two or more combined SLCO1B1 521C and UGT1A1*28 variant alleles were also at significantly increased risk of severe neutropenia (grade 3/4) compared with patients wild-type for both alleles (OR=4.154, 95% CI=1.06–16.36). Non-haematological toxicity was associated with drug transporters and metabolising enzymes (Table 4). Carriers of a common SNP in CES1 (rs2244613) had significantly lower risk of higher-grade diarrhoea (OR=0.29, 95% CI=0.09–0.89). Patients with ABCB1 3435 C/T and T/T genotypes were much more likely to experience higher-grade nausea/vomiting (OR=9.06, 95% CI=1.03–79.41 and OR=10.52, 95% CI=1.10–100.2, respectively). Higher-grade oral mucositis was observed in patients expressing CYP3A5 (OR=8.10, 95% CI=1.57–41.90), whereas those heterozygous for SLCO1B1 388A/G had a significantly lower risk (OR=0.19, 95% CI=0.05–0.72). No significant adjusting covariates were found in the binary logistic or multinomial logistic regression models. Biomarkers of PFS: Approximately 73% of patients were considered responders having radiographic evidence of no change or tumour shrinkage during CPT-11-based therapy. At the time of analysis (censor date: 13 August 2013), disease progression was evident in the majority of patients with a median PFS time of 10.5 months (range, 0.2–43 months; Table 2). Cox-regression analysis for PFS was performed on patients (excluding pancreatic cancer patients) treated with BEV-FOLFIRI or FOLFIRI regimens only (n=103, Table 5). Patients homozygous for SLCO1B1 388G/G alleles had a significantly increased PFS compared with wild-type patients (HR=1.60, 95% CI=1.04–2.46). Patients carrying two ABCC2 c.-24T alleles had decreased PFS (HR=0.62, 95% CI=0.40–0.95). Reduced PFS was also observed for UGT1A1*28 (*1/*28, HR=0.69, 95% CI=0.52–0.92; *28/*28, HR=0.60, 95% CI=0.38–0.97). ABCC5 was associated with PFS in the univariate analysis (P=0.05), but did not remain significant following adjustment in the final model. No significant adjusting covariates were found in the multivariate analysis. Tumour OATP1B3 expression suggests poorer clinical outcomes: The role of OATP1B3 expression, known to transport SN-38, in response to CPT-11-based chemotherapy is unknown. Recently, OATP1B3 expression within colon tumours has been shown to be a cancer-specific isoform that may be expressed primarily as an intracellular protein calling into question the functional relevance of this transporter within the tumour (Thakkar et al, 2013). Paired normal and tumour samples were available for a subset of patients treated with CPT-11-based chemotherapy regimens (n=30: BEV-FOLFIRI, n=22; FOLFIRI, n=5; FOLFIRINOX, n=1; BEV-irinotecan, n=1; cetuximab-irinotecan, n=1). Staining for OATP1B3 was performed and intensity of expression was scored. Tumour tissue had a significantly higher OATP1B3 score compared with paired normal tissue (P<0.05, Figure 2A and C). Progression-free survival was significantly reduced in patients with high (score, 2 or 3) OATP1B3 expression compared with patients with low (score, 0 or 1) expression (Figure 2B), suggesting that OATP1B3 expression may correlate with poorer clinical response to CPT-11 therapy. Discussion: Pharmacogenetics of irinotecan therapy has been widely studied and UGT1A1*28 in particular, shown to be of potential clinical relevance especially for patients prescribed high-dose CPT-11 therapy (Hoskins et al, 2007). As severe toxicities are often observed in patients treated with combination regimens at lower CPT-11 doses, we examined pharmacogenetic determinants of CPT-11 and metabolite exposure in this setting. SN-38 plasma exposure was significantly increased SLCO1B1 521C (*5) allele carriers (P<0.001) (Figure 3) consistent with studies correlating SLCO1B1 521C to higher SN-38 and CPT-11 AUC in both colorectal cancer and non-small-lung cell cancer patients (Han et al, 2008; Innocenti et al, 2009). UGT1A1*28 carriers had significantly increased SN-38 levels compared with non-carriers (P<0.01) consistent with previous reports, which can result in greater toxicity risk (Iyer et al, 2002; Innocenti et al, 2009; Hu et al, 2010; Sai et al, 2010; Cai et al, 2013). Importantly, we demonstrate an additive effect of SLCO1B1 521C and UGT1A1*28 on SN-38 exposure as patients with an increasing number of combined variant alleles had higher plasma levels and a corresponding decrease in the SN-38G/SN-38 ratio. To our knowledge this is the first report of an additive effect in a primarily Caucasian (94%) population, which remained significant upon exclusion of non-Caucasian patients. A similar additive effect has been noted for SN-38 AUC with combined SLCO1B1*15 and UGT1A1*6 or *28 polymorphisms in Japanese patients, likely due to the presence of the SLCO1B1 521C allele (Sai et al, 2010). Together, these results suggest that heterozygous carriers of both SLCO1B1 521C and UGT1A1*28 may reach a toxicity risk comparable to homozygous UGT1A1*28 patients as evidenced by several case reports of SLCO521C and UGT1A1*28 carriers presenting with life-threatening toxicities (Sakaguchi et al, 2009; Takane et al, 2009). Here we observed an OR of 4.15 (95%CI=1.06–16.36) for grade 3/4 neutropenia in patients with two or more combined variant allele. The clinical relevance of the additive effect of these two genes may be underestimated and need to be further examined as a better strategy for personalising CPT-11 therapy. Although plasma SN-38 levels appear to be the most predictive of toxicity risk, secondary metabolites, SN-38G and APC, may also be important contributors. We observed decreased plasma levels of these metabolites associated with an ABCC5 polymorphism (rs562) that was recently identified as a significant predictor of GI toxicity (Di Martino et al, 2011). These effects may be due to reduced ABCC5 hepatic efflux leading to SN-38G and APC accumulation within the liver (Figure 3). Higher hepatic concentrations may ultimately lead to increased intestinal SN-38G levels, through increased biliary excretion via other ABC transporters, which may then undergo β-glucuronidase-dependent reconversion to SN-38 thereby augmenting GI toxicity risk (Figure 3). In this study, CPT-11 levels were lower in homozygous ABCB1 3435T patients. ABCB1 is a well-established transporter of CPT-11 and SN-38, but its clinical relevance for CPT-11 therapy remains inconclusive. ABCB1 polymorphisms have been associated with both increased or reduced exposure, decreased clearance and increased toxicity and response, whereas other studies were unable to confirm these results (Mathijssen et al, 2003; Sai et al, 2003; Mathijssen et al, 2004; Cote et al, 2007; Innocenti et al, 2009; Sai et al, 2010; Glimelius et al, 2011). This may be due to the redundancy of other transporters, including ABCC2 and ABCG2, capable of biliary excretion of CPT-11 and metabolites. The lack of consistent evidence suggests that ABCB1 may not be useful in personalising CPT-11 therapy. As a secondary objective we examined the impact of pharmacogenetic factors on adverse events and PFS. Here, the majority (97%) of patients were prescribed 180 mg m−2 CPT-11 with 17% experiencing severe neutropenia. Neutropenia was associated with UGT1A1*28 as has been reported (Hoskins et al, 2007). More importantly, patients carrying two or more combined SLCO1B1 521C and UGT1A1*28 alleles were at significantly increased risk of myelosuppression, suggesting these SNPs together may provide a more comprehensive strategy for assessing haematological toxicity risk. Heterozygous ABCC2–24C/T but not TT carriers predicted reduced neutropenia risk indicating further validation is needed. Non-haematological toxicities including diarrhoea, nausea/vomiting, and oral mucositis were assessed. Most notably, a common SNP in CES1 (rs2244613) was associated with lower risk of diarrhoea, which may be due to reduced conversion of CPT-11 to SN-38. CES1 (rs2244613) was recently associated with reduced trough levels of dabigatran, a new oral anticoagulant drug, and a significantly decreased bleeding risk in patients, suggesting this SNP may be clinically relevant for many CES-dependent drugs (Pare et al, 2013). Interestingly, homozygous ABCB1 3435T carriers had a much higher likelihood of experiencing higher-grade nausea/vomiting (OR=10.52, P<0.05) and patients expressing CYP3A5 had a significantly increased risk of oral mucositis, suggesting higher levels of M4, APC, or NPC metabolites may contribute its development. A potential limitation to this analysis is the concurrent use of 5-fluorouracil with CPT-11 in most patients. Although the side-effect profile of the two drugs is similar, the majority of markers assessed are not specific to 5-fluorouracil, suggesting these correlations are likely due to modulation of irinotecan disposition but require confirmation in a study designed to assess toxicity as a primary objective. PFS analysis was limited to patients treated with FOLFIRI with or without BEV. SLCO1B1 388(G/G) was associated with longer PFS, suggesting variant carriers may experience better response to FOLFIRI-based regimens. This effect may be due to increased OATP1B1 expression as this variant has recently been correlated to increased expression in Caucasian liver samples (Nies et al, 2013). ABCC2–24TT was associated with reduced PFS, consistent with lower response rates and shorter PFS observed in Japanese patients, but contrary to the lack of association in Caucasian metastatic colorectal cancer patients treated with FOLFIRI regimes (Akiyama et al, 2012). UGT1A1*28 was also associated with reduced PFS, but a recent meta-analysis suggested that UGT1A1*28 status may not be a reliable predictor of PFS (Liu et al, 2013). Our evaluation of PFS may be confounded by patients still undergoing active therapy or surveillance at the date of censoring. Although the genes investigated are not thought to have a role in the other drugs in the FOLFIRI regimen, we cannot rule out the effect of these drugs on PFS. Importantly, we analysed OATP1B3 tumour expression in a subset of patients. To date, tumour biomarkers of CPT-11 response remain unknown. Our group was the first to note OATP1B3 overexpression in colon tumours and this expression was recently identified exclusively as a cancer-specific (cs)OATP1B3 splice variant (Lee et al, 2008; Han et al, 2013; Imai et al, 2013; Thakkar et al, 2013). The localisation (thought to be intracellular), function, and clinical relevance of csOATP1B3 are under investigation (Imai et al, 2013; Thakkar et al, 2013). Here we show the first evidence, to our knowledge, that higher OATP1B3 expression in colon tumours was significantly associated with reduced PFS (Figure 3). Lack of membrane expression of a functional OATP1B3 transporter may lead to reduced SN-38 tumour influx leading to a poorer clinical response or alternatively, overexpression of csOATP1B3 within the tumour may induce a p53-dependent survival mechanism, which has been demonstrated in WT-OATP1B3 colon cancer cell lines (Lee et al, 2008). Although we are unable to definitively confirm OATP1B3 to be the cancer-specific isoform due to use of an antibody recognising the common c-terminal region, the reported lack of wild-type OATP1B3 expression within colon tumours suggests this is the form expressed in these tumours (Thakkar et al, 2013). Owing to the complexity of CPT-11 disposition, it is questionable that one gene alone will be useful for personalising therapy and will likely require assessing the right combination of genes. Our data provide new insight regarding transporters, particularly members of the OATP1B subfamily, to the disposition and clinical effects of CPT-11 (summarised in Figure 3). The additive effect of SLCO1B1 521C and UGT1A1*28 on SN-38 exposure and neutropenia risk seen here in patients carrying two or more combined alleles (24.8% of this population) provides rationale for examining the utility of combined genotyping to better predict toxicity risk in CPT-11-based regimens. Future prospective studies should be designed to compare combined genotyping to UGT1A1*28 genotyping alone to advance the goal of personalising irinotecan.

Background: Treatment of advanced and metastatic colorectal cancer with irinotecan is hampered by severe toxicities. The active metabolite of irinotecan, SN-38, is a known substrate of drug-metabolising enzymes, including UGT1A1, as well as OATP and ABC drug transporters. Methods: Blood samples (n=127) and tumour tissue (n=30) were obtained from advanced cancer patients treated with irinotecan-based regimens for pharmacogenetic and drug level analysis and transporter expression. Clinical variables, toxicity, and outcomes data were collected. Results: SLCO1B1 521C was significantly associated with increased SN-38 exposure (P<0.001), which was additive with UGT1A1*28. ABCC5 (rs562) carriers had significantly reduced SN-38 glucuronide and APC metabolite levels. Reduced risk of neutropenia and diarrhoea was associated with ABCC2-24C/T (odds ratio (OR)=0.22, 0.06-0.85) and CES1 (rs2244613; OR=0.29, 0.09-0.89), respectively. Progression-free survival (PFS) was significantly longer in SLCO1B1 388G/G patients and reduced in ABCC2-24T/T and UGT1A1*28 carriers. Notably, higher OATP1B3 tumour expression was associated with reduced PFS. Conclusions: Clarifying the association of host genetic variation in OATP and ABC transporters to SN-38 exposure, toxicity and PFS provides rationale for personalising irinotecan-based chemotherapy. Our findings suggest that OATP polymorphisms and expression in tumour tissue may serve as important new biomarkers.

[CONTENT] OATP1B1 | OATP1B3 | Irinotecan | SN-38 | colorectal cancer [SUMMARY]

[CONTENT] OATP1B1 | OATP1B3 | Irinotecan | SN-38 | colorectal cancer [SUMMARY]

[CONTENT] Adult | Aged | Aged, 80 and over | Antineoplastic Agents, Phytogenic | Camptothecin | Colorectal Neoplasms | Disease-Free Survival | Female | Humans | Irinotecan | Liver-Specific Organic Anion Transporter 1 | Male | Middle Aged | Multidrug Resistance-Associated Protein 2 | Multidrug Resistance-Associated Proteins | Organic Anion Transporters | Organic Anion Transporters, Sodium-Independent | Polymorphism, Single Nucleotide | Solute Carrier Organic Anion Transporter Family Member 1B3 [SUMMARY]

[CONTENT] Adult | Aged | Aged, 80 and over | Antineoplastic Agents, Phytogenic | Camptothecin | Colorectal Neoplasms | Disease-Free Survival | Female | Humans | Irinotecan | Liver-Specific Organic Anion Transporter 1 | Male | Middle Aged | Multidrug Resistance-Associated Protein 2 | Multidrug Resistance-Associated Proteins | Organic Anion Transporters | Organic Anion Transporters, Sodium-Independent | Polymorphism, Single Nucleotide | Solute Carrier Organic Anion Transporter Family Member 1B3 [SUMMARY]

[CONTENT] chemotherapy regimens scored | chemotherapy majority patients | folfiri chemotherapy regimens | metastatic colorectal cancer | 11 based chemotherapy [SUMMARY]

[CONTENT] chemotherapy regimens scored | chemotherapy majority patients | folfiri chemotherapy regimens | metastatic colorectal cancer | 11 based chemotherapy [SUMMARY]

[CONTENT] patients | 11 | cpt 11 | cpt | sn | 38 | sn 38 | grade | analysis | 28 [SUMMARY]

[CONTENT] patients | 11 | cpt 11 | cpt | sn | 38 | sn 38 | grade | analysis | 28 [SUMMARY]

[CONTENT] patients | sn | 95 | table | ci | 95 ci | significantly | levels | 28 | compared [SUMMARY]

[CONTENT] patients | 11 | cpt | cpt 11 | sn | table | grade | oatp1b3 | 95 | analysis [SUMMARY]

[CONTENT] 521C | 28 ||| ABCC5 | rs562 | SN-38 | APC ||| ABCC2-24C/T | 0.06-0.85 | CES1 | OR=0.29 | 0.09-0.89 ||| 388G | ABCC2-24T | 28 ||| [SUMMARY]

[CONTENT] irinotecan ||| irinotecan | SN-38 | ABC ||| irinotecan ||| ||| ||| 521C | 28 ||| ABCC5 | rs562 | SN-38 | APC ||| ABCC2-24C/T | 0.06-0.85 | CES1 | OR=0.29 | 0.09-0.89 ||| 388G | ABCC2-24T | 28 ||| ||| ABC | irinotecan ||| [SUMMARY]

A model to explain the challenges of emergency medical technicians' decision making process in emergency situations: a grounded theory.

35067498

To manage life-threatening conditions and reduce morbidity and mortality, pre-hospital's on-scene decision making is an influential factor. Since pre-hospital's decision making is a challenging process, it is necessary to be identified this process. This study was conducted to explore the model of Iranian emergency medical technicians' decision making in emergency situations.

BACKGROUND

This study was applied through grounded theory method using direct field observations and semi-structured interviews. Purposeful sampling with 26 participants including 17 emergency medical technicians including dispatchers, physicians of medical directions, managers and 1 representative for court affairs was performed. Interviews were lasted from October 2018 to July 2019. Corbin and Strauss approach, 2015 (open, axial and selective coding) were used to analyze data.

METHODS

A paradigm model was developed to explain the relationships among the main categories. Decision making in the context of fear and concern was emerged as the core category. Unclear duties, insufficient authorities and competencies as well as lack of enough decision making's protocols and guidelines were categorized as casual conditions. Other important categories linked to the core category were interactions, feelings and "customer focus approach". Action-interaction strategies were taken by Emergency Medical technicians lead to some negative consequences that can threaten clinical outcome and patient safety.

RESULTS

Based on the finding of this study, Emergency Medical technicians' decision making in the context of fear and concern, as the core concept of this model, lead to decrease in quality of the pre-hospital services, stakeholders' dissatisfaction, hospital emergency units' overload, decrease in reputation of the Emergency Medical Technicians, threat to patient clinical outcome and patient safety. To prevent of these negative consequences, facilitation of the Emergency Medical Technicians' on-scene decision making is recommended.

CONCLUSIONS

9115813

Introduction

Emergency Medical Services (EMS) centers were established to provide on-time and rapid services to the patients and injured from the scene to the hospital.1 These centers can reduce significantly morbidity and mortality with providing pre-hospital cares in life-threatening conditions.2 To manage life-threatening conditions, save lives and reduce morbidity and mortality, rapid and accurate on-scene decision making is an important and influential factor.3 Decisions about the type and priority of emergency medical interventions require triage.4 As the conditions of some patients are complex and variable, the technicians' decision changes frequently, which can lead to technician error.5,6 Decision making errors threaten patient’s safety and Emergency Medical Technicians (EMTs)’ reputation.7-10 However, proper emergency medical decision-making leads to prevention of hospital emergency unit overcrowding, increased quality of health services and ultimately reduced morbidity and mortality.9,11 EMTs' on-scene decision making is influenced by incredible barriers and challenges.7,12,13 These factors influence the thinking, feeling and performance of the EMTs and so, increase their decision error, especially in some interventions including endotracheal intubation, mediation and no transportation.14,15 Facing these barriers and challenges lead to the EMTs’ concern and fear which of these fears have a profound impact on patient outcomes.16,17 As stress and anxiety often arise at the time of providing on-scene health services,18 identifying these factors are necessary. There is a paucity of studies in the field of EMTs’ on-scene decision making that have provided a good overview of the factors affecting dispatch services and transmission, but the decision-making process at the scene was not explored yet.19-21 While EMTs’ decision-making is a dynamic and on- going process starting from the scene until the patient is delivered to the hospital.20,22 This process will be influenced by multiple factors, so will have multiple outcomes. Understanding this process is crucial and can significantly address challenges and improve patient outcomes. Therefore, this study was conducted to design EMTs' on-scene decision making model in order to simplify the relationship between the components of this process.

Methods

Study setting In Iran, national EMS system provides free services for patients and injured people from scene to hospital. In the provinces and cities, EMS centers are under supervision of universities medical sciences and national EMS system. EMTs including diploma(basic), nurses with associated degrees(intermediate), bachelor’s and master’s degrees (advanced/ paramedic) provide services in dispatch centers or on scene. There is a dispatch center in each provincial center or cities which response to calls, provide counselling, dispatch EMTs and coordinate dispatched ambulances. Two EMTs in each ambulance were dispatched to the scene who should get advice from general physician as Medical Direction’s Physicians (MDPs) of Dispatch in the provinces’ EMS center. In the cities’ EMS centers. EMTs typically operate independently due to lack of MDPs and communicating challenges. General physicians did not attend the ambulance and most of them as a medical director provide online radio or telephone counselling to EMTs at proveniences’ EMS centers and some of them are as head of emergency centers. There is no general physician in the cities’ EMS center. Study design A grounded theory method based on Corbin and Strauss approach in 2015 was conducted for data gathering. This method is useful for achieving new area or exploring new perspectives of known field.23,24 Participants’ selection Maximum variety sampling method was performed, of which 26 participants including Emergency Medical Technicians, Medical Direction’s Physicians (MDPs), dispatchers, representative for court affairs and EMS’ managers were chosen according to purposeful sampling (Table 1). Having practical or theoretical experience, being verbal and participations’ willingness were inclusion criteria. Moreover, observational field note by means of triangulation was used for data collection as well as data validation. Saturation principle was used for concept saturation, of which data saturation was reached after 26 interviews and 10 filed observations. Data collection Data were collected through in-depth interviews and field observations. Interviews were done in Farsi at the participants’ workplaces. Initially three unstructured interviews were conducted. These first three interviews helped to identify interview guideline and important concepts that effect on-scene decision makings. Following that 26 semi-structured interviews were conducted using interview guideline. Interviews were started with general question about participants’ experiences of on-scene decision making process. Following that, to explore participants’ experiences probing questions were performed. Examples of questions were including “What are your challenges at the time of on-scene decision making?”, “What factors effect on your on-scene decision making in emergency situations?”, “What strategies do you use to deal with on-scene decision making’s challenges?” The interviews’ duration lasted between 45 and 75 minutes from October 2018 to July 2019. Each recorded interview was listened several times and transcribed verbatim by the principal investigator (PI). The PI started observation to take field notes after three interviews. Moreover, in order to saturate the concept by means of triangulation, ten observations also conducted from November 2018 to March 2019. In order to do that, PI took part in the EMS’ mission as the participant as observer. In this regard, all relations and interactions were taken as the field notes. Data analysis Data analysis was carried out simultaneously and immediately after data collection. The PI listened the audio files several times and compared transcribed interviews with recorded digital audios file. Based on Corbin and Strauss recommendations in 2015,23 continuously during data analysis. Regarding open coding, transcribed interviews and filed note observations were analyzed line by line. In this regard, MSK and DKZ discussed interviews’ process and codes extracted to provide the guide of future interviews and analysis. Both are experts in the field of health in emergencies and disasters. Translate and back-translate was also carried out by KB who is expert in the field of emergency medicine and English language. As a result of the researchers’ engagement and their consensus, the extracted codes were integrated in sub-categories and categories. Following that, via axial coding, the sub-categories and categories were compared and categorized based on their similarities and differences. Finally, based on selective coding, the link of categories and core categories were obtained and after that conceptual model was designed based on initial theoretical structure. Following that the paradigm model was presented to visualize the core category that is the central and main phenomenon of study.25 This model presents the decision making in the context of fear and concern. This core category is affected by Causal condition as a group of situations that influence decision making in the context of fear and concern. Context conditions as a component of the model, refer to a group of conditions that the phenomenon is raised and people respond to it via some strategies. Intervening conditions are facilitators or barriers action-interaction strategies which are purposeful acts that are taken by people to resolve a problem and lead to a number of consequences.25 Trustworthiness Four strategies including credibility, confirmability, dependability and transferability were applied to achieve trustworthiness.26 To achieve credibility, data collection triangulations including interviews and observational field, researchers’ eligibility in the field of health in emergencies and disasters as well as their prolonged engagement with data was employed. In addition, member checking, peer checking and finally expert checking were applied to improve credibility of the findings. To determine the data consistency, beside the PI other team members, as external auditors, checked generated codes, categories and sub-categories. Expert check was carried out by the opinions of research supervisor, DKZ and KB on the findings. To establish Confirmability, triangulation for data collection including face-to-face interview, bracketing principle and peer evaluation were used. Transferability was achieved using detailed explanations of the data collection, analysis and result interpretation. Finally, dependability was met thought triangulation, code-recode strategy, and peer evaluation.

Results

This study explored the EMTS 'on-scene decision making process. In the initial analysis, 1352 codes were extracted and finally classified into six categories and 21 sub-categories. “Decision making in the context of fear and concern” was defined as a core category. This type of decision means that most EMTs are concerned about the consequences of their decisions due to ambiguous tasks and responsibilities, insufficient authority, competency and decision-making protocols. These conditions are in the setting concluding lack of trust and inadequate supportive rules for EMTs. Intervening conditions including interactions, fillings, and “Customer focus approach” intervene in the decision making. The participant number were included (by P) for each quote and (FNO) for each field notes observation. Based on the findings of this study, in dealing with fear and concern, EMTs adopt two types of action-interaction strategies in order to manage their stress and anxiety. They may first consult with their co-worker, MDP, or dispatch operator. If this strategy does not solve the problem, they will refer to the existing protocols. Due to the findings of this study, there are many weaknesses and barriers to these strategies, so due to frustration, some EMTs may take the strategies including irresponsibility, only transportation approach and leave the scene and transfer the patients without medical intervention. All of these strategies lead to negative consequences include reduced quality of EMS services, reduced stakeholders' satisfaction, hospital emergency unit overload, reduced EMT reputation and threat to patient safety and their outcomes (Figure 1). Causal condition In this study, factors including unclear duties and ambiguous tasks, insufficient EMTs’ authorities, inadequate some EMTs’ competencies, and lack of appropriate on-scene decision-making protocols and guidelines lead to decision making in the context of fear and concern. According to EMTs, unclear duties and responsibilities are one of the most important factors that cause fear and concern at the time of decisions making. The EMTs are not completely aware of their duties and often their intervention is not legal. They described that multiple and sometimes indefinite responsibilities not only lead to neglect of necessary interventions, but also lead to excessive stress and anxiety. "…I don't know exactly what is my duty and responsibility. If occurred complication in a patient due to my intervention and its possible failure, would we be prosecuted and compensated? ... What duty do we have for patients who refuse to be cared? Whether a patient's endotracheal intubation or medication is illegal without a MDP’s advice…? ". (P3) Inadequate authority in contrast with delegated responsibility, inadequate competencies including knowledge, skills, experience and accurate judgment were identified as four important factors influencing the decision-making process of EMTs. The EMTs were concerned about the lack of sufficient authority. They believed that delegating multiple responsibilities without giving them sufficient authority would deprive many patients of emergency medical interventions and cares. According to participants and observation field notes, some EMTs were insufficiently competent to provide emergency interventions and cares. They declared that clinical judgment, experience, knowledge and skills of effective communication and scene management ability are critical to making the right decision in a timely manner. Field observations showed that there were cases of poor emergency scene management when assisting the patients that their family were agitated, which resulted in violence against EMTs. "Some co-workers have enough experience and great skill to manage the scene. Judging right in spite of the patient's anger and illogical laypeople’s interference is of important. ...sometimes a patient does not have need an intervention and special care, but unfortunately, we don't have enough authority to leave such a patient. Sometimes the patient doesn't have to be taken to the hospital, but because we don't have the enough authorization, we are afraid of not taking the patient to the hospital." (FNO2) According to the EMTs, sometimes making a decision to manage the patients is very difficult and EMT's consultation may be ineffective, so guidelines and protocols are very helpful. If decisions are made on the basis of scientific protocols and guidelines, in addition to appropriate medical interventions, concerns and fears of EMTs about the negative consequences of their interventions will be eliminated. In other words, EMTs can make better decisions with these supporting documents. The lack of protocols for some emergencies and deficiencies in existing protocols, was a major challenge of decision making. "... sometimes I talk to my colleague about the patient, but we really don't know what to do ... a guidance can be very helpful, but the number of our pre-hospital emergency guidelines and protocols are limited ... the available guidelines and protocols because of their simplicity or complexity are unusable." (p16) Contextual conditions In this paradigm model, community setting, trust, supportive and supervisory rules were identified as contextual conditions. The EMTs were extremely unsatisfied with the irrational and sometimes violent intervention of laypeople on the scene. Distress caused by irrational interferences disrupts emergency medical cares and results in rapid transmission without medical care. Another important aspect of community setting is the level of public education and awareness. The participants believed that community perception and its’ attitudes toward the importance of pre-hospital cares have an important role in their cooperation and collaboration. In the field observation, the involvement of laypeople with severe verbal violence were observed. "Unfortunately, sometimes people angrily disrespect us. Although we do our best to help a patient with life-threatening conditions, they prevent us by irrational interferences. I am sure that the main reason for these behaviors and interferences is the low understanding and awareness. “(P10) The findings of the study showed that trust had a significant impact on EMT's decision making. All participants believed that lack of trust is a major barrier to participatory decision making. They mentioned that there is no sufficient trust between EMTs, dispatchers, MDPs, and hospital emergency physicians. Some MDPs do not trust the technician's description and only recommend transferring instead of providing medication and medical advice. Lack of proper decision-making condition due to insufficient trust leads to fear and anxiety of EMTs. The EMTs believed that there is no a " Chain of trust " to facilitate the decision-making process, so people do not have trust to knowledge, skills and importance of EMTs’ medical approaches and cares. "I believe that team-working in the pre-hospital emergency services is not possible without trust. I think there is no sufficient trust among our staff. When a physician does not have enough trust to the technician's report, how to order medical intervention. (P9) ... People do not trust our EMTs' capability ... because of the weak trust chain, sometimes some patients and their companions are not honest... ". (P6) This study addressed that supervisory and supportive laws play an important role in emergency medical decision making. Lack of clear rules and regulations to determine the scopes of EMTs’ duties and lack of proper professional liability coverage by insurance companies were obtained as one of the most important factors leading to EMTs’ concern and fear. The EMTs stated that they were constantly afraid of the possible negative consequences of their decisions and interventions. Uncertainties at the time of dealing with an alone elderly or child patient who refuse to receive care, lack of clear rules related to end-stage patient and cardiopulmonary arrest who had clinical death were the most important stressors for decision making. ".... We don't know what the legal consequences of leaving the alone patient who refuse to receive care or transport ...” (P1) …there are many missions that the patient is dead and there is nothing to do, but we transfer them to hospitals due to lack of supportive laws and insurance." (P21) Intervening conditions Interactions, feelings, and the “Customer focus approach” are three important intervening factors in EMTs’ on-scene decision-making that lead to Action-Interaction strategies. According to participants, although some patients don’t not have emergency medical problems, they or families tend to exaggerate their problem to use free services. On the other hand, some people behave violently with the EMTs that these verbal and non-verbal violence were identified as the most important disrupting interactions factor and consequently making decisions with fear. "Sometimes some patients pretend to be ill that we find out thorough physical and psychological examination...” (P13) Sometimes we face violent and inappropriate behavior from laypeople and bystanders that lead to stress, fear and wrong decisions." (P4) The EMTs declared that their decisions and actions are also influenced by their feelings. The findings of this study indicated that feelings such as altruism and empathy play an important role in the decisions of EMTs. Seeing patients' suffering affects the feelings and focus of the EMTs. Another negative feeling reported by EMTs was that they felt unhelpful and sense of inadequacy. They were sometimes unable to do something, in spite of being aware of patient's needs. "... Pain and groaning of some patients, especially those who are end-stage and those who are really poor affect me. We put all our focus and efforts on patients. Nevertheless, when I feel like some managers don't care, I'll be disappointed... “(P16) Excessive attention to the satisfaction of patients and their families by EMS system managers and policymakers, was noted repeatedly by the EMTs as "Customer focus approach”. Most participants claimed that some EMS Managers and health policy makers only want to obtain the satisfaction of caregivers. The data extracted from observations and interviews showed that some managers neglected the EMTs’ satisfaction that had negative effect on EMTs’ motivation and so their decisions. "... There are many cases with no emergencies that request an ambulance several times a week. Some patient or families threaten us very badly ...” (P14) … in the face violence, the executives say, in any case, they are sick and your duty is providing services ...” (P20) Action-Interaction strategies Based on the participants and fieldnotes, action strategies including protocol and guideline-based decision making, consultancy, and interaction strategies including “why me culture?”, only transportation approach and leaving scene without emergency medical intervention were obtained. EMTs act based on action strategies to overcome decision-making challenges and barriers. If consulting with another EMT could not resolve decision-making problems, they will ask dispatcher to consult with a physician, but most EMS centers do not have a physician. Protocol and guideline-based decision-making is another action strategy that face challenges such as the difficulty of using paper-based protocols, forgetfulness of protocols, the complexity of some protocols. "Sometimes I can't make a decision. I consult with my colleague. Sometimes my colleagues ask me what to do. We sometimes ask dispatcher to consult with a physician, but most of the time it is not possible to consult with...” (P2) In the case of EMTs’ frustration to adopt an action strategy, they adopt a strategy of irresponsibility that was coded as "Why me? Culture”. Most participants declared that because of distress caused by bystanders’ irrational interference, they refused to medical emergency interventions and their responsibility. Fear of the decision’s negative consequences, feelings, people's interference and scene conditions are factors that lead to the only transportation approach so that EMTs sometimes leave the scene without care or intervention. "In my opinion, the uncertainty, the pressure of the scene, the fear of the negative consequences of the decision have led some of our colleagues to not take responsibility...”. (P11) Consequences According to the most participants, the EMTs’ frustration in on-scene decision-making process lead to decrease in quality of the pre-hospital services, stakeholders’ dissatisfaction, hospital emergency units’ overload, decrease in reputation of the EMTs and threat to patient clinical outcome and patient safety. The participants believed that strategies such as "why me? Culture", only transportation approach, and leaving the scene without providing emergency medical cares would lead to decrease in quality of the pre-hospital services. The EMTs’ concern and fear lead to make hasty or incorrect decisions and so, threat to patients’ safety. Rapid and correct on-scene decision-making lead to on time medical intervention and emergency care so it concluded to caregivers’ satisfaction. Furthermore, as a result of correct decisions, on time, enough and adequate pre-hospital cares prevent from transportation of all cases to the hospital and hospital emergency unit burden. "A lot of times we have to move patients as soon as we arrived to the scene…” (P10) …transporting all cases to the hospital, cause overcrowding and overloading of the hospital…” (P22) … inadequate pre-hospital cares lead to threat to patients’ safety and patients’ dissatisfaction”. (P10) Another negative consequence was the reduction of EMT reputation. In other words, some people do not trust the knowledge and competence of EMTs. The EMTs claimed that some people do not believe in them. Furthermore, attitudes and perceptions of hospital staff including nurses and physicians based on inadequacy of EMTs' competence were unbearable to EMTs. "People think we are just drivers. I think, the main reason of this attitude, is transporting the patients without enough emergency cares”. (P1)

Conclusion

This study designed a paradigm model to explain the process and to explore the relationships between different components and affecting factors on EMTs' on-scene decision making process. Technicians' decision making in the context of fear and concern leads to action-interactive strategies that will ultimately lead to stockholders’ dissatisfaction, decrease in EMTs’ reputation, threat to patient safety and related outcome. According to the model, unclear tasks, and insufficient EMTs’ authority, inadequate some EMTs’ competencies, and lack of sufficient protocols and guidelines were categorized as Casual Conditions. Promoting EMTs’ competencies, delegating sufficient authority, and passing supportive rules for emergency personnel are suggested. Based on this model, there are several contextual and intervening conditions that effect on EMTs’ decisions. Finally, to facilitate on-scene decision making process, it is necessary that some strategies were implicated in national EMS system. Abbreviations EMS: Emergency Medical Services; EMTs: Emergency Medical Technicians; MDPs: Medical Direction’s Physicians; DKZ: Davoud Khorasani-Zavareh; MSK: Meysam Safi-Keykaleh; ZG: Zohreh Ghomian; KB: Katarina Bohm. Acknowledgements This study is part of the Ph.D. thesis in School of Public Health and Safety. The authors would like to thank Shahid Beheshti University of Medical Sciences as well as all the participants in this study especially the EMTs.

Introduction: Emergency Medical Services (EMS) centers were established to provide on-time and rapid services to the patients and injured from the scene to the hospital.1 These centers can reduce significantly morbidity and mortality with providing pre-hospital cares in life-threatening conditions.2 To manage life-threatening conditions, save lives and reduce morbidity and mortality, rapid and accurate on-scene decision making is an important and influential factor.3 Decisions about the type and priority of emergency medical interventions require triage.4 As the conditions of some patients are complex and variable, the technicians' decision changes frequently, which can lead to technician error.5,6 Decision making errors threaten patient’s safety and Emergency Medical Technicians (EMTs)’ reputation.7-10 However, proper emergency medical decision-making leads to prevention of hospital emergency unit overcrowding, increased quality of health services and ultimately reduced morbidity and mortality.9,11 EMTs' on-scene decision making is influenced by incredible barriers and challenges.7,12,13 These factors influence the thinking, feeling and performance of the EMTs and so, increase their decision error, especially in some interventions including endotracheal intubation, mediation and no transportation.14,15 Facing these barriers and challenges lead to the EMTs’ concern and fear which of these fears have a profound impact on patient outcomes.16,17 As stress and anxiety often arise at the time of providing on-scene health services,18 identifying these factors are necessary. There is a paucity of studies in the field of EMTs’ on-scene decision making that have provided a good overview of the factors affecting dispatch services and transmission, but the decision-making process at the scene was not explored yet.19-21 While EMTs’ decision-making is a dynamic and on- going process starting from the scene until the patient is delivered to the hospital.20,22 This process will be influenced by multiple factors, so will have multiple outcomes. Understanding this process is crucial and can significantly address challenges and improve patient outcomes. Therefore, this study was conducted to design EMTs' on-scene decision making model in order to simplify the relationship between the components of this process. Methods : Study setting In Iran, national EMS system provides free services for patients and injured people from scene to hospital. In the provinces and cities, EMS centers are under supervision of universities medical sciences and national EMS system. EMTs including diploma(basic), nurses with associated degrees(intermediate), bachelor’s and master’s degrees (advanced/ paramedic) provide services in dispatch centers or on scene. There is a dispatch center in each provincial center or cities which response to calls, provide counselling, dispatch EMTs and coordinate dispatched ambulances. Two EMTs in each ambulance were dispatched to the scene who should get advice from general physician as Medical Direction’s Physicians (MDPs) of Dispatch in the provinces’ EMS center. In the cities’ EMS centers. EMTs typically operate independently due to lack of MDPs and communicating challenges. General physicians did not attend the ambulance and most of them as a medical director provide online radio or telephone counselling to EMTs at proveniences’ EMS centers and some of them are as head of emergency centers. There is no general physician in the cities’ EMS center. Study design A grounded theory method based on Corbin and Strauss approach in 2015 was conducted for data gathering. This method is useful for achieving new area or exploring new perspectives of known field.23,24 Participants’ selection Maximum variety sampling method was performed, of which 26 participants including Emergency Medical Technicians, Medical Direction’s Physicians (MDPs), dispatchers, representative for court affairs and EMS’ managers were chosen according to purposeful sampling (Table 1). Having practical or theoretical experience, being verbal and participations’ willingness were inclusion criteria. Moreover, observational field note by means of triangulation was used for data collection as well as data validation. Saturation principle was used for concept saturation, of which data saturation was reached after 26 interviews and 10 filed observations. Data collection Data were collected through in-depth interviews and field observations. Interviews were done in Farsi at the participants’ workplaces. Initially three unstructured interviews were conducted. These first three interviews helped to identify interview guideline and important concepts that effect on-scene decision makings. Following that 26 semi-structured interviews were conducted using interview guideline. Interviews were started with general question about participants’ experiences of on-scene decision making process. Following that, to explore participants’ experiences probing questions were performed. Examples of questions were including “What are your challenges at the time of on-scene decision making?”, “What factors effect on your on-scene decision making in emergency situations?”, “What strategies do you use to deal with on-scene decision making’s challenges?” The interviews’ duration lasted between 45 and 75 minutes from October 2018 to July 2019. Each recorded interview was listened several times and transcribed verbatim by the principal investigator (PI). The PI started observation to take field notes after three interviews. Moreover, in order to saturate the concept by means of triangulation, ten observations also conducted from November 2018 to March 2019. In order to do that, PI took part in the EMS’ mission as the participant as observer. In this regard, all relations and interactions were taken as the field notes. Data analysis Data analysis was carried out simultaneously and immediately after data collection. The PI listened the audio files several times and compared transcribed interviews with recorded digital audios file. Based on Corbin and Strauss recommendations in 2015,23 continuously during data analysis. Regarding open coding, transcribed interviews and filed note observations were analyzed line by line. In this regard, MSK and DKZ discussed interviews’ process and codes extracted to provide the guide of future interviews and analysis. Both are experts in the field of health in emergencies and disasters. Translate and back-translate was also carried out by KB who is expert in the field of emergency medicine and English language. As a result of the researchers’ engagement and their consensus, the extracted codes were integrated in sub-categories and categories. Following that, via axial coding, the sub-categories and categories were compared and categorized based on their similarities and differences. Finally, based on selective coding, the link of categories and core categories were obtained and after that conceptual model was designed based on initial theoretical structure. Following that the paradigm model was presented to visualize the core category that is the central and main phenomenon of study.25 This model presents the decision making in the context of fear and concern. This core category is affected by Causal condition as a group of situations that influence decision making in the context of fear and concern. Context conditions as a component of the model, refer to a group of conditions that the phenomenon is raised and people respond to it via some strategies. Intervening conditions are facilitators or barriers action-interaction strategies which are purposeful acts that are taken by people to resolve a problem and lead to a number of consequences.25 Trustworthiness Four strategies including credibility, confirmability, dependability and transferability were applied to achieve trustworthiness.26 To achieve credibility, data collection triangulations including interviews and observational field, researchers’ eligibility in the field of health in emergencies and disasters as well as their prolonged engagement with data was employed. In addition, member checking, peer checking and finally expert checking were applied to improve credibility of the findings. To determine the data consistency, beside the PI other team members, as external auditors, checked generated codes, categories and sub-categories. Expert check was carried out by the opinions of research supervisor, DKZ and KB on the findings. To establish Confirmability, triangulation for data collection including face-to-face interview, bracketing principle and peer evaluation were used. Transferability was achieved using detailed explanations of the data collection, analysis and result interpretation. Finally, dependability was met thought triangulation, code-recode strategy, and peer evaluation. Results: This study explored the EMTS 'on-scene decision making process. In the initial analysis, 1352 codes were extracted and finally classified into six categories and 21 sub-categories. “Decision making in the context of fear and concern” was defined as a core category. This type of decision means that most EMTs are concerned about the consequences of their decisions due to ambiguous tasks and responsibilities, insufficient authority, competency and decision-making protocols. These conditions are in the setting concluding lack of trust and inadequate supportive rules for EMTs. Intervening conditions including interactions, fillings, and “Customer focus approach” intervene in the decision making. The participant number were included (by P) for each quote and (FNO) for each field notes observation. Based on the findings of this study, in dealing with fear and concern, EMTs adopt two types of action-interaction strategies in order to manage their stress and anxiety. They may first consult with their co-worker, MDP, or dispatch operator. If this strategy does not solve the problem, they will refer to the existing protocols. Due to the findings of this study, there are many weaknesses and barriers to these strategies, so due to frustration, some EMTs may take the strategies including irresponsibility, only transportation approach and leave the scene and transfer the patients without medical intervention. All of these strategies lead to negative consequences include reduced quality of EMS services, reduced stakeholders' satisfaction, hospital emergency unit overload, reduced EMT reputation and threat to patient safety and their outcomes (Figure 1). Causal condition In this study, factors including unclear duties and ambiguous tasks, insufficient EMTs’ authorities, inadequate some EMTs’ competencies, and lack of appropriate on-scene decision-making protocols and guidelines lead to decision making in the context of fear and concern. According to EMTs, unclear duties and responsibilities are one of the most important factors that cause fear and concern at the time of decisions making. The EMTs are not completely aware of their duties and often their intervention is not legal. They described that multiple and sometimes indefinite responsibilities not only lead to neglect of necessary interventions, but also lead to excessive stress and anxiety. "…I don't know exactly what is my duty and responsibility. If occurred complication in a patient due to my intervention and its possible failure, would we be prosecuted and compensated? ... What duty do we have for patients who refuse to be cared? Whether a patient's endotracheal intubation or medication is illegal without a MDP’s advice…? ". (P3) Inadequate authority in contrast with delegated responsibility, inadequate competencies including knowledge, skills, experience and accurate judgment were identified as four important factors influencing the decision-making process of EMTs. The EMTs were concerned about the lack of sufficient authority. They believed that delegating multiple responsibilities without giving them sufficient authority would deprive many patients of emergency medical interventions and cares. According to participants and observation field notes, some EMTs were insufficiently competent to provide emergency interventions and cares. They declared that clinical judgment, experience, knowledge and skills of effective communication and scene management ability are critical to making the right decision in a timely manner. Field observations showed that there were cases of poor emergency scene management when assisting the patients that their family were agitated, which resulted in violence against EMTs. "Some co-workers have enough experience and great skill to manage the scene. Judging right in spite of the patient's anger and illogical laypeople’s interference is of important. ...sometimes a patient does not have need an intervention and special care, but unfortunately, we don't have enough authority to leave such a patient. Sometimes the patient doesn't have to be taken to the hospital, but because we don't have the enough authorization, we are afraid of not taking the patient to the hospital." (FNO2) According to the EMTs, sometimes making a decision to manage the patients is very difficult and EMT's consultation may be ineffective, so guidelines and protocols are very helpful. If decisions are made on the basis of scientific protocols and guidelines, in addition to appropriate medical interventions, concerns and fears of EMTs about the negative consequences of their interventions will be eliminated. In other words, EMTs can make better decisions with these supporting documents. The lack of protocols for some emergencies and deficiencies in existing protocols, was a major challenge of decision making. "... sometimes I talk to my colleague about the patient, but we really don't know what to do ... a guidance can be very helpful, but the number of our pre-hospital emergency guidelines and protocols are limited ... the available guidelines and protocols because of their simplicity or complexity are unusable." (p16) Contextual conditions In this paradigm model, community setting, trust, supportive and supervisory rules were identified as contextual conditions. The EMTs were extremely unsatisfied with the irrational and sometimes violent intervention of laypeople on the scene. Distress caused by irrational interferences disrupts emergency medical cares and results in rapid transmission without medical care. Another important aspect of community setting is the level of public education and awareness. The participants believed that community perception and its’ attitudes toward the importance of pre-hospital cares have an important role in their cooperation and collaboration. In the field observation, the involvement of laypeople with severe verbal violence were observed. "Unfortunately, sometimes people angrily disrespect us. Although we do our best to help a patient with life-threatening conditions, they prevent us by irrational interferences. I am sure that the main reason for these behaviors and interferences is the low understanding and awareness. “(P10) The findings of the study showed that trust had a significant impact on EMT's decision making. All participants believed that lack of trust is a major barrier to participatory decision making. They mentioned that there is no sufficient trust between EMTs, dispatchers, MDPs, and hospital emergency physicians. Some MDPs do not trust the technician's description and only recommend transferring instead of providing medication and medical advice. Lack of proper decision-making condition due to insufficient trust leads to fear and anxiety of EMTs. The EMTs believed that there is no a " Chain of trust " to facilitate the decision-making process, so people do not have trust to knowledge, skills and importance of EMTs’ medical approaches and cares. "I believe that team-working in the pre-hospital emergency services is not possible without trust. I think there is no sufficient trust among our staff. When a physician does not have enough trust to the technician's report, how to order medical intervention. (P9) ... People do not trust our EMTs' capability ... because of the weak trust chain, sometimes some patients and their companions are not honest... ". (P6) This study addressed that supervisory and supportive laws play an important role in emergency medical decision making. Lack of clear rules and regulations to determine the scopes of EMTs’ duties and lack of proper professional liability coverage by insurance companies were obtained as one of the most important factors leading to EMTs’ concern and fear. The EMTs stated that they were constantly afraid of the possible negative consequences of their decisions and interventions. Uncertainties at the time of dealing with an alone elderly or child patient who refuse to receive care, lack of clear rules related to end-stage patient and cardiopulmonary arrest who had clinical death were the most important stressors for decision making. ".... We don't know what the legal consequences of leaving the alone patient who refuse to receive care or transport ...” (P1) …there are many missions that the patient is dead and there is nothing to do, but we transfer them to hospitals due to lack of supportive laws and insurance." (P21) Intervening conditions Interactions, feelings, and the “Customer focus approach” are three important intervening factors in EMTs’ on-scene decision-making that lead to Action-Interaction strategies. According to participants, although some patients don’t not have emergency medical problems, they or families tend to exaggerate their problem to use free services. On the other hand, some people behave violently with the EMTs that these verbal and non-verbal violence were identified as the most important disrupting interactions factor and consequently making decisions with fear. "Sometimes some patients pretend to be ill that we find out thorough physical and psychological examination...” (P13) Sometimes we face violent and inappropriate behavior from laypeople and bystanders that lead to stress, fear and wrong decisions." (P4) The EMTs declared that their decisions and actions are also influenced by their feelings. The findings of this study indicated that feelings such as altruism and empathy play an important role in the decisions of EMTs. Seeing patients' suffering affects the feelings and focus of the EMTs. Another negative feeling reported by EMTs was that they felt unhelpful and sense of inadequacy. They were sometimes unable to do something, in spite of being aware of patient's needs. "... Pain and groaning of some patients, especially those who are end-stage and those who are really poor affect me. We put all our focus and efforts on patients. Nevertheless, when I feel like some managers don't care, I'll be disappointed... “(P16) Excessive attention to the satisfaction of patients and their families by EMS system managers and policymakers, was noted repeatedly by the EMTs as "Customer focus approach”. Most participants claimed that some EMS Managers and health policy makers only want to obtain the satisfaction of caregivers. The data extracted from observations and interviews showed that some managers neglected the EMTs’ satisfaction that had negative effect on EMTs’ motivation and so their decisions. "... There are many cases with no emergencies that request an ambulance several times a week. Some patient or families threaten us very badly ...” (P14) … in the face violence, the executives say, in any case, they are sick and your duty is providing services ...” (P20) Action-Interaction strategies Based on the participants and fieldnotes, action strategies including protocol and guideline-based decision making, consultancy, and interaction strategies including “why me culture?”, only transportation approach and leaving scene without emergency medical intervention were obtained. EMTs act based on action strategies to overcome decision-making challenges and barriers. If consulting with another EMT could not resolve decision-making problems, they will ask dispatcher to consult with a physician, but most EMS centers do not have a physician. Protocol and guideline-based decision-making is another action strategy that face challenges such as the difficulty of using paper-based protocols, forgetfulness of protocols, the complexity of some protocols. "Sometimes I can't make a decision. I consult with my colleague. Sometimes my colleagues ask me what to do. We sometimes ask dispatcher to consult with a physician, but most of the time it is not possible to consult with...” (P2) In the case of EMTs’ frustration to adopt an action strategy, they adopt a strategy of irresponsibility that was coded as "Why me? Culture”. Most participants declared that because of distress caused by bystanders’ irrational interference, they refused to medical emergency interventions and their responsibility. Fear of the decision’s negative consequences, feelings, people's interference and scene conditions are factors that lead to the only transportation approach so that EMTs sometimes leave the scene without care or intervention. "In my opinion, the uncertainty, the pressure of the scene, the fear of the negative consequences of the decision have led some of our colleagues to not take responsibility...”. (P11) Consequences According to the most participants, the EMTs’ frustration in on-scene decision-making process lead to decrease in quality of the pre-hospital services, stakeholders’ dissatisfaction, hospital emergency units’ overload, decrease in reputation of the EMTs and threat to patient clinical outcome and patient safety. The participants believed that strategies such as "why me? Culture", only transportation approach, and leaving the scene without providing emergency medical cares would lead to decrease in quality of the pre-hospital services. The EMTs’ concern and fear lead to make hasty or incorrect decisions and so, threat to patients’ safety. Rapid and correct on-scene decision-making lead to on time medical intervention and emergency care so it concluded to caregivers’ satisfaction. Furthermore, as a result of correct decisions, on time, enough and adequate pre-hospital cares prevent from transportation of all cases to the hospital and hospital emergency unit burden. "A lot of times we have to move patients as soon as we arrived to the scene…” (P10) …transporting all cases to the hospital, cause overcrowding and overloading of the hospital…” (P22) … inadequate pre-hospital cares lead to threat to patients’ safety and patients’ dissatisfaction”. (P10) Another negative consequence was the reduction of EMT reputation. In other words, some people do not trust the knowledge and competence of EMTs. The EMTs claimed that some people do not believe in them. Furthermore, attitudes and perceptions of hospital staff including nurses and physicians based on inadequacy of EMTs' competence were unbearable to EMTs. "People think we are just drivers. I think, the main reason of this attitude, is transporting the patients without enough emergency cares”. (P1) Discussion: The findings of this study indicated that some EMTs face fear and concern due to unclear duties and ambiguous tasks, insufficient authority, inadequate competencies, and deficiencies of medical protocols. These factors are in the setting of inadequate trust, inadequate supervision and lack of supportive laws. The findings suggested that intervening factors such as interactions, EMT feelings, and “Customer focus approach” increase EMTs' fear and concern at the time of decision making. Technicians may adopt strategies such as consultancy or protocol referencing to overcome this fear. If these strategies fail, they take the irresponsibility approach and leave the scene without medical intervention. Based on the extracted model, these strategies concluded to decrease in quality of pre-hospital services, reduced stockholders’' satisfaction, hospital emergency unit overload, decrease in EMT’s reputation and threat to patient’s safety and outcomes. One of the most important factors affecting the fear and concern of EMTs was insufficient authority. The EMTs believed that despite numerous and critical tasks, they did not have sufficient authority to provide medical emergency interventions. sufficient authority, especially in pre-hospital setting where patients are critical and in a life-threatening condition, is essential because it is not possible to delegate vital responsibility and provide lifesaving care without sufficient authority.27 Lack of sufficient and usable protocols and guidelines were categorized as the casual conditions. The EMTs have to make decision based on their judgment that is not evidence-based. So, the EMTs fear of legal consequences due to possible threat to patients' safety and dissatisfaction. The studies suggested that protocols and guidelines not only facilitate the physicians’ and nurses’ work but also provide better outcomes for patients and supportive space for health providers.28-30 A study that explored the experiences and perceptions of health providers in Iran, showed that decision-making protocols were complement to dispatchers’ experience.31 Therefore, Based on the experiences of the participants, some EMTs did not have sufficient competency including knowledge, skills, experience and the power of self-management, which were important causes for fear and concern. However, the findings of other studies that explored the competencies of pre-hospital providers, emphasized that having enough competencies is vital to provide on-time and correct clinical judgment.32,33 In order to improve EMTs competencies, it needs to continuous educational plan for both knowledge and particularly their skills, which is in line with previous study in pre-hospital phase improvement in Iran.34 As a contextual condition, community setting is one of the affecting factors on EMTs' on-scene decision making process. One of the aspects related to community setting is the bystanders and laypeople interferences in relief and rescue operations. The reason for people's interference was lack of their awareness of the EMTs’ duties and importance of pre-hospital cares. Other studies in the field of road traffic injuries in Iran also mentioned that bystanders’ interference has a negative role in pre-hospital emergency services21,35 because it led to over/under triage and EMT's confusion.36,37 The findings of this study showed that pre-hospital emergency services should be provided in a context of mutual trust. This chain of trust should exist between providers and caregivers as well as intra-organizational EMS staff. The studies claimed that trust is a vital component in the provision of health services and the promotion of stockholders’ collaboration.38,39 Based on this study findings, the cause of the malingering was getting rid of or obtaining specific conditions. Because malinger patients are resistant to EMTs’ intervention and cares and their families are believed in them, technicians face serious problems when dealing with these people.40,41 Empathy with such a patient and his family as well as respectful communication were recommended.42,43 Threats and violence against emergency personnel were identified as an aspect of negative and stressful interactions. The studies showed that fear, anxiety, sleep disorders, decrease in self-esteem are the negative consequences of exposure to violence.44-46 According to the findings of this study, most violent technicians were dissatisfied with poor support and follow-up of their organization. However, according to the studies, organizations should strive to satisfy their employees.47-49 Based on that, it is recommended for passing supportive law for EMTs in their duties and particularly on-scene protection of EMTs. The findings of this study revealed that EMTs have adopted various strategies when faced with decision-making barriers that one of the most important was use of protocols, but most technicians stated that the protocols are inadequate and complex. A study was conducted to explore the feedback system between dispatch centers and ambulances in Sweden, also described that some EMS personnel did not use decision-making and dispatching protocols because of their complexity.50 A medical protocol should take all the clinical aspects of an emergency situation in order to make the best decision and thus the best possible service.51 Consulting with Dispatch Operators and MDPs was another EMTs’ strategy to take guidance and sometimes get rid of the legal consequences of decisions thorough delegating responsibility to others. Responsibility is known as the cornerstone of provision of medical service.52,53 However, the findings of this study showed that technicians face the mistrust of some physicians and communication barriers. In line with our findings, similar studies mentioned legal Issues for the medical director.54,55 The analysis of the findings of this study indicated that some EMTs adopt an only transportation strategy with minimal and sometimes no care and medical intervention due to irrational interference of laypeople and bystanders. Although technicians provide essential services to patients in the ambulance during transportation, deprivation of critical care on the scene results in reduced quality of service, threat to patient safety. In line with the findings of this study, the studies highlighted that hospital emergency units’ overload, wasting time and facilities to transfer non-emergency patients is one of the negative consequences of this strategy.15,56-58 Study strengthens and limitations The qualitative approach was used to explore the model of EMTs’ on-scene decision making in a middle-income setting. To our best knowledge, this study is the first study on EMTs’ on-scene decision making in Iran that was employed both interview and observation. To have maximum variation, the participants were selected among EMTs, dispatchers, managers, MDPs and policymakers. To achieve consistency and the credibility of the data, different methods including constant comparative analysis, member check, and peer review were used. One potential limitation of the current study is that, this study conducted in the Iranian context that can be limited in generalizability to other countries. In qualitative study, the scope is not generalization, however, it seems that given the shared problems, the results can be applicable in low- and middle-income countries. Conclusion: This study designed a paradigm model to explain the process and to explore the relationships between different components and affecting factors on EMTs' on-scene decision making process. Technicians' decision making in the context of fear and concern leads to action-interactive strategies that will ultimately lead to stockholders’ dissatisfaction, decrease in EMTs’ reputation, threat to patient safety and related outcome. According to the model, unclear tasks, and insufficient EMTs’ authority, inadequate some EMTs’ competencies, and lack of sufficient protocols and guidelines were categorized as Casual Conditions. Promoting EMTs’ competencies, delegating sufficient authority, and passing supportive rules for emergency personnel are suggested. Based on this model, there are several contextual and intervening conditions that effect on EMTs’ decisions. Finally, to facilitate on-scene decision making process, it is necessary that some strategies were implicated in national EMS system. Abbreviations EMS: Emergency Medical Services; EMTs: Emergency Medical Technicians; MDPs: Medical Direction’s Physicians; DKZ: Davoud Khorasani-Zavareh; MSK: Meysam Safi-Keykaleh; ZG: Zohreh Ghomian; KB: Katarina Bohm. Acknowledgements This study is part of the Ph.D. thesis in School of Public Health and Safety. The authors would like to thank Shahid Beheshti University of Medical Sciences as well as all the participants in this study especially the EMTs.

Background: To manage life-threatening conditions and reduce morbidity and mortality, pre-hospital's on-scene decision making is an influential factor. Since pre-hospital's decision making is a challenging process, it is necessary to be identified this process. This study was conducted to explore the model of Iranian emergency medical technicians' decision making in emergency situations. Methods: This study was applied through grounded theory method using direct field observations and semi-structured interviews. Purposeful sampling with 26 participants including 17 emergency medical technicians including dispatchers, physicians of medical directions, managers and 1 representative for court affairs was performed. Interviews were lasted from October 2018 to July 2019. Corbin and Strauss approach, 2015 (open, axial and selective coding) were used to analyze data. Results: A paradigm model was developed to explain the relationships among the main categories. Decision making in the context of fear and concern was emerged as the core category. Unclear duties, insufficient authorities and competencies as well as lack of enough decision making's protocols and guidelines were categorized as casual conditions. Other important categories linked to the core category were interactions, feelings and "customer focus approach". Action-interaction strategies were taken by Emergency Medical technicians lead to some negative consequences that can threaten clinical outcome and patient safety. Conclusions: Based on the finding of this study, Emergency Medical technicians' decision making in the context of fear and concern, as the core concept of this model, lead to decrease in quality of the pre-hospital services, stakeholders' dissatisfaction, hospital emergency units' overload, decrease in reputation of the Emergency Medical Technicians, threat to patient clinical outcome and patient safety. To prevent of these negative consequences, facilitation of the Emergency Medical Technicians' on-scene decision making is recommended.

Introduction: Emergency Medical Services (EMS) centers were established to provide on-time and rapid services to the patients and injured from the scene to the hospital.1 These centers can reduce significantly morbidity and mortality with providing pre-hospital cares in life-threatening conditions.2 To manage life-threatening conditions, save lives and reduce morbidity and mortality, rapid and accurate on-scene decision making is an important and influential factor.3 Decisions about the type and priority of emergency medical interventions require triage.4 As the conditions of some patients are complex and variable, the technicians' decision changes frequently, which can lead to technician error.5,6 Decision making errors threaten patient’s safety and Emergency Medical Technicians (EMTs)’ reputation.7-10 However, proper emergency medical decision-making leads to prevention of hospital emergency unit overcrowding, increased quality of health services and ultimately reduced morbidity and mortality.9,11 EMTs' on-scene decision making is influenced by incredible barriers and challenges.7,12,13 These factors influence the thinking, feeling and performance of the EMTs and so, increase their decision error, especially in some interventions including endotracheal intubation, mediation and no transportation.14,15 Facing these barriers and challenges lead to the EMTs’ concern and fear which of these fears have a profound impact on patient outcomes.16,17 As stress and anxiety often arise at the time of providing on-scene health services,18 identifying these factors are necessary. There is a paucity of studies in the field of EMTs’ on-scene decision making that have provided a good overview of the factors affecting dispatch services and transmission, but the decision-making process at the scene was not explored yet.19-21 While EMTs’ decision-making is a dynamic and on- going process starting from the scene until the patient is delivered to the hospital.20,22 This process will be influenced by multiple factors, so will have multiple outcomes. Understanding this process is crucial and can significantly address challenges and improve patient outcomes. Therefore, this study was conducted to design EMTs' on-scene decision making model in order to simplify the relationship between the components of this process. Conclusion: This study designed a paradigm model to explain the process and to explore the relationships between different components and affecting factors on EMTs' on-scene decision making process. Technicians' decision making in the context of fear and concern leads to action-interactive strategies that will ultimately lead to stockholders’ dissatisfaction, decrease in EMTs’ reputation, threat to patient safety and related outcome. According to the model, unclear tasks, and insufficient EMTs’ authority, inadequate some EMTs’ competencies, and lack of sufficient protocols and guidelines were categorized as Casual Conditions. Promoting EMTs’ competencies, delegating sufficient authority, and passing supportive rules for emergency personnel are suggested. Based on this model, there are several contextual and intervening conditions that effect on EMTs’ decisions. Finally, to facilitate on-scene decision making process, it is necessary that some strategies were implicated in national EMS system. Abbreviations EMS: Emergency Medical Services; EMTs: Emergency Medical Technicians; MDPs: Medical Direction’s Physicians; DKZ: Davoud Khorasani-Zavareh; MSK: Meysam Safi-Keykaleh; ZG: Zohreh Ghomian; KB: Katarina Bohm. Acknowledgements This study is part of the Ph.D. thesis in School of Public Health and Safety. The authors would like to thank Shahid Beheshti University of Medical Sciences as well as all the participants in this study especially the EMTs.

Background: To manage life-threatening conditions and reduce morbidity and mortality, pre-hospital's on-scene decision making is an influential factor. Since pre-hospital's decision making is a challenging process, it is necessary to be identified this process. This study was conducted to explore the model of Iranian emergency medical technicians' decision making in emergency situations. Methods: This study was applied through grounded theory method using direct field observations and semi-structured interviews. Purposeful sampling with 26 participants including 17 emergency medical technicians including dispatchers, physicians of medical directions, managers and 1 representative for court affairs was performed. Interviews were lasted from October 2018 to July 2019. Corbin and Strauss approach, 2015 (open, axial and selective coding) were used to analyze data. Results: A paradigm model was developed to explain the relationships among the main categories. Decision making in the context of fear and concern was emerged as the core category. Unclear duties, insufficient authorities and competencies as well as lack of enough decision making's protocols and guidelines were categorized as casual conditions. Other important categories linked to the core category were interactions, feelings and "customer focus approach". Action-interaction strategies were taken by Emergency Medical technicians lead to some negative consequences that can threaten clinical outcome and patient safety. Conclusions: Based on the finding of this study, Emergency Medical technicians' decision making in the context of fear and concern, as the core concept of this model, lead to decrease in quality of the pre-hospital services, stakeholders' dissatisfaction, hospital emergency units' overload, decrease in reputation of the Emergency Medical Technicians, threat to patient clinical outcome and patient safety. To prevent of these negative consequences, facilitation of the Emergency Medical Technicians' on-scene decision making is recommended.

[CONTENT] Emergency medical services | Emergency medical technician | Decision making | Pre-hospital | Iran [SUMMARY]

[CONTENT] Emergency medical services | Emergency medical technician | Decision making | Pre-hospital | Iran [SUMMARY]

[CONTENT] Emergency medical services | Emergency medical technician | Decision making | Pre-hospital | Iran [SUMMARY]

[CONTENT] Emergency medical services | Emergency medical technician | Decision making | Pre-hospital | Iran [SUMMARY]

[CONTENT] Emergency medical services | Emergency medical technician | Decision making | Pre-hospital | Iran [SUMMARY]

[CONTENT] Emergency medical services | Emergency medical technician | Decision making | Pre-hospital | Iran [SUMMARY]

[CONTENT] Decision Making | Emergency Medical Services | Emergency Medical Technicians | Grounded Theory | Humans | Iran | Qualitative Research [SUMMARY]

[CONTENT] Decision Making | Emergency Medical Services | Emergency Medical Technicians | Grounded Theory | Humans | Iran | Qualitative Research [SUMMARY]

[CONTENT] Decision Making | Emergency Medical Services | Emergency Medical Technicians | Grounded Theory | Humans | Iran | Qualitative Research [SUMMARY]

[CONTENT] Decision Making | Emergency Medical Services | Emergency Medical Technicians | Grounded Theory | Humans | Iran | Qualitative Research [SUMMARY]

[CONTENT] Decision Making | Emergency Medical Services | Emergency Medical Technicians | Grounded Theory | Humans | Iran | Qualitative Research [SUMMARY]

[CONTENT] Decision Making | Emergency Medical Services | Emergency Medical Technicians | Grounded Theory | Humans | Iran | Qualitative Research [SUMMARY]

[CONTENT] effect emts decisions | emergency care concluded | services emts emergency | emt decision making | decision making emergency [SUMMARY]

[CONTENT] effect emts decisions | emergency care concluded | services emts emergency | emt decision making | decision making emergency [SUMMARY]

[CONTENT] effect emts decisions | emergency care concluded | services emts emergency | emt decision making | decision making emergency [SUMMARY]

[CONTENT] effect emts decisions | emergency care concluded | services emts emergency | emt decision making | decision making emergency [SUMMARY]

[CONTENT] effect emts decisions | emergency care concluded | services emts emergency | emt decision making | decision making emergency [SUMMARY]

[CONTENT] effect emts decisions | emergency care concluded | services emts emergency | emt decision making | decision making emergency [SUMMARY]

[CONTENT] emts | decision | making | decision making | scene | emergency | medical | hospital | study | patient [SUMMARY]

[CONTENT] emts | decision | making | decision making | scene | emergency | medical | hospital | study | patient [SUMMARY]

[CONTENT] emts | decision | making | decision making | scene | emergency | medical | hospital | study | patient [SUMMARY]

[CONTENT] emts | decision | making | decision making | scene | emergency | medical | hospital | study | patient [SUMMARY]

[CONTENT] emts | decision | making | decision making | scene | emergency | medical | hospital | study | patient [SUMMARY]

[CONTENT] emts | decision | making | decision making | scene | emergency | medical | hospital | study | patient [SUMMARY]

[CONTENT] decision | scene | making | decision making | emts | morbidity | mortality | morbidity mortality | process | emergency [SUMMARY]

[CONTENT] interviews | data | categories | data collection | collection | field | pi | ems | center | general [SUMMARY]

[CONTENT] emts | decision | trust | making | patient | decision making | hospital | patients | emergency | protocols [SUMMARY]

[CONTENT] emts | medical | study | model | decision | decision making | process | emergency | making | sufficient [SUMMARY]

[CONTENT] emts | decision | making | decision making | scene | emergency | medical | study | hospital | patient [SUMMARY]

[CONTENT] emts | decision | making | decision making | scene | emergency | medical | study | hospital | patient [SUMMARY]

[CONTENT] ||| ||| Iranian [SUMMARY]

[CONTENT] ||| 26 | 17 | 1 ||| Interviews | October 2018 to July 2019 ||| Strauss | 2015 [SUMMARY]

[CONTENT] ||| ||| ||| ||| Emergency Medical [SUMMARY]

[CONTENT] Emergency Medical | the Emergency Medical Technicians ||| the Emergency Medical Technicians' [SUMMARY]

[CONTENT] ||| ||| Iranian ||| ||| 26 | 17 | 1 ||| Interviews | October 2018 to July 2019 ||| Strauss | 2015 ||| ||| ||| ||| ||| ||| Emergency Medical ||| Emergency Medical | the Emergency Medical Technicians ||| the Emergency Medical Technicians' [SUMMARY]

[CONTENT] ||| ||| Iranian ||| ||| 26 | 17 | 1 ||| Interviews | October 2018 to July 2019 ||| Strauss | 2015 ||| ||| ||| ||| ||| ||| Emergency Medical ||| Emergency Medical | the Emergency Medical Technicians ||| the Emergency Medical Technicians' [SUMMARY]

Developing a Prediction Score for the Diagnosis of Malignant Pleural Effusion: MPE Score.

35092368

The objective of this study was to develop a diagnostic prediction model for diagnosis of malignant pleural effusion (MPE) from pleural fluid cytology (MPE score).

BACKGROUND

Retrospective analysis of pleural fluid cytology was conducted in patients with MPE between 2018 and 2020. Multivariable logistic regression was used to explore the potential predictors. The selected logistic coefficients were transformed into a diagnostic predictive scoring system. Internal validation was done using the bootstrapping procedure.

MATERIALS AND METHODS

The data of pleural fluid cytology from 155 MPE patients were analyzed. Seventy-eight positive pleural cytology patients were found (50.32%). Lung cancer was the cancer most commonly sent for pleural fluid testing, with 66.67% positive cytology.  The predictive indicators included pleural fluid protein > 4.64 g/dL, pleural fluid LDH > 555 IU/L, and pleural fluid sugar > 60 mg/dL. Lung mass from imaging and double tap for pleural cytology were used for the derivation of the diagnostic prediction model. The score-based model showed that the area under the receiver operating characteristic curve was 0.74 (95% CI 0.66-0.82). The developed MPE score ranged from zero to 17. The cut-off point was 15 with 88.31% of specificity, 37.18% of sensitivity, positive predictive value of 0.76, and negative predictive value of 0.58. The measurement of the calibration was illustrated using a calibration plot (p-value = 0.49 for the Hosmer-Lemeshow based goodness of fit). Internal validation with 1,000 bootstrap resampling showed a good discrimination.

RESULTS

The MPE score, as the diagnostic prediction model can be used in planning for more efficient diagnosis of MPE in patients with cancer under MPE.

CONCLUSIONS

9258676

Introduction

Malignant pleural effusion (MPE) refers to finding cytology pleural fluid caused by the metastasis of malignant mesothelioma, which is mostly due to lung cancer in men and breast cancer in women (Psallidas et al., 2016; Agrawal et al., 2015; Mongardon et al., 2011; Aydin et al., 2009). The MPE is also the cause of exudative pleural effusion from 42% up to 77% (Valdes et al., 1996). Diagnosis of MPE through cytology initially showed 60% of positive cytology depending on the type of cancer cells and cancer severity (Antonangelo et al., 2015; Loddenkemper and Boutin, 1993). Later, the diagnostic accuracy of MPE was improved by using pleuroscopy to enhance the efficiency of testing metastasis to the pleura (Ali et al., 2019; Ferreiro et al., 2017). However, this method is an invasive procedure. Therefore, less invasive ones are used such as metabolic imaging with 18-fluoro-deoxy glucose positron emission tomography (FDG-PET). The sensitivity was increased to 90% (Nakajima et al.,2015; Toaff et al.,2005); nonetheless, this method could not determine the types of cancer cells. In addition, epigenetic analysis of the pleural fluid was used to distinguish malignant DNA from methylation-specific PCR (MSP). This could help the diagnosis of MPE and efficiently specify the types of cancer cells (Herman et al., 1996; Brock et al., 2005; Zhang et al., 2007). Nevertheless, this method is expensive and is not used widely. Cytology is still a key method with 60% sensitivity depending on the type of cancer (Johnston, 1985; Starr and Sherman, 1991; and Hsu, 1987). Mostly, positive cytology pleural fluid is found in lung cancer and breast cancer. According to studies (Garcia et al., 1994; Desai and Lee, 2017), the repetition pleural fluid cytology can increase the diagnostic opportunities by 24%. However, more than double of the repetition is impractical for the diagnosis of MPE (Garcia et al., 1994). Thus, pleuroscopy is also required for confirmation to conduct a pleural biopsy, which is an invasive procedure. The clinical features and pleural fluid profile should be used to assist MPE diagnosis as a routine clinical practice and a diagnostic prediction score to facilitate decision-making on whether to wait for cytology results or perform an invasive procedure for efficient and rapid MPE diagnosis. This is because some hospitals still have limited diagnosing capabilities, or patients may have to wait for the cytologic results for weeks. Moreover, not all hospitals have the facilities to perform pleuroscopy, which make the test inaccessible for many patients. Therefore, this diagnostic research aimed to develop a diagnostic prediction model to help the decision-making in the diagnosis of MPE and plan appropriate and efficient diagnostic guidelines in the future.

Results

The data of the cytologic results of pleural fluid of 166 patients were collected. Eleven patients were excluded; including six patients with incomplete data of the biochemical tests and five patients did not have the pathological results to confirm the malignancy diagnosis. Therefore, the data were collected for 155 patients. Seventy-eight patients (50.32%) had positive cytology whereas seventy-seven patients (49.68%) had negative cytology. In terms of the pathological diagnosis and among different cancers, lung cancer was the most frequent cancer (61.9%) that needed the pleural fluid test. It was also the cancer with 66.67% of positive cytology (Table1). Based on the univariate analysis of the clinical characteristics and pleural fluid profile on MPE, it was found that lung mass detected by clinical imaging, lung cancer, breast cancer, and lung cancer with extrathoracic metastasis were the factors significantly affecting the predictive variables on MPE (Table2). Model development After analyzing the variable factors by univariate logistic regression analysis, the potential predictors affecting the diagnosis of MPE were selected for the multivariate logistic regression analysis of the scoring system derivation. The area under the receiver operating characteristic curve (AUC) for the final model was equal to 0.74 (95% CI 0.66-0.82). Score transformation Each potential predictor in the multivariable model was assigned with a specific score derived from the logistic regression coefficient (Table 3). The scoring scheme had a total score ranging from zero to 17. For the discriminative ability, the area under the parametric ROC curve for the score-based logistic regression model was equal to 0.74 (95% CI 0.66-0.82) (Figure 1). The measurement of the calibration is illustrated with a calibration plot, and the p-value via the Hosmer-Lemeshow goodness of fit test is equal to 0.49 (Figure 2). According to the sensitivity and specificity in each cut-off point, the point at 15 had 88.31% of specificity and 37.18% of sensitivity. This point displayed appropriate specificity that could be used as a diagnosis tool (Table 4). With the cut-off point of the MPE score at 15, to help the diagnosis of MPE, the odds ratio was equal to 3.18 (95% CI 1.35-8.11; p-value 0.004), positive predictive value (PPV) was equal to 0.76, negative predictive value (NPV) was equal to 0.58, positive likelihood ratio (LR+) was equal to 3.18, and negative likelihood ratio (LR-) was equal to 0.71. Internal validation Through conducting the internal validation using the predictive model with 1,000 resampling bootstrap method data set, the mean of the AUC of the apparent curve was obtained equal to 0.75, the test curve was equal to 0.72 (bootstrap estimator), and average estimates of the optimism curve was equal to 0.03 (Table 5). Types of Cancer Confirming the Pathological Diagnosis with the Cytologic Results of the Pleural Fluid Performance of the Clinical Risk Score, Area under the Receiver Operating Characteristics Curve (AUC), and 95% Confidence Band (Above). The calibration plots (pmcalplot) comparing the observed probabilities (y) and predicted probabilities (x) of the use of the MPE score to predict MPE (Below). Univariate Logistic Regression Analysis of MPE and Variable Factors protein ratio, Pleural fluid protein / serum protein; LDH ratio, Pleural fluid LDH / serum LDH; *, statistical significant Observed risk (circle) versus the score predicted risk (solid line) of the positive cytology pleural effusion (malignant pleural effusion). The size of the circle represents the frequency of MPE in each score (Left). Well-fitting model shows non-significance difference between the model and the observed data on the Hosmer-Lemeshow goodness of fit test (p-value 0.49) (Right) Risk Score Derivation Using Multivariate Logistic Regression Coefficients The Sensitivity, Specificity, Positive Likelihood Ratio (LR+), and Negative Likelihood Ratio (LR-) of Each Cut-Off Point Value of the MPE Score Internal Validation via 1,000 Resampling Bootstrap Method

Introduction: Malignant pleural effusion (MPE) refers to finding cytology pleural fluid caused by the metastasis of malignant mesothelioma, which is mostly due to lung cancer in men and breast cancer in women (Psallidas et al., 2016; Agrawal et al., 2015; Mongardon et al., 2011; Aydin et al., 2009). The MPE is also the cause of exudative pleural effusion from 42% up to 77% (Valdes et al., 1996). Diagnosis of MPE through cytology initially showed 60% of positive cytology depending on the type of cancer cells and cancer severity (Antonangelo et al., 2015; Loddenkemper and Boutin, 1993). Later, the diagnostic accuracy of MPE was improved by using pleuroscopy to enhance the efficiency of testing metastasis to the pleura (Ali et al., 2019; Ferreiro et al., 2017). However, this method is an invasive procedure. Therefore, less invasive ones are used such as metabolic imaging with 18-fluoro-deoxy glucose positron emission tomography (FDG-PET). The sensitivity was increased to 90% (Nakajima et al.,2015; Toaff et al.,2005); nonetheless, this method could not determine the types of cancer cells. In addition, epigenetic analysis of the pleural fluid was used to distinguish malignant DNA from methylation-specific PCR (MSP). This could help the diagnosis of MPE and efficiently specify the types of cancer cells (Herman et al., 1996; Brock et al., 2005; Zhang et al., 2007). Nevertheless, this method is expensive and is not used widely. Cytology is still a key method with 60% sensitivity depending on the type of cancer (Johnston, 1985; Starr and Sherman, 1991; and Hsu, 1987). Mostly, positive cytology pleural fluid is found in lung cancer and breast cancer. According to studies (Garcia et al., 1994; Desai and Lee, 2017), the repetition pleural fluid cytology can increase the diagnostic opportunities by 24%. However, more than double of the repetition is impractical for the diagnosis of MPE (Garcia et al., 1994). Thus, pleuroscopy is also required for confirmation to conduct a pleural biopsy, which is an invasive procedure. The clinical features and pleural fluid profile should be used to assist MPE diagnosis as a routine clinical practice and a diagnostic prediction score to facilitate decision-making on whether to wait for cytology results or perform an invasive procedure for efficient and rapid MPE diagnosis. This is because some hospitals still have limited diagnosing capabilities, or patients may have to wait for the cytologic results for weeks. Moreover, not all hospitals have the facilities to perform pleuroscopy, which make the test inaccessible for many patients. Therefore, this diagnostic research aimed to develop a diagnostic prediction model to help the decision-making in the diagnosis of MPE and plan appropriate and efficient diagnostic guidelines in the future. Materials and Methods: Clinical characteristics of patients with suspected MPE and cytology results of pleural fluid in Buddhasothorn Hospital, Chachoengsao, and Thailand were collected between 2018 and 2020.. The inclusion criteria were as follows: 1) Patients older than 18 years. 2) The results of the pleural fluid comprised of biochemical tests and key serum tests, i.e., lactate dehydrogenase (LDH) and protein. 3) Cancer data based on the radiology findings. The exclusion criteria consisted of: 1) No results of a pathological diagnosis in the case of negative cytology pleural fluid based on the cytologic results. Data analysis Step 1: The data were analyzed to find the potential factors in the diagnosis of positive cytology pleural fluid (MPE) using univariate regression analysis and multivariate logistic regression analysis. Step 2: Multiple imputation for missing data: Three predictor variables (pleural fluid white blood cell, pleural fluid lymphocyte, pleural fluid sugar) had more than 10% missing values, which could lead to biased estimates of the diagnostic model with the complete-case analysis. Multiple imputation with chained equation via mi impute chained command was used to generate missing values prior to model derivation. The logit model was chosen for the imputation of multivariable missing predictors. Step 3: The predictive variables from the multivariate logistic regression analysis were brought into the transformation of the risk score. A logistic regression coefficient was used to develop the MPE score to help diagnosing MPE. Step 4: The area under the receiver operating characteristic curve (AUC) based on the MPE score for the diagnosis of MPE was calculated and showed the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), likelihood ratio for a positive test (LR +) and likelihood ratio for a negative test (LR -). Step 5: The accuracy was tested through calibration curve and the Hosmer-Lemeshow goodness of fit test. Internal validation was tested by using the bootstrapping procedure (1,000 replicates). This study was approved by the Institutional Review Board of Buddhasothorn Hospital under the codes BSH-IRB 036/2563 Results: The data of the cytologic results of pleural fluid of 166 patients were collected. Eleven patients were excluded; including six patients with incomplete data of the biochemical tests and five patients did not have the pathological results to confirm the malignancy diagnosis. Therefore, the data were collected for 155 patients. Seventy-eight patients (50.32%) had positive cytology whereas seventy-seven patients (49.68%) had negative cytology. In terms of the pathological diagnosis and among different cancers, lung cancer was the most frequent cancer (61.9%) that needed the pleural fluid test. It was also the cancer with 66.67% of positive cytology (Table1). Based on the univariate analysis of the clinical characteristics and pleural fluid profile on MPE, it was found that lung mass detected by clinical imaging, lung cancer, breast cancer, and lung cancer with extrathoracic metastasis were the factors significantly affecting the predictive variables on MPE (Table2). Model development After analyzing the variable factors by univariate logistic regression analysis, the potential predictors affecting the diagnosis of MPE were selected for the multivariate logistic regression analysis of the scoring system derivation. The area under the receiver operating characteristic curve (AUC) for the final model was equal to 0.74 (95% CI 0.66-0.82). Score transformation Each potential predictor in the multivariable model was assigned with a specific score derived from the logistic regression coefficient (Table 3). The scoring scheme had a total score ranging from zero to 17. For the discriminative ability, the area under the parametric ROC curve for the score-based logistic regression model was equal to 0.74 (95% CI 0.66-0.82) (Figure 1). The measurement of the calibration is illustrated with a calibration plot, and the p-value via the Hosmer-Lemeshow goodness of fit test is equal to 0.49 (Figure 2). According to the sensitivity and specificity in each cut-off point, the point at 15 had 88.31% of specificity and 37.18% of sensitivity. This point displayed appropriate specificity that could be used as a diagnosis tool (Table 4). With the cut-off point of the MPE score at 15, to help the diagnosis of MPE, the odds ratio was equal to 3.18 (95% CI 1.35-8.11; p-value 0.004), positive predictive value (PPV) was equal to 0.76, negative predictive value (NPV) was equal to 0.58, positive likelihood ratio (LR+) was equal to 3.18, and negative likelihood ratio (LR-) was equal to 0.71. Internal validation Through conducting the internal validation using the predictive model with 1,000 resampling bootstrap method data set, the mean of the AUC of the apparent curve was obtained equal to 0.75, the test curve was equal to 0.72 (bootstrap estimator), and average estimates of the optimism curve was equal to 0.03 (Table 5). Types of Cancer Confirming the Pathological Diagnosis with the Cytologic Results of the Pleural Fluid Performance of the Clinical Risk Score, Area under the Receiver Operating Characteristics Curve (AUC), and 95% Confidence Band (Above). The calibration plots (pmcalplot) comparing the observed probabilities (y) and predicted probabilities (x) of the use of the MPE score to predict MPE (Below). Univariate Logistic Regression Analysis of MPE and Variable Factors protein ratio, Pleural fluid protein / serum protein; LDH ratio, Pleural fluid LDH / serum LDH; *, statistical significant Observed risk (circle) versus the score predicted risk (solid line) of the positive cytology pleural effusion (malignant pleural effusion). The size of the circle represents the frequency of MPE in each score (Left). Well-fitting model shows non-significance difference between the model and the observed data on the Hosmer-Lemeshow goodness of fit test (p-value 0.49) (Right) Risk Score Derivation Using Multivariate Logistic Regression Coefficients The Sensitivity, Specificity, Positive Likelihood Ratio (LR+), and Negative Likelihood Ratio (LR-) of Each Cut-Off Point Value of the MPE Score Internal Validation via 1,000 Resampling Bootstrap Method Discussion: Pleural fluid found in malignant disease was due to two major conditions, i.e., paramalignant pleural effusion (PMPE) and malignant pleural effusion (MPE) (Wong et al., 1963; Epelbaum and Rahman, 2019). The PMPE is not a consequence of a malignant disease spreading to the pleura. The probability that an effusion is paramalignant is higher when the effusion is transudative, while MPE is exudative. Therefore, understanding the differentiation between PMPE and MPE is necessary. There are studies on the use of the cancer ratio using the ratio of serum lactate dehydrogenase (LDH) to adenosine deaminase (ADA) in pleural fluid. The ratios used were based on the cut-off level > 20 to help diagnose the causes of exudative pleural fluid between benign and MPE. It was found that sensitivity and specificity were high because the relationship of the levels of serum LDH was usually high in a malignant disease (Verma et al., 2016; Korczyński et al., 2018; Verma et al., 2016). This was a result of using glycolysis for energy in tumor cells instead of oxidative phosphorylation, a switch in the adenosine triphosphate (ATP) generating pathways, which was mediated by LDH (Pfeiffer et al., 2001; Goldman et al., 1964; Mansouri et al., 2017). Likewise, the infection caused by tuberculosis in the pleural fluid usually had a higher ADA secreted by mononuclear cells, lymphocytes, neutrophils, and red blood cells (Liang et al., 2008; Jiménez Castro et al., 2003). However, the meta-analysis of using the cancer ratio for the diagnosis of MPE was based on the data from the PubMed and EMBASE databases The cancer ratio had a high diagnostic accuracy for predicting MPE. The pooled sensitivity and specificity of the cancer ratio were equal to 0.97 (95% CI 0.92-0.99) and 0.89 (0.69-0.97) respectively; with AUC equal to 0.98 (95% CI 0.97-0.99). Nevertheless, there were some limitations due to the bias of patient selection and potential partial verification (Han et al., 2019). Yet, it was frequently found that serum LDH may not be raised in the case of cancer. A high LDH may be related to poorer overall survival. Some minor studies (Chantharakit, 2018) also found that high LDH was related to cancer under liver metastasis; however, the data still contained a few limitations. Porcel et al., (2004) used a panel of tumor markers, i.e., carcinoembryonic antigen (CEA), cancer antigen (CA) 125, carbohydrate antigen (CA) 15-3, and cytokeratin 19 fragments in pleural fluid for the differential diagnosis of benign and malignant effusions. The combination of the four tumor markers reached a sensitivity of 54%, whereas the combined use of the cytology and the tumor marker panel increased the diagnostic yield of the former by 18% (95% CI; 13-23%). Yang et al., (2017) reported about a updated meta-analysis of patients with undiagnosed pleural effusion and showed that the combinations of positive pleural CEA + CA 15-3 and CEA + CA 19-9 were highly suspicious for pleural malignancy. Still, the sensitivity of these tests was poor. Clive et al., (2014) studied prognostic indicators and found that the ones affecting the survival of MPE patients were pleural fluid LDH, the Eastern Cooperative Oncology Group (ECOG) performance status, and neutrophil-to-lymphocyte ratio (NLR). It was also found that the tumor type could be developed by the LENT scoring system as a prognostic prediction model. The levels of pleural fluid LDH were key markers of inflammation or cellular injury. LDH levels greater than three times the upper limit of normal (often >1,000 U/L) are often indicative of pleural infection. This can also be associated with rheumatoid pleurisy, tuberculous pleurisy or malignancy. This study is different from previous studies as it focused only on PME. Pleural fluid cytology results may be positive malignant cells or negative malignant cells and it is not the intention of this study to distinguish MPE from benign disease. Therefore, all the pleural fluid was exudative pleural fluid according to Light’s criteria. The clinical information fitting the malignant disease was used to find the predictive indicators affecting the diagnosis of MPE using pleural fluid cytology to develop a diagnostic prediction model (MPE score). This was the first diagnostic prediction model used to assist in the diagnosis of MPE in patients with cancer. The data from the pleural fluid biomarkers were used along with the clinical data of the patients rather than using only the data from the biomarkers. However, the standard diagnosis of MPE features pleural fluid cytology supported by testing for confirmation by pleural biopsy in the case of negative pleural fluid cytology; still, with suspected MPE. This is an invasive procedure. Despite the effort to use a non-invasive technique, e.g., biomarker tests or molecular analysis from the pleural fluid for the diagnosis of MPE, the less invasive methods have not become popular yet. The validated clinical data found that there were still some limitations of use; thus, further studies are required. Therefore, using the MPE score at the cut-off point of 15, which has high specificity, may help in predicting MPE diagnosis to make decisions about planning for investigation while waiting for pleural fluid cytology results.This would enhance better efficiency of the diagnosis of MPE. Author Contribution Statement: Chaichana Chantharakhit: Designed the study, reviewed the paper, collected data, analyzed data, and edited the final version. Nantapa Sujaritvanichpong: Collected data. All authors read and approved the final version. Funding Statement: The authors confirm that there are no relevant financial or non-financial competing interests to report and no conflicts of interest to declare. Data Availability: The data used to support the findings of this study have been deposited in the repository [https://drive.google.com/drive/folders/1Enrqg5Zq_co3rY73-epiqymNe_njLL7W?usp=sharing]. We confirm that there are no relevant financial or non-financial competing interests to report and no conflicts of interest to declare

Background: The objective of this study was to develop a diagnostic prediction model for diagnosis of malignant pleural effusion (MPE) from pleural fluid cytology (MPE score). Methods: Retrospective analysis of pleural fluid cytology was conducted in patients with MPE between 2018 and 2020. Multivariable logistic regression was used to explore the potential predictors. The selected logistic coefficients were transformed into a diagnostic predictive scoring system. Internal validation was done using the bootstrapping procedure. Results: The data of pleural fluid cytology from 155 MPE patients were analyzed. Seventy-eight positive pleural cytology patients were found (50.32%). Lung cancer was the cancer most commonly sent for pleural fluid testing, with 66.67% positive cytology.  The predictive indicators included pleural fluid protein > 4.64 g/dL, pleural fluid LDH > 555 IU/L, and pleural fluid sugar > 60 mg/dL. Lung mass from imaging and double tap for pleural cytology were used for the derivation of the diagnostic prediction model. The score-based model showed that the area under the receiver operating characteristic curve was 0.74 (95% CI 0.66-0.82). The developed MPE score ranged from zero to 17. The cut-off point was 15 with 88.31% of specificity, 37.18% of sensitivity, positive predictive value of 0.76, and negative predictive value of 0.58. The measurement of the calibration was illustrated using a calibration plot (p-value = 0.49 for the Hosmer-Lemeshow based goodness of fit). Internal validation with 1,000 bootstrap resampling showed a good discrimination. Conclusions: The MPE score, as the diagnostic prediction model can be used in planning for more efficient diagnosis of MPE in patients with cancer under MPE.

[CONTENT] MPE score | pleural fluid cytology | diagnostic prediction model | malignant pleural effusion [SUMMARY]

[CONTENT] MPE score | pleural fluid cytology | diagnostic prediction model | malignant pleural effusion [SUMMARY]

[CONTENT] MPE score | pleural fluid cytology | diagnostic prediction model | malignant pleural effusion [SUMMARY]

[CONTENT] Adult | Area Under Curve | Biomarkers, Tumor | Body Fluids | Clinical Decision Rules | Female | Humans | Logistic Models | Male | Middle Aged | Pleura | Pleural Effusion, Malignant | Predictive Value of Tests | ROC Curve | Retrospective Studies | Risk Assessment | Risk Factors | Sensitivity and Specificity [SUMMARY]

[CONTENT] Adult | Area Under Curve | Biomarkers, Tumor | Body Fluids | Clinical Decision Rules | Female | Humans | Logistic Models | Male | Middle Aged | Pleura | Pleural Effusion, Malignant | Predictive Value of Tests | ROC Curve | Retrospective Studies | Risk Assessment | Risk Factors | Sensitivity and Specificity [SUMMARY]

[CONTENT] Adult | Area Under Curve | Biomarkers, Tumor | Body Fluids | Clinical Decision Rules | Female | Humans | Logistic Models | Male | Middle Aged | Pleura | Pleural Effusion, Malignant | Predictive Value of Tests | ROC Curve | Retrospective Studies | Risk Assessment | Risk Factors | Sensitivity and Specificity [SUMMARY]

[CONTENT] diagnosis mpe cytology | mpe improved pleuroscopy | cytology pleural effusion | mpe pleural fluid | testing metastasis pleura [SUMMARY]

[CONTENT] diagnosis mpe cytology | mpe improved pleuroscopy | cytology pleural effusion | mpe pleural fluid | testing metastasis pleura [SUMMARY]

[CONTENT] diagnosis mpe cytology | mpe improved pleuroscopy | cytology pleural effusion | mpe pleural fluid | testing metastasis pleura [SUMMARY]

[CONTENT] pleural | mpe | fluid | pleural fluid | cancer | diagnosis | data | cytology | score | ratio [SUMMARY]

[CONTENT] pleural | mpe | fluid | pleural fluid | cancer | diagnosis | data | cytology | score | ratio [SUMMARY]

[CONTENT] pleural | mpe | fluid | pleural fluid | cancer | diagnosis | data | cytology | score | ratio [SUMMARY]

[CONTENT] mpe | cancer | pleural | cytology | diagnostic | diagnosis | invasive | method | fluid | pleural fluid [SUMMARY]

[CONTENT] equal | score | mpe | curve | value | logistic regression | logistic | regression | ratio | pleural [SUMMARY]

[CONTENT] pleural | mpe | fluid | pleural fluid | financial | data | cancer | diagnosis | cytology | score [SUMMARY]

[CONTENT] MPE | pleural fluid | MPE [SUMMARY]

[CONTENT] pleural fluid | 155 | MPE ||| Seventy-eight | 50.32% ||| 66.67% ||| 4.64 | 555 ||| 60 | dL. Lung ||| 0.74 | 95% | CI | 0.66-0.82 ||| MPE | zero | 17 ||| 15 | 88.31% | 37.18% | 0.76 | 0.58 ||| 0.49 | the Hosmer-Lemeshow ||| 1,000 [SUMMARY]

[CONTENT] MPE | pleural fluid | MPE ||| MPE | between 2018 and 2020 ||| ||| ||| ||| ||| pleural fluid | 155 | MPE ||| Seventy-eight | 50.32% ||| 66.67% ||| 4.64 | 555 ||| 60 | dL. Lung ||| 0.74 | 95% | CI | 0.66-0.82 ||| MPE | zero | 17 ||| 15 | 88.31% | 37.18% | 0.76 | 0.58 ||| 0.49 | the Hosmer-Lemeshow ||| 1,000 ||| MPE | MPE | MPE [SUMMARY]

A deep neural network using audio files for detection of aortic stenosis.

35438211

Although aortic stenosis (AS) is the most common valvular heart disease in the western world, many affected patients remain undiagnosed. Auscultation is a readily available screening tool for AS. However, it requires a high level of professional expertise.

BACKGROUND

A deep neural network (DNN) was trained by preprocessed audio files of 100 patients with AS and 100 controls. The DNN's performance was evaluated with a test data set of 40 patients. The primary outcome measures were sensitivity, specificity, and F1-score. Results of the DNN were compared with the performance of cardiologists, residents, and medical students.

METHODS

Eighteen percent of patients without AS and 22% of patients with AS showed an additional moderate or severe mitral regurgitation. The DNN showed a sensitivity of 0.90 (0.81-0.99), a specificity of 1, and an F1-score of 0.95 (0.89-1.0) for the detection of AS. In comparison, we calculated an F1-score of 0.94 (0.86-1.0) for cardiologists, 0.88 (0.78-0.98) for residents, and 0.88 (0.78-0.98) for students.

RESULTS

The present study shows that deep learning-guided auscultation predicts significant AS with similar accuracy as cardiologists. The results of this pilot study suggest that AI-assisted auscultation may help general practitioners without special cardiology training in daily practice.

CONCLUSIONS

9175247

BACKGROUND

After mitral regurgitation, aortic valve stenosis (AS) is the second most common valvular disease. 1 Its prevalence in the general population is 0.4%, with a sharp age‐dependent increase in the older population. A prevalence of 3.4% in the age cohort >65 years can be observed. 2 It leads more often to hospitalization than other heart valve diseases and accounts for 45% of patients operated for valvular disease. 3 It is well known that symptomatic AS has high mortality when no valve replacement therapy is performed. 4 However, recent data suggest that even asymptomatic severe AS is associated with a mortality of up to 58% within 8 years. 5 Besides, it has been shown that aortic valve replacement in patients with severe asymptomatic AS significantly improves outcome. 6 Unfortunately, a high number of patients with significant AS remains undetected. The OxVALVE Population Cohort Study shows that unrecognized, significant AS was present in 1.6% of patients ≥65 years. 7 Thus, a reliable, readily available, and cheap screening tool is necessary. Since the invention of the stethoscope by Laennec in 1816, cardiac auscultation has been one of the pillars of cardiovascular examination. 8 However, this method's significant drawbacks are that heart murmurs are variable, and auscultation skills are highly performer‐dependent. 9 Deep learning is a branch of machine learning using artificial neural networks that models human brains' architecture. Heart sound classification can be principally done using a convolutional neural network (CNN), a subform of deep neural networks (DNNs). 10 We hypothesized that we could train a DNN to identify heart murmurs suspicious for significant AS with an accuracy comparable to experienced cardiologists.

METHODS

The study consists of two parts. In the first part, we trained a neural network to classify auscultation findings of patients who have significant AS or not. In the second part, we compared the performance of the trained DNN with the auscultatory skills of 10 experienced cardiologists, 10 residents, and 10 medical students by using a test data set that consisted of a completely disjointed set of patients. For training, we used auscultation audio files from 100 patients with significant AS and 100 patients without AS. The ground truth was defined by echocardiography. Significant AS was defined as V max of >3.5 m/s measured by continuous‐wave Doppler. Although the definition of high‐grade AS has not yet been reached, we chose this cut‐off value, as these patients require close monitoring. Patients admitted for suspicious coronary artery disease or other cardiac diseases were taken as a control group. We used an electronic stethoscope (Eko) connected to a smartphone interface via Bluetooth for auscultation. Auscultation was performed at the aortic auscultation point (second intercostal space, right sternal border) and the mitral auscultation point (fifth intercostal space, midclavicular line). Thus, from each patient, we included two auscultation files. At each auscultation point, audio files with an interval of 15 s and a sampling rate of 40 kHz were recorded. The audio files were recorded as part of the clinical routine in a tertiary teaching hospital with a large valve unit specialized in transcatheter aortic valve implantation (TAVI). Only data from patients of this database were included who got echocardiography within 7 days before or after auscultation. Since the study was retrospective, an explicit ethics vote was not necessary according to the regulations of the responsible ethics committee. We preprocessed the data in our study before using them to train the network. In the first step, the 15 s sound files were divided into three equal parts of 5 s. This was done to overcome the risk that small portions of an auscultation file falsified by respiration 11 contribute disproportionately to the training of the whole network. Consequently, six sound files per patient (three files for each of the 2 auscultation points) contribute to the network's training. In the second step, we performed Mel Frequency Cepstral Coefficients transformation of the audio files (MFCC‐transformation). This transformation maps the perception of human hearing and has been proposed for audio data analysis of heart and lung auscultation. 12 , 13 , 14 Subsequently, we trained the DNN with 1200 processed audio files (6 files per patient, 100 patients with AS, 100 patients without AS). We developed a CNN for classification (Figure 1), which takes as input MFCCs. The sequential model has three two‐dimensional‐convolutional layers and one max pooling layer. As an activation function, we used “ReLU”. Before each convolutional layer, we applied batch normalization. To combat overfitting, we used a dropout layer between the convolutional layers that sets a random portion of the weights equal to a probability of 0.2 to 0. Thereby the network has to learn different aspects of the data each time. 15 Data processing and analysis were principally done in two steps. In the first step, MFCC feature extraction was done. In the second step, the preprocessed data were fed to the convolutional part of the DNN. After the convolutional layers, the output is flattened to a one‐dimensional tensor. Data are then fed to a fully connected layer using the ReLU (rectified linear unit) activation function. To overcome overfitting, which means that the network is too much adapted to the training data set, the regulizer and dropout techniques were applied. In the softmax function, the input values are transformed to a probability distribution that gives the probability of AS or no AS in the present case. AS, aortic valve stenosis; MFCC, Mel frequency cepstral coefficients. Hyperparameter tuning was done iteratively for learning rate, batch size, number of epochs, number of kernels, and grid size of the convolutional layers. Model comparison was made using K‐fold cross‐validation. After training the network, it was applied to the test set that the model had not seen before. The test set consists of 20 patients with AS and 20 without AS. Accuracy, sensitivity, specificity, receiver operating characteristic curves (ROC), and F1 score were calculated. F1 score was calculated using the following formula: F1 score = 2 × (recall × precision)/(recall + precision). Then the same test set was classified by 10 experienced cardiologists, 10 residents, and 10 final year medical students. The performance parameters were averaged in cardiologists, residents, and students. Audio file processing, training the DNN, and making predictions were made with the general‐purpose programming language Python. Preprocessing audio files was done by the audio analysis library Librosa. We generated Mel Frequency Cepstral Coefficients (MFCC) with a hopelength of 10 and 13 coefficients. 16 A CNN was implemented using the Keras framework with a TensorFlow (Google) backend. 17 Continuous baseline characteristics are given as mean ± SD. Continuous variables were compared using t‐test, and categorical variables were compared using chi‐quadrat test. Accuracy, sensitivity, specificity, and F1‐value for cardiologists, residents, and students are given as mean with a 95% confidence interval. The confidence intervals for the specific parameters of the DNN were calculated with the formula = z × sqrt((parameter × (1 − parameter))/n), where z is the corresponding parameter and n is the size of the test sample, 40 in the present case. Inter‐rater reliability was assessed by calculating Fleiss' kappa. 18

RESULTS

Data from 120 patients with and 120 patients without AS were taken for the present study. From each group, audio files from 100 patients were allocated to the training group and 20 to the test group, respectively. Of the 120 patients with AS, in 99 patients femoral TAVI, in 5 transapical aortic valve implantation, in 13 patients open‐heart surgery, and in 1 patient only valvuloplasty were performed. Two patients were treated conservatively. Patients with AS were older than control patients in the training and test patients. A significant proportion of patients had mitral or tricuspid valve disease. Atrioventricular valve defects were more frequent in patients with AS. However, this difference was not significant. For details on patient characteristics, see Table 1. Patient characteristics Note: Data are given as mean ± SD. Moderate or severe heart valve disease. *p < .05 no AS versus AS. **p < .01 no AS versus AS. DNN's diagnostic accuracy Hyperparameter tuning was done using K‐fold cross‐validation. The training data were split into fourfolds, and while iterating through the folds, each iteration uses onefold as the validation set. Using this approach, the optimized DNN consists of three convolutional layers with 32 kernels with a grid size of 3 × 3 (convolutional layer 1 + 2) and 2 × 2 (convolutional layer 3). The best results could be achieved with 40 epochs, a batch size of 4, and a learning rate of 0.001. After flattening the tensor, a fully connected layer with 150 nodes followed. The final fully connected softmax layer produces a distribution over the two output classes. A schematic of the neural network is shown in Figure 1. We analyzed whether the algorithm showed different results depending on the auscultation point. As expected, worse prediction accuracy is shown when the neural network was trained only with audio files from one auscultation point. This is most likely due to the smaller number of audio files. However, the cause might also be that both auscultation points provide complementary information. Actually, the performance of a model trained only with data from the mitral point was the same as when only trained with data from the aortic point. This result contradicts common clinical experience. However, DNNs are able to identify patterns that are not recognizable to humans. The ROC‐AUC of the DNN trained only with audio files from the aortic auscultation point was 0.93, only trained with audio files from the mitral auscultation point was 0.83 and for merged data was 0.99. The DNN did not detect two patients with AS. No patient without AS was misclassified. The positive predictive value for the DNN using both auscultation points was 1.0, for students 0.83 (0.81–0.86), for residents 0.86 (0.82–0.89), and for cardiologists 0.93 (0.91–0.94). The negative predictive value for the DNN using both auscultation points was 0.91 (0.82–1.0), for students 0.94 (0.93–0.95), for residents 0.94 (0.93–0.95), and for cardiologists 0.96 (0.94–0.97). Hyperparameter tuning was done using K‐fold cross‐validation. The training data were split into fourfolds, and while iterating through the folds, each iteration uses onefold as the validation set. Using this approach, the optimized DNN consists of three convolutional layers with 32 kernels with a grid size of 3 × 3 (convolutional layer 1 + 2) and 2 × 2 (convolutional layer 3). The best results could be achieved with 40 epochs, a batch size of 4, and a learning rate of 0.001. After flattening the tensor, a fully connected layer with 150 nodes followed. The final fully connected softmax layer produces a distribution over the two output classes. A schematic of the neural network is shown in Figure 1. We analyzed whether the algorithm showed different results depending on the auscultation point. As expected, worse prediction accuracy is shown when the neural network was trained only with audio files from one auscultation point. This is most likely due to the smaller number of audio files. However, the cause might also be that both auscultation points provide complementary information. Actually, the performance of a model trained only with data from the mitral point was the same as when only trained with data from the aortic point. This result contradicts common clinical experience. However, DNNs are able to identify patterns that are not recognizable to humans. The ROC‐AUC of the DNN trained only with audio files from the aortic auscultation point was 0.93, only trained with audio files from the mitral auscultation point was 0.83 and for merged data was 0.99. The DNN did not detect two patients with AS. No patient without AS was misclassified. The positive predictive value for the DNN using both auscultation points was 1.0, for students 0.83 (0.81–0.86), for residents 0.86 (0.82–0.89), and for cardiologists 0.93 (0.91–0.94). The negative predictive value for the DNN using both auscultation points was 0.91 (0.82–1.0), for students 0.94 (0.93–0.95), for residents 0.94 (0.93–0.95), and for cardiologists 0.96 (0.94–0.97). Diagnostic accuracy of the CNN versus cardiologists, residents, and students Ten students, 10 residents in an advanced stage of training, and 10 consultant cardiologists participated in the study. Participants were blinded for the results of the DNN. They were asked to classify patients whether to have AS using two audio files for each patient. For inter‐rater reliability, Fleiss' kappa was 0.69 in students, 0.64 in residents, and 0.84 in cardiologists. This shows that the agreement in the group of cardiologists is much higher than in the group of residents and students. The F1‐score is a parameter to compare the performance of different models or rater groups when seeking a balance between precision and recall. In Figure 2 ROC curves for deep learning model, students, residents, and cardiologists are shown. The DNN showed a higher F1‐score than the mean score of cardiologists, residents, and students. Values for accuracy, sensitivity, specificity, and F1‐score are given in Table 2. ROC curve (orange line) achieved by the model in comparison to students (A), residents (B), and cardiologists (C). Individual rater performance is indicated by the black crosses, and averaged cardiologist performance is indicated by the red dot. Comparison between the DNN and humans in the detection of AS in the test set Note: Data are given as mean and confidence intervals. Abbreviations: DNN, deep neural network; MR, mitral regurgitation. Ten students, 10 residents in an advanced stage of training, and 10 consultant cardiologists participated in the study. Participants were blinded for the results of the DNN. They were asked to classify patients whether to have AS using two audio files for each patient. For inter‐rater reliability, Fleiss' kappa was 0.69 in students, 0.64 in residents, and 0.84 in cardiologists. This shows that the agreement in the group of cardiologists is much higher than in the group of residents and students. The F1‐score is a parameter to compare the performance of different models or rater groups when seeking a balance between precision and recall. In Figure 2 ROC curves for deep learning model, students, residents, and cardiologists are shown. The DNN showed a higher F1‐score than the mean score of cardiologists, residents, and students. Values for accuracy, sensitivity, specificity, and F1‐score are given in Table 2. ROC curve (orange line) achieved by the model in comparison to students (A), residents (B), and cardiologists (C). Individual rater performance is indicated by the black crosses, and averaged cardiologist performance is indicated by the red dot. Comparison between the DNN and humans in the detection of AS in the test set Note: Data are given as mean and confidence intervals. Abbreviations: DNN, deep neural network; MR, mitral regurgitation.

BACKGROUND: After mitral regurgitation, aortic valve stenosis (AS) is the second most common valvular disease. 1 Its prevalence in the general population is 0.4%, with a sharp age‐dependent increase in the older population. A prevalence of 3.4% in the age cohort >65 years can be observed. 2 It leads more often to hospitalization than other heart valve diseases and accounts for 45% of patients operated for valvular disease. 3 It is well known that symptomatic AS has high mortality when no valve replacement therapy is performed. 4 However, recent data suggest that even asymptomatic severe AS is associated with a mortality of up to 58% within 8 years. 5 Besides, it has been shown that aortic valve replacement in patients with severe asymptomatic AS significantly improves outcome. 6 Unfortunately, a high number of patients with significant AS remains undetected. The OxVALVE Population Cohort Study shows that unrecognized, significant AS was present in 1.6% of patients ≥65 years. 7 Thus, a reliable, readily available, and cheap screening tool is necessary. Since the invention of the stethoscope by Laennec in 1816, cardiac auscultation has been one of the pillars of cardiovascular examination. 8 However, this method's significant drawbacks are that heart murmurs are variable, and auscultation skills are highly performer‐dependent. 9 Deep learning is a branch of machine learning using artificial neural networks that models human brains' architecture. Heart sound classification can be principally done using a convolutional neural network (CNN), a subform of deep neural networks (DNNs). 10 We hypothesized that we could train a DNN to identify heart murmurs suspicious for significant AS with an accuracy comparable to experienced cardiologists. METHODS: The study consists of two parts. In the first part, we trained a neural network to classify auscultation findings of patients who have significant AS or not. In the second part, we compared the performance of the trained DNN with the auscultatory skills of 10 experienced cardiologists, 10 residents, and 10 medical students by using a test data set that consisted of a completely disjointed set of patients. For training, we used auscultation audio files from 100 patients with significant AS and 100 patients without AS. The ground truth was defined by echocardiography. Significant AS was defined as V max of >3.5 m/s measured by continuous‐wave Doppler. Although the definition of high‐grade AS has not yet been reached, we chose this cut‐off value, as these patients require close monitoring. Patients admitted for suspicious coronary artery disease or other cardiac diseases were taken as a control group. We used an electronic stethoscope (Eko) connected to a smartphone interface via Bluetooth for auscultation. Auscultation was performed at the aortic auscultation point (second intercostal space, right sternal border) and the mitral auscultation point (fifth intercostal space, midclavicular line). Thus, from each patient, we included two auscultation files. At each auscultation point, audio files with an interval of 15 s and a sampling rate of 40 kHz were recorded. The audio files were recorded as part of the clinical routine in a tertiary teaching hospital with a large valve unit specialized in transcatheter aortic valve implantation (TAVI). Only data from patients of this database were included who got echocardiography within 7 days before or after auscultation. Since the study was retrospective, an explicit ethics vote was not necessary according to the regulations of the responsible ethics committee. We preprocessed the data in our study before using them to train the network. In the first step, the 15 s sound files were divided into three equal parts of 5 s. This was done to overcome the risk that small portions of an auscultation file falsified by respiration 11 contribute disproportionately to the training of the whole network. Consequently, six sound files per patient (three files for each of the 2 auscultation points) contribute to the network's training. In the second step, we performed Mel Frequency Cepstral Coefficients transformation of the audio files (MFCC‐transformation). This transformation maps the perception of human hearing and has been proposed for audio data analysis of heart and lung auscultation. 12 , 13 , 14 Subsequently, we trained the DNN with 1200 processed audio files (6 files per patient, 100 patients with AS, 100 patients without AS). We developed a CNN for classification (Figure 1), which takes as input MFCCs. The sequential model has three two‐dimensional‐convolutional layers and one max pooling layer. As an activation function, we used “ReLU”. Before each convolutional layer, we applied batch normalization. To combat overfitting, we used a dropout layer between the convolutional layers that sets a random portion of the weights equal to a probability of 0.2 to 0. Thereby the network has to learn different aspects of the data each time. 15 Data processing and analysis were principally done in two steps. In the first step, MFCC feature extraction was done. In the second step, the preprocessed data were fed to the convolutional part of the DNN. After the convolutional layers, the output is flattened to a one‐dimensional tensor. Data are then fed to a fully connected layer using the ReLU (rectified linear unit) activation function. To overcome overfitting, which means that the network is too much adapted to the training data set, the regulizer and dropout techniques were applied. In the softmax function, the input values are transformed to a probability distribution that gives the probability of AS or no AS in the present case. AS, aortic valve stenosis; MFCC, Mel frequency cepstral coefficients. Hyperparameter tuning was done iteratively for learning rate, batch size, number of epochs, number of kernels, and grid size of the convolutional layers. Model comparison was made using K‐fold cross‐validation. After training the network, it was applied to the test set that the model had not seen before. The test set consists of 20 patients with AS and 20 without AS. Accuracy, sensitivity, specificity, receiver operating characteristic curves (ROC), and F1 score were calculated. F1 score was calculated using the following formula: F1 score = 2 × (recall × precision)/(recall + precision). Then the same test set was classified by 10 experienced cardiologists, 10 residents, and 10 final year medical students. The performance parameters were averaged in cardiologists, residents, and students. Audio file processing, training the DNN, and making predictions were made with the general‐purpose programming language Python. Preprocessing audio files was done by the audio analysis library Librosa. We generated Mel Frequency Cepstral Coefficients (MFCC) with a hopelength of 10 and 13 coefficients. 16 A CNN was implemented using the Keras framework with a TensorFlow (Google) backend. 17 Continuous baseline characteristics are given as mean ± SD. Continuous variables were compared using t‐test, and categorical variables were compared using chi‐quadrat test. Accuracy, sensitivity, specificity, and F1‐value for cardiologists, residents, and students are given as mean with a 95% confidence interval. The confidence intervals for the specific parameters of the DNN were calculated with the formula = z × sqrt((parameter × (1 − parameter))/n), where z is the corresponding parameter and n is the size of the test sample, 40 in the present case. Inter‐rater reliability was assessed by calculating Fleiss' kappa. 18 RESULTS: Data from 120 patients with and 120 patients without AS were taken for the present study. From each group, audio files from 100 patients were allocated to the training group and 20 to the test group, respectively. Of the 120 patients with AS, in 99 patients femoral TAVI, in 5 transapical aortic valve implantation, in 13 patients open‐heart surgery, and in 1 patient only valvuloplasty were performed. Two patients were treated conservatively. Patients with AS were older than control patients in the training and test patients. A significant proportion of patients had mitral or tricuspid valve disease. Atrioventricular valve defects were more frequent in patients with AS. However, this difference was not significant. For details on patient characteristics, see Table 1. Patient characteristics Note: Data are given as mean ± SD. Moderate or severe heart valve disease. *p < .05 no AS versus AS. **p < .01 no AS versus AS. DNN's diagnostic accuracy Hyperparameter tuning was done using K‐fold cross‐validation. The training data were split into fourfolds, and while iterating through the folds, each iteration uses onefold as the validation set. Using this approach, the optimized DNN consists of three convolutional layers with 32 kernels with a grid size of 3 × 3 (convolutional layer 1 + 2) and 2 × 2 (convolutional layer 3). The best results could be achieved with 40 epochs, a batch size of 4, and a learning rate of 0.001. After flattening the tensor, a fully connected layer with 150 nodes followed. The final fully connected softmax layer produces a distribution over the two output classes. A schematic of the neural network is shown in Figure 1. We analyzed whether the algorithm showed different results depending on the auscultation point. As expected, worse prediction accuracy is shown when the neural network was trained only with audio files from one auscultation point. This is most likely due to the smaller number of audio files. However, the cause might also be that both auscultation points provide complementary information. Actually, the performance of a model trained only with data from the mitral point was the same as when only trained with data from the aortic point. This result contradicts common clinical experience. However, DNNs are able to identify patterns that are not recognizable to humans. The ROC‐AUC of the DNN trained only with audio files from the aortic auscultation point was 0.93, only trained with audio files from the mitral auscultation point was 0.83 and for merged data was 0.99. The DNN did not detect two patients with AS. No patient without AS was misclassified. The positive predictive value for the DNN using both auscultation points was 1.0, for students 0.83 (0.81–0.86), for residents 0.86 (0.82–0.89), and for cardiologists 0.93 (0.91–0.94). The negative predictive value for the DNN using both auscultation points was 0.91 (0.82–1.0), for students 0.94 (0.93–0.95), for residents 0.94 (0.93–0.95), and for cardiologists 0.96 (0.94–0.97). Hyperparameter tuning was done using K‐fold cross‐validation. The training data were split into fourfolds, and while iterating through the folds, each iteration uses onefold as the validation set. Using this approach, the optimized DNN consists of three convolutional layers with 32 kernels with a grid size of 3 × 3 (convolutional layer 1 + 2) and 2 × 2 (convolutional layer 3). The best results could be achieved with 40 epochs, a batch size of 4, and a learning rate of 0.001. After flattening the tensor, a fully connected layer with 150 nodes followed. The final fully connected softmax layer produces a distribution over the two output classes. A schematic of the neural network is shown in Figure 1. We analyzed whether the algorithm showed different results depending on the auscultation point. As expected, worse prediction accuracy is shown when the neural network was trained only with audio files from one auscultation point. This is most likely due to the smaller number of audio files. However, the cause might also be that both auscultation points provide complementary information. Actually, the performance of a model trained only with data from the mitral point was the same as when only trained with data from the aortic point. This result contradicts common clinical experience. However, DNNs are able to identify patterns that are not recognizable to humans. The ROC‐AUC of the DNN trained only with audio files from the aortic auscultation point was 0.93, only trained with audio files from the mitral auscultation point was 0.83 and for merged data was 0.99. The DNN did not detect two patients with AS. No patient without AS was misclassified. The positive predictive value for the DNN using both auscultation points was 1.0, for students 0.83 (0.81–0.86), for residents 0.86 (0.82–0.89), and for cardiologists 0.93 (0.91–0.94). The negative predictive value for the DNN using both auscultation points was 0.91 (0.82–1.0), for students 0.94 (0.93–0.95), for residents 0.94 (0.93–0.95), and for cardiologists 0.96 (0.94–0.97). Diagnostic accuracy of the CNN versus cardiologists, residents, and students Ten students, 10 residents in an advanced stage of training, and 10 consultant cardiologists participated in the study. Participants were blinded for the results of the DNN. They were asked to classify patients whether to have AS using two audio files for each patient. For inter‐rater reliability, Fleiss' kappa was 0.69 in students, 0.64 in residents, and 0.84 in cardiologists. This shows that the agreement in the group of cardiologists is much higher than in the group of residents and students. The F1‐score is a parameter to compare the performance of different models or rater groups when seeking a balance between precision and recall. In Figure 2 ROC curves for deep learning model, students, residents, and cardiologists are shown. The DNN showed a higher F1‐score than the mean score of cardiologists, residents, and students. Values for accuracy, sensitivity, specificity, and F1‐score are given in Table 2. ROC curve (orange line) achieved by the model in comparison to students (A), residents (B), and cardiologists (C). Individual rater performance is indicated by the black crosses, and averaged cardiologist performance is indicated by the red dot. Comparison between the DNN and humans in the detection of AS in the test set Note: Data are given as mean and confidence intervals. Abbreviations: DNN, deep neural network; MR, mitral regurgitation. Ten students, 10 residents in an advanced stage of training, and 10 consultant cardiologists participated in the study. Participants were blinded for the results of the DNN. They were asked to classify patients whether to have AS using two audio files for each patient. For inter‐rater reliability, Fleiss' kappa was 0.69 in students, 0.64 in residents, and 0.84 in cardiologists. This shows that the agreement in the group of cardiologists is much higher than in the group of residents and students. The F1‐score is a parameter to compare the performance of different models or rater groups when seeking a balance between precision and recall. In Figure 2 ROC curves for deep learning model, students, residents, and cardiologists are shown. The DNN showed a higher F1‐score than the mean score of cardiologists, residents, and students. Values for accuracy, sensitivity, specificity, and F1‐score are given in Table 2. ROC curve (orange line) achieved by the model in comparison to students (A), residents (B), and cardiologists (C). Individual rater performance is indicated by the black crosses, and averaged cardiologist performance is indicated by the red dot. Comparison between the DNN and humans in the detection of AS in the test set Note: Data are given as mean and confidence intervals. Abbreviations: DNN, deep neural network; MR, mitral regurgitation. DNN's diagnostic accuracy: Hyperparameter tuning was done using K‐fold cross‐validation. The training data were split into fourfolds, and while iterating through the folds, each iteration uses onefold as the validation set. Using this approach, the optimized DNN consists of three convolutional layers with 32 kernels with a grid size of 3 × 3 (convolutional layer 1 + 2) and 2 × 2 (convolutional layer 3). The best results could be achieved with 40 epochs, a batch size of 4, and a learning rate of 0.001. After flattening the tensor, a fully connected layer with 150 nodes followed. The final fully connected softmax layer produces a distribution over the two output classes. A schematic of the neural network is shown in Figure 1. We analyzed whether the algorithm showed different results depending on the auscultation point. As expected, worse prediction accuracy is shown when the neural network was trained only with audio files from one auscultation point. This is most likely due to the smaller number of audio files. However, the cause might also be that both auscultation points provide complementary information. Actually, the performance of a model trained only with data from the mitral point was the same as when only trained with data from the aortic point. This result contradicts common clinical experience. However, DNNs are able to identify patterns that are not recognizable to humans. The ROC‐AUC of the DNN trained only with audio files from the aortic auscultation point was 0.93, only trained with audio files from the mitral auscultation point was 0.83 and for merged data was 0.99. The DNN did not detect two patients with AS. No patient without AS was misclassified. The positive predictive value for the DNN using both auscultation points was 1.0, for students 0.83 (0.81–0.86), for residents 0.86 (0.82–0.89), and for cardiologists 0.93 (0.91–0.94). The negative predictive value for the DNN using both auscultation points was 0.91 (0.82–1.0), for students 0.94 (0.93–0.95), for residents 0.94 (0.93–0.95), and for cardiologists 0.96 (0.94–0.97). Diagnostic accuracy of the CNN versus cardiologists, residents, and students: Ten students, 10 residents in an advanced stage of training, and 10 consultant cardiologists participated in the study. Participants were blinded for the results of the DNN. They were asked to classify patients whether to have AS using two audio files for each patient. For inter‐rater reliability, Fleiss' kappa was 0.69 in students, 0.64 in residents, and 0.84 in cardiologists. This shows that the agreement in the group of cardiologists is much higher than in the group of residents and students. The F1‐score is a parameter to compare the performance of different models or rater groups when seeking a balance between precision and recall. In Figure 2 ROC curves for deep learning model, students, residents, and cardiologists are shown. The DNN showed a higher F1‐score than the mean score of cardiologists, residents, and students. Values for accuracy, sensitivity, specificity, and F1‐score are given in Table 2. ROC curve (orange line) achieved by the model in comparison to students (A), residents (B), and cardiologists (C). Individual rater performance is indicated by the black crosses, and averaged cardiologist performance is indicated by the red dot. Comparison between the DNN and humans in the detection of AS in the test set Note: Data are given as mean and confidence intervals. Abbreviations: DNN, deep neural network; MR, mitral regurgitation. DISCUSSION: The present study shows that deep learning‐guided auscultation predicts significant AS with similar accuracy as board‐certified cardiologists. These results suggest that artificial intelligence‐assisted auscultation may help general practitioners without special cardiology training. Auscultation is one of the pillars of clinical investigation. It is readily available, and no sophisticated technical requirements are necessary. On the other hand, a high level of expertise is essential, and the skills, once acquired, need to be used continuously. These circumstances may explain errors in auscultation between 20% and 80% in residents, primary care physicians, and cardiologists. 19 , 20 For this reason, computer‐assisted auscultation was already proposed for clinical use at the beginning of this century. The developed algorithms decompose the cyclical heart sound, and hand‐engineered processing is applied for classification. In young patients with hypertrophic cardiomyopathy and children with congenital heart disease, sufficient sensitivity and specificity could be achieved with these techniques. 21 However, these studies were done in young patients with no conditions complicating the auscultation results like adiposity and lung emphysema. Deep neural networks (DNNs) take a completely different direction. A DNN is a machine learning algorithm that models brain architecture. They consist of perceptrons with adjustable weights and activation thresholds. In supervised learning, labeled data, often called ground truth, are propagated through a network of perceptrons that allows the DNN to adjust weights. In this step, the DNN learns how to assign known data to predefined categories. After this training phase, the performance of the final model is evaluated on a test data set. In this step, the trained DNN makes predictions in the form of probabilities for so far unknown data. In first applications, the focus was directed on standard diagnostic techniques used by doctors daily but cannot be provided on an expert level in any case. 22 Deep learning systems were developed for ECG interpretation, 23 skin cancer identification, 24 and papilledema detection. 25 In this context, it has been recognized that AI may also be a valuable tool to support doctors in identifying valvular heart disease. In a recently published study by Chorba et al., physicians assigned 5878 auscultation findings to the labels “heart murmur”, “no heart murmur”, or “inadequate signal”. The DNN was trained with these data using an end‐to‐end (E2E) network design. In a second step, the DNN was then validated on a test data set of 1774 recordings annotated by separate expert clinicians. 26 In contrast, the ground truth was not defined by physician assignment but by echocardiography in the present study. By using this gold standard, the well‐known erroneous annotation of auscultation findings by physicians was avoided. This is a crucial point as machine learning is based on detecting subtle patterns in data. Only when using high‐quality training data, noise that masks these patterns can be sufficiently reduced. Furthermore, the data in our study were pre‐processed before they were used to train the DNN. In this pre‐processing, attributes of the audio data were isolated that have been shown to be essential for pattern recognition in audio files. 27 , 28 With this hybrid approach, our DNN showed similar accuracy as board‐certified cardiologists. The precisely defined ground truth in conjunction with the preprocessing of the audio data compensates for a comparatively low patient number. A potential limitation of our study is that we only included a few patients with moderate aortic stenosis. This is due to the fact that the study was conducted with data from patients admitted to a tertiary teaching hospital for specialized valve therapy. Moreover, patients with only moderate valve disease are challenging to identify because they are often completely asymptomatic and therefore not under medical surveillance. Similarly, the DNN was trained predominantly with normal‐flow, high‐gradient AS. Probably the present algorithm will have worse performance in low‐flow, low‐gradient AS because the flow properties over the valve and thereby the sound characteristics are entirely different. This fact naturally lowers the sensitivity, but this condition affects only a small portion of the patients. In the present study, a machine learning algorithm was trained to detect patients with aortic stenosis. Patients with other valvular diseases like hypertrophic obstructive cardiomyopathy were not included, and few patients with pure mitral regurgitation participated. Thus, the presented algorithm is far from perfect, and this study can only be the first step in introducing artificial intelligence for valvular heart disease into everyday clinical practice. On the other side, it also shows that artificial intelligence can, in principle, be helpful in the auscultation of heart sounds. Clinical perspectives The present study gives proof of concept that AI‐assisted auscultation can provide results on a high expert level concerning the detection of aortic stenosis. Integrating this technology in an electronic stethoscope could be the next step to upgrade this system for everyday clinical use. A stethoscope that indicates a warning in the event of certain valve defects would be conceivable in the foreseeable future. Introducing such a device in countries that cannot provide comprehensive medical care has already been proposed years ago. 29 The present study gives proof of concept that AI‐assisted auscultation can provide results on a high expert level concerning the detection of aortic stenosis. Integrating this technology in an electronic stethoscope could be the next step to upgrade this system for everyday clinical use. A stethoscope that indicates a warning in the event of certain valve defects would be conceivable in the foreseeable future. Introducing such a device in countries that cannot provide comprehensive medical care has already been proposed years ago. 29 Clinical perspectives: The present study gives proof of concept that AI‐assisted auscultation can provide results on a high expert level concerning the detection of aortic stenosis. Integrating this technology in an electronic stethoscope could be the next step to upgrade this system for everyday clinical use. A stethoscope that indicates a warning in the event of certain valve defects would be conceivable in the foreseeable future. Introducing such a device in countries that cannot provide comprehensive medical care has already been proposed years ago. 29 CONFLICTS OF INTEREST: The authors declare no conflicts of interest.

Background: Although aortic stenosis (AS) is the most common valvular heart disease in the western world, many affected patients remain undiagnosed. Auscultation is a readily available screening tool for AS. However, it requires a high level of professional expertise. Methods: A deep neural network (DNN) was trained by preprocessed audio files of 100 patients with AS and 100 controls. The DNN's performance was evaluated with a test data set of 40 patients. The primary outcome measures were sensitivity, specificity, and F1-score. Results of the DNN were compared with the performance of cardiologists, residents, and medical students. Results: Eighteen percent of patients without AS and 22% of patients with AS showed an additional moderate or severe mitral regurgitation. The DNN showed a sensitivity of 0.90 (0.81-0.99), a specificity of 1, and an F1-score of 0.95 (0.89-1.0) for the detection of AS. In comparison, we calculated an F1-score of 0.94 (0.86-1.0) for cardiologists, 0.88 (0.78-0.98) for residents, and 0.88 (0.78-0.98) for students. Conclusions: The present study shows that deep learning-guided auscultation predicts significant AS with similar accuracy as cardiologists. The results of this pilot study suggest that AI-assisted auscultation may help general practitioners without special cardiology training in daily practice.

[CONTENT] aortic stenosis | artificial intelligence | auscultation | deep neural network | machine learning | valvular heart disease [SUMMARY]

[CONTENT] aortic stenosis | artificial intelligence | auscultation | deep neural network | machine learning | valvular heart disease [SUMMARY]

[CONTENT] aortic stenosis | artificial intelligence | auscultation | deep neural network | machine learning | valvular heart disease [SUMMARY]

[CONTENT] aortic stenosis | artificial intelligence | auscultation | deep neural network | machine learning | valvular heart disease [SUMMARY]

[CONTENT] Aortic Valve Stenosis | Cardiologists | Cardiology | Humans | Neural Networks, Computer | Pilot Projects [SUMMARY]

[CONTENT] Aortic Valve Stenosis | Cardiologists | Cardiology | Humans | Neural Networks, Computer | Pilot Projects [SUMMARY]

[CONTENT] Aortic Valve Stenosis | Cardiologists | Cardiology | Humans | Neural Networks, Computer | Pilot Projects [SUMMARY]

[CONTENT] Aortic Valve Stenosis | Cardiologists | Cardiology | Humans | Neural Networks, Computer | Pilot Projects [SUMMARY]

[CONTENT] valve stenosis second | aortic valve replacement | stenosis patients valvular | aortic stenosis fact | moderate aortic stenosis [SUMMARY]

[CONTENT] valve stenosis second | aortic valve replacement | stenosis patients valvular | aortic stenosis fact | moderate aortic stenosis [SUMMARY]

[CONTENT] valve stenosis second | aortic valve replacement | stenosis patients valvular | aortic stenosis fact | moderate aortic stenosis [SUMMARY]

[CONTENT] valve stenosis second | aortic valve replacement | stenosis patients valvular | aortic stenosis fact | moderate aortic stenosis [SUMMARY]

[CONTENT] auscultation | dnn | patients | data | cardiologists | residents | students | audio | files | audio files [SUMMARY]

[CONTENT] auscultation | dnn | patients | data | cardiologists | residents | students | audio | files | audio files [SUMMARY]

[CONTENT] auscultation | dnn | patients | data | cardiologists | residents | students | audio | files | audio files [SUMMARY]

[CONTENT] auscultation | dnn | patients | data | cardiologists | residents | students | audio | files | audio files [SUMMARY]

[CONTENT] population | heart | significant | valve | years | 65 | replacement | mortality | valve replacement | murmurs [SUMMARY]

[CONTENT] files | auscultation | audio | patients | data | 10 | test | coefficients | mfcc | convolutional [SUMMARY]

[CONTENT] students | residents | dnn | cardiologists | patients | point | auscultation | 94 | 93 | trained [SUMMARY]

[CONTENT] auscultation | dnn | students | patients | data | residents | cardiologists | files | audio | point [SUMMARY]

[CONTENT] ||| ||| [SUMMARY]

[CONTENT] 100 | AS and 100 ||| 40 ||| F1 ||| [SUMMARY]

[CONTENT] Eighteen percent | 22% ||| 0.90 | 0.81 | 1 | F1 | 0.95 | 0.89-1.0 ||| F1 | 0.94 | 0.86-1.0 | 0.88 | 0.78-0.98 | 0.88 | 0.78-0.98 [SUMMARY]

[CONTENT] ||| ||| ||| 100 | AS and 100 ||| 40 ||| F1 ||| ||| Eighteen percent | 22% ||| 0.90 | 0.81 | 1 | F1 | 0.95 | 0.89-1.0 ||| F1 | 0.94 | 0.86-1.0 | 0.88 | 0.78-0.98 | 0.88 | 0.78-0.98 ||| ||| AI | daily [SUMMARY]

Skin barrier dysfunction measured by transepidermal water loss at 2 days and 2 months predates and predicts atopic dermatitis at 1 year.

25618747

Loss-of-function mutations in the skin barrier protein filaggrin (FLG) are a major risk for atopic dermatitis (AD). The pathogenic sequence of disturbances in skin barrier function before or during the early development of AD is not fully understood. A more detailed understanding of these events is needed to develop a clearer picture of disease pathogenesis. A robust, noninvasive test to identify babies at high risk of AD would be important in planning early intervention and/or prevention studies.

BACKGROUND

A total of 1903 infants were enrolled in the Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints Birth Cohort study from July 2009 to October 2011. Measurements of TEWL were made at birth (day 2) and at 2 and 6 months. The presence of AD was ascertained at 6 and 12 months, and disease severity was assessed by using the SCORing Atopic Dermatitis clinical tool at 6 months and by using both the SCORing Atopic Dermatitis clinical tool and Nottingham Severity Score at 12 months. A total of 1300 infants were genotyped for FLG mutations.

METHODS

At 6 months, 18.7% of the children had AD, and at 12 months, 15.53%. In a logistic regression model, day 2 upper quartile TEWL measurement was significantly predictive of AD at 12 months (area under the receiver operating characteristic curve, 0.81; P < .05). Lowest quartile day 2 TEWL was protective against AD at 12 months. An upper quartile 2 month TEWL was also strongly predictive of AD at 12 months (area under the receiver operating characteristic curve, 0.84; P < .05). At both ages, this effect was independent of parental atopy, FLG status, or report of an itchy flexural rash at 2 months. Associations were increased when parental atopy status or child FLG mutation status was added into the linear regression model.

RESULTS

Impairment of skin barrier function at birth and at 2 months precedes clinical AD. In addition to providing important mechanistic insights into disease pathogenesis, these findings have implications for the optimal timing of interventions for the prevention of AD.

CONCLUSIONS

4382348

Methods

Study subjects The Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints (BASELINE) Birth Cohort study is the first birth cohort study in Ireland.27 It was developed as the pediatric follow on from the Cork Centre for the Screening for Pregnancy Endpoints (SCOPE) study, a multicenter international study evaluating diseases of pregnancy in primigravidous women.28 The Cork BASELINE Birth Cohort study recruited from August 2009 through to October 2011. Infants recruited from July 2009 had skin barrier assessment at birth and throughout early life; these 1903 infants are included in this study. A total of 1303 infants were recruited antenatally (stream1). These women were subject to the inclusion criteria of the SCOPE study, namely, first-time, low-risk mothers with singleton pregnancies delivered at or near term. Consent for recruitment was sought at 20 weeks of gestation and confirmed at birth of the live baby. Cord blood samples were taken at birth and stored for future use. A second recruitment stream began in July 2010 that recruited mothers and babies on the postnatal ward (stream2). A total of 600 infants were recruited postnatally. These mothers were enrolled independently of the SCOPE study. The sole inclusion criterion in stream 2 was a healthy term infant on the postnatal ward. Stream 1 and stream 2 infants were assessed and followed up in an identical fashion at birth and thereafter clinic visits at 2, 6, and 12 months. All infants had assessment at birth, 2 months, 6 months, and 12 months involving parental questionnaires and physical assessment. Parental questionnaires at 2, 6, and 12 months contained specific screening questions for AD. Experienced health care personnel diagnosed AD at 6 and 12 months in accordance with the UK Working Party diagnostic criteria.29-31 If AD was present, severity was assessed at these timepoints using the SCORing Atopic Dermatitis (SCORAD) clinical tool.32,33 At 12 months, the Nottingham Severity Score was also recorded to assess the severity of AD.34 The Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints (BASELINE) Birth Cohort study is the first birth cohort study in Ireland.27 It was developed as the pediatric follow on from the Cork Centre for the Screening for Pregnancy Endpoints (SCOPE) study, a multicenter international study evaluating diseases of pregnancy in primigravidous women.28 The Cork BASELINE Birth Cohort study recruited from August 2009 through to October 2011. Infants recruited from July 2009 had skin barrier assessment at birth and throughout early life; these 1903 infants are included in this study. A total of 1303 infants were recruited antenatally (stream1). These women were subject to the inclusion criteria of the SCOPE study, namely, first-time, low-risk mothers with singleton pregnancies delivered at or near term. Consent for recruitment was sought at 20 weeks of gestation and confirmed at birth of the live baby. Cord blood samples were taken at birth and stored for future use. A second recruitment stream began in July 2010 that recruited mothers and babies on the postnatal ward (stream2). A total of 600 infants were recruited postnatally. These mothers were enrolled independently of the SCOPE study. The sole inclusion criterion in stream 2 was a healthy term infant on the postnatal ward. Stream 1 and stream 2 infants were assessed and followed up in an identical fashion at birth and thereafter clinic visits at 2, 6, and 12 months. All infants had assessment at birth, 2 months, 6 months, and 12 months involving parental questionnaires and physical assessment. Parental questionnaires at 2, 6, and 12 months contained specific screening questions for AD. Experienced health care personnel diagnosed AD at 6 and 12 months in accordance with the UK Working Party diagnostic criteria.29-31 If AD was present, severity was assessed at these timepoints using the SCORing Atopic Dermatitis (SCORAD) clinical tool.32,33 At 12 months, the Nottingham Severity Score was also recorded to assess the severity of AD.34 TEWL measurements TEWL measurements were carried out using a widely validated open chamber system (Tewameter TM 300; Courage+Khazaka Electronic, Cologne, Germany). For newborns, TEWL was taken in the Cork University Maternity Hospital. The subject's arm was acclimated before measurement by exposing the arm in a nonenvironmentally controlled room for 10 minutes. This was typically done in the cot beside the mother's bed while an interview with the mother or parents was carried out. The infant was then brought to a windowless room in which both temperature and humidity were maintained constant by an air conditioning system. Temperature was set between 20°C and 25°C. Humidity was monitored by a manometer in the room and was maintained between 30% and 45%. TEWL was taken on the lower volar surface of the forearm by applying a probe to the exposed volar skin for approximately 15 seconds until the measurement was recorded. Three readings were taken, and the mean of the 3 readings was recorded. For TEWL readings at other timepoints, the same procedure was carried out at 2 months and 6 months in an environmentally controlled room in the Health Research Board Discovery Centre, the clinical research facility for children at Cork University Hospital. The parents were advised not to apply emollients to the infant's skin for 12 hours before the reading was taken. TEWL measurements were carried out using a widely validated open chamber system (Tewameter TM 300; Courage+Khazaka Electronic, Cologne, Germany). For newborns, TEWL was taken in the Cork University Maternity Hospital. The subject's arm was acclimated before measurement by exposing the arm in a nonenvironmentally controlled room for 10 minutes. This was typically done in the cot beside the mother's bed while an interview with the mother or parents was carried out. The infant was then brought to a windowless room in which both temperature and humidity were maintained constant by an air conditioning system. Temperature was set between 20°C and 25°C. Humidity was monitored by a manometer in the room and was maintained between 30% and 45%. TEWL was taken on the lower volar surface of the forearm by applying a probe to the exposed volar skin for approximately 15 seconds until the measurement was recorded. Three readings were taken, and the mean of the 3 readings was recorded. For TEWL readings at other timepoints, the same procedure was carried out at 2 months and 6 months in an environmentally controlled room in the Health Research Board Discovery Centre, the clinical research facility for children at Cork University Hospital. The parents were advised not to apply emollients to the infant's skin for 12 hours before the reading was taken. Statistical analysis A series of logistic regression (LR) models was used to estimate the factors at (1) 2 days and (2) 2 months that influence the diagnosis of AD at 12 months. The dependent variable that measures diagnosis at 12 months is AD “Yes” and is equal to 1, with AD “No” = 0. To examine the influence of each variable, with and without the presence of other significant predictors, we carried out both univariable and multivariable analyses. With regard to the latter, 3 models were produced. Model 1 included parental atopy, FLG mutation status, and presence of an itchy rash at 2 months as independent variables. Model 2 included parental atopy and TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3, respectively) as independent variables, but did not include FLG mutation status. This was done to determine the unique influence of TEWL, without the contribution of FLG mutations. Model 3 included all independent variables: parental atopy, TEWL test reading scores, and FLG mutation status. All models controlled for the variables “use of emollient before test reading,” “presence of an itchy rash,” “infant sex,” and “infant birthweight.” Receiver operating characteristic (ROC) curve output, which plots the true positives (sensitivity) vs false positives (1 − specificity), was then evaluated to compare the ability of the 3 models to produce a prognosis with high accuracy. Our rationale for variable selection was to include those predictors necessary for face validity but only if they were significant at a .05 level in the univariable analysis or if they altered the coefficient of the main variable by more than 10% in cases in which the main association was significant.35,36 A series of logistic regression (LR) models was used to estimate the factors at (1) 2 days and (2) 2 months that influence the diagnosis of AD at 12 months. The dependent variable that measures diagnosis at 12 months is AD “Yes” and is equal to 1, with AD “No” = 0. To examine the influence of each variable, with and without the presence of other significant predictors, we carried out both univariable and multivariable analyses. With regard to the latter, 3 models were produced. Model 1 included parental atopy, FLG mutation status, and presence of an itchy rash at 2 months as independent variables. Model 2 included parental atopy and TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3, respectively) as independent variables, but did not include FLG mutation status. This was done to determine the unique influence of TEWL, without the contribution of FLG mutations. Model 3 included all independent variables: parental atopy, TEWL test reading scores, and FLG mutation status. All models controlled for the variables “use of emollient before test reading,” “presence of an itchy rash,” “infant sex,” and “infant birthweight.” Receiver operating characteristic (ROC) curve output, which plots the true positives (sensitivity) vs false positives (1 − specificity), was then evaluated to compare the ability of the 3 models to produce a prognosis with high accuracy. Our rationale for variable selection was to include those predictors necessary for face validity but only if they were significant at a .05 level in the univariable analysis or if they altered the coefficient of the main variable by more than 10% in cases in which the main association was significant.35,36

Results

A total of 1903 infants were enrolled onto the study. The demographic details of the population studied are presented in Table I. Filaggrin mutation typing Stream 1 infants had cord blood samples taken at birth. Infants without cord blood samples had Oragene saliva samples taken. Those infants still enrolled in the study at 2 years had EDTA blood samples taken. FLG genotyping was carried out in 1300 infants with available DNA. The cumulative FLG mutation rate was 10.46% (136 of 1300). Four infants were homozygous for FLG mutation, with the remainder heterozygous for FLG mutation (Table II). Stream 1 infants had cord blood samples taken at birth. Infants without cord blood samples had Oragene saliva samples taken. Those infants still enrolled in the study at 2 years had EDTA blood samples taken. FLG genotyping was carried out in 1300 infants with available DNA. The cumulative FLG mutation rate was 10.46% (136 of 1300). Four infants were homozygous for FLG mutation, with the remainder heterozygous for FLG mutation (Table II). TEWL through early infancy TEWL was taken in the early newborn period in 1691 of 1903 (88.86%) infants of the cohort. The mean TEWL for newborns was 7.32 ± 3.33 gwater/m2/h. Using univariable analysis, we found that there was no significant association between TEWL measurements at birth and sex, gestation, or postnatal age at measurement or recruitment stream 1 or 2 (data not shown). A total of 1614 of 1638 (98.5%) infants who attended the 2-month appointment had TEWL measured (98.5%). Mean 2-month TEWL was 10.97 ± 7.98 gwater/m2/h. A total of 1516 of 1537 (98.6%) infants who attended the 6-month appointment had TEWL measured. Mean 6-month TEWL was 10.71 ± 7.10 gwater/m2/h. TEWL measurements by quartiles are presented in Table III. TEWL was taken in the early newborn period in 1691 of 1903 (88.86%) infants of the cohort. The mean TEWL for newborns was 7.32 ± 3.33 gwater/m2/h. Using univariable analysis, we found that there was no significant association between TEWL measurements at birth and sex, gestation, or postnatal age at measurement or recruitment stream 1 or 2 (data not shown). A total of 1614 of 1638 (98.5%) infants who attended the 2-month appointment had TEWL measured (98.5%). Mean 2-month TEWL was 10.97 ± 7.98 gwater/m2/h. A total of 1516 of 1537 (98.6%) infants who attended the 6-month appointment had TEWL measured. Mean 6-month TEWL was 10.71 ± 7.10 gwater/m2/h. TEWL measurements by quartiles are presented in Table III. Clinical diagnosis of AD AD was screened for at 6- and 12-month appointments. This diagnosis of AD was made according to the UK Working Party diagnostic criteria. At 6 months, 18.7% (299 of 1597) of the infants screened were diagnosed with AD. A total of 287 had a SCORAD completed, with a mean SCORAD score of 21.54 ± 16.29 (range, 0-88). At 12 months, 15.53% (232 of 1494) of the infants screened were diagnosed with AD. The mean SCORAD score at 12 months in affected children was 18.56 ± 14.92 (range, 0-77). AD was screened for at 6- and 12-month appointments. This diagnosis of AD was made according to the UK Working Party diagnostic criteria. At 6 months, 18.7% (299 of 1597) of the infants screened were diagnosed with AD. A total of 287 had a SCORAD completed, with a mean SCORAD score of 21.54 ± 16.29 (range, 0-88). At 12 months, 15.53% (232 of 1494) of the infants screened were diagnosed with AD. The mean SCORAD score at 12 months in affected children was 18.56 ± 14.92 (range, 0-77). LR analysis: Factors influencing the development of AD at 1 year Our primary outcome was the presence of AD at 1 year. We developed a model that explored the relative contributions of FLG mutation status, parental atopy, presence of an itchy rash at 2 months, TEWL at day 2 and month 2, and ΔTEWL (day 2 to month 2 and day 2 to month 6) on AD at 1 year (Table IV). FLG mutation status was not associated with an elevated TEWL at birth. However, it was associated with an elevated TEWL at 2 and 6 months and with elevated ΔTEWL day 2 to 2 months and ΔTEWL day 2 to 6 months. Notably, FLG mutation carriers did not have elevated ΔTEWL 2 months to 6 months, implying that the major changes in skin barrier function in those who develop AD by 12 months start in the first 2 months of life. In our univariable modeling, the odds ratio (OR) for AD at 12 months conferred by both parents being atopic was 2.5 compared with an OR of 3.1 for an upper quartile TEWL at 2 months (see Table E1 in this article's Online Repository at www.jacionline.org). LR models were then used to examine the factors at birth and at 2 months that influence a diagnosis of AD at 12 months. We used 3 separate models to explore the relative contribution of parental atopy, FLG mutation status, and TEWL (at day 2 and at 2 months) toward the development of AD at 12 months. All models controlled for the variables use of emollient before test reading, presence of an itchy rash, infant sex, and infant birthweight. These covariates were not significant in any of the 3 multivariable models. Our primary outcome was the presence of AD at 1 year. We developed a model that explored the relative contributions of FLG mutation status, parental atopy, presence of an itchy rash at 2 months, TEWL at day 2 and month 2, and ΔTEWL (day 2 to month 2 and day 2 to month 6) on AD at 1 year (Table IV). FLG mutation status was not associated with an elevated TEWL at birth. However, it was associated with an elevated TEWL at 2 and 6 months and with elevated ΔTEWL day 2 to 2 months and ΔTEWL day 2 to 6 months. Notably, FLG mutation carriers did not have elevated ΔTEWL 2 months to 6 months, implying that the major changes in skin barrier function in those who develop AD by 12 months start in the first 2 months of life. In our univariable modeling, the odds ratio (OR) for AD at 12 months conferred by both parents being atopic was 2.5 compared with an OR of 3.1 for an upper quartile TEWL at 2 months (see Table E1 in this article's Online Repository at www.jacionline.org). LR models were then used to examine the factors at birth and at 2 months that influence a diagnosis of AD at 12 months. We used 3 separate models to explore the relative contribution of parental atopy, FLG mutation status, and TEWL (at day 2 and at 2 months) toward the development of AD at 12 months. All models controlled for the variables use of emollient before test reading, presence of an itchy rash, infant sex, and infant birthweight. These covariates were not significant in any of the 3 multivariable models. LR modeling of AD risk at 12 months including day 2 TEWL The LR model incorporating day 2 TEWL is summarized in Table V. In our first model, we included parental atopy and FLG mutation status as independent variables and excluded day 2 TEWL. Parental atopy and FLG mutation status significantly increase the likelihood of a positive diagnosis of AD at 12 months by 13.7 and 8.6 times, respectively, compared with infants without FLG mutation and without an atopic parent. The ROC, which plots the true positives (sensitivity) vs false positives (1 −specificity), is 0.8 for this model. Our second model includes day 2 TEWL test reading scores at 25th, 50th, and 75th percentiles as independent variables. We did not include FLG mutation status in this second model. The area under the ROC curve (AUC) remains at 0.8 for this model. Infants with a TEWL reading of 9.0 or above are 7.1 times more likely to be diagnosed with AD at 1 year than are infants with a reading below this point, controlling for all other variables in the model. This suggests that day 2 TEWL can be used as a sole indicator of likelihood of AD at 12 months. Our third model includes all independent variables, including both TEWL and FLG status. The LR model incorporating day 2 TEWL is summarized in Table V. In our first model, we included parental atopy and FLG mutation status as independent variables and excluded day 2 TEWL. Parental atopy and FLG mutation status significantly increase the likelihood of a positive diagnosis of AD at 12 months by 13.7 and 8.6 times, respectively, compared with infants without FLG mutation and without an atopic parent. The ROC, which plots the true positives (sensitivity) vs false positives (1 −specificity), is 0.8 for this model. Our second model includes day 2 TEWL test reading scores at 25th, 50th, and 75th percentiles as independent variables. We did not include FLG mutation status in this second model. The area under the ROC curve (AUC) remains at 0.8 for this model. Infants with a TEWL reading of 9.0 or above are 7.1 times more likely to be diagnosed with AD at 1 year than are infants with a reading below this point, controlling for all other variables in the model. This suggests that day 2 TEWL can be used as a sole indicator of likelihood of AD at 12 months. Our third model includes all independent variables, including both TEWL and FLG status. LR modeling of AD risk at 12 months including TEWL at 2 months We repeated the LR approach for month 2 TEWL values, summarized in Table VI. Again, we used 3 models. In the first model, we included parental atopy and FLG mutation status as independent variables but did not include TEWL. Parental atopy and FLG mutation carriage significantly increase the likelihood of a positive AD diagnosis at 12 months by 2.7 and 3.0 times, respectively, compared with infants without an FLG mutation and without an atopic parent. The ROC is 0.66 for this model. In the second model, we included the month 2 TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3 gwater/m2/h, respectively) as independent variables. We did not include FLG mutation status in this second model, as explained above. The AUC improved from 0.66 to 0.82 in this model. Infants with a TEWL reading of 12.3 or above were 5.6 times more likely to be diagnosed with AD at 12 months than were infants with a reading below this point, controlling for all other variables in the model. Model 3 included all independent variables. In a pattern similar to that found at 2 days, the AUC improved slightly to 0.84, although here again it is important to note that model 2 demonstrates that FLG status need not be measured to produce a prognosis with high accuracy. We modeled these in a traditional stepwise LR analysis. The significant finding was that after controlling for all other possible influencing factors, month 2 TEWL was the strongest independent predictor of AD at 12 months. We repeated the LR approach for month 2 TEWL values, summarized in Table VI. Again, we used 3 models. In the first model, we included parental atopy and FLG mutation status as independent variables but did not include TEWL. Parental atopy and FLG mutation carriage significantly increase the likelihood of a positive AD diagnosis at 12 months by 2.7 and 3.0 times, respectively, compared with infants without an FLG mutation and without an atopic parent. The ROC is 0.66 for this model. In the second model, we included the month 2 TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3 gwater/m2/h, respectively) as independent variables. We did not include FLG mutation status in this second model, as explained above. The AUC improved from 0.66 to 0.82 in this model. Infants with a TEWL reading of 12.3 or above were 5.6 times more likely to be diagnosed with AD at 12 months than were infants with a reading below this point, controlling for all other variables in the model. Model 3 included all independent variables. In a pattern similar to that found at 2 days, the AUC improved slightly to 0.84, although here again it is important to note that model 2 demonstrates that FLG status need not be measured to produce a prognosis with high accuracy. We modeled these in a traditional stepwise LR analysis. The significant finding was that after controlling for all other possible influencing factors, month 2 TEWL was the strongest independent predictor of AD at 12 months.

Methods: Study subjects The Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints (BASELINE) Birth Cohort study is the first birth cohort study in Ireland.27 It was developed as the pediatric follow on from the Cork Centre for the Screening for Pregnancy Endpoints (SCOPE) study, a multicenter international study evaluating diseases of pregnancy in primigravidous women.28 The Cork BASELINE Birth Cohort study recruited from August 2009 through to October 2011. Infants recruited from July 2009 had skin barrier assessment at birth and throughout early life; these 1903 infants are included in this study. A total of 1303 infants were recruited antenatally (stream1). These women were subject to the inclusion criteria of the SCOPE study, namely, first-time, low-risk mothers with singleton pregnancies delivered at or near term. Consent for recruitment was sought at 20 weeks of gestation and confirmed at birth of the live baby. Cord blood samples were taken at birth and stored for future use. A second recruitment stream began in July 2010 that recruited mothers and babies on the postnatal ward (stream2). A total of 600 infants were recruited postnatally. These mothers were enrolled independently of the SCOPE study. The sole inclusion criterion in stream 2 was a healthy term infant on the postnatal ward. Stream 1 and stream 2 infants were assessed and followed up in an identical fashion at birth and thereafter clinic visits at 2, 6, and 12 months. All infants had assessment at birth, 2 months, 6 months, and 12 months involving parental questionnaires and physical assessment. Parental questionnaires at 2, 6, and 12 months contained specific screening questions for AD. Experienced health care personnel diagnosed AD at 6 and 12 months in accordance with the UK Working Party diagnostic criteria.29-31 If AD was present, severity was assessed at these timepoints using the SCORing Atopic Dermatitis (SCORAD) clinical tool.32,33 At 12 months, the Nottingham Severity Score was also recorded to assess the severity of AD.34 The Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints (BASELINE) Birth Cohort study is the first birth cohort study in Ireland.27 It was developed as the pediatric follow on from the Cork Centre for the Screening for Pregnancy Endpoints (SCOPE) study, a multicenter international study evaluating diseases of pregnancy in primigravidous women.28 The Cork BASELINE Birth Cohort study recruited from August 2009 through to October 2011. Infants recruited from July 2009 had skin barrier assessment at birth and throughout early life; these 1903 infants are included in this study. A total of 1303 infants were recruited antenatally (stream1). These women were subject to the inclusion criteria of the SCOPE study, namely, first-time, low-risk mothers with singleton pregnancies delivered at or near term. Consent for recruitment was sought at 20 weeks of gestation and confirmed at birth of the live baby. Cord blood samples were taken at birth and stored for future use. A second recruitment stream began in July 2010 that recruited mothers and babies on the postnatal ward (stream2). A total of 600 infants were recruited postnatally. These mothers were enrolled independently of the SCOPE study. The sole inclusion criterion in stream 2 was a healthy term infant on the postnatal ward. Stream 1 and stream 2 infants were assessed and followed up in an identical fashion at birth and thereafter clinic visits at 2, 6, and 12 months. All infants had assessment at birth, 2 months, 6 months, and 12 months involving parental questionnaires and physical assessment. Parental questionnaires at 2, 6, and 12 months contained specific screening questions for AD. Experienced health care personnel diagnosed AD at 6 and 12 months in accordance with the UK Working Party diagnostic criteria.29-31 If AD was present, severity was assessed at these timepoints using the SCORing Atopic Dermatitis (SCORAD) clinical tool.32,33 At 12 months, the Nottingham Severity Score was also recorded to assess the severity of AD.34 TEWL measurements TEWL measurements were carried out using a widely validated open chamber system (Tewameter TM 300; Courage+Khazaka Electronic, Cologne, Germany). For newborns, TEWL was taken in the Cork University Maternity Hospital. The subject's arm was acclimated before measurement by exposing the arm in a nonenvironmentally controlled room for 10 minutes. This was typically done in the cot beside the mother's bed while an interview with the mother or parents was carried out. The infant was then brought to a windowless room in which both temperature and humidity were maintained constant by an air conditioning system. Temperature was set between 20°C and 25°C. Humidity was monitored by a manometer in the room and was maintained between 30% and 45%. TEWL was taken on the lower volar surface of the forearm by applying a probe to the exposed volar skin for approximately 15 seconds until the measurement was recorded. Three readings were taken, and the mean of the 3 readings was recorded. For TEWL readings at other timepoints, the same procedure was carried out at 2 months and 6 months in an environmentally controlled room in the Health Research Board Discovery Centre, the clinical research facility for children at Cork University Hospital. The parents were advised not to apply emollients to the infant's skin for 12 hours before the reading was taken. TEWL measurements were carried out using a widely validated open chamber system (Tewameter TM 300; Courage+Khazaka Electronic, Cologne, Germany). For newborns, TEWL was taken in the Cork University Maternity Hospital. The subject's arm was acclimated before measurement by exposing the arm in a nonenvironmentally controlled room for 10 minutes. This was typically done in the cot beside the mother's bed while an interview with the mother or parents was carried out. The infant was then brought to a windowless room in which both temperature and humidity were maintained constant by an air conditioning system. Temperature was set between 20°C and 25°C. Humidity was monitored by a manometer in the room and was maintained between 30% and 45%. TEWL was taken on the lower volar surface of the forearm by applying a probe to the exposed volar skin for approximately 15 seconds until the measurement was recorded. Three readings were taken, and the mean of the 3 readings was recorded. For TEWL readings at other timepoints, the same procedure was carried out at 2 months and 6 months in an environmentally controlled room in the Health Research Board Discovery Centre, the clinical research facility for children at Cork University Hospital. The parents were advised not to apply emollients to the infant's skin for 12 hours before the reading was taken. Statistical analysis A series of logistic regression (LR) models was used to estimate the factors at (1) 2 days and (2) 2 months that influence the diagnosis of AD at 12 months. The dependent variable that measures diagnosis at 12 months is AD “Yes” and is equal to 1, with AD “No” = 0. To examine the influence of each variable, with and without the presence of other significant predictors, we carried out both univariable and multivariable analyses. With regard to the latter, 3 models were produced. Model 1 included parental atopy, FLG mutation status, and presence of an itchy rash at 2 months as independent variables. Model 2 included parental atopy and TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3, respectively) as independent variables, but did not include FLG mutation status. This was done to determine the unique influence of TEWL, without the contribution of FLG mutations. Model 3 included all independent variables: parental atopy, TEWL test reading scores, and FLG mutation status. All models controlled for the variables “use of emollient before test reading,” “presence of an itchy rash,” “infant sex,” and “infant birthweight.” Receiver operating characteristic (ROC) curve output, which plots the true positives (sensitivity) vs false positives (1 − specificity), was then evaluated to compare the ability of the 3 models to produce a prognosis with high accuracy. Our rationale for variable selection was to include those predictors necessary for face validity but only if they were significant at a .05 level in the univariable analysis or if they altered the coefficient of the main variable by more than 10% in cases in which the main association was significant.35,36 A series of logistic regression (LR) models was used to estimate the factors at (1) 2 days and (2) 2 months that influence the diagnosis of AD at 12 months. The dependent variable that measures diagnosis at 12 months is AD “Yes” and is equal to 1, with AD “No” = 0. To examine the influence of each variable, with and without the presence of other significant predictors, we carried out both univariable and multivariable analyses. With regard to the latter, 3 models were produced. Model 1 included parental atopy, FLG mutation status, and presence of an itchy rash at 2 months as independent variables. Model 2 included parental atopy and TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3, respectively) as independent variables, but did not include FLG mutation status. This was done to determine the unique influence of TEWL, without the contribution of FLG mutations. Model 3 included all independent variables: parental atopy, TEWL test reading scores, and FLG mutation status. All models controlled for the variables “use of emollient before test reading,” “presence of an itchy rash,” “infant sex,” and “infant birthweight.” Receiver operating characteristic (ROC) curve output, which plots the true positives (sensitivity) vs false positives (1 − specificity), was then evaluated to compare the ability of the 3 models to produce a prognosis with high accuracy. Our rationale for variable selection was to include those predictors necessary for face validity but only if they were significant at a .05 level in the univariable analysis or if they altered the coefficient of the main variable by more than 10% in cases in which the main association was significant.35,36 Study subjects: The Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints (BASELINE) Birth Cohort study is the first birth cohort study in Ireland.27 It was developed as the pediatric follow on from the Cork Centre for the Screening for Pregnancy Endpoints (SCOPE) study, a multicenter international study evaluating diseases of pregnancy in primigravidous women.28 The Cork BASELINE Birth Cohort study recruited from August 2009 through to October 2011. Infants recruited from July 2009 had skin barrier assessment at birth and throughout early life; these 1903 infants are included in this study. A total of 1303 infants were recruited antenatally (stream1). These women were subject to the inclusion criteria of the SCOPE study, namely, first-time, low-risk mothers with singleton pregnancies delivered at or near term. Consent for recruitment was sought at 20 weeks of gestation and confirmed at birth of the live baby. Cord blood samples were taken at birth and stored for future use. A second recruitment stream began in July 2010 that recruited mothers and babies on the postnatal ward (stream2). A total of 600 infants were recruited postnatally. These mothers were enrolled independently of the SCOPE study. The sole inclusion criterion in stream 2 was a healthy term infant on the postnatal ward. Stream 1 and stream 2 infants were assessed and followed up in an identical fashion at birth and thereafter clinic visits at 2, 6, and 12 months. All infants had assessment at birth, 2 months, 6 months, and 12 months involving parental questionnaires and physical assessment. Parental questionnaires at 2, 6, and 12 months contained specific screening questions for AD. Experienced health care personnel diagnosed AD at 6 and 12 months in accordance with the UK Working Party diagnostic criteria.29-31 If AD was present, severity was assessed at these timepoints using the SCORing Atopic Dermatitis (SCORAD) clinical tool.32,33 At 12 months, the Nottingham Severity Score was also recorded to assess the severity of AD.34 TEWL measurements: TEWL measurements were carried out using a widely validated open chamber system (Tewameter TM 300; Courage+Khazaka Electronic, Cologne, Germany). For newborns, TEWL was taken in the Cork University Maternity Hospital. The subject's arm was acclimated before measurement by exposing the arm in a nonenvironmentally controlled room for 10 minutes. This was typically done in the cot beside the mother's bed while an interview with the mother or parents was carried out. The infant was then brought to a windowless room in which both temperature and humidity were maintained constant by an air conditioning system. Temperature was set between 20°C and 25°C. Humidity was monitored by a manometer in the room and was maintained between 30% and 45%. TEWL was taken on the lower volar surface of the forearm by applying a probe to the exposed volar skin for approximately 15 seconds until the measurement was recorded. Three readings were taken, and the mean of the 3 readings was recorded. For TEWL readings at other timepoints, the same procedure was carried out at 2 months and 6 months in an environmentally controlled room in the Health Research Board Discovery Centre, the clinical research facility for children at Cork University Hospital. The parents were advised not to apply emollients to the infant's skin for 12 hours before the reading was taken. Statistical analysis: A series of logistic regression (LR) models was used to estimate the factors at (1) 2 days and (2) 2 months that influence the diagnosis of AD at 12 months. The dependent variable that measures diagnosis at 12 months is AD “Yes” and is equal to 1, with AD “No” = 0. To examine the influence of each variable, with and without the presence of other significant predictors, we carried out both univariable and multivariable analyses. With regard to the latter, 3 models were produced. Model 1 included parental atopy, FLG mutation status, and presence of an itchy rash at 2 months as independent variables. Model 2 included parental atopy and TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3, respectively) as independent variables, but did not include FLG mutation status. This was done to determine the unique influence of TEWL, without the contribution of FLG mutations. Model 3 included all independent variables: parental atopy, TEWL test reading scores, and FLG mutation status. All models controlled for the variables “use of emollient before test reading,” “presence of an itchy rash,” “infant sex,” and “infant birthweight.” Receiver operating characteristic (ROC) curve output, which plots the true positives (sensitivity) vs false positives (1 − specificity), was then evaluated to compare the ability of the 3 models to produce a prognosis with high accuracy. Our rationale for variable selection was to include those predictors necessary for face validity but only if they were significant at a .05 level in the univariable analysis or if they altered the coefficient of the main variable by more than 10% in cases in which the main association was significant.35,36 Results: A total of 1903 infants were enrolled onto the study. The demographic details of the population studied are presented in Table I. Filaggrin mutation typing Stream 1 infants had cord blood samples taken at birth. Infants without cord blood samples had Oragene saliva samples taken. Those infants still enrolled in the study at 2 years had EDTA blood samples taken. FLG genotyping was carried out in 1300 infants with available DNA. The cumulative FLG mutation rate was 10.46% (136 of 1300). Four infants were homozygous for FLG mutation, with the remainder heterozygous for FLG mutation (Table II). Stream 1 infants had cord blood samples taken at birth. Infants without cord blood samples had Oragene saliva samples taken. Those infants still enrolled in the study at 2 years had EDTA blood samples taken. FLG genotyping was carried out in 1300 infants with available DNA. The cumulative FLG mutation rate was 10.46% (136 of 1300). Four infants were homozygous for FLG mutation, with the remainder heterozygous for FLG mutation (Table II). TEWL through early infancy TEWL was taken in the early newborn period in 1691 of 1903 (88.86%) infants of the cohort. The mean TEWL for newborns was 7.32 ± 3.33 gwater/m2/h. Using univariable analysis, we found that there was no significant association between TEWL measurements at birth and sex, gestation, or postnatal age at measurement or recruitment stream 1 or 2 (data not shown). A total of 1614 of 1638 (98.5%) infants who attended the 2-month appointment had TEWL measured (98.5%). Mean 2-month TEWL was 10.97 ± 7.98 gwater/m2/h. A total of 1516 of 1537 (98.6%) infants who attended the 6-month appointment had TEWL measured. Mean 6-month TEWL was 10.71 ± 7.10 gwater/m2/h. TEWL measurements by quartiles are presented in Table III. TEWL was taken in the early newborn period in 1691 of 1903 (88.86%) infants of the cohort. The mean TEWL for newborns was 7.32 ± 3.33 gwater/m2/h. Using univariable analysis, we found that there was no significant association between TEWL measurements at birth and sex, gestation, or postnatal age at measurement or recruitment stream 1 or 2 (data not shown). A total of 1614 of 1638 (98.5%) infants who attended the 2-month appointment had TEWL measured (98.5%). Mean 2-month TEWL was 10.97 ± 7.98 gwater/m2/h. A total of 1516 of 1537 (98.6%) infants who attended the 6-month appointment had TEWL measured. Mean 6-month TEWL was 10.71 ± 7.10 gwater/m2/h. TEWL measurements by quartiles are presented in Table III. Clinical diagnosis of AD AD was screened for at 6- and 12-month appointments. This diagnosis of AD was made according to the UK Working Party diagnostic criteria. At 6 months, 18.7% (299 of 1597) of the infants screened were diagnosed with AD. A total of 287 had a SCORAD completed, with a mean SCORAD score of 21.54 ± 16.29 (range, 0-88). At 12 months, 15.53% (232 of 1494) of the infants screened were diagnosed with AD. The mean SCORAD score at 12 months in affected children was 18.56 ± 14.92 (range, 0-77). AD was screened for at 6- and 12-month appointments. This diagnosis of AD was made according to the UK Working Party diagnostic criteria. At 6 months, 18.7% (299 of 1597) of the infants screened were diagnosed with AD. A total of 287 had a SCORAD completed, with a mean SCORAD score of 21.54 ± 16.29 (range, 0-88). At 12 months, 15.53% (232 of 1494) of the infants screened were diagnosed with AD. The mean SCORAD score at 12 months in affected children was 18.56 ± 14.92 (range, 0-77). LR analysis: Factors influencing the development of AD at 1 year Our primary outcome was the presence of AD at 1 year. We developed a model that explored the relative contributions of FLG mutation status, parental atopy, presence of an itchy rash at 2 months, TEWL at day 2 and month 2, and ΔTEWL (day 2 to month 2 and day 2 to month 6) on AD at 1 year (Table IV). FLG mutation status was not associated with an elevated TEWL at birth. However, it was associated with an elevated TEWL at 2 and 6 months and with elevated ΔTEWL day 2 to 2 months and ΔTEWL day 2 to 6 months. Notably, FLG mutation carriers did not have elevated ΔTEWL 2 months to 6 months, implying that the major changes in skin barrier function in those who develop AD by 12 months start in the first 2 months of life. In our univariable modeling, the odds ratio (OR) for AD at 12 months conferred by both parents being atopic was 2.5 compared with an OR of 3.1 for an upper quartile TEWL at 2 months (see Table E1 in this article's Online Repository at www.jacionline.org). LR models were then used to examine the factors at birth and at 2 months that influence a diagnosis of AD at 12 months. We used 3 separate models to explore the relative contribution of parental atopy, FLG mutation status, and TEWL (at day 2 and at 2 months) toward the development of AD at 12 months. All models controlled for the variables use of emollient before test reading, presence of an itchy rash, infant sex, and infant birthweight. These covariates were not significant in any of the 3 multivariable models. Our primary outcome was the presence of AD at 1 year. We developed a model that explored the relative contributions of FLG mutation status, parental atopy, presence of an itchy rash at 2 months, TEWL at day 2 and month 2, and ΔTEWL (day 2 to month 2 and day 2 to month 6) on AD at 1 year (Table IV). FLG mutation status was not associated with an elevated TEWL at birth. However, it was associated with an elevated TEWL at 2 and 6 months and with elevated ΔTEWL day 2 to 2 months and ΔTEWL day 2 to 6 months. Notably, FLG mutation carriers did not have elevated ΔTEWL 2 months to 6 months, implying that the major changes in skin barrier function in those who develop AD by 12 months start in the first 2 months of life. In our univariable modeling, the odds ratio (OR) for AD at 12 months conferred by both parents being atopic was 2.5 compared with an OR of 3.1 for an upper quartile TEWL at 2 months (see Table E1 in this article's Online Repository at www.jacionline.org). LR models were then used to examine the factors at birth and at 2 months that influence a diagnosis of AD at 12 months. We used 3 separate models to explore the relative contribution of parental atopy, FLG mutation status, and TEWL (at day 2 and at 2 months) toward the development of AD at 12 months. All models controlled for the variables use of emollient before test reading, presence of an itchy rash, infant sex, and infant birthweight. These covariates were not significant in any of the 3 multivariable models. LR modeling of AD risk at 12 months including day 2 TEWL The LR model incorporating day 2 TEWL is summarized in Table V. In our first model, we included parental atopy and FLG mutation status as independent variables and excluded day 2 TEWL. Parental atopy and FLG mutation status significantly increase the likelihood of a positive diagnosis of AD at 12 months by 13.7 and 8.6 times, respectively, compared with infants without FLG mutation and without an atopic parent. The ROC, which plots the true positives (sensitivity) vs false positives (1 −specificity), is 0.8 for this model. Our second model includes day 2 TEWL test reading scores at 25th, 50th, and 75th percentiles as independent variables. We did not include FLG mutation status in this second model. The area under the ROC curve (AUC) remains at 0.8 for this model. Infants with a TEWL reading of 9.0 or above are 7.1 times more likely to be diagnosed with AD at 1 year than are infants with a reading below this point, controlling for all other variables in the model. This suggests that day 2 TEWL can be used as a sole indicator of likelihood of AD at 12 months. Our third model includes all independent variables, including both TEWL and FLG status. The LR model incorporating day 2 TEWL is summarized in Table V. In our first model, we included parental atopy and FLG mutation status as independent variables and excluded day 2 TEWL. Parental atopy and FLG mutation status significantly increase the likelihood of a positive diagnosis of AD at 12 months by 13.7 and 8.6 times, respectively, compared with infants without FLG mutation and without an atopic parent. The ROC, which plots the true positives (sensitivity) vs false positives (1 −specificity), is 0.8 for this model. Our second model includes day 2 TEWL test reading scores at 25th, 50th, and 75th percentiles as independent variables. We did not include FLG mutation status in this second model. The area under the ROC curve (AUC) remains at 0.8 for this model. Infants with a TEWL reading of 9.0 or above are 7.1 times more likely to be diagnosed with AD at 1 year than are infants with a reading below this point, controlling for all other variables in the model. This suggests that day 2 TEWL can be used as a sole indicator of likelihood of AD at 12 months. Our third model includes all independent variables, including both TEWL and FLG status. LR modeling of AD risk at 12 months including TEWL at 2 months We repeated the LR approach for month 2 TEWL values, summarized in Table VI. Again, we used 3 models. In the first model, we included parental atopy and FLG mutation status as independent variables but did not include TEWL. Parental atopy and FLG mutation carriage significantly increase the likelihood of a positive AD diagnosis at 12 months by 2.7 and 3.0 times, respectively, compared with infants without an FLG mutation and without an atopic parent. The ROC is 0.66 for this model. In the second model, we included the month 2 TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3 gwater/m2/h, respectively) as independent variables. We did not include FLG mutation status in this second model, as explained above. The AUC improved from 0.66 to 0.82 in this model. Infants with a TEWL reading of 12.3 or above were 5.6 times more likely to be diagnosed with AD at 12 months than were infants with a reading below this point, controlling for all other variables in the model. Model 3 included all independent variables. In a pattern similar to that found at 2 days, the AUC improved slightly to 0.84, although here again it is important to note that model 2 demonstrates that FLG status need not be measured to produce a prognosis with high accuracy. We modeled these in a traditional stepwise LR analysis. The significant finding was that after controlling for all other possible influencing factors, month 2 TEWL was the strongest independent predictor of AD at 12 months. We repeated the LR approach for month 2 TEWL values, summarized in Table VI. Again, we used 3 models. In the first model, we included parental atopy and FLG mutation status as independent variables but did not include TEWL. Parental atopy and FLG mutation carriage significantly increase the likelihood of a positive AD diagnosis at 12 months by 2.7 and 3.0 times, respectively, compared with infants without an FLG mutation and without an atopic parent. The ROC is 0.66 for this model. In the second model, we included the month 2 TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3 gwater/m2/h, respectively) as independent variables. We did not include FLG mutation status in this second model, as explained above. The AUC improved from 0.66 to 0.82 in this model. Infants with a TEWL reading of 12.3 or above were 5.6 times more likely to be diagnosed with AD at 12 months than were infants with a reading below this point, controlling for all other variables in the model. Model 3 included all independent variables. In a pattern similar to that found at 2 days, the AUC improved slightly to 0.84, although here again it is important to note that model 2 demonstrates that FLG status need not be measured to produce a prognosis with high accuracy. We modeled these in a traditional stepwise LR analysis. The significant finding was that after controlling for all other possible influencing factors, month 2 TEWL was the strongest independent predictor of AD at 12 months. Filaggrin mutation typing: Stream 1 infants had cord blood samples taken at birth. Infants without cord blood samples had Oragene saliva samples taken. Those infants still enrolled in the study at 2 years had EDTA blood samples taken. FLG genotyping was carried out in 1300 infants with available DNA. The cumulative FLG mutation rate was 10.46% (136 of 1300). Four infants were homozygous for FLG mutation, with the remainder heterozygous for FLG mutation (Table II). TEWL through early infancy: TEWL was taken in the early newborn period in 1691 of 1903 (88.86%) infants of the cohort. The mean TEWL for newborns was 7.32 ± 3.33 gwater/m2/h. Using univariable analysis, we found that there was no significant association between TEWL measurements at birth and sex, gestation, or postnatal age at measurement or recruitment stream 1 or 2 (data not shown). A total of 1614 of 1638 (98.5%) infants who attended the 2-month appointment had TEWL measured (98.5%). Mean 2-month TEWL was 10.97 ± 7.98 gwater/m2/h. A total of 1516 of 1537 (98.6%) infants who attended the 6-month appointment had TEWL measured. Mean 6-month TEWL was 10.71 ± 7.10 gwater/m2/h. TEWL measurements by quartiles are presented in Table III. Clinical diagnosis of AD: AD was screened for at 6- and 12-month appointments. This diagnosis of AD was made according to the UK Working Party diagnostic criteria. At 6 months, 18.7% (299 of 1597) of the infants screened were diagnosed with AD. A total of 287 had a SCORAD completed, with a mean SCORAD score of 21.54 ± 16.29 (range, 0-88). At 12 months, 15.53% (232 of 1494) of the infants screened were diagnosed with AD. The mean SCORAD score at 12 months in affected children was 18.56 ± 14.92 (range, 0-77). LR analysis: Factors influencing the development of AD at 1 year: Our primary outcome was the presence of AD at 1 year. We developed a model that explored the relative contributions of FLG mutation status, parental atopy, presence of an itchy rash at 2 months, TEWL at day 2 and month 2, and ΔTEWL (day 2 to month 2 and day 2 to month 6) on AD at 1 year (Table IV). FLG mutation status was not associated with an elevated TEWL at birth. However, it was associated with an elevated TEWL at 2 and 6 months and with elevated ΔTEWL day 2 to 2 months and ΔTEWL day 2 to 6 months. Notably, FLG mutation carriers did not have elevated ΔTEWL 2 months to 6 months, implying that the major changes in skin barrier function in those who develop AD by 12 months start in the first 2 months of life. In our univariable modeling, the odds ratio (OR) for AD at 12 months conferred by both parents being atopic was 2.5 compared with an OR of 3.1 for an upper quartile TEWL at 2 months (see Table E1 in this article's Online Repository at www.jacionline.org). LR models were then used to examine the factors at birth and at 2 months that influence a diagnosis of AD at 12 months. We used 3 separate models to explore the relative contribution of parental atopy, FLG mutation status, and TEWL (at day 2 and at 2 months) toward the development of AD at 12 months. All models controlled for the variables use of emollient before test reading, presence of an itchy rash, infant sex, and infant birthweight. These covariates were not significant in any of the 3 multivariable models. LR modeling of AD risk at 12 months including day 2 TEWL: The LR model incorporating day 2 TEWL is summarized in Table V. In our first model, we included parental atopy and FLG mutation status as independent variables and excluded day 2 TEWL. Parental atopy and FLG mutation status significantly increase the likelihood of a positive diagnosis of AD at 12 months by 13.7 and 8.6 times, respectively, compared with infants without FLG mutation and without an atopic parent. The ROC, which plots the true positives (sensitivity) vs false positives (1 −specificity), is 0.8 for this model. Our second model includes day 2 TEWL test reading scores at 25th, 50th, and 75th percentiles as independent variables. We did not include FLG mutation status in this second model. The area under the ROC curve (AUC) remains at 0.8 for this model. Infants with a TEWL reading of 9.0 or above are 7.1 times more likely to be diagnosed with AD at 1 year than are infants with a reading below this point, controlling for all other variables in the model. This suggests that day 2 TEWL can be used as a sole indicator of likelihood of AD at 12 months. Our third model includes all independent variables, including both TEWL and FLG status. LR modeling of AD risk at 12 months including TEWL at 2 months: We repeated the LR approach for month 2 TEWL values, summarized in Table VI. Again, we used 3 models. In the first model, we included parental atopy and FLG mutation status as independent variables but did not include TEWL. Parental atopy and FLG mutation carriage significantly increase the likelihood of a positive AD diagnosis at 12 months by 2.7 and 3.0 times, respectively, compared with infants without an FLG mutation and without an atopic parent. The ROC is 0.66 for this model. In the second model, we included the month 2 TEWL test reading scores at 25th, 50th, and 75th percentiles (7.0, 9.4, and 12.3 gwater/m2/h, respectively) as independent variables. We did not include FLG mutation status in this second model, as explained above. The AUC improved from 0.66 to 0.82 in this model. Infants with a TEWL reading of 12.3 or above were 5.6 times more likely to be diagnosed with AD at 12 months than were infants with a reading below this point, controlling for all other variables in the model. Model 3 included all independent variables. In a pattern similar to that found at 2 days, the AUC improved slightly to 0.84, although here again it is important to note that model 2 demonstrates that FLG status need not be measured to produce a prognosis with high accuracy. We modeled these in a traditional stepwise LR analysis. The significant finding was that after controlling for all other possible influencing factors, month 2 TEWL was the strongest independent predictor of AD at 12 months. Discussion: This study involves a large, unselected birth cohort study and is the first study of this scale to assess skin barrier function in the newborn period and early infancy. We have shown that changes in skin barrier predate clinical AD with a signal for barrier impairment detected in asymptomatic infants at day 2 and more markedly at 2 months. In our univariable modeling, the OR for AD at 12 months conferred by both parents being atopic was 2.5 compared with an OR of 3.1 for an upper quartile TEWL at 2 months (see Table E1), showing the additional value of an TEWL reading at this stage. These changes are seen in both high-risk and low-risk infants and, crucially, are independent of FLG status. Although in our multivariable analysis the AUC improves slightly to 0.83 when all available variables are included, importantly our second model (which excludes FLG genotyping) demonstrates that FLG status need not be measured to produce an accurate 12-month AD prognosis for babies with upper quartile day 2 TEWL readings. Some children will develop AD before 6 months. In our study, we did not formally diagnose AD at 2 months but all parents were asked about the presence of itchy rash on face or skin folds. These infants were designated as “itchy rash,” seen in 8% of all cases at 2 months. There are no widely recognized diagnostic criteria for AD at this age; however, we controlled for an itchy red rash at the 2-month visit in our model. AD was screened for in all participants at 6 and 12 months and formally diagnosed using the UK Working Party Criteria. When compared with other studies in the field, our study shows the earliest timepoint for a signal for impaired TEWL in asymptomatic infants: at day 2. A further, stronger signal is seen at 2 months. Previous studies, of smaller scale have shown a relationship between FLG mutation status and TEWL.19 However, our study is much larger, by several orders of magnitude, and is not selected for high risk of atopy. In our study, 30% of FLG mutation carriers developed AD by age 12 months and 70% did not, consistent with large previous studies. At birth, there was no difference in mean TEWL reading between FLG mutation and FLG wild-type groups (7.33 ± 3.62 gwater/m2/h vs 7.3 ± 3.38 gwater/m2/h). However, by 2 months, FLG mutation carrying infants have a significantly higher mean TEWL than do FLG wild-type infants. This change persisted at 6 months. Selecting out children in the upper quartile regardless of FLG status allows for a noninvasive, discriminating, and targeted means of prediction of AD in the first days of life, allowing for possible preventive measures to be put in place. Conversely, selecting out infants in the lower quartile of TEWL, which has a protective effect in relation to AD at 12 months, could be an important piece of positive information for their families and would reduce the “number needed to treat” in an intervention study of potential protective interventions. Detection of an increased TEWL as early as day 2 or at 2 months, before developing clinical features of AD, is a novel and unique finding in a general pediatric population. Accurate identification of individuals at high risk for AD, using family history in combination with a noninvasive measurement, with the optional addition of FLG mutation status has great potential for intervention studies. Stratified interventions could be made on the basis of these 2 or 3 variables. Pilot studies for the primary prevention of AD by use of liberal emollients have shown the potential of this approach.24 We believe that the novel findings in this study will facilitate and inform stratification of future intervention studies. A lowest quartile TEWL at birth is protective against AD. A highest quartile TEWL at 2 days and at 2 months is strongly associated with increased prevalence of AD at 12 months. These changes predate the development of clinically apparent AD. The mechanistic implications of these observations are that abnormalities in the skin barrier predate symptomatic or clinically detectable AD. These changes occur very early in the postnatal environment after transition from the intrauterine aqueous environment to the xerotic postnatal environment. It is notable that this effect is not dependent on FLG loss-of-function mutations but is enhanced by and is interactive with these genetic factors and presumably with environmental influences encountered postnatally. TEWL measurement in early life is therefore an effective tool to detect infants at risk of AD, especially when combined in a model with parental history of atopy and/or FLG mutation status. The obvious next step would be to assess whether intervention studies between birth and age 2 months in those with raised TEWL at birth would be effective in maintaining the skin barrier and reducing the incidence of AD.Clinical implicationsA signal for the development of AD is seen at 2 days and at 2 months in asymptomatic infants. Interventions to potentially prevent AD could be targeted toward such infants. A signal for the development of AD is seen at 2 days and at 2 months in asymptomatic infants. Interventions to potentially prevent AD could be targeted toward such infants.

Background: Loss-of-function mutations in the skin barrier protein filaggrin (FLG) are a major risk for atopic dermatitis (AD). The pathogenic sequence of disturbances in skin barrier function before or during the early development of AD is not fully understood. A more detailed understanding of these events is needed to develop a clearer picture of disease pathogenesis. A robust, noninvasive test to identify babies at high risk of AD would be important in planning early intervention and/or prevention studies. Methods: A total of 1903 infants were enrolled in the Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints Birth Cohort study from July 2009 to October 2011. Measurements of TEWL were made at birth (day 2) and at 2 and 6 months. The presence of AD was ascertained at 6 and 12 months, and disease severity was assessed by using the SCORing Atopic Dermatitis clinical tool at 6 months and by using both the SCORing Atopic Dermatitis clinical tool and Nottingham Severity Score at 12 months. A total of 1300 infants were genotyped for FLG mutations. Results: At 6 months, 18.7% of the children had AD, and at 12 months, 15.53%. In a logistic regression model, day 2 upper quartile TEWL measurement was significantly predictive of AD at 12 months (area under the receiver operating characteristic curve, 0.81; P < .05). Lowest quartile day 2 TEWL was protective against AD at 12 months. An upper quartile 2 month TEWL was also strongly predictive of AD at 12 months (area under the receiver operating characteristic curve, 0.84; P < .05). At both ages, this effect was independent of parental atopy, FLG status, or report of an itchy flexural rash at 2 months. Associations were increased when parental atopy status or child FLG mutation status was added into the linear regression model. Conclusions: Impairment of skin barrier function at birth and at 2 months precedes clinical AD. In addition to providing important mechanistic insights into disease pathogenesis, these findings have implications for the optimal timing of interventions for the prevention of AD.

[CONTENT] Infant | skin barrier | TEWL | atopic dermatitis | filaggrin | predictor | biomarker | AD, Atopic dermatitis | AUC, Area under the ROC curve | BASELINE, Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints | FLG, Filaggrin gene | LR, Logistic regression | OR, Odds ratio | ROC, Receiver operating characteristic | SCOPE, Screening for Pregnancy Endpoints | TEWL, Transepidermal water loss | ΔTEWL, Change in TEWL measurement between one timepoint and another [SUMMARY]

[CONTENT] Infant | skin barrier | TEWL | atopic dermatitis | filaggrin | predictor | biomarker | AD, Atopic dermatitis | AUC, Area under the ROC curve | BASELINE, Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints | FLG, Filaggrin gene | LR, Logistic regression | OR, Odds ratio | ROC, Receiver operating characteristic | SCOPE, Screening for Pregnancy Endpoints | TEWL, Transepidermal water loss | ΔTEWL, Change in TEWL measurement between one timepoint and another [SUMMARY]

[CONTENT] Infant | skin barrier | TEWL | atopic dermatitis | filaggrin | predictor | biomarker | AD, Atopic dermatitis | AUC, Area under the ROC curve | BASELINE, Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional Endpoints | FLG, Filaggrin gene | LR, Logistic regression | OR, Odds ratio | ROC, Receiver operating characteristic | SCOPE, Screening for Pregnancy Endpoints | TEWL, Transepidermal water loss | ΔTEWL, Change in TEWL measurement between one timepoint and another [SUMMARY]

[CONTENT] DNA Mutational Analysis | Dermatitis, Atopic | Female | Filaggrin Proteins | Genotype | Humans | Infant | Infant, Newborn | Intermediate Filament Proteins | Male | Prognosis | Risk | Skin | Time Factors [SUMMARY]

[CONTENT] DNA Mutational Analysis | Dermatitis, Atopic | Female | Filaggrin Proteins | Genotype | Humans | Infant | Infant, Newborn | Intermediate Filament Proteins | Male | Prognosis | Risk | Skin | Time Factors [SUMMARY]

[CONTENT] DNA Mutational Analysis | Dermatitis, Atopic | Female | Filaggrin Proteins | Genotype | Humans | Infant | Infant, Newborn | Intermediate Filament Proteins | Male | Prognosis | Risk | Skin | Time Factors [SUMMARY]

[CONTENT] term infant postnatal | cork university maternity | birth cohort study | cork baseline birth | infants included study [SUMMARY]

[CONTENT] term infant postnatal | cork university maternity | birth cohort study | cork baseline birth | infants included study [SUMMARY]

[CONTENT] term infant postnatal | cork university maternity | birth cohort study | cork baseline birth | infants included study [SUMMARY]

[CONTENT] months | tewl | ad | flg | 12 | infants | 12 months | model | mutation | flg mutation [SUMMARY]

[CONTENT] months | tewl | ad | flg | 12 | infants | 12 months | model | mutation | flg mutation [SUMMARY]

[CONTENT] months | tewl | ad | flg | 12 | infants | 12 months | model | mutation | flg mutation [SUMMARY]

[CONTENT] months | study | recruited | birth | cork | 12 | room | variable | scope | tewl [SUMMARY]

[CONTENT] tewl | months | model | flg | ad | mutation | infants | day | flg mutation | month [SUMMARY]

[CONTENT] months | tewl | ad | flg | infants | model | 12 | mutation | flg mutation | 12 months [SUMMARY]

[CONTENT] 1903 | the Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional | July 2009 to October 2011 ||| TEWL | day 2 | 2 and 6 months ||| 6 and 12 months | Atopic Dermatitis | 6 months | Atopic Dermatitis | Nottingham Severity Score | 12 months ||| 1300 | FLG [SUMMARY]

[CONTENT] 6 months | 18.7% | 12 months | 15.53% ||| day 2 | TEWL | 12 months | 0.81 ||| day 2 | TEWL | AD | 12 months ||| 2 month | TEWL | 12 months | 0.84 ||| FLG | 2 months ||| FLG [SUMMARY]

[CONTENT] FLG ||| ||| ||| ||| 1903 | the Cork Babies After Scope: Evaluating the Longitudinal Impact Using Neurological and Nutritional | July 2009 to October 2011 ||| TEWL | day 2 | 2 and 6 months ||| 6 and 12 months | Atopic Dermatitis | 6 months | Atopic Dermatitis | Nottingham Severity Score | 12 months ||| 1300 | FLG ||| ||| 6 months | 18.7% | 12 months | 15.53% ||| day 2 | TEWL | 12 months | 0.81 ||| day 2 | TEWL | AD | 12 months ||| 2 month | TEWL | 12 months | 0.84 ||| FLG | 2 months ||| FLG ||| 2 months ||| [SUMMARY]

In vitro antioxidant and anticancer effects of solvent fractions from Prunella vulgaris var. lilacina.

24206840

Recently, considerable attention has been focused on exploring the potential antioxidant properties of plant extracts or isolated products of plant origin. Prunella vulgaris var. lilacina is widely distributed in Korea, Japan, China, and Europe, and it continues to be used to treat inflammation, eye pain, headache, and dizziness. However, reports on the antioxidant activities of P. vulgaris var. lilacina are limited, particularly concerning the relationship between its phenolic content and antioxidant capacity. In this study, we investigated the antioxidant and anticancer activities of an ethanol extract from P. vulgaris var. lilacina and its fractions.

BACKGROUND

Dried powder of P. vulgaris var. lilacina was extracted with ethanol, and the extract was fractionated to produce the hexane fraction, butanol fraction, chloroform fraction and residual water fraction. The phenolic content was assayed using the Folin-Ciocalteu colorimetric method. Subsequently, the antioxidant activities of the ethanol extract and its fractions were analyzed employing various antioxidant assay methods including DPPH, FRAP, ABTS, SOD activity and production of reactive oxygen species. Additionally, the extract and fractions were assayed for their ability to exert cytotoxic activities on various cancer cells using the MTT assay. We also investigated the expression of genes associated with apoptotic cell death by RT-PCR.

METHODS

The total phenolic contents of the ethanol extract and water fraction of P. vulgaris var. lilacina were 303.66 and 322.80 mg GAE/g dry weight (or fractions), respectively. The results showed that the ethanol extract and the water fraction of P. vulgaris var. lilacina had higher antioxidant content than other solvent fractions, similar to their total phenolic content. Anticancer activity was also tested using the HepG2, HT29, A549, MKN45 and HeLa cancer cell lines. The results clearly demonstrated that the P. vulgaris var. lilacina ethanol extract induced significant cytotoxic effects on the various cancer cell lines, and these effects were stronger than those induced by the P. vulgaris var. lilacina solvent fractions. We also investigated the expression of genes associated with apoptotic cell death. We confirmed that the P. vulgaris var. lilacina ethanol extract and water fraction significantly increased the expression of p53, Bax and Fas.

RESULTS

These results suggest that the ethanol extract from P. vulgaris var. lilacina and its fractions could be applied as natural sources of antioxidants and anticancer activities in food and in the pharmaceutical industry.

CONCLUSIONS

4226201

Background

Oxidative stress is caused by reactive oxygen species (ROS), which are associated with many pathological disorders such as atherosclerosis, diabetes, ageing and cancer [1-3]. In order to protect human beings against oxidative damage, synthetic antioxidants such as BHA and BHT were created due to demand [4]. However, there has been concern regarding the toxicity and carcinogenic effects of synthetic antioxidants [5,6]. Thus, it is important to identify new sources of safe and inexpensive antioxidants of natural origin. Natural antioxidants, especially plant phenolics, flavonoids, tannins and anthocyanidins, are safe and are also bioactive [7]. Therefore, in recent years, considerable attention has been focused on exploring the potential antioxidant properties of plant extracts or isolated products of plant origin [8]. Prunella vulgaris var. lilacina is widely distributed in Korea, Japan, China, and Europe, and it continues to be used to treat inflammation, eye pain, headache, and dizziness [9,10]. It is rich in active compounds known to significantly affect human health, such as triterpenoid, rosmarinic acid, hyperoside, ursolic acid, and flavonoids [11-16]. Furthermore, P. vulgaris var. lilacina has been shown to have anti-allergic, anti-inflammatory, anti-oxidative, anti-microbial, and anti-viral effects [17-19]. However, reports on the antioxidant activities of P. vulgaris var. lilacina are limited, particularly concerning the relationship between its phenolic content and antioxidant capacity. Therefore, the aims of this study were to identify new sources of antioxidants from P. vulgaris var. lilacina extract. Additionally, the effects of the extraction solvent (70% ethanol, hexane, butanol, chloroform, or water) on the total phenolic content and antioxidant activities of P. vulgaris var. lilacina were investigated.

Methods

Reagents The reagents 1,1-diphenyl-1-picrylhydrazyl (DPPH), 2,2′-azinibis 3-ethyl benzothiazoline-6-sulfonic acid (ABTS), α-tocopherol, 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ), iron(III) chloride hexahydrate, gallic acid, Folin and Ciocalteu’s phenol reagent, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and lipopolysaccharides (LPS) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Iron(II) sulfate heptahydrate and acetic acid were purchased from Junsei (Tokyo, Japan). A superoxide dismutase-WST kit was purchased from Dojindo (Kumamoto, Japan). Dulbecco’s Modified Eagle’s Medium (DMEM), RPMI 1640 medium, fetal bovine serum (FBS), and penicillin–streptomycin were obtained from Invitrogen (Carlsbad, CA, USA). The reagents 1,1-diphenyl-1-picrylhydrazyl (DPPH), 2,2′-azinibis 3-ethyl benzothiazoline-6-sulfonic acid (ABTS), α-tocopherol, 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ), iron(III) chloride hexahydrate, gallic acid, Folin and Ciocalteu’s phenol reagent, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and lipopolysaccharides (LPS) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Iron(II) sulfate heptahydrate and acetic acid were purchased from Junsei (Tokyo, Japan). A superoxide dismutase-WST kit was purchased from Dojindo (Kumamoto, Japan). Dulbecco’s Modified Eagle’s Medium (DMEM), RPMI 1640 medium, fetal bovine serum (FBS), and penicillin–streptomycin were obtained from Invitrogen (Carlsbad, CA, USA). Sample preparation and extraction Whole plants of P. vulgaris var. lilacina were purchased from the Plant Extract Bank (#007-017, Dae-Jeon, Korea). Dried P. vulgaris var. lilacina was milled into powder of 80-mesh particle size and stored at -70°C. The dried P. vulgaris var. lilacina was extracted three times with 70% ethanol. The 70% ethanol extract powder (10 g) was suspended in 500 mL of distilled water and extracted with 500 mL of the following solvents in a stepwise manner: hexane, chloroform, and butanol. Each fraction was filtered through Whatman filter paper No. 2 (Advantec, Tokyo, Japan). Subsequently, the filtrates were combined and evaporated under a vacuum and then lyophilized with a freeze dryer (Ilshine Lab, Suwon, Korea) at -70°C under reduced pressure (< 20 Pa). The dry residue was stored at -20°C. For further analysis, we reconstituted the dry extract and fractions with DMSO. Whole plants of P. vulgaris var. lilacina were purchased from the Plant Extract Bank (#007-017, Dae-Jeon, Korea). Dried P. vulgaris var. lilacina was milled into powder of 80-mesh particle size and stored at -70°C. The dried P. vulgaris var. lilacina was extracted three times with 70% ethanol. The 70% ethanol extract powder (10 g) was suspended in 500 mL of distilled water and extracted with 500 mL of the following solvents in a stepwise manner: hexane, chloroform, and butanol. Each fraction was filtered through Whatman filter paper No. 2 (Advantec, Tokyo, Japan). Subsequently, the filtrates were combined and evaporated under a vacuum and then lyophilized with a freeze dryer (Ilshine Lab, Suwon, Korea) at -70°C under reduced pressure (< 20 Pa). The dry residue was stored at -20°C. For further analysis, we reconstituted the dry extract and fractions with DMSO. Total phenolic content The total phenolic contents of P. vulgaris var. lilacina extract and its fractions were determined using the Folin-Ciocalteu method [20]. The extract and each fraction were oxidized with Folin–Ciocalteu’s reagents, and then, the reaction was neutralized with 10% sodium carbonate. After incubation at room temperature for 1 h, the absorbance of the reaction mixture was measured at 725 nm using a microplate reader (Molecular Devices, Sunnyvale, CA, USA). Quantification was performed based on a standard curve with gallic acid. Results were expressed as milligrams gallic acid equivalent (GAE) per gram of dry weight of extract (or fractions). The total phenolic contents of P. vulgaris var. lilacina extract and its fractions were determined using the Folin-Ciocalteu method [20]. The extract and each fraction were oxidized with Folin–Ciocalteu’s reagents, and then, the reaction was neutralized with 10% sodium carbonate. After incubation at room temperature for 1 h, the absorbance of the reaction mixture was measured at 725 nm using a microplate reader (Molecular Devices, Sunnyvale, CA, USA). Quantification was performed based on a standard curve with gallic acid. Results were expressed as milligrams gallic acid equivalent (GAE) per gram of dry weight of extract (or fractions). DPPH radical scavenging activity Analysis of DPPH radical-scavenging activity was carried out according to the Blois method [21]. 0.3 mM DPPH was added to each sample. After incubation for 30 min in the dark at room temperature, the absorbance was measured at 518 nm using a microplate reader. α-Tocopherol was used as a positive control. Percent reduction of the DPPH radical was calculated in the following way: inhibition concentration (%) = 100 - (Asample/Acontrol) × 100, where Acontrol is the absorbance of the control reaction (containing all reagents except the test sample), and Asample is the absorbance of the test sample. Tests were carried out in triplicate. For the final results, RC50 values (the concentrations required for 50% reduction of DPPH by 30 min after starting the reaction) were calculated from the absorbance diminished by 50%. The experiment was performed in triplicate. Analysis of DPPH radical-scavenging activity was carried out according to the Blois method [21]. 0.3 mM DPPH was added to each sample. After incubation for 30 min in the dark at room temperature, the absorbance was measured at 518 nm using a microplate reader. α-Tocopherol was used as a positive control. Percent reduction of the DPPH radical was calculated in the following way: inhibition concentration (%) = 100 - (Asample/Acontrol) × 100, where Acontrol is the absorbance of the control reaction (containing all reagents except the test sample), and Asample is the absorbance of the test sample. Tests were carried out in triplicate. For the final results, RC50 values (the concentrations required for 50% reduction of DPPH by 30 min after starting the reaction) were calculated from the absorbance diminished by 50%. The experiment was performed in triplicate. Ferric-reducing antioxidant power (FRAP) activity FRAP activity was determined using manual assay methods [22]. The working fluid was freshly prepared by mixing acetate buffer (300 mM, pH 3.6) with TPTZ in HCl and iron (III) chloride hexahydrate. Each sample solution or α-tocopherol was added to 3 mL of working fluid, and the mixture was left for 4 min at room temperature. The absorbance was measured at 593 nm. The results were expressed as iron (II) sulfate heptahydrate (FeSO4) equivalents. FRAP activity was determined using manual assay methods [22]. The working fluid was freshly prepared by mixing acetate buffer (300 mM, pH 3.6) with TPTZ in HCl and iron (III) chloride hexahydrate. Each sample solution or α-tocopherol was added to 3 mL of working fluid, and the mixture was left for 4 min at room temperature. The absorbance was measured at 593 nm. The results were expressed as iron (II) sulfate heptahydrate (FeSO4) equivalents. ABTS radical cation scavenging activity The ABTS assay was based on the ability of different fractions to scavenge the ABTS radical cation in comparison to a standard (α-tocopherol) [23]. The radical cation was prepared by mixing 7 mM ABTS with 2.45 mM potassium persulfate (1:1 v/v) and leaving the mixture for 24 h until the reaction was completed and the absorbance was stable. The ABTS radical solution was diluted with PBS to an absorbance of 0.7 ± 0.02 at 732 nm. The photometric assay was conducted with 180 μL of ABTS radical solution and 20 μL of samples; measurements were taken at 732 nm after 1 min. The antioxidative activity of the tested samples was calculated by determining the decrease in absorbance. The free radical scavenging capacity was expressed by RC50. The ABTS assay was based on the ability of different fractions to scavenge the ABTS radical cation in comparison to a standard (α-tocopherol) [23]. The radical cation was prepared by mixing 7 mM ABTS with 2.45 mM potassium persulfate (1:1 v/v) and leaving the mixture for 24 h until the reaction was completed and the absorbance was stable. The ABTS radical solution was diluted with PBS to an absorbance of 0.7 ± 0.02 at 732 nm. The photometric assay was conducted with 180 μL of ABTS radical solution and 20 μL of samples; measurements were taken at 732 nm after 1 min. The antioxidative activity of the tested samples was calculated by determining the decrease in absorbance. The free radical scavenging capacity was expressed by RC50. Superoxide dismutase (SOD) activity Superoxide dismutase activity was determined using the highly water-soluble tetrazolium salt WST-1, which produces a water-soluble formazan dye upon reduction with a superoxide anion. SOD activity was determined using an SOD assay kit (Dojindo, Kumamoto, Japan) in accordance with the manufacturer’s instructions. Briefly, WST working solution was made by diluting 1 mL of WST solution into 19 ml of buffer solution. Enzyme working solution was made after the enzyme solution tube was centrifuged for 5 sec. Fifteen microliters of enzyme solution were diluted with 2.5 mL of dilution buffer. SOD activity was expressed as the percentage of inhibition rate. Superoxide dismutase activity was determined using the highly water-soluble tetrazolium salt WST-1, which produces a water-soluble formazan dye upon reduction with a superoxide anion. SOD activity was determined using an SOD assay kit (Dojindo, Kumamoto, Japan) in accordance with the manufacturer’s instructions. Briefly, WST working solution was made by diluting 1 mL of WST solution into 19 ml of buffer solution. Enzyme working solution was made after the enzyme solution tube was centrifuged for 5 sec. Fifteen microliters of enzyme solution were diluted with 2.5 mL of dilution buffer. SOD activity was expressed as the percentage of inhibition rate. Cells and culture The mouse macrophage cell line RAW264.7, human liver cancer cell line HepG2, human colon cancer cell line HT29, human lung cancer cell line A549, human stomach cancer cell line MKN-45 and human cervical cancer cell line HeLa were purchased from the Korean Cell Line Bank (Seoul, Korea). The cell lines were grown in RPMI 1640 medium or DMEM with 10% FBS and 1% penicillin-streptomycin and incubated at 37°C in 5% CO2. The mouse macrophage cell line RAW264.7, human liver cancer cell line HepG2, human colon cancer cell line HT29, human lung cancer cell line A549, human stomach cancer cell line MKN-45 and human cervical cancer cell line HeLa were purchased from the Korean Cell Line Bank (Seoul, Korea). The cell lines were grown in RPMI 1640 medium or DMEM with 10% FBS and 1% penicillin-streptomycin and incubated at 37°C in 5% CO2. Cell cytotoxicity assay Exponentially growing cells were collected and plated at 5 × 103 - 1 × 104 cells/well. P. vulgaris var. lilacina ethanol extract and its solvent fractions in DMSO were diluted in PBS to obtain final concentrations of 10, 50 and 100 μg/mL. Cells were treated with samples for 24 h, and MTT solution was added. After 4 h, the media was removed, and DMSO was added to each well. The resulting absorbance was measured at 540 nm [24]. Exponentially growing cells were collected and plated at 5 × 103 - 1 × 104 cells/well. P. vulgaris var. lilacina ethanol extract and its solvent fractions in DMSO were diluted in PBS to obtain final concentrations of 10, 50 and 100 μg/mL. Cells were treated with samples for 24 h, and MTT solution was added. After 4 h, the media was removed, and DMSO was added to each well. The resulting absorbance was measured at 540 nm [24]. Intracellular reactive oxygen species (ROS) scavenging activity For microscopic detection of ROS formation, RAW264.7 cells were grown to 80% confluence in six-well plates and treated with samples for 24 h. After incubation, cells were incubated with dichlorofluorescein diacetate (DCF-DA) (25 μM) for 30 min at 37°C in the dark. After several washings with PBS, cells were observed with a fluorescence microscope (Carl ZEISS, Oberkochen, Germany). For microscopic detection of ROS formation, RAW264.7 cells were grown to 80% confluence in six-well plates and treated with samples for 24 h. After incubation, cells were incubated with dichlorofluorescein diacetate (DCF-DA) (25 μM) for 30 min at 37°C in the dark. After several washings with PBS, cells were observed with a fluorescence microscope (Carl ZEISS, Oberkochen, Germany). Real-time reverse transcription polymerase chain reaction analysis (RT-PCR) To determine the expression levels of p53, Bax, Bcl-2 and Fas, RT-PCR was performed using a Qiagen Rotor-Gene Q real-time thermal cycler (Valencia, CA, USA) in accordance with the manufacturer’s instructions. The cells were treated with P. vulgaris var. lilacina extracts and cultured for 24 h. Thereafter, cDNA was synthesized from the total RNA isolated from cells. The PCR reaction was performed using 2× SYBR Green mix (Qiagen, Valencia, CA, USA). All results were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression. The following primer sequences were used for the real-time RT-PCR: GAPDH, 5′-CGG AGT CAA CGG ATT TGG TCG TAT-3′ (forward), 5′-AGC CTT CTC CAT GGT GGT GAA GAC-3′ (reverse); p53, 5′-GCT CTG ACT GTA CCA CCA TCC-3′ (forward), 5′-CTC TCG GAA CAT CTC GAA GCG-3′ (reverse); Bax, 5′-ATG GAC GGG TCC GGG GAG-3′ (forward), 5′-TCA GCC CAT CTT CTT CCA-3′ (reverse); Bcl-2, 5′-CAG CTG CAC CTG ACG-3′ (forward), 5′-ATG CAC CTA CCC AGC-3′ (reverse); Fas, 5′- TCT AAC TTG GGG TGG CTT TGT CTT C -3′ (forward), 5′- GTG TCA TAC GCT TTC TTT CCA T-3′ (reverse). To determine the expression levels of p53, Bax, Bcl-2 and Fas, RT-PCR was performed using a Qiagen Rotor-Gene Q real-time thermal cycler (Valencia, CA, USA) in accordance with the manufacturer’s instructions. The cells were treated with P. vulgaris var. lilacina extracts and cultured for 24 h. Thereafter, cDNA was synthesized from the total RNA isolated from cells. The PCR reaction was performed using 2× SYBR Green mix (Qiagen, Valencia, CA, USA). All results were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression. The following primer sequences were used for the real-time RT-PCR: GAPDH, 5′-CGG AGT CAA CGG ATT TGG TCG TAT-3′ (forward), 5′-AGC CTT CTC CAT GGT GGT GAA GAC-3′ (reverse); p53, 5′-GCT CTG ACT GTA CCA CCA TCC-3′ (forward), 5′-CTC TCG GAA CAT CTC GAA GCG-3′ (reverse); Bax, 5′-ATG GAC GGG TCC GGG GAG-3′ (forward), 5′-TCA GCC CAT CTT CTT CCA-3′ (reverse); Bcl-2, 5′-CAG CTG CAC CTG ACG-3′ (forward), 5′-ATG CAC CTA CCC AGC-3′ (reverse); Fas, 5′- TCT AAC TTG GGG TGG CTT TGT CTT C -3′ (forward), 5′- GTG TCA TAC GCT TTC TTT CCA T-3′ (reverse). Gas chromatography-mass spectrum analysis (GC-MS) GC-MS analysis was carried out using an Agilent 6890 gas chromatograph equipped with a DB-5 ms capillary column (60 m × 0.25 mm; coating thickness 1.4 μm) and an Agilent 5975 MSD detector (Loveland, CO, USA). Analytical conditions were as follows: injector and transfer line temperatures of 250°C; oven temperature was programmed from 50°C to 150°C at 10°C/min, from 150°C to 200°C at 7°C/min, and from 200°C to 250°C at 5°C/min; carrier gas helium at 1 mL/min; and split ratio 1:10. Identification of the constituents was based on comparison of the retention times with those of authentic samples. GC-MS analysis was carried out using an Agilent 6890 gas chromatograph equipped with a DB-5 ms capillary column (60 m × 0.25 mm; coating thickness 1.4 μm) and an Agilent 5975 MSD detector (Loveland, CO, USA). Analytical conditions were as follows: injector and transfer line temperatures of 250°C; oven temperature was programmed from 50°C to 150°C at 10°C/min, from 150°C to 200°C at 7°C/min, and from 200°C to 250°C at 5°C/min; carrier gas helium at 1 mL/min; and split ratio 1:10. Identification of the constituents was based on comparison of the retention times with those of authentic samples. Statistical analysis Statistical analysis was performed with SPSS (version 17.0; SPSS Inc., Chicago, IL, USA). Descriptive statistics were used to calculate the mean and standard error of the mean (SEM). One-way analysis of variance was performed, and when the significance (p < 0.05) was determined, the differences of the mean values were identified using Duncan’s multiple range tests. Statistical analysis was performed with SPSS (version 17.0; SPSS Inc., Chicago, IL, USA). Descriptive statistics were used to calculate the mean and standard error of the mean (SEM). One-way analysis of variance was performed, and when the significance (p < 0.05) was determined, the differences of the mean values were identified using Duncan’s multiple range tests.

Conclusions

The present study determined that P. vulgaris var. lilacina extract and its fractions have strong antioxidant and anticancer activities in vitro. The correlation coefficients between antioxidant capacity and the phenolic content were very strong, and phenolic compounds were a major contributor to the antioxidant capacities of P. vulgaris var. lilacina. In addition, we confirmed the presence of 7 major components of P. vulgaris var. lilacina. However, further studies to study the mechanisms of these compounds and find the root of their antioxidative and anticancer activity are underway. On the basis of these results, P. vulgaris var. lilacina appears to be a good source of natural antioxidant and anticancer agents and could be of significance in the food industry and for the control of various human and animal diseases.

Background: Oxidative stress is caused by reactive oxygen species (ROS), which are associated with many pathological disorders such as atherosclerosis, diabetes, ageing and cancer [1-3]. In order to protect human beings against oxidative damage, synthetic antioxidants such as BHA and BHT were created due to demand [4]. However, there has been concern regarding the toxicity and carcinogenic effects of synthetic antioxidants [5,6]. Thus, it is important to identify new sources of safe and inexpensive antioxidants of natural origin. Natural antioxidants, especially plant phenolics, flavonoids, tannins and anthocyanidins, are safe and are also bioactive [7]. Therefore, in recent years, considerable attention has been focused on exploring the potential antioxidant properties of plant extracts or isolated products of plant origin [8]. Prunella vulgaris var. lilacina is widely distributed in Korea, Japan, China, and Europe, and it continues to be used to treat inflammation, eye pain, headache, and dizziness [9,10]. It is rich in active compounds known to significantly affect human health, such as triterpenoid, rosmarinic acid, hyperoside, ursolic acid, and flavonoids [11-16]. Furthermore, P. vulgaris var. lilacina has been shown to have anti-allergic, anti-inflammatory, anti-oxidative, anti-microbial, and anti-viral effects [17-19]. However, reports on the antioxidant activities of P. vulgaris var. lilacina are limited, particularly concerning the relationship between its phenolic content and antioxidant capacity. Therefore, the aims of this study were to identify new sources of antioxidants from P. vulgaris var. lilacina extract. Additionally, the effects of the extraction solvent (70% ethanol, hexane, butanol, chloroform, or water) on the total phenolic content and antioxidant activities of P. vulgaris var. lilacina were investigated. Methods: Reagents The reagents 1,1-diphenyl-1-picrylhydrazyl (DPPH), 2,2′-azinibis 3-ethyl benzothiazoline-6-sulfonic acid (ABTS), α-tocopherol, 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ), iron(III) chloride hexahydrate, gallic acid, Folin and Ciocalteu’s phenol reagent, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and lipopolysaccharides (LPS) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Iron(II) sulfate heptahydrate and acetic acid were purchased from Junsei (Tokyo, Japan). A superoxide dismutase-WST kit was purchased from Dojindo (Kumamoto, Japan). Dulbecco’s Modified Eagle’s Medium (DMEM), RPMI 1640 medium, fetal bovine serum (FBS), and penicillin–streptomycin were obtained from Invitrogen (Carlsbad, CA, USA). The reagents 1,1-diphenyl-1-picrylhydrazyl (DPPH), 2,2′-azinibis 3-ethyl benzothiazoline-6-sulfonic acid (ABTS), α-tocopherol, 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ), iron(III) chloride hexahydrate, gallic acid, Folin and Ciocalteu’s phenol reagent, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and lipopolysaccharides (LPS) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Iron(II) sulfate heptahydrate and acetic acid were purchased from Junsei (Tokyo, Japan). A superoxide dismutase-WST kit was purchased from Dojindo (Kumamoto, Japan). Dulbecco’s Modified Eagle’s Medium (DMEM), RPMI 1640 medium, fetal bovine serum (FBS), and penicillin–streptomycin were obtained from Invitrogen (Carlsbad, CA, USA). Sample preparation and extraction Whole plants of P. vulgaris var. lilacina were purchased from the Plant Extract Bank (#007-017, Dae-Jeon, Korea). Dried P. vulgaris var. lilacina was milled into powder of 80-mesh particle size and stored at -70°C. The dried P. vulgaris var. lilacina was extracted three times with 70% ethanol. The 70% ethanol extract powder (10 g) was suspended in 500 mL of distilled water and extracted with 500 mL of the following solvents in a stepwise manner: hexane, chloroform, and butanol. Each fraction was filtered through Whatman filter paper No. 2 (Advantec, Tokyo, Japan). Subsequently, the filtrates were combined and evaporated under a vacuum and then lyophilized with a freeze dryer (Ilshine Lab, Suwon, Korea) at -70°C under reduced pressure (< 20 Pa). The dry residue was stored at -20°C. For further analysis, we reconstituted the dry extract and fractions with DMSO. Whole plants of P. vulgaris var. lilacina were purchased from the Plant Extract Bank (#007-017, Dae-Jeon, Korea). Dried P. vulgaris var. lilacina was milled into powder of 80-mesh particle size and stored at -70°C. The dried P. vulgaris var. lilacina was extracted three times with 70% ethanol. The 70% ethanol extract powder (10 g) was suspended in 500 mL of distilled water and extracted with 500 mL of the following solvents in a stepwise manner: hexane, chloroform, and butanol. Each fraction was filtered through Whatman filter paper No. 2 (Advantec, Tokyo, Japan). Subsequently, the filtrates were combined and evaporated under a vacuum and then lyophilized with a freeze dryer (Ilshine Lab, Suwon, Korea) at -70°C under reduced pressure (< 20 Pa). The dry residue was stored at -20°C. For further analysis, we reconstituted the dry extract and fractions with DMSO. Total phenolic content The total phenolic contents of P. vulgaris var. lilacina extract and its fractions were determined using the Folin-Ciocalteu method [20]. The extract and each fraction were oxidized with Folin–Ciocalteu’s reagents, and then, the reaction was neutralized with 10% sodium carbonate. After incubation at room temperature for 1 h, the absorbance of the reaction mixture was measured at 725 nm using a microplate reader (Molecular Devices, Sunnyvale, CA, USA). Quantification was performed based on a standard curve with gallic acid. Results were expressed as milligrams gallic acid equivalent (GAE) per gram of dry weight of extract (or fractions). The total phenolic contents of P. vulgaris var. lilacina extract and its fractions were determined using the Folin-Ciocalteu method [20]. The extract and each fraction were oxidized with Folin–Ciocalteu’s reagents, and then, the reaction was neutralized with 10% sodium carbonate. After incubation at room temperature for 1 h, the absorbance of the reaction mixture was measured at 725 nm using a microplate reader (Molecular Devices, Sunnyvale, CA, USA). Quantification was performed based on a standard curve with gallic acid. Results were expressed as milligrams gallic acid equivalent (GAE) per gram of dry weight of extract (or fractions). DPPH radical scavenging activity Analysis of DPPH radical-scavenging activity was carried out according to the Blois method [21]. 0.3 mM DPPH was added to each sample. After incubation for 30 min in the dark at room temperature, the absorbance was measured at 518 nm using a microplate reader. α-Tocopherol was used as a positive control. Percent reduction of the DPPH radical was calculated in the following way: inhibition concentration (%) = 100 - (Asample/Acontrol) × 100, where Acontrol is the absorbance of the control reaction (containing all reagents except the test sample), and Asample is the absorbance of the test sample. Tests were carried out in triplicate. For the final results, RC50 values (the concentrations required for 50% reduction of DPPH by 30 min after starting the reaction) were calculated from the absorbance diminished by 50%. The experiment was performed in triplicate. Analysis of DPPH radical-scavenging activity was carried out according to the Blois method [21]. 0.3 mM DPPH was added to each sample. After incubation for 30 min in the dark at room temperature, the absorbance was measured at 518 nm using a microplate reader. α-Tocopherol was used as a positive control. Percent reduction of the DPPH radical was calculated in the following way: inhibition concentration (%) = 100 - (Asample/Acontrol) × 100, where Acontrol is the absorbance of the control reaction (containing all reagents except the test sample), and Asample is the absorbance of the test sample. Tests were carried out in triplicate. For the final results, RC50 values (the concentrations required for 50% reduction of DPPH by 30 min after starting the reaction) were calculated from the absorbance diminished by 50%. The experiment was performed in triplicate. Ferric-reducing antioxidant power (FRAP) activity FRAP activity was determined using manual assay methods [22]. The working fluid was freshly prepared by mixing acetate buffer (300 mM, pH 3.6) with TPTZ in HCl and iron (III) chloride hexahydrate. Each sample solution or α-tocopherol was added to 3 mL of working fluid, and the mixture was left for 4 min at room temperature. The absorbance was measured at 593 nm. The results were expressed as iron (II) sulfate heptahydrate (FeSO4) equivalents. FRAP activity was determined using manual assay methods [22]. The working fluid was freshly prepared by mixing acetate buffer (300 mM, pH 3.6) with TPTZ in HCl and iron (III) chloride hexahydrate. Each sample solution or α-tocopherol was added to 3 mL of working fluid, and the mixture was left for 4 min at room temperature. The absorbance was measured at 593 nm. The results were expressed as iron (II) sulfate heptahydrate (FeSO4) equivalents. ABTS radical cation scavenging activity The ABTS assay was based on the ability of different fractions to scavenge the ABTS radical cation in comparison to a standard (α-tocopherol) [23]. The radical cation was prepared by mixing 7 mM ABTS with 2.45 mM potassium persulfate (1:1 v/v) and leaving the mixture for 24 h until the reaction was completed and the absorbance was stable. The ABTS radical solution was diluted with PBS to an absorbance of 0.7 ± 0.02 at 732 nm. The photometric assay was conducted with 180 μL of ABTS radical solution and 20 μL of samples; measurements were taken at 732 nm after 1 min. The antioxidative activity of the tested samples was calculated by determining the decrease in absorbance. The free radical scavenging capacity was expressed by RC50. The ABTS assay was based on the ability of different fractions to scavenge the ABTS radical cation in comparison to a standard (α-tocopherol) [23]. The radical cation was prepared by mixing 7 mM ABTS with 2.45 mM potassium persulfate (1:1 v/v) and leaving the mixture for 24 h until the reaction was completed and the absorbance was stable. The ABTS radical solution was diluted with PBS to an absorbance of 0.7 ± 0.02 at 732 nm. The photometric assay was conducted with 180 μL of ABTS radical solution and 20 μL of samples; measurements were taken at 732 nm after 1 min. The antioxidative activity of the tested samples was calculated by determining the decrease in absorbance. The free radical scavenging capacity was expressed by RC50. Superoxide dismutase (SOD) activity Superoxide dismutase activity was determined using the highly water-soluble tetrazolium salt WST-1, which produces a water-soluble formazan dye upon reduction with a superoxide anion. SOD activity was determined using an SOD assay kit (Dojindo, Kumamoto, Japan) in accordance with the manufacturer’s instructions. Briefly, WST working solution was made by diluting 1 mL of WST solution into 19 ml of buffer solution. Enzyme working solution was made after the enzyme solution tube was centrifuged for 5 sec. Fifteen microliters of enzyme solution were diluted with 2.5 mL of dilution buffer. SOD activity was expressed as the percentage of inhibition rate. Superoxide dismutase activity was determined using the highly water-soluble tetrazolium salt WST-1, which produces a water-soluble formazan dye upon reduction with a superoxide anion. SOD activity was determined using an SOD assay kit (Dojindo, Kumamoto, Japan) in accordance with the manufacturer’s instructions. Briefly, WST working solution was made by diluting 1 mL of WST solution into 19 ml of buffer solution. Enzyme working solution was made after the enzyme solution tube was centrifuged for 5 sec. Fifteen microliters of enzyme solution were diluted with 2.5 mL of dilution buffer. SOD activity was expressed as the percentage of inhibition rate. Cells and culture The mouse macrophage cell line RAW264.7, human liver cancer cell line HepG2, human colon cancer cell line HT29, human lung cancer cell line A549, human stomach cancer cell line MKN-45 and human cervical cancer cell line HeLa were purchased from the Korean Cell Line Bank (Seoul, Korea). The cell lines were grown in RPMI 1640 medium or DMEM with 10% FBS and 1% penicillin-streptomycin and incubated at 37°C in 5% CO2. The mouse macrophage cell line RAW264.7, human liver cancer cell line HepG2, human colon cancer cell line HT29, human lung cancer cell line A549, human stomach cancer cell line MKN-45 and human cervical cancer cell line HeLa were purchased from the Korean Cell Line Bank (Seoul, Korea). The cell lines were grown in RPMI 1640 medium or DMEM with 10% FBS and 1% penicillin-streptomycin and incubated at 37°C in 5% CO2. Cell cytotoxicity assay Exponentially growing cells were collected and plated at 5 × 103 - 1 × 104 cells/well. P. vulgaris var. lilacina ethanol extract and its solvent fractions in DMSO were diluted in PBS to obtain final concentrations of 10, 50 and 100 μg/mL. Cells were treated with samples for 24 h, and MTT solution was added. After 4 h, the media was removed, and DMSO was added to each well. The resulting absorbance was measured at 540 nm [24]. Exponentially growing cells were collected and plated at 5 × 103 - 1 × 104 cells/well. P. vulgaris var. lilacina ethanol extract and its solvent fractions in DMSO were diluted in PBS to obtain final concentrations of 10, 50 and 100 μg/mL. Cells were treated with samples for 24 h, and MTT solution was added. After 4 h, the media was removed, and DMSO was added to each well. The resulting absorbance was measured at 540 nm [24]. Intracellular reactive oxygen species (ROS) scavenging activity For microscopic detection of ROS formation, RAW264.7 cells were grown to 80% confluence in six-well plates and treated with samples for 24 h. After incubation, cells were incubated with dichlorofluorescein diacetate (DCF-DA) (25 μM) for 30 min at 37°C in the dark. After several washings with PBS, cells were observed with a fluorescence microscope (Carl ZEISS, Oberkochen, Germany). For microscopic detection of ROS formation, RAW264.7 cells were grown to 80% confluence in six-well plates and treated with samples for 24 h. After incubation, cells were incubated with dichlorofluorescein diacetate (DCF-DA) (25 μM) for 30 min at 37°C in the dark. After several washings with PBS, cells were observed with a fluorescence microscope (Carl ZEISS, Oberkochen, Germany). Real-time reverse transcription polymerase chain reaction analysis (RT-PCR) To determine the expression levels of p53, Bax, Bcl-2 and Fas, RT-PCR was performed using a Qiagen Rotor-Gene Q real-time thermal cycler (Valencia, CA, USA) in accordance with the manufacturer’s instructions. The cells were treated with P. vulgaris var. lilacina extracts and cultured for 24 h. Thereafter, cDNA was synthesized from the total RNA isolated from cells. The PCR reaction was performed using 2× SYBR Green mix (Qiagen, Valencia, CA, USA). All results were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression. The following primer sequences were used for the real-time RT-PCR: GAPDH, 5′-CGG AGT CAA CGG ATT TGG TCG TAT-3′ (forward), 5′-AGC CTT CTC CAT GGT GGT GAA GAC-3′ (reverse); p53, 5′-GCT CTG ACT GTA CCA CCA TCC-3′ (forward), 5′-CTC TCG GAA CAT CTC GAA GCG-3′ (reverse); Bax, 5′-ATG GAC GGG TCC GGG GAG-3′ (forward), 5′-TCA GCC CAT CTT CTT CCA-3′ (reverse); Bcl-2, 5′-CAG CTG CAC CTG ACG-3′ (forward), 5′-ATG CAC CTA CCC AGC-3′ (reverse); Fas, 5′- TCT AAC TTG GGG TGG CTT TGT CTT C -3′ (forward), 5′- GTG TCA TAC GCT TTC TTT CCA T-3′ (reverse). To determine the expression levels of p53, Bax, Bcl-2 and Fas, RT-PCR was performed using a Qiagen Rotor-Gene Q real-time thermal cycler (Valencia, CA, USA) in accordance with the manufacturer’s instructions. The cells were treated with P. vulgaris var. lilacina extracts and cultured for 24 h. Thereafter, cDNA was synthesized from the total RNA isolated from cells. The PCR reaction was performed using 2× SYBR Green mix (Qiagen, Valencia, CA, USA). All results were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression. The following primer sequences were used for the real-time RT-PCR: GAPDH, 5′-CGG AGT CAA CGG ATT TGG TCG TAT-3′ (forward), 5′-AGC CTT CTC CAT GGT GGT GAA GAC-3′ (reverse); p53, 5′-GCT CTG ACT GTA CCA CCA TCC-3′ (forward), 5′-CTC TCG GAA CAT CTC GAA GCG-3′ (reverse); Bax, 5′-ATG GAC GGG TCC GGG GAG-3′ (forward), 5′-TCA GCC CAT CTT CTT CCA-3′ (reverse); Bcl-2, 5′-CAG CTG CAC CTG ACG-3′ (forward), 5′-ATG CAC CTA CCC AGC-3′ (reverse); Fas, 5′- TCT AAC TTG GGG TGG CTT TGT CTT C -3′ (forward), 5′- GTG TCA TAC GCT TTC TTT CCA T-3′ (reverse). Gas chromatography-mass spectrum analysis (GC-MS) GC-MS analysis was carried out using an Agilent 6890 gas chromatograph equipped with a DB-5 ms capillary column (60 m × 0.25 mm; coating thickness 1.4 μm) and an Agilent 5975 MSD detector (Loveland, CO, USA). Analytical conditions were as follows: injector and transfer line temperatures of 250°C; oven temperature was programmed from 50°C to 150°C at 10°C/min, from 150°C to 200°C at 7°C/min, and from 200°C to 250°C at 5°C/min; carrier gas helium at 1 mL/min; and split ratio 1:10. Identification of the constituents was based on comparison of the retention times with those of authentic samples. GC-MS analysis was carried out using an Agilent 6890 gas chromatograph equipped with a DB-5 ms capillary column (60 m × 0.25 mm; coating thickness 1.4 μm) and an Agilent 5975 MSD detector (Loveland, CO, USA). Analytical conditions were as follows: injector and transfer line temperatures of 250°C; oven temperature was programmed from 50°C to 150°C at 10°C/min, from 150°C to 200°C at 7°C/min, and from 200°C to 250°C at 5°C/min; carrier gas helium at 1 mL/min; and split ratio 1:10. Identification of the constituents was based on comparison of the retention times with those of authentic samples. Statistical analysis Statistical analysis was performed with SPSS (version 17.0; SPSS Inc., Chicago, IL, USA). Descriptive statistics were used to calculate the mean and standard error of the mean (SEM). One-way analysis of variance was performed, and when the significance (p < 0.05) was determined, the differences of the mean values were identified using Duncan’s multiple range tests. Statistical analysis was performed with SPSS (version 17.0; SPSS Inc., Chicago, IL, USA). Descriptive statistics were used to calculate the mean and standard error of the mean (SEM). One-way analysis of variance was performed, and when the significance (p < 0.05) was determined, the differences of the mean values were identified using Duncan’s multiple range tests. Reagents: The reagents 1,1-diphenyl-1-picrylhydrazyl (DPPH), 2,2′-azinibis 3-ethyl benzothiazoline-6-sulfonic acid (ABTS), α-tocopherol, 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ), iron(III) chloride hexahydrate, gallic acid, Folin and Ciocalteu’s phenol reagent, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and lipopolysaccharides (LPS) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Iron(II) sulfate heptahydrate and acetic acid were purchased from Junsei (Tokyo, Japan). A superoxide dismutase-WST kit was purchased from Dojindo (Kumamoto, Japan). Dulbecco’s Modified Eagle’s Medium (DMEM), RPMI 1640 medium, fetal bovine serum (FBS), and penicillin–streptomycin were obtained from Invitrogen (Carlsbad, CA, USA). Sample preparation and extraction: Whole plants of P. vulgaris var. lilacina were purchased from the Plant Extract Bank (#007-017, Dae-Jeon, Korea). Dried P. vulgaris var. lilacina was milled into powder of 80-mesh particle size and stored at -70°C. The dried P. vulgaris var. lilacina was extracted three times with 70% ethanol. The 70% ethanol extract powder (10 g) was suspended in 500 mL of distilled water and extracted with 500 mL of the following solvents in a stepwise manner: hexane, chloroform, and butanol. Each fraction was filtered through Whatman filter paper No. 2 (Advantec, Tokyo, Japan). Subsequently, the filtrates were combined and evaporated under a vacuum and then lyophilized with a freeze dryer (Ilshine Lab, Suwon, Korea) at -70°C under reduced pressure (< 20 Pa). The dry residue was stored at -20°C. For further analysis, we reconstituted the dry extract and fractions with DMSO. Total phenolic content: The total phenolic contents of P. vulgaris var. lilacina extract and its fractions were determined using the Folin-Ciocalteu method [20]. The extract and each fraction were oxidized with Folin–Ciocalteu’s reagents, and then, the reaction was neutralized with 10% sodium carbonate. After incubation at room temperature for 1 h, the absorbance of the reaction mixture was measured at 725 nm using a microplate reader (Molecular Devices, Sunnyvale, CA, USA). Quantification was performed based on a standard curve with gallic acid. Results were expressed as milligrams gallic acid equivalent (GAE) per gram of dry weight of extract (or fractions). DPPH radical scavenging activity: Analysis of DPPH radical-scavenging activity was carried out according to the Blois method [21]. 0.3 mM DPPH was added to each sample. After incubation for 30 min in the dark at room temperature, the absorbance was measured at 518 nm using a microplate reader. α-Tocopherol was used as a positive control. Percent reduction of the DPPH radical was calculated in the following way: inhibition concentration (%) = 100 - (Asample/Acontrol) × 100, where Acontrol is the absorbance of the control reaction (containing all reagents except the test sample), and Asample is the absorbance of the test sample. Tests were carried out in triplicate. For the final results, RC50 values (the concentrations required for 50% reduction of DPPH by 30 min after starting the reaction) were calculated from the absorbance diminished by 50%. The experiment was performed in triplicate. Ferric-reducing antioxidant power (FRAP) activity: FRAP activity was determined using manual assay methods [22]. The working fluid was freshly prepared by mixing acetate buffer (300 mM, pH 3.6) with TPTZ in HCl and iron (III) chloride hexahydrate. Each sample solution or α-tocopherol was added to 3 mL of working fluid, and the mixture was left for 4 min at room temperature. The absorbance was measured at 593 nm. The results were expressed as iron (II) sulfate heptahydrate (FeSO4) equivalents. ABTS radical cation scavenging activity: The ABTS assay was based on the ability of different fractions to scavenge the ABTS radical cation in comparison to a standard (α-tocopherol) [23]. The radical cation was prepared by mixing 7 mM ABTS with 2.45 mM potassium persulfate (1:1 v/v) and leaving the mixture for 24 h until the reaction was completed and the absorbance was stable. The ABTS radical solution was diluted with PBS to an absorbance of 0.7 ± 0.02 at 732 nm. The photometric assay was conducted with 180 μL of ABTS radical solution and 20 μL of samples; measurements were taken at 732 nm after 1 min. The antioxidative activity of the tested samples was calculated by determining the decrease in absorbance. The free radical scavenging capacity was expressed by RC50. Superoxide dismutase (SOD) activity: Superoxide dismutase activity was determined using the highly water-soluble tetrazolium salt WST-1, which produces a water-soluble formazan dye upon reduction with a superoxide anion. SOD activity was determined using an SOD assay kit (Dojindo, Kumamoto, Japan) in accordance with the manufacturer’s instructions. Briefly, WST working solution was made by diluting 1 mL of WST solution into 19 ml of buffer solution. Enzyme working solution was made after the enzyme solution tube was centrifuged for 5 sec. Fifteen microliters of enzyme solution were diluted with 2.5 mL of dilution buffer. SOD activity was expressed as the percentage of inhibition rate. Cells and culture: The mouse macrophage cell line RAW264.7, human liver cancer cell line HepG2, human colon cancer cell line HT29, human lung cancer cell line A549, human stomach cancer cell line MKN-45 and human cervical cancer cell line HeLa were purchased from the Korean Cell Line Bank (Seoul, Korea). The cell lines were grown in RPMI 1640 medium or DMEM with 10% FBS and 1% penicillin-streptomycin and incubated at 37°C in 5% CO2. Cell cytotoxicity assay: Exponentially growing cells were collected and plated at 5 × 103 - 1 × 104 cells/well. P. vulgaris var. lilacina ethanol extract and its solvent fractions in DMSO were diluted in PBS to obtain final concentrations of 10, 50 and 100 μg/mL. Cells were treated with samples for 24 h, and MTT solution was added. After 4 h, the media was removed, and DMSO was added to each well. The resulting absorbance was measured at 540 nm [24]. Intracellular reactive oxygen species (ROS) scavenging activity: For microscopic detection of ROS formation, RAW264.7 cells were grown to 80% confluence in six-well plates and treated with samples for 24 h. After incubation, cells were incubated with dichlorofluorescein diacetate (DCF-DA) (25 μM) for 30 min at 37°C in the dark. After several washings with PBS, cells were observed with a fluorescence microscope (Carl ZEISS, Oberkochen, Germany). Real-time reverse transcription polymerase chain reaction analysis (RT-PCR): To determine the expression levels of p53, Bax, Bcl-2 and Fas, RT-PCR was performed using a Qiagen Rotor-Gene Q real-time thermal cycler (Valencia, CA, USA) in accordance with the manufacturer’s instructions. The cells were treated with P. vulgaris var. lilacina extracts and cultured for 24 h. Thereafter, cDNA was synthesized from the total RNA isolated from cells. The PCR reaction was performed using 2× SYBR Green mix (Qiagen, Valencia, CA, USA). All results were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression. The following primer sequences were used for the real-time RT-PCR: GAPDH, 5′-CGG AGT CAA CGG ATT TGG TCG TAT-3′ (forward), 5′-AGC CTT CTC CAT GGT GGT GAA GAC-3′ (reverse); p53, 5′-GCT CTG ACT GTA CCA CCA TCC-3′ (forward), 5′-CTC TCG GAA CAT CTC GAA GCG-3′ (reverse); Bax, 5′-ATG GAC GGG TCC GGG GAG-3′ (forward), 5′-TCA GCC CAT CTT CTT CCA-3′ (reverse); Bcl-2, 5′-CAG CTG CAC CTG ACG-3′ (forward), 5′-ATG CAC CTA CCC AGC-3′ (reverse); Fas, 5′- TCT AAC TTG GGG TGG CTT TGT CTT C -3′ (forward), 5′- GTG TCA TAC GCT TTC TTT CCA T-3′ (reverse). Gas chromatography-mass spectrum analysis (GC-MS): GC-MS analysis was carried out using an Agilent 6890 gas chromatograph equipped with a DB-5 ms capillary column (60 m × 0.25 mm; coating thickness 1.4 μm) and an Agilent 5975 MSD detector (Loveland, CO, USA). Analytical conditions were as follows: injector and transfer line temperatures of 250°C; oven temperature was programmed from 50°C to 150°C at 10°C/min, from 150°C to 200°C at 7°C/min, and from 200°C to 250°C at 5°C/min; carrier gas helium at 1 mL/min; and split ratio 1:10. Identification of the constituents was based on comparison of the retention times with those of authentic samples. Statistical analysis: Statistical analysis was performed with SPSS (version 17.0; SPSS Inc., Chicago, IL, USA). Descriptive statistics were used to calculate the mean and standard error of the mean (SEM). One-way analysis of variance was performed, and when the significance (p < 0.05) was determined, the differences of the mean values were identified using Duncan’s multiple range tests. Results and discussion: Extraction yield The yield of the extract and each fraction obtained from dry plant material was measured (Additional file 1: Table S1). The highest solid residue yields were obtained using butanol as the extraction solvent. The yield of the extract and each fraction obtained from dry plant material was measured (Additional file 1: Table S1). The highest solid residue yields were obtained using butanol as the extraction solvent. Total phenolic content The total phenolic content of the P. vulgaris var. lilacina extract and its fractions were determined through a linear gallic acid standard curve and expressed as mg GAE/g dry weight of extract (or fractions). As shown Table 1, the total phenolic content of all fractions from P. vulgaris var. lilacina varied from 109.31 to 322.80 mg GAE/g. The highest total phenolic content was detected in the water fraction (322.80 ± 15.12 mg GAE/g), whereas the lowest content was found in the hexane fraction (109.31 ± 4.08 mg GAE/g). Phenolic compounds are reported to be associated with antioxidant activity, anticancer effects, and other biological functions and may prevent the development of aging and disease [25]. These results suggest that P. vulgaris var. lilacina extracts might have high antioxidant and anticancer activities. Total phenolic contents of various solvent fractions obtained from the ethanol extract of Prunella vulgaris var. lilacina 1)Total phenolic content expressed in mg of gallic acid equivalent (GAE) per gram of dry weight of extract (or fractions). Values are mean ± SEM. Values with different superscripts within same column are significantly different (p < 0.05). The total phenolic content of the P. vulgaris var. lilacina extract and its fractions were determined through a linear gallic acid standard curve and expressed as mg GAE/g dry weight of extract (or fractions). As shown Table 1, the total phenolic content of all fractions from P. vulgaris var. lilacina varied from 109.31 to 322.80 mg GAE/g. The highest total phenolic content was detected in the water fraction (322.80 ± 15.12 mg GAE/g), whereas the lowest content was found in the hexane fraction (109.31 ± 4.08 mg GAE/g). Phenolic compounds are reported to be associated with antioxidant activity, anticancer effects, and other biological functions and may prevent the development of aging and disease [25]. These results suggest that P. vulgaris var. lilacina extracts might have high antioxidant and anticancer activities. Total phenolic contents of various solvent fractions obtained from the ethanol extract of Prunella vulgaris var. lilacina 1)Total phenolic content expressed in mg of gallic acid equivalent (GAE) per gram of dry weight of extract (or fractions). Values are mean ± SEM. Values with different superscripts within same column are significantly different (p < 0.05). Antioxidant capacities of prunella vulgaris var. Lilacina Results of the radical scavenging capacities determined by DPPH, FRAP, ABTS and SOD assays are shown in Table 2. In the DPPH assay, the DPPH radical scavenging activity of all fractions from P. vulgaris var. lilacina extract increased as shown in Table 2; the RC50 values of radical scavenging activity for DPPH were found to be 73.05 ± 10.32, 1402.96 ± 194.46, 83.52 ± 7.01, 521.58 ± 8.10, 64.26 ± 2.22, and 12.82 ± 1.33 μg/mL for ethanol extract, hexane, butanol, chloroform, water fractions and α-tocopherol, respectively. The water fraction showed the highest DPPH radical-scavenging activity. The DPPH scavenging activity of all fractions showed a similar trend to the content of total phenolic compounds. The FRAP assay measures total antioxidant activity based on the reduction of the ferric tripyridyltriazine (Fe3+-TPTZ) complex to the ferrous form. The ferric complex reducing abilities of different fractions were similar to the results obtained for the radical scavenging assay; water fraction exhibited very strong ferric ion reducing activity, and the five fractions in descending order of strength of ferric ion reducing activity were water fraction > ethanol extract > butanol > chloroform > hexane fraction. In terms of the ABTS assay, the water fraction demonstrated the highest scavenging activity, followed by the ethanol extract, butanol fraction, chloroform fraction and hexane fraction, and these trends was similar to those of the DPPH and FRAP assays. For the results of the SOD activity assay, most fractions with the exception of hexane exhibited very high SOD activity similar to α-tocopherol (99.64 ± 2.45%). However, there was no significant different among the fractions. Total antioxidant capacities of various solvent fractions of the ethanol extract of Prunella vulgaris var. lilacina 1)Concentration of test sample required to produce 50% inhibition of the DPPH radical. 2)Expressed as mmol of Fe2+ equivalents per gram of dry plant weight. 3)Concentration of test sample required to produce 50% inhibition of ABTS radical. 4)Expressed as the superoxide inhibition rate of Prunella vulgaris var. lilacina ethanol extract and its fractions. Values are mean ± SEM. Values with different superscripts within same column are significantly different (p < 0.05). DPPH, FRAP, ABTS and SOD assays are widely used to determine the antioxidant capacity of plant extracts due to their simplicity, stability, and reproducibility [26]. In this study, the DPPH, FRAP, ABTS and SOD assays provided comparable results for the antioxidant capacity measured in P. vulgaris var. lilacina extract and its fractions. The P. vulgaris var. lilacina extract and its fractions exhibited strong antioxidant activities against various oxidative systems in vitro. The strong antioxidant activity of a plant extract is correlated with a high content of total phenols [27,28]. In our research, we observed that the P. vulgaris var. lilacina extract and its fractions that contained higher phenol content exerted stronger radical scavenging effects (Table 3). The correlations between the antioxidant assays, such as DPPH, FRAP, ABTS and SOD activity and phenolic content, were highly positive (0.759 < │r│ < 0.993, p < 0.01), indicating that the four assays provided comparable values when they were used to estimate the antioxidant capacity of P. vulgaris var. lilacina extract. Many studies have shown a good positive linear correlation between antioxidant capacity and the total phenolic content of spices, medicinal herbs, and other dietary plants. Moreover, these results have also suggested that phenolic compounds are responsible for their antioxidant capacity [29-31]. Correlation coefficients between the antioxidant capacity and phenolic content of various solvent fractions of the ethanol extract of Prunella vulgaris var. lilacina 1)TPC: total phenolic content, DPPH: DPPH radical scavenging activity, FRAP: ferric reducing antioxidant powers, ABTS: ABTS radical scavenging activity, SOD: Superoxide dismutase activity. *p<0.05, **p<0.01. Results of the radical scavenging capacities determined by DPPH, FRAP, ABTS and SOD assays are shown in Table 2. In the DPPH assay, the DPPH radical scavenging activity of all fractions from P. vulgaris var. lilacina extract increased as shown in Table 2; the RC50 values of radical scavenging activity for DPPH were found to be 73.05 ± 10.32, 1402.96 ± 194.46, 83.52 ± 7.01, 521.58 ± 8.10, 64.26 ± 2.22, and 12.82 ± 1.33 μg/mL for ethanol extract, hexane, butanol, chloroform, water fractions and α-tocopherol, respectively. The water fraction showed the highest DPPH radical-scavenging activity. The DPPH scavenging activity of all fractions showed a similar trend to the content of total phenolic compounds. The FRAP assay measures total antioxidant activity based on the reduction of the ferric tripyridyltriazine (Fe3+-TPTZ) complex to the ferrous form. The ferric complex reducing abilities of different fractions were similar to the results obtained for the radical scavenging assay; water fraction exhibited very strong ferric ion reducing activity, and the five fractions in descending order of strength of ferric ion reducing activity were water fraction > ethanol extract > butanol > chloroform > hexane fraction. In terms of the ABTS assay, the water fraction demonstrated the highest scavenging activity, followed by the ethanol extract, butanol fraction, chloroform fraction and hexane fraction, and these trends was similar to those of the DPPH and FRAP assays. For the results of the SOD activity assay, most fractions with the exception of hexane exhibited very high SOD activity similar to α-tocopherol (99.64 ± 2.45%). However, there was no significant different among the fractions. Total antioxidant capacities of various solvent fractions of the ethanol extract of Prunella vulgaris var. lilacina 1)Concentration of test sample required to produce 50% inhibition of the DPPH radical. 2)Expressed as mmol of Fe2+ equivalents per gram of dry plant weight. 3)Concentration of test sample required to produce 50% inhibition of ABTS radical. 4)Expressed as the superoxide inhibition rate of Prunella vulgaris var. lilacina ethanol extract and its fractions. Values are mean ± SEM. Values with different superscripts within same column are significantly different (p < 0.05). DPPH, FRAP, ABTS and SOD assays are widely used to determine the antioxidant capacity of plant extracts due to their simplicity, stability, and reproducibility [26]. In this study, the DPPH, FRAP, ABTS and SOD assays provided comparable results for the antioxidant capacity measured in P. vulgaris var. lilacina extract and its fractions. The P. vulgaris var. lilacina extract and its fractions exhibited strong antioxidant activities against various oxidative systems in vitro. The strong antioxidant activity of a plant extract is correlated with a high content of total phenols [27,28]. In our research, we observed that the P. vulgaris var. lilacina extract and its fractions that contained higher phenol content exerted stronger radical scavenging effects (Table 3). The correlations between the antioxidant assays, such as DPPH, FRAP, ABTS and SOD activity and phenolic content, were highly positive (0.759 < │r│ < 0.993, p < 0.01), indicating that the four assays provided comparable values when they were used to estimate the antioxidant capacity of P. vulgaris var. lilacina extract. Many studies have shown a good positive linear correlation between antioxidant capacity and the total phenolic content of spices, medicinal herbs, and other dietary plants. Moreover, these results have also suggested that phenolic compounds are responsible for their antioxidant capacity [29-31]. Correlation coefficients between the antioxidant capacity and phenolic content of various solvent fractions of the ethanol extract of Prunella vulgaris var. lilacina 1)TPC: total phenolic content, DPPH: DPPH radical scavenging activity, FRAP: ferric reducing antioxidant powers, ABTS: ABTS radical scavenging activity, SOD: Superoxide dismutase activity. *p<0.05, **p<0.01. Intracellular ROS scavenging activity To investigate the intracellular levels of ROS, the cell-permeable probe DCF-DA was utilized. Non-fluorescent DCF-DA, hydrolyzed to DCFH inside the cells, yields highly fluorescent DCF-DA in the presence of intracellular hydrogen peroxide and related peroxides [32]. We examined whether P. vulgaris var. lilacina extract and its fractions inhibited LPS-induced ROS generation. As shown in Figure 1, LPS treatment significantly increased ROS formation in RAW264.7 cells as determined by DCF fluorescence. However, treatment with P. vulgaris var. lilacina extract and its fractions blocked LPS-induced ROS generation similar to the results obtained for the antioxidant assays. Reactive oxygen species scavenging activities of Prunella vulgaris var. lilacina. Cells were treated with solvent fractions of Prunella vulgaris var. lilacina at 10 μg/mL. After treatment for 24 h, ROS was stained with a DCF-DA for 30 min, and the generation of ROS was analyzed with fluorescence microscopy. To investigate the intracellular levels of ROS, the cell-permeable probe DCF-DA was utilized. Non-fluorescent DCF-DA, hydrolyzed to DCFH inside the cells, yields highly fluorescent DCF-DA in the presence of intracellular hydrogen peroxide and related peroxides [32]. We examined whether P. vulgaris var. lilacina extract and its fractions inhibited LPS-induced ROS generation. As shown in Figure 1, LPS treatment significantly increased ROS formation in RAW264.7 cells as determined by DCF fluorescence. However, treatment with P. vulgaris var. lilacina extract and its fractions blocked LPS-induced ROS generation similar to the results obtained for the antioxidant assays. Reactive oxygen species scavenging activities of Prunella vulgaris var. lilacina. Cells were treated with solvent fractions of Prunella vulgaris var. lilacina at 10 μg/mL. After treatment for 24 h, ROS was stained with a DCF-DA for 30 min, and the generation of ROS was analyzed with fluorescence microscopy. Cell cytotoxicity activity In order to evaluate the cytotoxic effects of all samples, we performed a preliminary cytotoxicity study with RAW264.7 cells exposed to various sample concentrations (10, 50, or 100 μg/mL). The P. vulgaris var. lilacina ethanol extract (at 50 and 100 μg/mL), hexane fraction (at 100 μg/mL) and chloroform fraction (at 50 and 100 μg/mL) inhibited cell proliferation, but did not at a concentration of 10 μg/mL (Figure 2). Conversely, the groups treated with 10 μg/mL of ethanol extract or butanol fraction or treated with 10, 50 and 100 μg/mL of the water fraction showed a proliferative effect of over 10%. Macrophages are specialized phagocytic cells that attack foreign substances and cancer cells through destruction and ingestion. They also stimulate lymphocytes and other immune cells to respond to pathogens [33]. These results suggest that the ethanol extract, butanol and water fractions of P. vulgaris var. lilacina can be used in the treatment of cancer. Based on this result, we determined the appropriate concentration to be 10 μg/mL. Effects of Prunella vulgaris var. lilacina on RAW264.7 cells as determined by the MTT assay. Cells were treated with solvent fractions of Prunella vulgaris var. lilacina at different concentrations (10, 50 and 100 μg/mL). After treatment for 24 h, cell viability was measured with the MTT assay. Values are the mean ± SEM; different marks within treatments indicate significant differences at *p < 0.05 compared to the PBS group. Involvement of free radical-mediated cell damage in many different diseases, particularly cancer, led us to evaluate the cytotoxic activities of the ethanol extract and the water fraction of P. vulgaris var. lilacina against five human cancer cell lines (liver, HepG2; colon, HT29; lung, A549; stomach, MKN-45; and cervical, HeLa) (Figure 3). The ethanol extract and the water fraction of P. vulgaris var. lilacina were most effective on A549 out of all the cancer cell lines; their values were 32.4 and 28.7% at 10 μg/mL, respectively. The cell cytotoxicity of the ethanol extract and water fraction from Prunella vulgaris var. lilacina against various cancer cell lines. Cells were treated with the ethanol extract and water fraction of P. vulgaris var. lilacina at 10 μg/ml. After treatment for 24 h, cell viability was measured with the MTT assay. Values are the means of three determinations ± SEM. The different letters indicate a significant difference of p < 0.05. In the present study, the results clearly demonstrate that the P. vulgaris var. lilacina ethanol extract induced significant cytotoxic effects on the various cancer cell lines studies, and these effects were stronger than for the P. vulgaris var. lilacina solvent fractions. It may be difficult to determine the contribution of individual components on the overall anticancer effects. In the literature, it has been reported that P. vulgaris var. lilacina components such as ursolic acid and rosmarinic acid are responsible for anticancer activities. Woo et al. [34] reported significant apoptogenic activity of 2α,3α-dihydroxyurs-12-ene-28-oic acid in Jurkat T cells. Lee et al. [35] and Hsu et al. [36] explored the cytotoxic effects of ursolic acid. Psotova et al. [37] found that rosmarinic acid from P. vulgaris var. lilacina exhibited strong anticancer activity. In the present study, we have isolated various presumed active compounds from the ethanol extract of P. vulgaris var. lilacina (Figures 4 and 5). The spectrum profile of GC-MS confirmed the presence of 7 major components, which were hexadecanoic acid, ethyl palmitate, lioleic acid, (z,z,z)-9,12,15-octadecatrienoic acid, ethyl (9E, 12E)-9,12-octadecadienoate, (z,z,z)-ethyl ester-9,12,15-octadecatrienoic acid, and ethyl linoleolate. Lai et al. [38] reported significant antitumor effects of fatty acids such as hexadecanoic acid and ethyl palmitate obtained from plant extracts. Taken together, the anticancer activity of ethanol extract may be the result of the synergistic effects of various compounds in P. vulgaris var. lilacina, which suggests that P. vulgaris var. lilacina can be used as a biological agent in the treatment of cancer. Gas chromatogram of the ethanol extract of Prunella vulgaris var. lilacina . MS Spectrum of the ethanol extract of Prunella vulgaris var. lilacina. The X-axis and Y-axis of the chromatogram show mass/charge (m/z) and abundance, respectively. (A) Hexadecanoic acid; (B) ethyl palmitate; (C) linoleic acid; (D) (z,z,z)-9,12,15-octadecatrienoic acid; (E) ethyl(9E,12E)-9,12-octadecadienoate; (F) (z,z,z)-ethyl ester-9,12,15-octadecatrienoic acid; (G) ethyl linoleolate. In order to evaluate the cytotoxic effects of all samples, we performed a preliminary cytotoxicity study with RAW264.7 cells exposed to various sample concentrations (10, 50, or 100 μg/mL). The P. vulgaris var. lilacina ethanol extract (at 50 and 100 μg/mL), hexane fraction (at 100 μg/mL) and chloroform fraction (at 50 and 100 μg/mL) inhibited cell proliferation, but did not at a concentration of 10 μg/mL (Figure 2). Conversely, the groups treated with 10 μg/mL of ethanol extract or butanol fraction or treated with 10, 50 and 100 μg/mL of the water fraction showed a proliferative effect of over 10%. Macrophages are specialized phagocytic cells that attack foreign substances and cancer cells through destruction and ingestion. They also stimulate lymphocytes and other immune cells to respond to pathogens [33]. These results suggest that the ethanol extract, butanol and water fractions of P. vulgaris var. lilacina can be used in the treatment of cancer. Based on this result, we determined the appropriate concentration to be 10 μg/mL. Effects of Prunella vulgaris var. lilacina on RAW264.7 cells as determined by the MTT assay. Cells were treated with solvent fractions of Prunella vulgaris var. lilacina at different concentrations (10, 50 and 100 μg/mL). After treatment for 24 h, cell viability was measured with the MTT assay. Values are the mean ± SEM; different marks within treatments indicate significant differences at *p < 0.05 compared to the PBS group. Involvement of free radical-mediated cell damage in many different diseases, particularly cancer, led us to evaluate the cytotoxic activities of the ethanol extract and the water fraction of P. vulgaris var. lilacina against five human cancer cell lines (liver, HepG2; colon, HT29; lung, A549; stomach, MKN-45; and cervical, HeLa) (Figure 3). The ethanol extract and the water fraction of P. vulgaris var. lilacina were most effective on A549 out of all the cancer cell lines; their values were 32.4 and 28.7% at 10 μg/mL, respectively. The cell cytotoxicity of the ethanol extract and water fraction from Prunella vulgaris var. lilacina against various cancer cell lines. Cells were treated with the ethanol extract and water fraction of P. vulgaris var. lilacina at 10 μg/ml. After treatment for 24 h, cell viability was measured with the MTT assay. Values are the means of three determinations ± SEM. The different letters indicate a significant difference of p < 0.05. In the present study, the results clearly demonstrate that the P. vulgaris var. lilacina ethanol extract induced significant cytotoxic effects on the various cancer cell lines studies, and these effects were stronger than for the P. vulgaris var. lilacina solvent fractions. It may be difficult to determine the contribution of individual components on the overall anticancer effects. In the literature, it has been reported that P. vulgaris var. lilacina components such as ursolic acid and rosmarinic acid are responsible for anticancer activities. Woo et al. [34] reported significant apoptogenic activity of 2α,3α-dihydroxyurs-12-ene-28-oic acid in Jurkat T cells. Lee et al. [35] and Hsu et al. [36] explored the cytotoxic effects of ursolic acid. Psotova et al. [37] found that rosmarinic acid from P. vulgaris var. lilacina exhibited strong anticancer activity. In the present study, we have isolated various presumed active compounds from the ethanol extract of P. vulgaris var. lilacina (Figures 4 and 5). The spectrum profile of GC-MS confirmed the presence of 7 major components, which were hexadecanoic acid, ethyl palmitate, lioleic acid, (z,z,z)-9,12,15-octadecatrienoic acid, ethyl (9E, 12E)-9,12-octadecadienoate, (z,z,z)-ethyl ester-9,12,15-octadecatrienoic acid, and ethyl linoleolate. Lai et al. [38] reported significant antitumor effects of fatty acids such as hexadecanoic acid and ethyl palmitate obtained from plant extracts. Taken together, the anticancer activity of ethanol extract may be the result of the synergistic effects of various compounds in P. vulgaris var. lilacina, which suggests that P. vulgaris var. lilacina can be used as a biological agent in the treatment of cancer. Gas chromatogram of the ethanol extract of Prunella vulgaris var. lilacina . MS Spectrum of the ethanol extract of Prunella vulgaris var. lilacina. The X-axis and Y-axis of the chromatogram show mass/charge (m/z) and abundance, respectively. (A) Hexadecanoic acid; (B) ethyl palmitate; (C) linoleic acid; (D) (z,z,z)-9,12,15-octadecatrienoic acid; (E) ethyl(9E,12E)-9,12-octadecadienoate; (F) (z,z,z)-ethyl ester-9,12,15-octadecatrienoic acid; (G) ethyl linoleolate. Real-time RT-PCR analysis We assessed whether P. vulgaris var. lilacina ethanol extract and the water fraction affected the expression of genes associated with apoptotic cell death, including the tumor suppressor p53,pro-apoptotic Bax, the anti-apoptotic Bcl-2 and Fas genes in A549 cells. p53-mediated apoptosis primarily occurs through the intrinsic apoptotic program [39]. It was reported that p53 induces apoptosis by either increasing transcriptional activity of proapoptotic genes such as Bax or suppressing the activity of the antiapoptotic genes of the Bcl-2 family [40]. Our data show that P. vulgaris var. lilacina ethanol extract and the water fraction significantly increased the expression of p53, Bax and Fas compared to the control. However, the expression of Bcl-2 was not decreased compared to that of the control (Figure 6). Therefore, the treatments altered the expression of Bax/Bcl-2, resulting in a shift in their ratio favoring apoptosis. Several other groups have shown in various cancer cell lines that P. vulgaris var. lilacina can lead to cell death by inducing apoptosis through regulation of p53 and Bax/Bcl-2 expression [41]. In our study, the resulting elevation in p53 and Bax protein expression in lung cancer cells is consistent with our earlier proposed involvement of p53 and Bax-related response systems. Taken together, we suggest that P. vulgaris var. lilacina ethanol extract and water fraction induce apoptosis through the regulation of p53, Bax, and Fas expression. mRNA expression of apoptotic genes in A549 cells treated with ethanol extracts and water fractions from Prunella vulgaris var. lilacina. Cells were treated with the ethanol extract and water fraction of P. vulgaris var. lilacina at 10 μg/ml. Values are the mean ± SEM; different marks within treatments indicate significant differences (*p < 0.05, **p < 0.01, ***p < 0.001). We assessed whether P. vulgaris var. lilacina ethanol extract and the water fraction affected the expression of genes associated with apoptotic cell death, including the tumor suppressor p53,pro-apoptotic Bax, the anti-apoptotic Bcl-2 and Fas genes in A549 cells. p53-mediated apoptosis primarily occurs through the intrinsic apoptotic program [39]. It was reported that p53 induces apoptosis by either increasing transcriptional activity of proapoptotic genes such as Bax or suppressing the activity of the antiapoptotic genes of the Bcl-2 family [40]. Our data show that P. vulgaris var. lilacina ethanol extract and the water fraction significantly increased the expression of p53, Bax and Fas compared to the control. However, the expression of Bcl-2 was not decreased compared to that of the control (Figure 6). Therefore, the treatments altered the expression of Bax/Bcl-2, resulting in a shift in their ratio favoring apoptosis. Several other groups have shown in various cancer cell lines that P. vulgaris var. lilacina can lead to cell death by inducing apoptosis through regulation of p53 and Bax/Bcl-2 expression [41]. In our study, the resulting elevation in p53 and Bax protein expression in lung cancer cells is consistent with our earlier proposed involvement of p53 and Bax-related response systems. Taken together, we suggest that P. vulgaris var. lilacina ethanol extract and water fraction induce apoptosis through the regulation of p53, Bax, and Fas expression. mRNA expression of apoptotic genes in A549 cells treated with ethanol extracts and water fractions from Prunella vulgaris var. lilacina. Cells were treated with the ethanol extract and water fraction of P. vulgaris var. lilacina at 10 μg/ml. Values are the mean ± SEM; different marks within treatments indicate significant differences (*p < 0.05, **p < 0.01, ***p < 0.001). Extraction yield: The yield of the extract and each fraction obtained from dry plant material was measured (Additional file 1: Table S1). The highest solid residue yields were obtained using butanol as the extraction solvent. Total phenolic content: The total phenolic content of the P. vulgaris var. lilacina extract and its fractions were determined through a linear gallic acid standard curve and expressed as mg GAE/g dry weight of extract (or fractions). As shown Table 1, the total phenolic content of all fractions from P. vulgaris var. lilacina varied from 109.31 to 322.80 mg GAE/g. The highest total phenolic content was detected in the water fraction (322.80 ± 15.12 mg GAE/g), whereas the lowest content was found in the hexane fraction (109.31 ± 4.08 mg GAE/g). Phenolic compounds are reported to be associated with antioxidant activity, anticancer effects, and other biological functions and may prevent the development of aging and disease [25]. These results suggest that P. vulgaris var. lilacina extracts might have high antioxidant and anticancer activities. Total phenolic contents of various solvent fractions obtained from the ethanol extract of Prunella vulgaris var. lilacina 1)Total phenolic content expressed in mg of gallic acid equivalent (GAE) per gram of dry weight of extract (or fractions). Values are mean ± SEM. Values with different superscripts within same column are significantly different (p < 0.05). Antioxidant capacities of prunella vulgaris var. Lilacina: Results of the radical scavenging capacities determined by DPPH, FRAP, ABTS and SOD assays are shown in Table 2. In the DPPH assay, the DPPH radical scavenging activity of all fractions from P. vulgaris var. lilacina extract increased as shown in Table 2; the RC50 values of radical scavenging activity for DPPH were found to be 73.05 ± 10.32, 1402.96 ± 194.46, 83.52 ± 7.01, 521.58 ± 8.10, 64.26 ± 2.22, and 12.82 ± 1.33 μg/mL for ethanol extract, hexane, butanol, chloroform, water fractions and α-tocopherol, respectively. The water fraction showed the highest DPPH radical-scavenging activity. The DPPH scavenging activity of all fractions showed a similar trend to the content of total phenolic compounds. The FRAP assay measures total antioxidant activity based on the reduction of the ferric tripyridyltriazine (Fe3+-TPTZ) complex to the ferrous form. The ferric complex reducing abilities of different fractions were similar to the results obtained for the radical scavenging assay; water fraction exhibited very strong ferric ion reducing activity, and the five fractions in descending order of strength of ferric ion reducing activity were water fraction > ethanol extract > butanol > chloroform > hexane fraction. In terms of the ABTS assay, the water fraction demonstrated the highest scavenging activity, followed by the ethanol extract, butanol fraction, chloroform fraction and hexane fraction, and these trends was similar to those of the DPPH and FRAP assays. For the results of the SOD activity assay, most fractions with the exception of hexane exhibited very high SOD activity similar to α-tocopherol (99.64 ± 2.45%). However, there was no significant different among the fractions. Total antioxidant capacities of various solvent fractions of the ethanol extract of Prunella vulgaris var. lilacina 1)Concentration of test sample required to produce 50% inhibition of the DPPH radical. 2)Expressed as mmol of Fe2+ equivalents per gram of dry plant weight. 3)Concentration of test sample required to produce 50% inhibition of ABTS radical. 4)Expressed as the superoxide inhibition rate of Prunella vulgaris var. lilacina ethanol extract and its fractions. Values are mean ± SEM. Values with different superscripts within same column are significantly different (p < 0.05). DPPH, FRAP, ABTS and SOD assays are widely used to determine the antioxidant capacity of plant extracts due to their simplicity, stability, and reproducibility [26]. In this study, the DPPH, FRAP, ABTS and SOD assays provided comparable results for the antioxidant capacity measured in P. vulgaris var. lilacina extract and its fractions. The P. vulgaris var. lilacina extract and its fractions exhibited strong antioxidant activities against various oxidative systems in vitro. The strong antioxidant activity of a plant extract is correlated with a high content of total phenols [27,28]. In our research, we observed that the P. vulgaris var. lilacina extract and its fractions that contained higher phenol content exerted stronger radical scavenging effects (Table 3). The correlations between the antioxidant assays, such as DPPH, FRAP, ABTS and SOD activity and phenolic content, were highly positive (0.759 < │r│ < 0.993, p < 0.01), indicating that the four assays provided comparable values when they were used to estimate the antioxidant capacity of P. vulgaris var. lilacina extract. Many studies have shown a good positive linear correlation between antioxidant capacity and the total phenolic content of spices, medicinal herbs, and other dietary plants. Moreover, these results have also suggested that phenolic compounds are responsible for their antioxidant capacity [29-31]. Correlation coefficients between the antioxidant capacity and phenolic content of various solvent fractions of the ethanol extract of Prunella vulgaris var. lilacina 1)TPC: total phenolic content, DPPH: DPPH radical scavenging activity, FRAP: ferric reducing antioxidant powers, ABTS: ABTS radical scavenging activity, SOD: Superoxide dismutase activity. *p<0.05, **p<0.01. Intracellular ROS scavenging activity: To investigate the intracellular levels of ROS, the cell-permeable probe DCF-DA was utilized. Non-fluorescent DCF-DA, hydrolyzed to DCFH inside the cells, yields highly fluorescent DCF-DA in the presence of intracellular hydrogen peroxide and related peroxides [32]. We examined whether P. vulgaris var. lilacina extract and its fractions inhibited LPS-induced ROS generation. As shown in Figure 1, LPS treatment significantly increased ROS formation in RAW264.7 cells as determined by DCF fluorescence. However, treatment with P. vulgaris var. lilacina extract and its fractions blocked LPS-induced ROS generation similar to the results obtained for the antioxidant assays. Reactive oxygen species scavenging activities of Prunella vulgaris var. lilacina. Cells were treated with solvent fractions of Prunella vulgaris var. lilacina at 10 μg/mL. After treatment for 24 h, ROS was stained with a DCF-DA for 30 min, and the generation of ROS was analyzed with fluorescence microscopy. Cell cytotoxicity activity: In order to evaluate the cytotoxic effects of all samples, we performed a preliminary cytotoxicity study with RAW264.7 cells exposed to various sample concentrations (10, 50, or 100 μg/mL). The P. vulgaris var. lilacina ethanol extract (at 50 and 100 μg/mL), hexane fraction (at 100 μg/mL) and chloroform fraction (at 50 and 100 μg/mL) inhibited cell proliferation, but did not at a concentration of 10 μg/mL (Figure 2). Conversely, the groups treated with 10 μg/mL of ethanol extract or butanol fraction or treated with 10, 50 and 100 μg/mL of the water fraction showed a proliferative effect of over 10%. Macrophages are specialized phagocytic cells that attack foreign substances and cancer cells through destruction and ingestion. They also stimulate lymphocytes and other immune cells to respond to pathogens [33]. These results suggest that the ethanol extract, butanol and water fractions of P. vulgaris var. lilacina can be used in the treatment of cancer. Based on this result, we determined the appropriate concentration to be 10 μg/mL. Effects of Prunella vulgaris var. lilacina on RAW264.7 cells as determined by the MTT assay. Cells were treated with solvent fractions of Prunella vulgaris var. lilacina at different concentrations (10, 50 and 100 μg/mL). After treatment for 24 h, cell viability was measured with the MTT assay. Values are the mean ± SEM; different marks within treatments indicate significant differences at *p < 0.05 compared to the PBS group. Involvement of free radical-mediated cell damage in many different diseases, particularly cancer, led us to evaluate the cytotoxic activities of the ethanol extract and the water fraction of P. vulgaris var. lilacina against five human cancer cell lines (liver, HepG2; colon, HT29; lung, A549; stomach, MKN-45; and cervical, HeLa) (Figure 3). The ethanol extract and the water fraction of P. vulgaris var. lilacina were most effective on A549 out of all the cancer cell lines; their values were 32.4 and 28.7% at 10 μg/mL, respectively. The cell cytotoxicity of the ethanol extract and water fraction from Prunella vulgaris var. lilacina against various cancer cell lines. Cells were treated with the ethanol extract and water fraction of P. vulgaris var. lilacina at 10 μg/ml. After treatment for 24 h, cell viability was measured with the MTT assay. Values are the means of three determinations ± SEM. The different letters indicate a significant difference of p < 0.05. In the present study, the results clearly demonstrate that the P. vulgaris var. lilacina ethanol extract induced significant cytotoxic effects on the various cancer cell lines studies, and these effects were stronger than for the P. vulgaris var. lilacina solvent fractions. It may be difficult to determine the contribution of individual components on the overall anticancer effects. In the literature, it has been reported that P. vulgaris var. lilacina components such as ursolic acid and rosmarinic acid are responsible for anticancer activities. Woo et al. [34] reported significant apoptogenic activity of 2α,3α-dihydroxyurs-12-ene-28-oic acid in Jurkat T cells. Lee et al. [35] and Hsu et al. [36] explored the cytotoxic effects of ursolic acid. Psotova et al. [37] found that rosmarinic acid from P. vulgaris var. lilacina exhibited strong anticancer activity. In the present study, we have isolated various presumed active compounds from the ethanol extract of P. vulgaris var. lilacina (Figures 4 and 5). The spectrum profile of GC-MS confirmed the presence of 7 major components, which were hexadecanoic acid, ethyl palmitate, lioleic acid, (z,z,z)-9,12,15-octadecatrienoic acid, ethyl (9E, 12E)-9,12-octadecadienoate, (z,z,z)-ethyl ester-9,12,15-octadecatrienoic acid, and ethyl linoleolate. Lai et al. [38] reported significant antitumor effects of fatty acids such as hexadecanoic acid and ethyl palmitate obtained from plant extracts. Taken together, the anticancer activity of ethanol extract may be the result of the synergistic effects of various compounds in P. vulgaris var. lilacina, which suggests that P. vulgaris var. lilacina can be used as a biological agent in the treatment of cancer. Gas chromatogram of the ethanol extract of Prunella vulgaris var. lilacina . MS Spectrum of the ethanol extract of Prunella vulgaris var. lilacina. The X-axis and Y-axis of the chromatogram show mass/charge (m/z) and abundance, respectively. (A) Hexadecanoic acid; (B) ethyl palmitate; (C) linoleic acid; (D) (z,z,z)-9,12,15-octadecatrienoic acid; (E) ethyl(9E,12E)-9,12-octadecadienoate; (F) (z,z,z)-ethyl ester-9,12,15-octadecatrienoic acid; (G) ethyl linoleolate. Real-time RT-PCR analysis: We assessed whether P. vulgaris var. lilacina ethanol extract and the water fraction affected the expression of genes associated with apoptotic cell death, including the tumor suppressor p53,pro-apoptotic Bax, the anti-apoptotic Bcl-2 and Fas genes in A549 cells. p53-mediated apoptosis primarily occurs through the intrinsic apoptotic program [39]. It was reported that p53 induces apoptosis by either increasing transcriptional activity of proapoptotic genes such as Bax or suppressing the activity of the antiapoptotic genes of the Bcl-2 family [40]. Our data show that P. vulgaris var. lilacina ethanol extract and the water fraction significantly increased the expression of p53, Bax and Fas compared to the control. However, the expression of Bcl-2 was not decreased compared to that of the control (Figure 6). Therefore, the treatments altered the expression of Bax/Bcl-2, resulting in a shift in their ratio favoring apoptosis. Several other groups have shown in various cancer cell lines that P. vulgaris var. lilacina can lead to cell death by inducing apoptosis through regulation of p53 and Bax/Bcl-2 expression [41]. In our study, the resulting elevation in p53 and Bax protein expression in lung cancer cells is consistent with our earlier proposed involvement of p53 and Bax-related response systems. Taken together, we suggest that P. vulgaris var. lilacina ethanol extract and water fraction induce apoptosis through the regulation of p53, Bax, and Fas expression. mRNA expression of apoptotic genes in A549 cells treated with ethanol extracts and water fractions from Prunella vulgaris var. lilacina. Cells were treated with the ethanol extract and water fraction of P. vulgaris var. lilacina at 10 μg/ml. Values are the mean ± SEM; different marks within treatments indicate significant differences (*p < 0.05, **p < 0.01, ***p < 0.001). Conclusions: The present study determined that P. vulgaris var. lilacina extract and its fractions have strong antioxidant and anticancer activities in vitro. The correlation coefficients between antioxidant capacity and the phenolic content were very strong, and phenolic compounds were a major contributor to the antioxidant capacities of P. vulgaris var. lilacina. In addition, we confirmed the presence of 7 major components of P. vulgaris var. lilacina. However, further studies to study the mechanisms of these compounds and find the root of their antioxidative and anticancer activity are underway. On the basis of these results, P. vulgaris var. lilacina appears to be a good source of natural antioxidant and anticancer agents and could be of significance in the food industry and for the control of various human and animal diseases. Abbreviations: ABTS: 2, 2′-Azinibis 3-ethyl benzothiazoline-6-sulfonic acid; DCF-DA: Dichlorofluorescein diacetate; DMSO: Dimethyl sulfoxide; DPPH: 1, 1-diphenyl-1-picrylhydrazyl; FRAP: Ferric-reducing antioxidant power; GAE: Gallic acid equivalent; LPS: Lipopolysaccharide; ROS: Reactive oxygen species; SOD: Superoxide dismutase; TPTZ: 2, 4, 6-tris(2-pyridyl)-s-triazine. Competing interests: The authors declare that they have no competing interests. Authors’ contributions: KAH conceived this study and designed the experiments. YJH and EJL performed most of the experiments. All authors including HRK analyzed the data and discussed the results. KAH supervised the project and wrote the manuscript with the help of YJH, EJL and HRK, and all authors read and approved the final manuscript. Pre-publication history: The pre-publication history for this paper can be accessed here: http://www.biomedcentral.com/1472-6882/13/310/prepub Supplementary Material: Extraction yield of Prunella vulgaris var. lilacina. Click here for file

Background: Recently, considerable attention has been focused on exploring the potential antioxidant properties of plant extracts or isolated products of plant origin. Prunella vulgaris var. lilacina is widely distributed in Korea, Japan, China, and Europe, and it continues to be used to treat inflammation, eye pain, headache, and dizziness. However, reports on the antioxidant activities of P. vulgaris var. lilacina are limited, particularly concerning the relationship between its phenolic content and antioxidant capacity. In this study, we investigated the antioxidant and anticancer activities of an ethanol extract from P. vulgaris var. lilacina and its fractions. Methods: Dried powder of P. vulgaris var. lilacina was extracted with ethanol, and the extract was fractionated to produce the hexane fraction, butanol fraction, chloroform fraction and residual water fraction. The phenolic content was assayed using the Folin-Ciocalteu colorimetric method. Subsequently, the antioxidant activities of the ethanol extract and its fractions were analyzed employing various antioxidant assay methods including DPPH, FRAP, ABTS, SOD activity and production of reactive oxygen species. Additionally, the extract and fractions were assayed for their ability to exert cytotoxic activities on various cancer cells using the MTT assay. We also investigated the expression of genes associated with apoptotic cell death by RT-PCR. Results: The total phenolic contents of the ethanol extract and water fraction of P. vulgaris var. lilacina were 303.66 and 322.80 mg GAE/g dry weight (or fractions), respectively. The results showed that the ethanol extract and the water fraction of P. vulgaris var. lilacina had higher antioxidant content than other solvent fractions, similar to their total phenolic content. Anticancer activity was also tested using the HepG2, HT29, A549, MKN45 and HeLa cancer cell lines. The results clearly demonstrated that the P. vulgaris var. lilacina ethanol extract induced significant cytotoxic effects on the various cancer cell lines, and these effects were stronger than those induced by the P. vulgaris var. lilacina solvent fractions. We also investigated the expression of genes associated with apoptotic cell death. We confirmed that the P. vulgaris var. lilacina ethanol extract and water fraction significantly increased the expression of p53, Bax and Fas. Conclusions: These results suggest that the ethanol extract from P. vulgaris var. lilacina and its fractions could be applied as natural sources of antioxidants and anticancer activities in food and in the pharmaceutical industry.

Background: Oxidative stress is caused by reactive oxygen species (ROS), which are associated with many pathological disorders such as atherosclerosis, diabetes, ageing and cancer [1-3]. In order to protect human beings against oxidative damage, synthetic antioxidants such as BHA and BHT were created due to demand [4]. However, there has been concern regarding the toxicity and carcinogenic effects of synthetic antioxidants [5,6]. Thus, it is important to identify new sources of safe and inexpensive antioxidants of natural origin. Natural antioxidants, especially plant phenolics, flavonoids, tannins and anthocyanidins, are safe and are also bioactive [7]. Therefore, in recent years, considerable attention has been focused on exploring the potential antioxidant properties of plant extracts or isolated products of plant origin [8]. Prunella vulgaris var. lilacina is widely distributed in Korea, Japan, China, and Europe, and it continues to be used to treat inflammation, eye pain, headache, and dizziness [9,10]. It is rich in active compounds known to significantly affect human health, such as triterpenoid, rosmarinic acid, hyperoside, ursolic acid, and flavonoids [11-16]. Furthermore, P. vulgaris var. lilacina has been shown to have anti-allergic, anti-inflammatory, anti-oxidative, anti-microbial, and anti-viral effects [17-19]. However, reports on the antioxidant activities of P. vulgaris var. lilacina are limited, particularly concerning the relationship between its phenolic content and antioxidant capacity. Therefore, the aims of this study were to identify new sources of antioxidants from P. vulgaris var. lilacina extract. Additionally, the effects of the extraction solvent (70% ethanol, hexane, butanol, chloroform, or water) on the total phenolic content and antioxidant activities of P. vulgaris var. lilacina were investigated. Conclusions: The present study determined that P. vulgaris var. lilacina extract and its fractions have strong antioxidant and anticancer activities in vitro. The correlation coefficients between antioxidant capacity and the phenolic content were very strong, and phenolic compounds were a major contributor to the antioxidant capacities of P. vulgaris var. lilacina. In addition, we confirmed the presence of 7 major components of P. vulgaris var. lilacina. However, further studies to study the mechanisms of these compounds and find the root of their antioxidative and anticancer activity are underway. On the basis of these results, P. vulgaris var. lilacina appears to be a good source of natural antioxidant and anticancer agents and could be of significance in the food industry and for the control of various human and animal diseases.

Background: Recently, considerable attention has been focused on exploring the potential antioxidant properties of plant extracts or isolated products of plant origin. Prunella vulgaris var. lilacina is widely distributed in Korea, Japan, China, and Europe, and it continues to be used to treat inflammation, eye pain, headache, and dizziness. However, reports on the antioxidant activities of P. vulgaris var. lilacina are limited, particularly concerning the relationship between its phenolic content and antioxidant capacity. In this study, we investigated the antioxidant and anticancer activities of an ethanol extract from P. vulgaris var. lilacina and its fractions. Methods: Dried powder of P. vulgaris var. lilacina was extracted with ethanol, and the extract was fractionated to produce the hexane fraction, butanol fraction, chloroform fraction and residual water fraction. The phenolic content was assayed using the Folin-Ciocalteu colorimetric method. Subsequently, the antioxidant activities of the ethanol extract and its fractions were analyzed employing various antioxidant assay methods including DPPH, FRAP, ABTS, SOD activity and production of reactive oxygen species. Additionally, the extract and fractions were assayed for their ability to exert cytotoxic activities on various cancer cells using the MTT assay. We also investigated the expression of genes associated with apoptotic cell death by RT-PCR. Results: The total phenolic contents of the ethanol extract and water fraction of P. vulgaris var. lilacina were 303.66 and 322.80 mg GAE/g dry weight (or fractions), respectively. The results showed that the ethanol extract and the water fraction of P. vulgaris var. lilacina had higher antioxidant content than other solvent fractions, similar to their total phenolic content. Anticancer activity was also tested using the HepG2, HT29, A549, MKN45 and HeLa cancer cell lines. The results clearly demonstrated that the P. vulgaris var. lilacina ethanol extract induced significant cytotoxic effects on the various cancer cell lines, and these effects were stronger than those induced by the P. vulgaris var. lilacina solvent fractions. We also investigated the expression of genes associated with apoptotic cell death. We confirmed that the P. vulgaris var. lilacina ethanol extract and water fraction significantly increased the expression of p53, Bax and Fas. Conclusions: These results suggest that the ethanol extract from P. vulgaris var. lilacina and its fractions could be applied as natural sources of antioxidants and anticancer activities in food and in the pharmaceutical industry.

[CONTENT] Prunella vulgaris var. lilacina | Antioxidative activity | Anticancer activity [SUMMARY]

[CONTENT] Prunella vulgaris var. lilacina | Antioxidative activity | Anticancer activity [SUMMARY]

[CONTENT] Prunella vulgaris var. lilacina | Antioxidative activity | Anticancer activity [SUMMARY]

[CONTENT] Prunella vulgaris var. lilacina | Antioxidative activity | Anticancer activity [SUMMARY]

[CONTENT] Prunella vulgaris var. lilacina | Antioxidative activity | Anticancer activity [SUMMARY]

[CONTENT] Analysis of Variance | Animals | Antineoplastic Agents | Antioxidants | Cell Line, Tumor | Cell Survival | Humans | Intracellular Space | Mice | Phenols | Plant Extracts | Prunella | Reactive Oxygen Species | Real-Time Polymerase Chain Reaction [SUMMARY]

[CONTENT] Analysis of Variance | Animals | Antineoplastic Agents | Antioxidants | Cell Line, Tumor | Cell Survival | Humans | Intracellular Space | Mice | Phenols | Plant Extracts | Prunella | Reactive Oxygen Species | Real-Time Polymerase Chain Reaction [SUMMARY]

[CONTENT] Analysis of Variance | Animals | Antineoplastic Agents | Antioxidants | Cell Line, Tumor | Cell Survival | Humans | Intracellular Space | Mice | Phenols | Plant Extracts | Prunella | Reactive Oxygen Species | Real-Time Polymerase Chain Reaction [SUMMARY]

[CONTENT] Analysis of Variance | Animals | Antineoplastic Agents | Antioxidants | Cell Line, Tumor | Cell Survival | Humans | Intracellular Space | Mice | Phenols | Plant Extracts | Prunella | Reactive Oxygen Species | Real-Time Polymerase Chain Reaction [SUMMARY]

[CONTENT] Analysis of Variance | Animals | Antineoplastic Agents | Antioxidants | Cell Line, Tumor | Cell Survival | Humans | Intracellular Space | Mice | Phenols | Plant Extracts | Prunella | Reactive Oxygen Species | Real-Time Polymerase Chain Reaction [SUMMARY]

[CONTENT] strong antioxidant activity | antioxidants especially plant | antioxidants vulgaris | extracts high antioxidant | source natural antioxidant [SUMMARY]

[CONTENT] strong antioxidant activity | antioxidants especially plant | antioxidants vulgaris | extracts high antioxidant | source natural antioxidant [SUMMARY]

[CONTENT] strong antioxidant activity | antioxidants especially plant | antioxidants vulgaris | extracts high antioxidant | source natural antioxidant [SUMMARY]

[CONTENT] strong antioxidant activity | antioxidants especially plant | antioxidants vulgaris | extracts high antioxidant | source natural antioxidant [SUMMARY]

[CONTENT] strong antioxidant activity | antioxidants especially plant | antioxidants vulgaris | extracts high antioxidant | source natural antioxidant [SUMMARY]

[CONTENT] vulgaris var | var lilacina | var | lilacina | vulgaris | vulgaris var lilacina | extract | activity | fractions | ethanol [SUMMARY]

[CONTENT] vulgaris var | var lilacina | var | lilacina | vulgaris | vulgaris var lilacina | extract | activity | fractions | ethanol [SUMMARY]

[CONTENT] vulgaris var | var lilacina | var | lilacina | vulgaris | vulgaris var lilacina | extract | activity | fractions | ethanol [SUMMARY]

[CONTENT] vulgaris var | var lilacina | var | lilacina | vulgaris | vulgaris var lilacina | extract | activity | fractions | ethanol [SUMMARY]

[CONTENT] vulgaris var | var lilacina | var | lilacina | vulgaris | vulgaris var lilacina | extract | activity | fractions | ethanol [SUMMARY]

[CONTENT] antioxidants | anti | oxidative | antioxidant | vulgaris | var lilacina | var | vulgaris var | vulgaris var lilacina | lilacina [SUMMARY]

[CONTENT] solution | absorbance | line | cell line | cell | min | radical | cells | reverse | usa [SUMMARY]

[CONTENT] antioxidant | anticancer | vulgaris | var lilacina | var | vulgaris var lilacina | vulgaris var | lilacina | major | antioxidant anticancer [SUMMARY]

[CONTENT] vulgaris | lilacina | var lilacina | vulgaris var lilacina | var | vulgaris var | extract | cells | fractions | acid [SUMMARY]

[CONTENT] vulgaris | lilacina | var lilacina | vulgaris var lilacina | var | vulgaris var | extract | cells | fractions | acid [SUMMARY]

[CONTENT] ||| ||| Korea | Japan | China | Europe ||| P. vulgaris var ||| ||| P. vulgaris var ||| [SUMMARY]

[CONTENT] ||| the hexane fraction ||| Folin ||| FRAP | ABTS | SOD ||| MTT ||| RT-PCR [SUMMARY]

[CONTENT] P. vulgaris var ||| [SUMMARY]

[CONTENT] ||| ||| Korea | Japan | China | Europe ||| P. vulgaris var ||| ||| P. vulgaris var ||| ||| ||| the hexane fraction ||| Folin ||| FRAP | ABTS | SOD ||| MTT ||| RT-PCR ||| ||| 303.66 | 322.80 | GAE ||| ||| ||| HepG2 | A549 | HeLa ||| ||| lilacina ethanol extract ||| ||| ||| ||| lilacina ethanol extract | Bax ||| P. vulgaris var ||| [SUMMARY]

[CONTENT] ||| ||| Korea | Japan | China | Europe ||| P. vulgaris var ||| ||| P. vulgaris var ||| ||| ||| the hexane fraction ||| Folin ||| FRAP | ABTS | SOD ||| MTT ||| RT-PCR ||| ||| 303.66 | 322.80 | GAE ||| ||| ||| HepG2 | A549 | HeLa ||| ||| lilacina ethanol extract ||| ||| ||| ||| lilacina ethanol extract | Bax ||| P. vulgaris var ||| [SUMMARY]

REDUCE PORT LAPAROSCOPIC SPLENECTOMY FOR GIANT EPITELIAL CYST.

26734802

Delaitre and Maignien performed the first successful laparoscopic splenectomy in 1991. After that, laparoscopic splenectomy has become one of the most frequently performed laparoscopic solid organ procedures.

BACKGROUND

A reduce port laparoscopic splenectomy was performed by using a 10 mm and two 5 mm trocars. To entered the abdomen a trans-umbilical open technique was done and a 10 mm trocar was placed. A subcostal 5 mm trocar was placed under direct vision at the level of the anterior axillary line and another 5 mm port was inserted at the mid-epigastric region. Once it was completely dissected and freed from all of its attachments the hilum, splenic artery and vein, was clipped with hem-o-lock and divided with scissors. Then an endobag was used to retrieve the spleen after being morcellated trough the umbilical incision.

METHODS

This technique was used in a 15 years old female with epigastric and left upper quadrant pain. An abdominal ultrasound demonstrated a giant cyst located in the spleen. Laboratory tests findings were normal. The CT scan was also done, and showed a giant cyst, which squeeze the stomach. The patient tolerated well the procedure, with an unremarkable postoperative. She was discharge home 72 h after the surgery.

RESULTS

The use of reduce port minimizes abdominal trauma and has the hypothetical advantages of shorter postoperative stay, greater pain control, and better cosmesis. Laparoscopic splenectomy for giant cysts by using reduce port trocars is safe and feasible and less invasive.

CONCLUSION

4755184

INTRODUCTION

Splenectomy was initially described for hereditary spherocytosis by Sutherland and Burghard in 1910 and for idiopathic thrombocytopenic purpura by Kaznelson in 19168. It has been well recognized as an effective cure for hematologic disorders, better than medical treatment. The first successful laparoscopic splenectomy was performed by Delaitre and Maignien in 19917 , 8. After that, laparoscopic splenectomy has become one of the most frequently performed laparoscopic solid organ procedures. Laparoscopic splenectomy is emerging as the gold standard for the management of various hematologic disorders. Minimally invasive surgery has earned boundless acceptance. The enthusiasm to limit the trauma of large incisions has been the incentive of the development of minimally invasive surgery during the past century. Numerous technology and equipment in laparoscopy have been emerging. It started gradually reducing and repositioning trocars during laparoscopic surgery. It began with repositioning the subxiphoid trocar into the umbilicus. Less trocars equals less abdominal trauma. In small spleens a single port laparoscopic surgery can be performed. In large spleens, less trocars can be used. The aim of this artcile is to details the splenectomy procedure using only three trochars.

METHOD

Surgical technique The patient is placed in lateral decubitus, to enter the abdomen using a trans-umbilical open technique and a 12 mm trocar is placed. Through it a 10 mm 30o scope is inserted. A subcostal 5 mm trocar is placed under direct vision at the level of the anterior axillary line and another 5 mm port is inserted at the midepigastric region. Using a 5 mm harmonic scalpel (Harmonic Ace, Ethicon) and 5 mm instruments, access is gained to the lesser sac by dividing the gastrosplenic ligament and short vessels until the upper pole of the spleen. The splenic flexure of the colon is mobilized to get the lower pole of the spleen freed. The posterior splenorenal ligament is then freed. Once the spleen is completely dissected free from all of its hilum attachments, splenic artery and vein are clipped with hemolocks and divided with scissors. It is suggested clipping the artery first and then the vein, this can reduce the size of the spleen in an important percentage. This is specially useful when dealing with great size spleen. Then an endobag is used to retrieve the spleen after being morcellated through the umbilical incision. A drain, exteriorized through the lateral 5 mm trocar is used routinely. The patient is placed in lateral decubitus, to enter the abdomen using a trans-umbilical open technique and a 12 mm trocar is placed. Through it a 10 mm 30o scope is inserted. A subcostal 5 mm trocar is placed under direct vision at the level of the anterior axillary line and another 5 mm port is inserted at the midepigastric region. Using a 5 mm harmonic scalpel (Harmonic Ace, Ethicon) and 5 mm instruments, access is gained to the lesser sac by dividing the gastrosplenic ligament and short vessels until the upper pole of the spleen. The splenic flexure of the colon is mobilized to get the lower pole of the spleen freed. The posterior splenorenal ligament is then freed. Once the spleen is completely dissected free from all of its hilum attachments, splenic artery and vein are clipped with hemolocks and divided with scissors. It is suggested clipping the artery first and then the vein, this can reduce the size of the spleen in an important percentage. This is specially useful when dealing with great size spleen. Then an endobag is used to retrieve the spleen after being morcellated through the umbilical incision. A drain, exteriorized through the lateral 5 mm trocar is used routinely.

CONCLUSION

The use of reduce port minimizes abdominal trauma and has the hypothetical advantages of shorter postoperative stay, greater pain control, and better cosmesis. Laparoscopic splenectomy for giant cysts by using reduce port trocars is safe and feasible and less invasive.

INTRODUCTION: Splenectomy was initially described for hereditary spherocytosis by Sutherland and Burghard in 1910 and for idiopathic thrombocytopenic purpura by Kaznelson in 19168. It has been well recognized as an effective cure for hematologic disorders, better than medical treatment. The first successful laparoscopic splenectomy was performed by Delaitre and Maignien in 19917 , 8. After that, laparoscopic splenectomy has become one of the most frequently performed laparoscopic solid organ procedures. Laparoscopic splenectomy is emerging as the gold standard for the management of various hematologic disorders. Minimally invasive surgery has earned boundless acceptance. The enthusiasm to limit the trauma of large incisions has been the incentive of the development of minimally invasive surgery during the past century. Numerous technology and equipment in laparoscopy have been emerging. It started gradually reducing and repositioning trocars during laparoscopic surgery. It began with repositioning the subxiphoid trocar into the umbilicus. Less trocars equals less abdominal trauma. In small spleens a single port laparoscopic surgery can be performed. In large spleens, less trocars can be used. The aim of this artcile is to details the splenectomy procedure using only three trochars. METHOD: Surgical technique The patient is placed in lateral decubitus, to enter the abdomen using a trans-umbilical open technique and a 12 mm trocar is placed. Through it a 10 mm 30o scope is inserted. A subcostal 5 mm trocar is placed under direct vision at the level of the anterior axillary line and another 5 mm port is inserted at the midepigastric region. Using a 5 mm harmonic scalpel (Harmonic Ace, Ethicon) and 5 mm instruments, access is gained to the lesser sac by dividing the gastrosplenic ligament and short vessels until the upper pole of the spleen. The splenic flexure of the colon is mobilized to get the lower pole of the spleen freed. The posterior splenorenal ligament is then freed. Once the spleen is completely dissected free from all of its hilum attachments, splenic artery and vein are clipped with hemolocks and divided with scissors. It is suggested clipping the artery first and then the vein, this can reduce the size of the spleen in an important percentage. This is specially useful when dealing with great size spleen. Then an endobag is used to retrieve the spleen after being morcellated through the umbilical incision. A drain, exteriorized through the lateral 5 mm trocar is used routinely. The patient is placed in lateral decubitus, to enter the abdomen using a trans-umbilical open technique and a 12 mm trocar is placed. Through it a 10 mm 30o scope is inserted. A subcostal 5 mm trocar is placed under direct vision at the level of the anterior axillary line and another 5 mm port is inserted at the midepigastric region. Using a 5 mm harmonic scalpel (Harmonic Ace, Ethicon) and 5 mm instruments, access is gained to the lesser sac by dividing the gastrosplenic ligament and short vessels until the upper pole of the spleen. The splenic flexure of the colon is mobilized to get the lower pole of the spleen freed. The posterior splenorenal ligament is then freed. Once the spleen is completely dissected free from all of its hilum attachments, splenic artery and vein are clipped with hemolocks and divided with scissors. It is suggested clipping the artery first and then the vein, this can reduce the size of the spleen in an important percentage. This is specially useful when dealing with great size spleen. Then an endobag is used to retrieve the spleen after being morcellated through the umbilical incision. A drain, exteriorized through the lateral 5 mm trocar is used routinely. Surgical technique: The patient is placed in lateral decubitus, to enter the abdomen using a trans-umbilical open technique and a 12 mm trocar is placed. Through it a 10 mm 30o scope is inserted. A subcostal 5 mm trocar is placed under direct vision at the level of the anterior axillary line and another 5 mm port is inserted at the midepigastric region. Using a 5 mm harmonic scalpel (Harmonic Ace, Ethicon) and 5 mm instruments, access is gained to the lesser sac by dividing the gastrosplenic ligament and short vessels until the upper pole of the spleen. The splenic flexure of the colon is mobilized to get the lower pole of the spleen freed. The posterior splenorenal ligament is then freed. Once the spleen is completely dissected free from all of its hilum attachments, splenic artery and vein are clipped with hemolocks and divided with scissors. It is suggested clipping the artery first and then the vein, this can reduce the size of the spleen in an important percentage. This is specially useful when dealing with great size spleen. Then an endobag is used to retrieve the spleen after being morcellated through the umbilical incision. A drain, exteriorized through the lateral 5 mm trocar is used routinely. RESULT: This technique was used in a 15 years old female refered due to epigastric and left upper quadrant pain. An abdominal ultrasound was performed in order to find gallblader stones, but a giant cyst located in the spleen was found. Laboratory tests were normal. The CT scan showed a giant cyst which squeeze the stomach (Figure 1). The patient was scheduled for surgery, and a laparoscopic approach was performed. FIGURE 1- CT scan showed a giant cyst which squeeze the stomach Lateral positioning was used and with open technique the abdomen was entered. A 10 mm trocar was inserted at the umbilicus and two more trocars were placed in the left upper quadrant. Once the abdomen was entered a gian cyst located in the spleen was observed (Figure 2). The ligaments were divided with electronic shears (Figure 3 A) and the splenic vein and artery were ligated by using hemolocks (Figure 3 B-D) FIGURE 2- Gian cyst located in the spleen is observed when entering the abdomen FIGURE 3- Technical steps: A) ligaments were divided with electronic shears; B, C and D) splenic vein and artery being ligated by using hemolocks Finally the spleen was placed into the retrieval bag and removed from the abdominal cavity before morcellating it with ringed forceps. In the Figure 4, the spleen and cyst are observed after its resection. FIGURE 4- Spleen and cyst are observed after resection The patient tolerated well the procedure, with an unremarkable postoperative period. She was discharge home 72 h after the surgery with all the vaccines set (Figure 5). FIGURE 5- Final aspect of the operation, and a drain exteriorized through the lateral 5 mm trocar is used routinely DISCUSSION: The first successful laparoscopic splenectomy was performed by Delaitre and Maignien in 19917 , 8. After that, laparoscopic splenectomy has become one of the most frequently performed laparoscopic solid organ procedures. Regarding the surgical technique, the lateral positioning is the preferred approach for laparoscopic splenectomy because the gravity effect pulls down the colon, stomach and omentum and allows better visualization of the spleen. Three left subcostal ports are usually adequate for normal-sized spleens. In the presence of splenomegaly, ports are optimally positioned 4 cm below the inferior tip of the spleen, parallel to the left costal margin, but within reach of the diaphragm. If the spleen is extremely large, the trocars may have to be placed substantially more inferiorly than normal, creating the need for an additional port posteriorly. This port allows for lateral retraction of the spleen and can facilitate access to the diaphragmatic attachments. Additional ports are placed under laparoscopic guidance. Visualization and efficiency are optimized by exchanging the camera between the medial and lateral ports, while the surgeon operates with both hands8. Supermassive spleens are most effectively managed with a hand-assisted approach. This approach utilizes similar patient positioning but uses a hand-assisted device to facilitate insertion of the surgeon's nondominant hand into the abdominal cavity while maintaining pneumoperitoneum. This technique allows for improved tissue handling and atraumatic manipulation of the enlarged spleen. For patients with supermassive spleens, lateral positioning is altered slightly. In these cases, the patient is placed supine with the left side elevated at 45°. This allows the surgeon to take advantage of gravity while also providing comfortable access through the hand-assist incision. Depending upon the hand dominance of the surgeon, the hand-assist device can be placed in either a midline (right-hand dominant) or a subcostal position (left-hand dominant) through a 7 to 8 cm incision which is located 2 to 4 cm caudal to the inferior pole of the enlarged spleen. In both situations, the surgeon stands on the right side of the patient. The nondominant hand is inserted through the hand-assist device and provides medial retraction and rotation of the spleen. The hilar pedicle is transected with an endoscopic gastrointestinal anastomotic stapler utilizing a vascular cartridge. The spleen is placed into an appropriately-sized impermeable retrieval bag. This bag must be strong enough to avoid rupture during morcellation and extraction of the specimen. Placing the spleen into the retrieval bag can be one of the most time-consuming and challenging aspects of the operation but it's necessary in order to avoid possible splenosis which is unwanted in many hematologic disease. The patient is placed into Trendelenburg's position and the spleen is gradually directed into the bag. After the spleen is within the retrieval bag, the opening of the bag is delivered through the largest port site or the hand-assist incision and the spleen is morcellated with ringed forceps. The most common hematologic disorder that requires splenectomy is the idiopathic thrombocytopenic purpura where surgery is indicated in patients with refractory symptomatic thrombocytopenia after 4 to 6 weeks of medical therapy, patients requiring toxic doses of steroids to achieve remission, and patients who relapse following an initial response to steroid therapy. Hereditary spherocytosis (hereditary hemolytic anemia), is curable in approximately 90% of patients after splenectomy. Surgery is indicated for all patients with hereditary spherocytosis and splenomegaly, for patients with symptoms of severe hemolytic anemia or mild hemolytic anemia and concomitant gallstones, and for patients with cholelithiasis in siblings8. Other indications for splenectomy are HIV-related, thrombocytopenic purpura systemic, lupus erythematosus-related, thrombocytopenic purpura, thrombotic thrombocytopenic purpura, autoimmune hemolytic anemias, Hodgkin's disease, non-Hodgkin's lymphoma, chronic lymphocytic leukemia and Hairy cell leukemia8. The contraindications to laparoscopic splenectomy can be divided into: absolute contraindications like, severe cardiopulmonary disease or cirrhosis with portal hypertension. And relative contraindication are previous abdominal surgery (open Hasson technique is mandatory) and some authors include massive splenectomy1. Laparoscopic splenectomy is emerging as the gold standard for the management of various hematologic disorders. Since the first laparoscopic splenectomies were performed in adults (1991) and children (1993), laparoscopic splenectomy has become the gold standard for the elective removal of normal sized spleens. It clearly has been shown to have less morbidity and postoperative pain, shorter hospital stay, earlier return of bowel function, and superior cosmesis compared with open splenectomy5 , 6. Although laparoscopic splenectomy is superior to open splenectomy for normal sized spleens, splenomegaly has been considered a contraindication to the laparoscopic approach because of difficulties with bleeding and removal from the abdomen. Over the last years, however, several authors have reported laparoscopic removal of enlarged spleens. With the development of hand-assisted laparoscopic surgery, the retraction of massive spleens was technically feasible. Recent studies, have shown that, aside from operative time, there is no difference in transfusions, length of stay, morbidity, or conversion rate with laparoscopic splenectomy for large spleens compared with normal sized spleens. Laparoscopic techniques, surgical skills, and instrumentation have improved, so are safety and efficacy even in the presence of splenomegaly. Authors like Kent W. Kercher et al.8, made a study on 177 patients who underwent laparoscopic splenectomy where forty-nine patients (28%) were identified as having massive splenomegaly. They defined massive splenomegaly as a craniocaudal length ≥17 cm or a weight ≥600 g. Spleens greater than 22 cm in craniocaudal length, 19 cm in width, or a weight greater than 1600 g were defined as "supermassive." And in children, splenic size greater than four times normal for age was defined as massive. They recommend that most surgeons acquire their initial laparoscopic experience with normal-sized spleens prior to attempting laparoscopic splenectomy of splenomegaly. And the author conclude that laparoscopic splenectomy has become the gold standard for elective splenectomy in patients with normal-sized spleens and laparoscopic splenectomy in the setting of massive splenomegaly is safe and effective providing distinct advantages over the open operation. In the presence of supermassive splenomegaly, the use of hand-assisted laparoscopic surgery maintains the benefits of a minimally invasive approach8. Other authors like Arin K. Greene 7, have used the Lahey bag to remove a massive spleen. The author describe that this technique facilitate the removal of massively enlarged spleens laparoscopically (>1,000 g), because a large abdominal incision to remove the spleen is not required. The spleen is broken up while in the Lahey bag so the risk of splenosis is eliminated. Because of suitable single port laparoscopy equipment is not available to may centers due to cost, Colon et al.3 developed a technique where additional instrumentation was kept to a minimum. Boone et al.2 demonstrate in their series that single port splenectomy is safe and feasible in an unselected patient population, their patient population included patients with prior surgery, obese patients, medical comorbidities, splenomegaly, and severe thrombocytopenia. In their study they also compared it to standard laparoscopic splenectomy. The results showed no statistically significance in morbidity and mortality between both groups. Analysis of postoperative pain medication requirement revealed that the single incision patients required fewer narcotics, but this did not reach statistical significance either. Single port splenectomy was associates with a significantly lower open conversion, shorter operative time, and similar median estimated blood loss. Overall, the study demonstrated that it is at least equivalent to standard laparoscopic splenectomy. They enounced that single port splenectomy is an appropriate procedure that can be done safely and may lead to higher patient satisfaction compared to laparoscopic splenectomy. Also, they stated, while articulating instruments and laparoscopes may offer technical advantages, they are not completely necessary for performing single port splenectomy. It is a safe and feasible technique to perform splenectomy for small spleens, but in giant and massive spleens, the reduce port laparoscopic surgery technique is a better choice. Laparoscopic splenectomy has become the gold standard for normal size spleens. Massive splenomegaly is not a contraindication for laparoscopic surgery moreover with the develop of the hand assisted technique that is feasible and safe. Morbidity and mortality are equivalent when coparing with regular laparoscopic splenectomy. CONCLUSION: The use of reduce port minimizes abdominal trauma and has the hypothetical advantages of shorter postoperative stay, greater pain control, and better cosmesis. Laparoscopic splenectomy for giant cysts by using reduce port trocars is safe and feasible and less invasive.

Background: Delaitre and Maignien performed the first successful laparoscopic splenectomy in 1991. After that, laparoscopic splenectomy has become one of the most frequently performed laparoscopic solid organ procedures. Methods: A reduce port laparoscopic splenectomy was performed by using a 10 mm and two 5 mm trocars. To entered the abdomen a trans-umbilical open technique was done and a 10 mm trocar was placed. A subcostal 5 mm trocar was placed under direct vision at the level of the anterior axillary line and another 5 mm port was inserted at the mid-epigastric region. Once it was completely dissected and freed from all of its attachments the hilum, splenic artery and vein, was clipped with hem-o-lock and divided with scissors. Then an endobag was used to retrieve the spleen after being morcellated trough the umbilical incision. Results: This technique was used in a 15 years old female with epigastric and left upper quadrant pain. An abdominal ultrasound demonstrated a giant cyst located in the spleen. Laboratory tests findings were normal. The CT scan was also done, and showed a giant cyst, which squeeze the stomach. The patient tolerated well the procedure, with an unremarkable postoperative. She was discharge home 72 h after the surgery. Conclusions: The use of reduce port minimizes abdominal trauma and has the hypothetical advantages of shorter postoperative stay, greater pain control, and better cosmesis. Laparoscopic splenectomy for giant cysts by using reduce port trocars is safe and feasible and less invasive.

INTRODUCTION: Splenectomy was initially described for hereditary spherocytosis by Sutherland and Burghard in 1910 and for idiopathic thrombocytopenic purpura by Kaznelson in 19168. It has been well recognized as an effective cure for hematologic disorders, better than medical treatment. The first successful laparoscopic splenectomy was performed by Delaitre and Maignien in 19917 , 8. After that, laparoscopic splenectomy has become one of the most frequently performed laparoscopic solid organ procedures. Laparoscopic splenectomy is emerging as the gold standard for the management of various hematologic disorders. Minimally invasive surgery has earned boundless acceptance. The enthusiasm to limit the trauma of large incisions has been the incentive of the development of minimally invasive surgery during the past century. Numerous technology and equipment in laparoscopy have been emerging. It started gradually reducing and repositioning trocars during laparoscopic surgery. It began with repositioning the subxiphoid trocar into the umbilicus. Less trocars equals less abdominal trauma. In small spleens a single port laparoscopic surgery can be performed. In large spleens, less trocars can be used. The aim of this artcile is to details the splenectomy procedure using only three trochars. CONCLUSION: The use of reduce port minimizes abdominal trauma and has the hypothetical advantages of shorter postoperative stay, greater pain control, and better cosmesis. Laparoscopic splenectomy for giant cysts by using reduce port trocars is safe and feasible and less invasive.

Background: Delaitre and Maignien performed the first successful laparoscopic splenectomy in 1991. After that, laparoscopic splenectomy has become one of the most frequently performed laparoscopic solid organ procedures. Methods: A reduce port laparoscopic splenectomy was performed by using a 10 mm and two 5 mm trocars. To entered the abdomen a trans-umbilical open technique was done and a 10 mm trocar was placed. A subcostal 5 mm trocar was placed under direct vision at the level of the anterior axillary line and another 5 mm port was inserted at the mid-epigastric region. Once it was completely dissected and freed from all of its attachments the hilum, splenic artery and vein, was clipped with hem-o-lock and divided with scissors. Then an endobag was used to retrieve the spleen after being morcellated trough the umbilical incision. Results: This technique was used in a 15 years old female with epigastric and left upper quadrant pain. An abdominal ultrasound demonstrated a giant cyst located in the spleen. Laboratory tests findings were normal. The CT scan was also done, and showed a giant cyst, which squeeze the stomach. The patient tolerated well the procedure, with an unremarkable postoperative. She was discharge home 72 h after the surgery. Conclusions: The use of reduce port minimizes abdominal trauma and has the hypothetical advantages of shorter postoperative stay, greater pain control, and better cosmesis. Laparoscopic splenectomy for giant cysts by using reduce port trocars is safe and feasible and less invasive.

[CONTENT] Laparoscopy | Splenectomy | Surgery [SUMMARY]

[CONTENT] Laparoscopy | Splenectomy | Surgery [SUMMARY]

[CONTENT] Laparoscopy | Splenectomy | Surgery [SUMMARY]

[CONTENT] Laparoscopy | Splenectomy | Surgery [SUMMARY]

[CONTENT] Laparoscopy | Splenectomy | Surgery [SUMMARY]

[CONTENT] Adolescent | Cysts | Epithelium | Female | Humans | Laparoscopy | Splenectomy | Splenic Diseases [SUMMARY]

[CONTENT] Adolescent | Cysts | Epithelium | Female | Humans | Laparoscopy | Splenectomy | Splenic Diseases [SUMMARY]

[CONTENT] Adolescent | Cysts | Epithelium | Female | Humans | Laparoscopy | Splenectomy | Splenic Diseases [SUMMARY]

[CONTENT] Adolescent | Cysts | Epithelium | Female | Humans | Laparoscopy | Splenectomy | Splenic Diseases [SUMMARY]

[CONTENT] Adolescent | Cysts | Epithelium | Female | Humans | Laparoscopy | Splenectomy | Splenic Diseases [SUMMARY]

[CONTENT] spleens laparoscopic splenectomy | compared laparoscopic splenectomy | attempting laparoscopic splenectomy | laparoscopic splenectomy stated | laparoscopic splenectomy emerging [SUMMARY]

[CONTENT] spleens laparoscopic splenectomy | compared laparoscopic splenectomy | attempting laparoscopic splenectomy | laparoscopic splenectomy stated | laparoscopic splenectomy emerging [SUMMARY]

[CONTENT] spleens laparoscopic splenectomy | compared laparoscopic splenectomy | attempting laparoscopic splenectomy | laparoscopic splenectomy stated | laparoscopic splenectomy emerging [SUMMARY]

[CONTENT] spleens laparoscopic splenectomy | compared laparoscopic splenectomy | attempting laparoscopic splenectomy | laparoscopic splenectomy stated | laparoscopic splenectomy emerging [SUMMARY]

[CONTENT] spleens laparoscopic splenectomy | compared laparoscopic splenectomy | attempting laparoscopic splenectomy | laparoscopic splenectomy stated | laparoscopic splenectomy emerging [SUMMARY]

[CONTENT] spleen | laparoscopic | splenectomy | mm | laparoscopic splenectomy | spleens | placed | technique | port | patients [SUMMARY]

[CONTENT] spleen | laparoscopic | splenectomy | mm | laparoscopic splenectomy | spleens | placed | technique | port | patients [SUMMARY]

[CONTENT] spleen | laparoscopic | splenectomy | mm | laparoscopic splenectomy | spleens | placed | technique | port | patients [SUMMARY]

[CONTENT] spleen | laparoscopic | splenectomy | mm | laparoscopic splenectomy | spleens | placed | technique | port | patients [SUMMARY]

[CONTENT] spleen | laparoscopic | splenectomy | mm | laparoscopic splenectomy | spleens | placed | technique | port | patients [SUMMARY]

[CONTENT] laparoscopic | splenectomy | surgery | performed | laparoscopic splenectomy | invasive surgery | minimally invasive surgery | repositioning | trocars | minimally invasive [SUMMARY]

[CONTENT] mm | spleen | mm trocar | trocar | placed | harmonic | artery vein | size spleen | trocar placed | pole spleen [SUMMARY]

[CONTENT] reduce port | reduce | port | minimizes | abdominal trauma hypothetical advantages | feasible invasive | control better cosmesis laparoscopic | minimizes abdominal trauma hypothetical | minimizes abdominal trauma | minimizes abdominal [SUMMARY]

[CONTENT] spleen | mm | splenectomy | laparoscopic | figure | laparoscopic splenectomy | placed | mm trocar | trocar | surgery [SUMMARY]

[CONTENT] spleen | mm | splenectomy | laparoscopic | figure | laparoscopic splenectomy | placed | mm trocar | trocar | surgery [SUMMARY]

[CONTENT] Delaitre | Maignien | first | 1991 ||| [SUMMARY]

[CONTENT] 10 mm and ||| 10 mm ||| 5 mm | 5 mm ||| ||| [SUMMARY]

[CONTENT] ||| [SUMMARY]

[CONTENT] Delaitre | Maignien | first | 1991 ||| ||| 10 mm and ||| 10 mm ||| 5 mm | 5 mm ||| ||| ||| ||| 15 years old ||| ||| ||| CT ||| ||| 72 ||| ||| [SUMMARY]

[CONTENT] Delaitre | Maignien | first | 1991 ||| ||| 10 mm and ||| 10 mm ||| 5 mm | 5 mm ||| ||| ||| ||| 15 years old ||| ||| ||| CT ||| ||| 72 ||| ||| [SUMMARY]

Association between depression, happiness, and sleep duration: data from the UAE healthy future pilot study.

36271400

The United Arab Emirates Healthy Future Study (UAEHFS) is one of the first large prospective cohort studies and one of the few studies in the region which examines causes and risk factors for chronic diseases among the nationals of the United Arab Emirates (UAE). The aim of this study is to investigate the eight-item Patient Health Questionnaire (PHQ-8) as a screening instrument for depression among the UAEHFS pilot participants.

BACKGROUND

The UAEHFS pilot data were analyzed to examine the relationship between the PHQ-8 and possible confounding factors, such as self-reported happiness, and self-reported sleep duration (hours) after adjusting for age, body mass index (BMI), and gender.

METHODS

Out of 517 participants who met the inclusion criteria, 487 (94.2%) participants filled out the questionnaire and were included in the statistical analysis using 100 multiple imputations. 231 (44.7%) were included in the primary statistical analysis after omitting the missing values. Participants' median age was 32.0 years (Interquartile Range: 24.0, 39.0). In total, 22 (9.5%) of the participant reported depression. Females have shown significantly higher odds of reporting depression than males with an odds ratio = 3.2 (95% CI:1.17, 8.88), and there were approximately 5-fold higher odds of reporting depression for unhappy than for happy individuals. For one interquartile-range increase in age and BMI, the odds ratio of reporting depression was 0.34 (95% CI: 0.1, 1.0) and 1.8 (95% CI: 0.97, 3.32) respectively.

RESULTS

Females are more likely to report depression compared to males. Increasing age may decrease the risk of reporting depression. Unhappy individuals have approximately 5-fold higher odds of reporting depression compared to happy individuals. A higher BMI was associated with a higher risk of reporting depression. In a sensitivity analysis, individuals who reported less than 6 h of sleep per 24 h were more likely to report depression than those who reported 7 h of sleep.

CONCLUSION

9587590

Introduction

Depression is defined as a set of disorders ranging from mild to moderate to severe [1]. It is widely recognized as a major public health problem worldwide [2]. Reports from the Global Burden of Diseases declared major depressive disorder as one of the top three causes of disability-adjusted life years [3, 4]. The effect of depression has been extensively studied on the individual’s daily functioning and productivity. This is directly reflected in increased economic costs per capita [3, 4]. For example, in Catalonia in 2006, the average annual cost of an adult with depression was close to 1800 Euros, and the total annual cost of depression was 735.4 million Euros. These costs were linked directly to primary care, mental health specialized care, hospitalization, and pharmacological care, as well as indirect costs due to productivity loss, temporary and permanent disability [4]. To measure the level of depression in non-clinical populations; clinical and epidemiological studies have often used the established and validated eight-item Patient Health Questionnaire scale (PHQ-8) instead of the nine- item Patient Health Questionnaire scale (PHQ-9) [5]. As has been confirmed by previous studies, the scale can detect major depression with sensitivity and a specificity of 88%, to classify subjects into depressed or non-depressed, respectively [6–8]. Regionally, in Jordan, Lebanon, Syria and Afghanistan, the cutoff point of 10 was used [9–12]. Similarly, in UAE, studies have mostly used the cutoff point of 10 [13–16]. Recent studies have focused on exploring the methods of improving individual and environmental effects on disability related to depression by investigating its triggers and associations [2, 10]. Factors such as sleep duration and self-reported happiness are evidenced to be predictors for depression [17, 18]. Sleep duration can be considered a risk factor for lower well-being [19]. The number of sleep hours has also been found to have a causal relationship with depression as well as self-reported happiness [20]. For example, shorter sleep duration might lead to lower positive emotions such as self-reported happiness and showed stronger associations with negative emotional affect [21, 22]. One question that needs to be asked, however, is the nature and strength of the association between these three variables respectively. Sleep duration is determined by how many hours an individual sleeps over 24 h. Individuals with depression often have poor sleep status including abnormal REM (rapid eye movement), and insomnia (difficulty falling asleep or staying asleep). Abnormal REM may contribute to the development of altered emotional processing in depression [23]. It was reported that people with insomnia might have a ten-fold higher risk of developing depression in contrast to people who get a good night’s sleep [24]. Also, 75% of depressed individuals will have trouble falling asleep or staying asleep [25]. A bidirectional relationship between sleep duration and reported depression has been investigated [24, 26]. Such studies are unsatisfactory because they do not explore the association between major depressive disorder and sleep duration based on specific population demographics. In this study we are considering other sociodemographic factors such as age, gender, and marital status. Interest in studying happiness in the context of mental health status (such as depression and anxiety) has been growing recently [27]. The findings of some research papers suggest that self-reported happiness is a potential factor in the prevention and management of depression [27, 34]. Happiness is correlated to a person’s ability to approach situations in a less stressful manner and to an individual’s capacity to perceive and control their own feelings. This indicates that higher happiness levels may have a protective effect on depression [27]. Well-being has been defined as the combination of feeling good and functioning well [28]; conversely, quality of life could be defined as an individual’s satisfaction with his or her actual life compared with his or her ideal life. Evaluation of the quality of life depends on one’s value system [29]. Moreover, studies have shown that individuals use various chronically accessible and stable sources of information when making life satisfaction judgments [30]. Yet, well-being, quality of life and life satisfaction are multidimensional constructs which include complex cognitive evaluation processes and cannot be adequately assessed by using a single item or question [31]. Unlike quality of life, which typically requires detailed assessments to ascertain [32], happiness is easier to evaluate using a single item question [33]. Investigating the association between happiness and depression would add valuable information to public health research as the relationship between happiness and depression is observed to be bidirectional (i.e., one variable can predict the other, and people might not report depression but are more likely to report feeling unhappy) [34]. In addition, there are some factors which confound with sleep duration and self-reported happiness, such as age, gender, and marital status [36, 37]. For example, reviews reported that a prognosis of depression was improved with increasing age [38]. Conversely, another study showed that older age was associated with worsening of depressive symptoms [37]. Moreover, depressive symptoms could worsen among widowed individuals as their age increased [40]. Some research findings reflected that females are more likely to perceive depression compared to males [38, 41]. Other studies showed significant interplay between marital status and depressive symptoms [40, 42]. For example, being unmarried could lead to perceiving and developing depressive symptoms [43], while a worsening in depression symptoms among married individuals could lead to separation or being unmarried [42, 44]. So, further exploration is required to study the interplay of these variables together. The United Arab Emirates (UAE) is a high-income developed country which has undergone a rapid epidemiological transition from a traditional semi-nomadic society to a modern affluent society with a lifestyle characterized by over-consumption of energy-dense foods and low physical activity [47]. The UAE was ranked 15th out of 157 countries included in the WHO World Happiness Report with a score of 7.06 [48]. Despite this, depression has been identified as the third leading cause of disability in the UAE [3]. Nevertheless, there are few studies in the Gulf region examining the relationship between happiness, sleep duration and depression [36, 49, 50]. Therefore, studying the association between depression, happiness, and sleep in the United Arab Emirates (UAE) population would be of interest to the public health field in this part of the world. This study aimed to examine the relationship between depression, self-reported happiness, and self-reported sleep duration after adjusting for age and gender using the UAE Healthy Future Study (UAEHFS) pilot data.

Results

Of 517 participants who consented to participate in the UAEHF pilot study, 487 (94.2%) had completed questionnaire data [47]. Figure 1 describes all key phases from recruitment up to inclusion in the study and in the statistical analysis. After omitting missing values, 231 (44.7%) participants were included in the main statistical analysis. However, 487 (94.2%) participants were included in the sensitivity analysis (using 100 multiple imputations) [73]. The median age of the UAEHFS pilot data participants was 32.0 years (Interquartile Range: 24.0–39.0). The percentage of females included in the study was 32%, which represented the UAE population well [59]. Therefore, we did not make any adjustments for gender bias. Fig. 1Flow chart of UAEHF pilot study (describes all the key phases from the recruitment up to the inclusion in the study analysis) Flow chart of UAEHF pilot study (describes all the key phases from the recruitment up to the inclusion in the study analysis) Note: This Figure represents the data of the included participants in the main statistical analysis after omitting missing values. The number of observed values (%) of each categorical variable was presented by the PHQ-8 categories in Table 1, where the majority of the study participants (90.5%) had PHQ-8 score less than 10. When categorical groups were compared within the PHQ-8 depression group (< 10 versus ≥ 10), there was only a statistically significant difference between females versus males with a Fisher exact p-value = 0.028. Table 1 shows that there was a statistically significant difference in the age distribution across the depression groups (p-value = 0.012). There was no statistically significant difference in the BMI measurements between the PHQ-8 groups (p-value = 0.44). The result of the primary analysis showed that females have statistically significantly greater odds of reporting depression compared to males OR = 3.2 (95% CI:1.1, 8.9), p-value = 0.024 (Table 2). Furthermore, there was an approximately 5-fold greater odds of reporting depression for unhappy than for happy individuals. However, this was not statistically significant (p-value = 0.10). Similarly, the sleep duration and marital status were both not statistically significant (p-value of 0.284 and 0.205, respectively; Table 2). However, in a sensitivity analysis, people who reported sleep duration of less than 6 h were more likely to report depression as compared to the people who reported sleep of 7 h OR (95% CI) of 2.6 (1.0, 6.4), p-value = 0.040 and 1.136 (1.015, 1.272), p-value = 0.027. Table 2 The results of the primary analysis of the fitted multivariate logistic regression model VariableOR (95% CI)PbHappiness - HappyReferenceHappiness - Unhappy5.5 (0.7, 42.2)0.100Age (linear)0.3 (0.1, 1.0)a0.056BMI (linear)1.8 (1.0, 3.3)a0.064Gender = MalesReferenceGender = Females3.2 (1.2, 8.9)0.024Sleep = 7ReferenceSleep < 64.0 (0.9, 17.4)0.061Sleep = 61.3 (0.2, 6.9)0.775Sleep = 81.5 (0.3, 7.3)0.585Sleep > 81.2 (0.2, 9.2)0.848MarriedReferenceSingle0.9 (0.3, 3.0)0.8Others0.2 (0.02, 3.3)0.3Note: aInterquartile-range odds ratio for age and BMI, which compares the 3rd quartile with the 1st quartile, and simple odds ratios for the categorical predictors (happiness, gender, sleep and marital status) compare each group with the reference group (the largest group) with corresponding 95% confidence interval (95% CI). bWald’s p-values are presented for each variable The results of the primary analysis of the fitted multivariate logistic regression model Note: aInterquartile-range odds ratio for age and BMI, which compares the 3rd quartile with the 1st quartile, and simple odds ratios for the categorical predictors (happiness, gender, sleep and marital status) compare each group with the reference group (the largest group) with corresponding 95% confidence interval (95% CI). bWald’s p-values are presented for each variable To compute Interquartile Range (IQR) odds ratios for continuous variables, the age and BMI variables were divided by their IQR values of 15 and 8.7, respectively [54, 56]. Table 2 shows that for one interquartile-range increase in the age and BMI, the IQR-OR was 0.34 (95% CI: 0.12, 1.0) and 1.8 (95% CI: 0.97, 3.3), respectively, where both results were statistically significant with a p-value of 0.056 and 0.064 individually. The results of the sensitivity analysis using 100 multiple imputations (Table 3) were approximately similar to the result of the multivariate logistic ordinal regression analyses using omitted data (Table 2). However, the happiness variable was statistically significant OR = 5.1 (95%CI: 1.7, 15.7, p-value = 0.005), and there was a statistically significant difference between the sleep variable for less than six hours as compared with seven hours OR = 2.6 (95% CI: 1.0, 6.4, p-value = 0.04). Supplementary Fig. 1 illustrates the percentages of missing values in the variables included in this statistical analysis. Marital status had the lowest number of missing values (5.3%), followed by BMI (12.7%) and the Sleep variable with 12.9% missing values. The Happiness variable had 21.1% missing values. The PHQ-8 variables had 32.6–35.1% missing values. Supplementary Table 1 shows the percentages of the “Prefer not to answer (DA)” and “Do not know (UN)” in the eight PHQ questions, where the percentages vary between 16 and 18.5% between the eight questions. The percentages of “Prefer not to answer (DA)” and “Do not know (UN)” in the happiness variable was 2.5% and 2.3%, respectively, which were also considered missing values, although these were not missing at random and can be correlated with one of the PHQ-8 answers. However, this has been addressed in the multiple imputation analysis. Table 3 Results of the sensitivity analysis using multivariate logistic regression models with a sample size N = 487 VariableOR (95% CI)PbHappiness – HappyReferenceHappiness - Unhappy5.1 (1.7, 15.7)0.004Age (linear)0.5 (0.3, 1.0)a0.060BMI (linear)1.7 (1.1, 2.5)a0.021Gender = MalesReferenceGender = Females1.6 (0.9, 3.0)0.154Sleep = 7ReferenceSleep < 62.6 (1.0, 6.4)0.040Sleep = 60.9 (0.3, 2.6)0.887Sleep = 80.9 (0.3, 2.4)0.779Sleep > 81.1 (0.3, 3.7)0.853MarriedReferenceSingle0.9 (0.4, 2.1)0.885Others0.7 (0.2, 2.9)0.597Note: aInterquartile-range odds ratio for Age and BMI, which compares the 3rd quartile with the 1st quartile, and simple odds ratios for the categorical predictors (happiness, gender, sleep and marital status) compare each group with the reference group (the largest group) with corresponding 95% confidence interval (95% CI). bWald’s p-values are presented for each variable. Results were summarized using Rubin’s rules Results of the sensitivity analysis using multivariate logistic regression models with a sample size N = 487 Note: aInterquartile-range odds ratio for Age and BMI, which compares the 3rd quartile with the 1st quartile, and simple odds ratios for the categorical predictors (happiness, gender, sleep and marital status) compare each group with the reference group (the largest group) with corresponding 95% confidence interval (95% CI). bWald’s p-values are presented for each variable. Results were summarized using Rubin’s rules An additional sensitivity analysis was performed using the ordinal PHQ-8 score as an outcome in a multivariate linear regression model (see supplementary Table 1). The result of this sensitivity analysis was very similar to the sensitivity analysis in Table 3.

Conclusion

The results of this study indicate that females are more likely to report depression compared to males. Older age is associated with a decrease in self-reported depression. Unhappy individuals are approximately 5-fold higher odds to report depression compared to happy individuals. BMI was positively associated with reporting depression. The result of the sensitivity analysis shows that individuals who sleep less than 6 h per day are more likely to report depression compared to those who sleep 7 h per day. The results of this study have a potential value for researchers and public health professionals as it presents novel data on the PHQ-8 score in a healthy UAE population, which has not been explored before. Furthermore, the results of this study can help contribute to the knowledge base on current and potential population mental health impact in the UAE and Gulf Region.

Introduction: Depression is defined as a set of disorders ranging from mild to moderate to severe [1]. It is widely recognized as a major public health problem worldwide [2]. Reports from the Global Burden of Diseases declared major depressive disorder as one of the top three causes of disability-adjusted life years [3, 4]. The effect of depression has been extensively studied on the individual’s daily functioning and productivity. This is directly reflected in increased economic costs per capita [3, 4]. For example, in Catalonia in 2006, the average annual cost of an adult with depression was close to 1800 Euros, and the total annual cost of depression was 735.4 million Euros. These costs were linked directly to primary care, mental health specialized care, hospitalization, and pharmacological care, as well as indirect costs due to productivity loss, temporary and permanent disability [4]. To measure the level of depression in non-clinical populations; clinical and epidemiological studies have often used the established and validated eight-item Patient Health Questionnaire scale (PHQ-8) instead of the nine- item Patient Health Questionnaire scale (PHQ-9) [5]. As has been confirmed by previous studies, the scale can detect major depression with sensitivity and a specificity of 88%, to classify subjects into depressed or non-depressed, respectively [6–8]. Regionally, in Jordan, Lebanon, Syria and Afghanistan, the cutoff point of 10 was used [9–12]. Similarly, in UAE, studies have mostly used the cutoff point of 10 [13–16]. Recent studies have focused on exploring the methods of improving individual and environmental effects on disability related to depression by investigating its triggers and associations [2, 10]. Factors such as sleep duration and self-reported happiness are evidenced to be predictors for depression [17, 18]. Sleep duration can be considered a risk factor for lower well-being [19]. The number of sleep hours has also been found to have a causal relationship with depression as well as self-reported happiness [20]. For example, shorter sleep duration might lead to lower positive emotions such as self-reported happiness and showed stronger associations with negative emotional affect [21, 22]. One question that needs to be asked, however, is the nature and strength of the association between these three variables respectively. Sleep duration is determined by how many hours an individual sleeps over 24 h. Individuals with depression often have poor sleep status including abnormal REM (rapid eye movement), and insomnia (difficulty falling asleep or staying asleep). Abnormal REM may contribute to the development of altered emotional processing in depression [23]. It was reported that people with insomnia might have a ten-fold higher risk of developing depression in contrast to people who get a good night’s sleep [24]. Also, 75% of depressed individuals will have trouble falling asleep or staying asleep [25]. A bidirectional relationship between sleep duration and reported depression has been investigated [24, 26]. Such studies are unsatisfactory because they do not explore the association between major depressive disorder and sleep duration based on specific population demographics. In this study we are considering other sociodemographic factors such as age, gender, and marital status. Interest in studying happiness in the context of mental health status (such as depression and anxiety) has been growing recently [27]. The findings of some research papers suggest that self-reported happiness is a potential factor in the prevention and management of depression [27, 34]. Happiness is correlated to a person’s ability to approach situations in a less stressful manner and to an individual’s capacity to perceive and control their own feelings. This indicates that higher happiness levels may have a protective effect on depression [27]. Well-being has been defined as the combination of feeling good and functioning well [28]; conversely, quality of life could be defined as an individual’s satisfaction with his or her actual life compared with his or her ideal life. Evaluation of the quality of life depends on one’s value system [29]. Moreover, studies have shown that individuals use various chronically accessible and stable sources of information when making life satisfaction judgments [30]. Yet, well-being, quality of life and life satisfaction are multidimensional constructs which include complex cognitive evaluation processes and cannot be adequately assessed by using a single item or question [31]. Unlike quality of life, which typically requires detailed assessments to ascertain [32], happiness is easier to evaluate using a single item question [33]. Investigating the association between happiness and depression would add valuable information to public health research as the relationship between happiness and depression is observed to be bidirectional (i.e., one variable can predict the other, and people might not report depression but are more likely to report feeling unhappy) [34]. In addition, there are some factors which confound with sleep duration and self-reported happiness, such as age, gender, and marital status [36, 37]. For example, reviews reported that a prognosis of depression was improved with increasing age [38]. Conversely, another study showed that older age was associated with worsening of depressive symptoms [37]. Moreover, depressive symptoms could worsen among widowed individuals as their age increased [40]. Some research findings reflected that females are more likely to perceive depression compared to males [38, 41]. Other studies showed significant interplay between marital status and depressive symptoms [40, 42]. For example, being unmarried could lead to perceiving and developing depressive symptoms [43], while a worsening in depression symptoms among married individuals could lead to separation or being unmarried [42, 44]. So, further exploration is required to study the interplay of these variables together. The United Arab Emirates (UAE) is a high-income developed country which has undergone a rapid epidemiological transition from a traditional semi-nomadic society to a modern affluent society with a lifestyle characterized by over-consumption of energy-dense foods and low physical activity [47]. The UAE was ranked 15th out of 157 countries included in the WHO World Happiness Report with a score of 7.06 [48]. Despite this, depression has been identified as the third leading cause of disability in the UAE [3]. Nevertheless, there are few studies in the Gulf region examining the relationship between happiness, sleep duration and depression [36, 49, 50]. Therefore, studying the association between depression, happiness, and sleep in the United Arab Emirates (UAE) population would be of interest to the public health field in this part of the world. This study aimed to examine the relationship between depression, self-reported happiness, and self-reported sleep duration after adjusting for age and gender using the UAE Healthy Future Study (UAEHFS) pilot data. Materials and methods: Study design This was a pilot prospective cohort study conducted from January 2015 to May 2015. The participants were recruited from two health care centers in Abu Dhabi. Participants completed an online questionnaire including questions on demographic data, PHQ-8 score, self-reported sleep duration, and self-reported happiness score. Physical measurements such as Body Mass Index (BMI) were collected during the recruitment visit. This was a pilot prospective cohort study conducted from January 2015 to May 2015. The participants were recruited from two health care centers in Abu Dhabi. Participants completed an online questionnaire including questions on demographic data, PHQ-8 score, self-reported sleep duration, and self-reported happiness score. Physical measurements such as Body Mass Index (BMI) were collected during the recruitment visit. Participants’ eligibility criteria and recruitment Seven hundred and sixty-nine UAE nationals aged ≥ 18 years were invited to participate voluntarily in the pilot study. Volunteers from the general population with inclusion criteria of age 18 or greater; able to consent; UAE nationals, resident in Abu Dhabi Emirate. All potential participants were given participant information leaflets in either Arabic or English to read and had the opportunity to ask questions prior to completion of the recruitment process. Participants signed an informed consent and were asked to complete a detailed questionnaire. However, 243 invited subjects did not respond, and their reasons for not participating were recorded [45]. Out of 517 participants who met the inclusion criteria, 487 (94.2%) participants filled out the questionnaire and were included in the statistical analysis [47]. Seven hundred and sixty-nine UAE nationals aged ≥ 18 years were invited to participate voluntarily in the pilot study. Volunteers from the general population with inclusion criteria of age 18 or greater; able to consent; UAE nationals, resident in Abu Dhabi Emirate. All potential participants were given participant information leaflets in either Arabic or English to read and had the opportunity to ask questions prior to completion of the recruitment process. Participants signed an informed consent and were asked to complete a detailed questionnaire. However, 243 invited subjects did not respond, and their reasons for not participating were recorded [45]. Out of 517 participants who met the inclusion criteria, 487 (94.2%) participants filled out the questionnaire and were included in the statistical analysis [47]. Study design: This was a pilot prospective cohort study conducted from January 2015 to May 2015. The participants were recruited from two health care centers in Abu Dhabi. Participants completed an online questionnaire including questions on demographic data, PHQ-8 score, self-reported sleep duration, and self-reported happiness score. Physical measurements such as Body Mass Index (BMI) were collected during the recruitment visit. Participants’ eligibility criteria and recruitment: Seven hundred and sixty-nine UAE nationals aged ≥ 18 years were invited to participate voluntarily in the pilot study. Volunteers from the general population with inclusion criteria of age 18 or greater; able to consent; UAE nationals, resident in Abu Dhabi Emirate. All potential participants were given participant information leaflets in either Arabic or English to read and had the opportunity to ask questions prior to completion of the recruitment process. Participants signed an informed consent and were asked to complete a detailed questionnaire. However, 243 invited subjects did not respond, and their reasons for not participating were recorded [45]. Out of 517 participants who met the inclusion criteria, 487 (94.2%) participants filled out the questionnaire and were included in the statistical analysis [47]. Measures: The PHQ-8 questionnaire was used to measure the participants’ depression levels [6, 39, 51]. This study used the cutoff point of 10 as well based on the common local practice considering that there are no specific related research guidelines in UAE [13–16]. The PHQ-8 score was dichotomized into no-depression (total PHQ-8 < 10) versus depression (≥ 10) [6, 51]. Happiness was measured using a one-question item that asked participants, “In general, how happy you are?’’ Those who responded as extremely happy, very happy and moderately happy were grouped as “happy” while those responding as moderately unhappy, very unhappy, and extremely unhappy were grouped as “unhappy”. Demographic variables such as age, gender, and marital status (single, married, and others) were also included in the reference. Sleep duration data was collected as an ordinal variable (number of hours) by asking how many hours of sleep the participant gets in a 24-hour period, including naps. Sleep duration was categorized into five categories (see Table 1) to avoid the linearity assumption between sleep duration and depression status as well as to be able to compare average sleepers to shorter and longer sleepers. The questionnaire was translated from English into Arabic and back-translated into English to check for linguistic validity. Table 1 Number (percentages) of the analyzed variables and median (IQR) for Age and BMI. VariableGroupPHQ-8 < 10PHQ-8 ≥ 10P-valueGenderFemale62 (83.8)12 (16.2)0.028aMale147 (93.6)10 (6.4)Sleep (hours)< 639 (83)8 (17)0.284a= 653 (93)4 (7)= 757 (95)3 (5)= 844 (89.8)5 (10.2)> 816 (88.9)2 (11.1)Marital StatusSingle120 (93)9 (7)0.205aMarried15 (93.8)1 (6.2)Other74 (86)12 (14)HappinessHappy204 (91.1)20 (8.9)0.136aUnhappy5 (71.4)2 (28.6)Total209 (90.5)22 (9.5)Median (IQR)Median (IQR)P-valueAgeyears33.0 (25.0, 40.0)24.5 (22.0, 35.0)0.012bBMIkg/m227.6 (23.4, 31.2)28.8 (24.2, 33.0)0.444bNote: aFisher’s exact test p-values for categorical data and bWilcoxon rank sum test for continuous data Number (percentages) of the analyzed variables and median (IQR) for Age and BMI. Note: aFisher’s exact test p-values for categorical data and bWilcoxon rank sum test for continuous data Body mass index (BMI) was obtained via physical measurement using Tanita MC-780 MA Segmental Body Composition Analyzer [52]. All physical measurements were collected by a clinical research nurse. Additional details of the study recruitment have been previously described [47]. Statistical analysis: All eligible participants 487 (94.2%) were included in a sensitivity analysis (using 100 multiple imputations). After excluding participants with missing values, 231 (44.7%) were included in the primary statistical analysis. The PHQ-8 questionnaire was used in this statistical analysis with two additional possible options to select for each question (i.e., P2A – P2H). These were “Do not know (UN)” and “Prefer not to answer (DA)”, which were treated as missing values in the statistical analysis. Fisher’s exact test was used to investigate the association between depression and categorical variables, such as Sleep (hours), Happiness, Marital Status, and Gender. Wilcoxon rank-sum test was performed to investigate differences in the distribution of age and BMI within the depressed versus non-depressed group respectively. To examine the factors associated with depression, a multivariate logistic regression model was performed with the dichotomized PHQ-8 score as the outcome. The predictors were Happiness score (Happy vs. Unhappy), Age (linear), Gender (Females vs. Males), BMI (linear), and Marital status (categorical). Interquartile-range odds ratios (IQR-OR’s) for continuous predictors and simple odds ratios (OR’s) for categorical predictors with 95% confidence intervals (Cis) were estimated for continuous and categorical variables respectively; corresponding p-values were calculated [53]. In sensitivity analysis, a multivariate logistic regression model and a multivariate linear regression model were conducted using multiple imputations (see sensitivity analysis section). All applied statistical tests were two-sided; p < 0.05 were considered statistically significant. No adjustment for multiple comparisons was made. Statistical analyses were performed in R version 4.0.2 [54]. Sensitivity analysis: The primary statistical analysis included subjects with at least one non-missing value. However, in a sensitivity analysis, a multivariate imputation by chained equations (MICE) procedure was applied with Classification and Regression Trees (CART) to impute missing values [55]. 100 multiple imputations were used [56]. Rubin’s rules were used to combine the multiple imputed estimates [57]. The pattern of missing values was investigated, and it was found that subjects who “did not want to answer” were not systematically different from those who answered the questionnaire. Therefore, “prefer not to answer” was recorded as missing value in the statistical analysis and was considered a missing variable in the sensitivity analysis [55]. Results: Of 517 participants who consented to participate in the UAEHF pilot study, 487 (94.2%) had completed questionnaire data [47]. Figure 1 describes all key phases from recruitment up to inclusion in the study and in the statistical analysis. After omitting missing values, 231 (44.7%) participants were included in the main statistical analysis. However, 487 (94.2%) participants were included in the sensitivity analysis (using 100 multiple imputations) [73]. The median age of the UAEHFS pilot data participants was 32.0 years (Interquartile Range: 24.0–39.0). The percentage of females included in the study was 32%, which represented the UAE population well [59]. Therefore, we did not make any adjustments for gender bias. Fig. 1Flow chart of UAEHF pilot study (describes all the key phases from the recruitment up to the inclusion in the study analysis) Flow chart of UAEHF pilot study (describes all the key phases from the recruitment up to the inclusion in the study analysis) Note: This Figure represents the data of the included participants in the main statistical analysis after omitting missing values. The number of observed values (%) of each categorical variable was presented by the PHQ-8 categories in Table 1, where the majority of the study participants (90.5%) had PHQ-8 score less than 10. When categorical groups were compared within the PHQ-8 depression group (< 10 versus ≥ 10), there was only a statistically significant difference between females versus males with a Fisher exact p-value = 0.028. Table 1 shows that there was a statistically significant difference in the age distribution across the depression groups (p-value = 0.012). There was no statistically significant difference in the BMI measurements between the PHQ-8 groups (p-value = 0.44). The result of the primary analysis showed that females have statistically significantly greater odds of reporting depression compared to males OR = 3.2 (95% CI:1.1, 8.9), p-value = 0.024 (Table 2). Furthermore, there was an approximately 5-fold greater odds of reporting depression for unhappy than for happy individuals. However, this was not statistically significant (p-value = 0.10). Similarly, the sleep duration and marital status were both not statistically significant (p-value of 0.284 and 0.205, respectively; Table 2). However, in a sensitivity analysis, people who reported sleep duration of less than 6 h were more likely to report depression as compared to the people who reported sleep of 7 h OR (95% CI) of 2.6 (1.0, 6.4), p-value = 0.040 and 1.136 (1.015, 1.272), p-value = 0.027. Table 2 The results of the primary analysis of the fitted multivariate logistic regression model VariableOR (95% CI)PbHappiness - HappyReferenceHappiness - Unhappy5.5 (0.7, 42.2)0.100Age (linear)0.3 (0.1, 1.0)a0.056BMI (linear)1.8 (1.0, 3.3)a0.064Gender = MalesReferenceGender = Females3.2 (1.2, 8.9)0.024Sleep = 7ReferenceSleep < 64.0 (0.9, 17.4)0.061Sleep = 61.3 (0.2, 6.9)0.775Sleep = 81.5 (0.3, 7.3)0.585Sleep > 81.2 (0.2, 9.2)0.848MarriedReferenceSingle0.9 (0.3, 3.0)0.8Others0.2 (0.02, 3.3)0.3Note: aInterquartile-range odds ratio for age and BMI, which compares the 3rd quartile with the 1st quartile, and simple odds ratios for the categorical predictors (happiness, gender, sleep and marital status) compare each group with the reference group (the largest group) with corresponding 95% confidence interval (95% CI). bWald’s p-values are presented for each variable The results of the primary analysis of the fitted multivariate logistic regression model Note: aInterquartile-range odds ratio for age and BMI, which compares the 3rd quartile with the 1st quartile, and simple odds ratios for the categorical predictors (happiness, gender, sleep and marital status) compare each group with the reference group (the largest group) with corresponding 95% confidence interval (95% CI). bWald’s p-values are presented for each variable To compute Interquartile Range (IQR) odds ratios for continuous variables, the age and BMI variables were divided by their IQR values of 15 and 8.7, respectively [54, 56]. Table 2 shows that for one interquartile-range increase in the age and BMI, the IQR-OR was 0.34 (95% CI: 0.12, 1.0) and 1.8 (95% CI: 0.97, 3.3), respectively, where both results were statistically significant with a p-value of 0.056 and 0.064 individually. The results of the sensitivity analysis using 100 multiple imputations (Table 3) were approximately similar to the result of the multivariate logistic ordinal regression analyses using omitted data (Table 2). However, the happiness variable was statistically significant OR = 5.1 (95%CI: 1.7, 15.7, p-value = 0.005), and there was a statistically significant difference between the sleep variable for less than six hours as compared with seven hours OR = 2.6 (95% CI: 1.0, 6.4, p-value = 0.04). Supplementary Fig. 1 illustrates the percentages of missing values in the variables included in this statistical analysis. Marital status had the lowest number of missing values (5.3%), followed by BMI (12.7%) and the Sleep variable with 12.9% missing values. The Happiness variable had 21.1% missing values. The PHQ-8 variables had 32.6–35.1% missing values. Supplementary Table 1 shows the percentages of the “Prefer not to answer (DA)” and “Do not know (UN)” in the eight PHQ questions, where the percentages vary between 16 and 18.5% between the eight questions. The percentages of “Prefer not to answer (DA)” and “Do not know (UN)” in the happiness variable was 2.5% and 2.3%, respectively, which were also considered missing values, although these were not missing at random and can be correlated with one of the PHQ-8 answers. However, this has been addressed in the multiple imputation analysis. Table 3 Results of the sensitivity analysis using multivariate logistic regression models with a sample size N = 487 VariableOR (95% CI)PbHappiness – HappyReferenceHappiness - Unhappy5.1 (1.7, 15.7)0.004Age (linear)0.5 (0.3, 1.0)a0.060BMI (linear)1.7 (1.1, 2.5)a0.021Gender = MalesReferenceGender = Females1.6 (0.9, 3.0)0.154Sleep = 7ReferenceSleep < 62.6 (1.0, 6.4)0.040Sleep = 60.9 (0.3, 2.6)0.887Sleep = 80.9 (0.3, 2.4)0.779Sleep > 81.1 (0.3, 3.7)0.853MarriedReferenceSingle0.9 (0.4, 2.1)0.885Others0.7 (0.2, 2.9)0.597Note: aInterquartile-range odds ratio for Age and BMI, which compares the 3rd quartile with the 1st quartile, and simple odds ratios for the categorical predictors (happiness, gender, sleep and marital status) compare each group with the reference group (the largest group) with corresponding 95% confidence interval (95% CI). bWald’s p-values are presented for each variable. Results were summarized using Rubin’s rules Results of the sensitivity analysis using multivariate logistic regression models with a sample size N = 487 Note: aInterquartile-range odds ratio for Age and BMI, which compares the 3rd quartile with the 1st quartile, and simple odds ratios for the categorical predictors (happiness, gender, sleep and marital status) compare each group with the reference group (the largest group) with corresponding 95% confidence interval (95% CI). bWald’s p-values are presented for each variable. Results were summarized using Rubin’s rules An additional sensitivity analysis was performed using the ordinal PHQ-8 score as an outcome in a multivariate linear regression model (see supplementary Table 1). The result of this sensitivity analysis was very similar to the sensitivity analysis in Table 3. Discussion: Overall, there is a lack of quantitative research examining the relationship between depression, perceived happiness, and sleep duration and other confounding factors in the Gulf region [35,46, 36, 49, 60]. The UAE Healthy Future Study is the first prospective cohort study of the UAE population and one of the few studies in the region which examines such relationships between happiness, sleep duration and quality, and depression using PHQ-8. The evidence collected from this study confirms what has been published in the literature. For instance, in this study, it has been found that males were less likely to report depression symptoms than females, which is similar to what has been documented by other studies [38, 41]; where women reported more depressive symptoms than men [60]. In addition, the results of this study revealed that older people have a lower odds of reporting depression as compared with younger people, suggesting a possible protective age effect. This study has used the pilot data of the UAE Healthy Future Study, as recruitment into the main cohort study is still ongoing. However, we plan to use the main UAEHFS data to confirm the results and provide further understanding of these findings. The association between sleep duration and depression has been intensively studied in the literature [23, 24]. The findings of this study revealed that individuals who reported less than 6 h of sleep per 24 h were more likely to report being depressed compared to those who reported 7 h of sleep. This is similar to what has been reported in the literature that participants who reported insufficient sleep showed a 62–179% increase in the prevalence of depression versus those sleeping 6 to 8 h per day and reporting sufficient sleep (P < 0.05) [61, 62]. Furthermore, self-reported happiness has been identified as a potential protective and management factor for depression [27]. The result of this study shows that unhappy individuals have approximately 5-fold higher odds of reporting depression compared to happy individuals, and this aligns with what has been found in the literature [27]. Additionally, a small number of studies found that insufficient sleep is associated with lower happiness in healthy adults using a self-reported questionnaire as a single item for measuring happiness [22]. Moreover, some longitudinal studies have found that the next-day happiness is lowered following a shorter sleep duration [63]. However, the associations between sleep and happiness have not been well-explored in adults. Our findings will add to the available evidence and will help to bring novel data from the Gulf Cooperation Council (GCC) countries about the association between sleep duration, self-reported happiness, and depression. A further finding is that marital status can contribute to health and self-reported happiness in a bidirectional way [49]. Several studies have found that there is an association between depression and marital status [63, 64]. There is an increased risk of reporting depression for divorced and separated people. It is frequently asserted that marriage is more beneficial for the mental health of men than women, but the evidence for this is far from clear-cut [65]. Single people have a higher level of depression as compared to married people and some studies have found that married people have a better mood than single people considering factors of age, gender, and education level [37, 40, 41, 66]. Research does not yet clarify whether gender differences in the prevalence of anxiety-mood disorders are greater among the married than the never-married or the previously married [65]. However, the mechanisms underlying the relationship between depression and marital status are not yet entirely clear and require further exploration. An association has been reported between BMI categories and depression [67, 68]. Higher BMI is a risk factor for the likelihood of developing depression in individuals [69]. Underweight increases the risk of depression as well [70]. Our study considered the BMI factor in the data analysis of this study. This statistical analysis shows that as BMI increases, the odds of reporting depression also increase. This result is comparable to other research finding that participants with central obesity had an increased chance of depression [70,71,72]. The UAEHFS is a unique cohort study in the UAE and Gulf region as it allows researchers to investigate the association between disease outcomes and related risk factors [58]. In this study, we investigated the association between depression and sleep duration, self-reported happiness, BMI, and sociodemographic status using the UAEHFS pilot data, which presents a population that has not been studied. The result of our study needs to be confirmed in the main UAEHFS data. Strengths and Limitations: Missing values were omitted in the primary statistical analysis, which decreased the sample size and can lead to overfitting in the main finding [55]. Therefore, the number of participants with the PHQ-8 ≥ 10 (i.e. – events) was 22 which is a limitation in this data set because it does not allow us to fit a complex multivariate statistical regression model if we would follow the statistical rule of thumb “ten events per predictor” [56, 57, 73]. Thus, sensitivity analysis was performed, and the result of the sensitivity analysis (Table 3) was approximately the same as the main finding (Table 2). Although some limitations were found in this study, the findings provide future direction to mental health research. Further investigation is needed with a larger sample size using the main UAEHFS data to have a better picture of the role of happiness, marital status, sleep, and social demographic variables in association with depression and mental health disorders. Conclusion: The results of this study indicate that females are more likely to report depression compared to males. Older age is associated with a decrease in self-reported depression. Unhappy individuals are approximately 5-fold higher odds to report depression compared to happy individuals. BMI was positively associated with reporting depression. The result of the sensitivity analysis shows that individuals who sleep less than 6 h per day are more likely to report depression compared to those who sleep 7 h per day. The results of this study have a potential value for researchers and public health professionals as it presents novel data on the PHQ-8 score in a healthy UAE population, which has not been explored before. Furthermore, the results of this study can help contribute to the knowledge base on current and potential population mental health impact in the UAE and Gulf Region. Electronic supplementary material: Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 : Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1

Background: The United Arab Emirates Healthy Future Study (UAEHFS) is one of the first large prospective cohort studies and one of the few studies in the region which examines causes and risk factors for chronic diseases among the nationals of the United Arab Emirates (UAE). The aim of this study is to investigate the eight-item Patient Health Questionnaire (PHQ-8) as a screening instrument for depression among the UAEHFS pilot participants. Methods: The UAEHFS pilot data were analyzed to examine the relationship between the PHQ-8 and possible confounding factors, such as self-reported happiness, and self-reported sleep duration (hours) after adjusting for age, body mass index (BMI), and gender. Results: Out of 517 participants who met the inclusion criteria, 487 (94.2%) participants filled out the questionnaire and were included in the statistical analysis using 100 multiple imputations. 231 (44.7%) were included in the primary statistical analysis after omitting the missing values. Participants' median age was 32.0 years (Interquartile Range: 24.0, 39.0). In total, 22 (9.5%) of the participant reported depression. Females have shown significantly higher odds of reporting depression than males with an odds ratio = 3.2 (95% CI:1.17, 8.88), and there were approximately 5-fold higher odds of reporting depression for unhappy than for happy individuals. For one interquartile-range increase in age and BMI, the odds ratio of reporting depression was 0.34 (95% CI: 0.1, 1.0) and 1.8 (95% CI: 0.97, 3.32) respectively. Conclusions: Females are more likely to report depression compared to males. Increasing age may decrease the risk of reporting depression. Unhappy individuals have approximately 5-fold higher odds of reporting depression compared to happy individuals. A higher BMI was associated with a higher risk of reporting depression. In a sensitivity analysis, individuals who reported less than 6 h of sleep per 24 h were more likely to report depression than those who reported 7 h of sleep.

Introduction: Depression is defined as a set of disorders ranging from mild to moderate to severe [1]. It is widely recognized as a major public health problem worldwide [2]. Reports from the Global Burden of Diseases declared major depressive disorder as one of the top three causes of disability-adjusted life years [3, 4]. The effect of depression has been extensively studied on the individual’s daily functioning and productivity. This is directly reflected in increased economic costs per capita [3, 4]. For example, in Catalonia in 2006, the average annual cost of an adult with depression was close to 1800 Euros, and the total annual cost of depression was 735.4 million Euros. These costs were linked directly to primary care, mental health specialized care, hospitalization, and pharmacological care, as well as indirect costs due to productivity loss, temporary and permanent disability [4]. To measure the level of depression in non-clinical populations; clinical and epidemiological studies have often used the established and validated eight-item Patient Health Questionnaire scale (PHQ-8) instead of the nine- item Patient Health Questionnaire scale (PHQ-9) [5]. As has been confirmed by previous studies, the scale can detect major depression with sensitivity and a specificity of 88%, to classify subjects into depressed or non-depressed, respectively [6–8]. Regionally, in Jordan, Lebanon, Syria and Afghanistan, the cutoff point of 10 was used [9–12]. Similarly, in UAE, studies have mostly used the cutoff point of 10 [13–16]. Recent studies have focused on exploring the methods of improving individual and environmental effects on disability related to depression by investigating its triggers and associations [2, 10]. Factors such as sleep duration and self-reported happiness are evidenced to be predictors for depression [17, 18]. Sleep duration can be considered a risk factor for lower well-being [19]. The number of sleep hours has also been found to have a causal relationship with depression as well as self-reported happiness [20]. For example, shorter sleep duration might lead to lower positive emotions such as self-reported happiness and showed stronger associations with negative emotional affect [21, 22]. One question that needs to be asked, however, is the nature and strength of the association between these three variables respectively. Sleep duration is determined by how many hours an individual sleeps over 24 h. Individuals with depression often have poor sleep status including abnormal REM (rapid eye movement), and insomnia (difficulty falling asleep or staying asleep). Abnormal REM may contribute to the development of altered emotional processing in depression [23]. It was reported that people with insomnia might have a ten-fold higher risk of developing depression in contrast to people who get a good night’s sleep [24]. Also, 75% of depressed individuals will have trouble falling asleep or staying asleep [25]. A bidirectional relationship between sleep duration and reported depression has been investigated [24, 26]. Such studies are unsatisfactory because they do not explore the association between major depressive disorder and sleep duration based on specific population demographics. In this study we are considering other sociodemographic factors such as age, gender, and marital status. Interest in studying happiness in the context of mental health status (such as depression and anxiety) has been growing recently [27]. The findings of some research papers suggest that self-reported happiness is a potential factor in the prevention and management of depression [27, 34]. Happiness is correlated to a person’s ability to approach situations in a less stressful manner and to an individual’s capacity to perceive and control their own feelings. This indicates that higher happiness levels may have a protective effect on depression [27]. Well-being has been defined as the combination of feeling good and functioning well [28]; conversely, quality of life could be defined as an individual’s satisfaction with his or her actual life compared with his or her ideal life. Evaluation of the quality of life depends on one’s value system [29]. Moreover, studies have shown that individuals use various chronically accessible and stable sources of information when making life satisfaction judgments [30]. Yet, well-being, quality of life and life satisfaction are multidimensional constructs which include complex cognitive evaluation processes and cannot be adequately assessed by using a single item or question [31]. Unlike quality of life, which typically requires detailed assessments to ascertain [32], happiness is easier to evaluate using a single item question [33]. Investigating the association between happiness and depression would add valuable information to public health research as the relationship between happiness and depression is observed to be bidirectional (i.e., one variable can predict the other, and people might not report depression but are more likely to report feeling unhappy) [34]. In addition, there are some factors which confound with sleep duration and self-reported happiness, such as age, gender, and marital status [36, 37]. For example, reviews reported that a prognosis of depression was improved with increasing age [38]. Conversely, another study showed that older age was associated with worsening of depressive symptoms [37]. Moreover, depressive symptoms could worsen among widowed individuals as their age increased [40]. Some research findings reflected that females are more likely to perceive depression compared to males [38, 41]. Other studies showed significant interplay between marital status and depressive symptoms [40, 42]. For example, being unmarried could lead to perceiving and developing depressive symptoms [43], while a worsening in depression symptoms among married individuals could lead to separation or being unmarried [42, 44]. So, further exploration is required to study the interplay of these variables together. The United Arab Emirates (UAE) is a high-income developed country which has undergone a rapid epidemiological transition from a traditional semi-nomadic society to a modern affluent society with a lifestyle characterized by over-consumption of energy-dense foods and low physical activity [47]. The UAE was ranked 15th out of 157 countries included in the WHO World Happiness Report with a score of 7.06 [48]. Despite this, depression has been identified as the third leading cause of disability in the UAE [3]. Nevertheless, there are few studies in the Gulf region examining the relationship between happiness, sleep duration and depression [36, 49, 50]. Therefore, studying the association between depression, happiness, and sleep in the United Arab Emirates (UAE) population would be of interest to the public health field in this part of the world. This study aimed to examine the relationship between depression, self-reported happiness, and self-reported sleep duration after adjusting for age and gender using the UAE Healthy Future Study (UAEHFS) pilot data. Conclusion: The results of this study indicate that females are more likely to report depression compared to males. Older age is associated with a decrease in self-reported depression. Unhappy individuals are approximately 5-fold higher odds to report depression compared to happy individuals. BMI was positively associated with reporting depression. The result of the sensitivity analysis shows that individuals who sleep less than 6 h per day are more likely to report depression compared to those who sleep 7 h per day. The results of this study have a potential value for researchers and public health professionals as it presents novel data on the PHQ-8 score in a healthy UAE population, which has not been explored before. Furthermore, the results of this study can help contribute to the knowledge base on current and potential population mental health impact in the UAE and Gulf Region.

Background: The United Arab Emirates Healthy Future Study (UAEHFS) is one of the first large prospective cohort studies and one of the few studies in the region which examines causes and risk factors for chronic diseases among the nationals of the United Arab Emirates (UAE). The aim of this study is to investigate the eight-item Patient Health Questionnaire (PHQ-8) as a screening instrument for depression among the UAEHFS pilot participants. Methods: The UAEHFS pilot data were analyzed to examine the relationship between the PHQ-8 and possible confounding factors, such as self-reported happiness, and self-reported sleep duration (hours) after adjusting for age, body mass index (BMI), and gender. Results: Out of 517 participants who met the inclusion criteria, 487 (94.2%) participants filled out the questionnaire and were included in the statistical analysis using 100 multiple imputations. 231 (44.7%) were included in the primary statistical analysis after omitting the missing values. Participants' median age was 32.0 years (Interquartile Range: 24.0, 39.0). In total, 22 (9.5%) of the participant reported depression. Females have shown significantly higher odds of reporting depression than males with an odds ratio = 3.2 (95% CI:1.17, 8.88), and there were approximately 5-fold higher odds of reporting depression for unhappy than for happy individuals. For one interquartile-range increase in age and BMI, the odds ratio of reporting depression was 0.34 (95% CI: 0.1, 1.0) and 1.8 (95% CI: 0.97, 3.32) respectively. Conclusions: Females are more likely to report depression compared to males. Increasing age may decrease the risk of reporting depression. Unhappy individuals have approximately 5-fold higher odds of reporting depression compared to happy individuals. A higher BMI was associated with a higher risk of reporting depression. In a sensitivity analysis, individuals who reported less than 6 h of sleep per 24 h were more likely to report depression than those who reported 7 h of sleep.

[CONTENT] PHQ-8 | Depression | Sleep duration | Happiness | Self-reported happiness | Sociodemographic and marital status [SUMMARY]

[CONTENT] PHQ-8 | Depression | Sleep duration | Happiness | Self-reported happiness | Sociodemographic and marital status [SUMMARY]

[CONTENT] PHQ-8 | Depression | Sleep duration | Happiness | Self-reported happiness | Sociodemographic and marital status [SUMMARY]

[CONTENT] PHQ-8 | Depression | Sleep duration | Happiness | Self-reported happiness | Sociodemographic and marital status [SUMMARY]

[CONTENT] PHQ-8 | Depression | Sleep duration | Happiness | Self-reported happiness | Sociodemographic and marital status [SUMMARY]

[CONTENT] Male | Female | Humans | Adult | Happiness | Pilot Projects | Prospective Studies | Depression | United Arab Emirates | Sleep [SUMMARY]

[CONTENT] Male | Female | Humans | Adult | Happiness | Pilot Projects | Prospective Studies | Depression | United Arab Emirates | Sleep [SUMMARY]

[CONTENT] Male | Female | Humans | Adult | Happiness | Pilot Projects | Prospective Studies | Depression | United Arab Emirates | Sleep [SUMMARY]

[CONTENT] Male | Female | Humans | Adult | Happiness | Pilot Projects | Prospective Studies | Depression | United Arab Emirates | Sleep [SUMMARY]

[CONTENT] Male | Female | Humans | Adult | Happiness | Pilot Projects | Prospective Studies | Depression | United Arab Emirates | Sleep [SUMMARY]

[CONTENT] measure participants depression | depression total | cost adult depression | prevalence depression versus | annual cost depression [SUMMARY]

[CONTENT] measure participants depression | depression total | cost adult depression | prevalence depression versus | annual cost depression [SUMMARY]

[CONTENT] measure participants depression | depression total | cost adult depression | prevalence depression versus | annual cost depression [SUMMARY]

[CONTENT] measure participants depression | depression total | cost adult depression | prevalence depression versus | annual cost depression [SUMMARY]

[CONTENT] measure participants depression | depression total | cost adult depression | prevalence depression versus | annual cost depression [SUMMARY]

[CONTENT] depression | sleep | study | analysis | happiness | participants | reported | age | sleep duration | duration [SUMMARY]

[CONTENT] depression | sleep | study | analysis | happiness | participants | reported | age | sleep duration | duration [SUMMARY]

[CONTENT] depression | sleep | study | analysis | happiness | participants | reported | age | sleep duration | duration [SUMMARY]

[CONTENT] depression | sleep | study | analysis | happiness | participants | reported | age | sleep duration | duration [SUMMARY]

[CONTENT] depression | sleep | study | analysis | happiness | participants | reported | age | sleep duration | duration [SUMMARY]

[CONTENT] depression | life | happiness | studies | sleep | reported | depressive | sleep duration | duration | individual [SUMMARY]

[CONTENT] 95 | 95 ci | ci | group | table | analysis | values | quartile | statistically | value [SUMMARY]

[CONTENT] depression | report depression compared | results study | results | individuals | depression compared | report | report depression | compared | sleep day [SUMMARY]

[CONTENT] material | supplementary material | depression | supplementary | material supplementary | material supplementary material | supplementary material supplementary material | supplementary material supplementary | participants | analysis [SUMMARY]

[CONTENT] material | supplementary material | depression | supplementary | material supplementary | material supplementary material | supplementary material supplementary material | supplementary material supplementary | participants | analysis [SUMMARY]

[CONTENT] The United Arab Emirates Healthy Future Study | first | one | the United Arab Emirates ||| eight | PHQ-8 | UAEHFS [SUMMARY]

[CONTENT] 517 | 487 | 94.2% | 100 ||| 231 | 44.7% ||| 32.0 years | 24.0 | 39.0 ||| 22 | 9.5% ||| 3.2 | 95% | 8.88 | approximately 5-fold ||| one | BMI | 0.34 | 95% | CI | 0.1 | 1.0 | 1.8 | 95% | CI | 0.97 | 3.32 [SUMMARY]

[CONTENT] ||| ||| approximately 5-fold ||| BMI ||| less than 6 | 24 [SUMMARY]

[CONTENT] The United Arab Emirates Healthy Future Study | first | one | the United Arab Emirates ||| eight | PHQ-8 | UAEHFS ||| UAEHFS | PHQ-8 | hours | BMI ||| 517 | 487 | 94.2% | 100 ||| 231 | 44.7% ||| 32.0 years | 24.0 | 39.0 ||| 22 | 9.5% ||| 3.2 | 95% | 8.88 | approximately 5-fold ||| one | BMI | 0.34 | 95% | CI | 0.1 | 1.0 | 1.8 | 95% | CI | 0.97 | 3.32 ||| ||| ||| approximately 5-fold ||| BMI ||| less than 6 | 24 [SUMMARY]

[CONTENT] The United Arab Emirates Healthy Future Study | first | one | the United Arab Emirates ||| eight | PHQ-8 | UAEHFS ||| UAEHFS | PHQ-8 | hours | BMI ||| 517 | 487 | 94.2% | 100 ||| 231 | 44.7% ||| 32.0 years | 24.0 | 39.0 ||| 22 | 9.5% ||| 3.2 | 95% | 8.88 | approximately 5-fold ||| one | BMI | 0.34 | 95% | CI | 0.1 | 1.0 | 1.8 | 95% | CI | 0.97 | 3.32 ||| ||| ||| approximately 5-fold ||| BMI ||| less than 6 | 24 [SUMMARY]

Prevalence and predisposing factors of brachial plexus birth palsy in a regional hospital in Ghana: a five year retrospective study.

31312323

Brachial plexus birth injury is one of the challenges associated with maternal delivery, with varying prevalence between countries. Brachial plexus birth injury poses negative health implications to children and also has socio-economic implications on families and the community as a whole. To treat brachial plexus birth injury, a multi-disciplinary treatment approach is recommended. Brachial plexus birth palsy (BPBP) is categorised into two-upper plexus injury (Erb's palsy) and lower plexus injury (Klumpke's palsy). These categories present with various degrees of injuries, with less severe injuries responding well to treatment and in most instances may resolve on their own, but serious and complicated injuries will require a multi-disciplinary treatment approach to treat and/or manage. Effective treatment and management depends on adequate knowledge of the disease condition. These include the risk factors and prevalence of brachial plexus birth palsy within a particular population at a specific period in time. The aim of this study was to determine the risk factors and the prevalence of a hospital based brachial plexus birth palsy within a five-year period (2013-2017).

INTRODUCTION

A five-year retrospective study design was used. The study involved selection of all clients' diagnosed with brachial plexus birth palsy, where their gender, birth weight, complications at birth, type of brachial plexus suffered, mothers' diabetes status, mother's age, birth attendant, side of affectation, presentation at birth and mode of delivery were recorded.

METHODS

The prevalence rate of brachial plexus birth palsy was 14.7% out of a total of three hundred and twenty (320) cases reviewed over the study period in the Volta Regional Hospital. Erb's palsy was found to be the modal type of BPBP in this population (93.6%).

RESULTS

There is the need to provide a nationwide education on the risk factors that predispose babies to brachial plexus birth palsy. There is also the need for frequent antenatal visit by pregnant women; this will help in the provision of best antenatal history, diagnostic investigation in determining the birth weight and safe mode of delivery.

CONCLUSION

6620083

Introduction

Brachial plexus birth palsy (BPBP) is a neurological condition that results from nerve injury to the brachial plexus; C5-T1, which supply the upper extremities [1, 2]. Al-Qattan et al. [3] classified BPBP into four main groups; group one shows injury to C5-C6 and results in paralysis of the shoulder and biceps muscles. Group two affects C5-C7, and typically results in paralysis of the shoulder, biceps and the forearm extensors [4]. Group three indicates injury to C5-T1 and results in complete paralysis of the affected upper limb. Group four affects C5-T1 and results in complete paralysis of the affected upper limb with Horner's syndrome. Zafeiriou and Psychogiou [5] further indicated that group one shows the least severity of injury whereas group four indicates the most severity of injury. Upper plexus injury is an injury to C5, C6 and sometimes C7 nerve of the cervical spine and lower plexus injury are injury to C8, and T1 nerves of the thoracic spine [6, 7]. The management of BPBP has many physical, psychological, financial, social and emotional implications on families and the country as a whole [8]. The prevalence of BPBP varies among countries. According to Foad et al. [9], the prevalence of BPBP in the United States is 1.5 in 1000 live births; shoulder dystocia had 100 times greater risk, an exceptionally large baby (>4.5 kg) had a fourteen times greater risk, and forceps delivery had a nine times greater risk for having a child with brachial plexus birth palsy. Having a twin or multiple birth mates and delivery by cesarean section had a protective effect against the occurrence of neonatal brachial plexus palsy. A study conducted by Evans-Jones and colleagues [10], reported the prevalence in the United Kingdom to be 0.42 per 1000 live births and the associated risk factors for BPBP was found to be shoulder dystocia, high birth weight and assisted delivery, but a considerably lower risk in infants delivered by caesarean section. The prevalence of children with BPBP in Nigeria over a ten year study period showed a persistent high prevalence, averaging 15.3% per year with associated problems such as birth asphyxia, humeral fracture, clavicular fracture and shoulder dislocation [4]. A study by Hamzat and colleagues [11] also reported the prevalence of BPBP in Accra, Ghana to be 27%, the results of the study further indicated that birth weight exceeding 4.0 kg, vertex presentation and vaginal delivery were the noticeable co-existing factors for BPBP in Accra. From the study done by Hamzat et al. [11] it was reported that only 55.2% of BPBP cases were referred for physiotherapy within one month after diagnosis and the treatment disposition for majority (88.1%) of the children were not documented and only 4.8% were formally discharged from physiotherapy. The prevalence of BPBP reported in Ghana focused on the Accra metropolis and there is no known data of BPBP in the Ho municipality, Volta region. The Ho municipality is a cosmopolitan urban city with health facilities serving the southern and central parts of the Volta region. BPBP is preventable when the risk factors and causes are known [12]. This work seeks to find out the prevalence and risk factors of BPBP in the Ho municipality of Ghana so as to advice policy makers on how to prevent the injury which will reduce the cost of treatment and increase productivity of the mothers. Managing of some children suffering from BPBP may need surgery and rehabilitation which poses financial issues to the family and loved ones [8]. Some children may have residual functional deficits and thereby affecting the way they function, as well psychological or emotional problems to the child [8]. This present study was undertaken to retrospectively investigate the prevalence of children who presented with BPBP and their predisposing factors in a regional hospital in Ghana. The clinical implication for this study is that there is not much information concerning the predisposing factors of BPBP in Ghana, apart from the study done by Hamzat et al. [11] that was done 10 years ago in Accra, an urban city in Ghana. Moreover the study by Hamzat et al. [11] did not provide information on the risk factors of BPBP in Ghana. As a result, there is paucity of information on the predisposing factors and the current prevalence of BPBP in a peri-urban city in Ghana. The aim of this study was to determine the prevalence and predisposing factors of BPBP over a five year period (January 2013-December 2017) at a regional hospital in Ghana.

Methods

Study site: the study was conducted at the physiotherapy department of the Volta Regional Hospital (VRH), the main referral centre for the Volta region of Ghana. The hospital attends to over 177,281 inhabitants in the Ho municipality and its environs [13]. Study design: the study design employed was a retrospective quantitative study over a five-year period (January 2013 to December 2017). Ethical approval: the study was approved by the Research Ethics Committees of the University of Health and Allied Sciences and the VRH with protocol identification number UHAS-REC/A.5 (62) 17-18. Inclusion and exclusion criterion: all paediatric cases under 15 years who attended the physiotherapy clinic of the VRH over the study period were recruited. This age category is chosen because children within this age group are treated in the paediatric unit of the department. Paediatric cases above 15 years of age who reported for treatment over the study period were excluded. Also acute brachial plexus injuries not associated with delivery were also excluded. Procedure for data collection: records of all newly diagnosed clients' were retrieved to identify all paediatric conditions. Sampling of all children diagnosed with birth brachial injuries were done, and this provided a means of recording their folder numbers for further investigation. The folder numbers obtained aided in the retrieval of individual folders. The study involved all children 15 years and below who attended the physiotherapy clinic of the VRH within the study period. The second stage of data collection included children that have been diagnosed with brachial plexus birth palsy within the study year period. Demographic data and clinical profile of children and their mothers were recorded. Information such as mother's history of birth, age at delivery, maternal diabetic status, parity status as well as the mothers' occupation was recorded. Other information retrieved included the diagnosis, side of affectation, birth weight, place of delivery, complication at birth, mothers' occupation, antenatal care history, type of BPBP, presentation at birth and others were retrieved from the client folders. The clients were grouped in two categories of Erb's and Klumpke's palsy. Those who had good wrist control were in Erb's classification while those with minimal to less wrist control but good shoulder control were classified as Klumpke's palsy. Those with no control of shoulders and wrist were classified as having total BPBP. Each participant was assigned a code to ensure anonymity. The data was kept under lock and key, accessible only by the researchers. Data analysis: the data collected was entered into IBM SPSS version 20.0 analysed using descriptive statistics.

Results

In all, the records of 47 participants who presented for BPBP were analysed. There were 28 (59.6%) female children with brachial plexus injuries, 17 (36.2%) were male children, and the gender for two (4.2%) of the cases were not documented. The average birth weight was 3.9 ± 0.5 kg. Table 1 presents the gender distribution for children diagnosed with BPBP in the five-year period with average birth weight and side of affectation. Table 2 shows the total number of paediatric conditions seen at the physiotherapy department of the Volta Regional Hospital during the study period. The weight category for children diagnosed with brachial plexus injuries was also recorded during the 5-year study period. Thirty-seven (37) of the children involved in the study had their birth weights recorded and their weights were categorised as shown in Table 2. The remaining 17.8% (n=8) of the children involved in the study had their birth weight not recorded. There was about half of brachial plexus injuries 48.7% (n=18) falling within the weight category (3.5 - 3.9 kg), 13.5% (n=5) falling within 3.0-3.4 kg and 24.3% (n=9) falling within the 4.5-5.0 kg weight category. Out of 320 paediatric cases reviewed during the study period, 47 (14.7%) of them were diagnosed with brachial plexus birth palsy. The rest of the paediatric cases constituted cerebral palsy (30.3%), musculoskeletal injuries (26.6%), acute flaccid paralysis (24%) and burns (4.4%). Brachial plexus birth palsy was the fourth (4th) leading cause for paediatric consultation and treatment at the physiotherapy department of the hospital. Gender distribution with their average birth weight and side of affectation Paediatric cases seen over the study period The clinical profile of the babies diagnosed with BPBP indicated that, the vast majority 44 (93%) had cephalic presentation at birth, 1 (2.1%) had breech presentation and 2 (4.2%) of the cases were not documented. Majority of the children 44 (93.6%) were delivered through spontaneous/normal vaginal delivery, 2 (4.3%) through caesarean section. However, the mode of delivery of one baby was not documented. 44 (93.6%) of the cases were diagnosed with Erb's palsy whereas 3 (6.4%) had Klumpke's palsy. Shoulder dystocia and prolonged labour were the most common complications documented; 13 (27.7%) and 20 (42.5%) respectively. 2 (4.3%) of the children had fractures of the clavicle and humerus. Unfortunately, there were no records on 12 (5.5%) of the children. In Table 3, the ages of the mothers were compared to the average birth weight of their children and complications such as shoulder dystocia and prolonged labour. Out of a total of nineteen (19) of mothers' whose age were documented, only thirteen (13) of their children's birth weight were documented. Within the age categories of 15-24, 25-34 and above 34 years of mothers' age, 71%, 75% and 80% of children's birth weight respectively were documented. Mother's whose ages were above 34 years recorded the highest average birth weight (4.07 ± 0.60 kg) for their children together with shoulder dystocia, n=6 and prolonged labour n=6. The number of children born with shoulder dystocia was directly proportional to the mothers' age. Mothers’ age category versus average birth weight, shoulder dystocia and prolonged labour Data is presented in mean ± SD

Conclusion

In conclusion, the results of this study showed a BPBP prevalence of 15% and most of the Erb's palsy cases had right arm affectation. The study also showed that majority of the babies had a birth weight >4.0 kg and this may accounts for the high number of shoulder dystocia cases. The presentations at birth were predominantly cephalic and this may pose an increased risk of upper plexus injuries among the macrosomic children. Majority of babies diagnosed with Erb's palsy were macrosomic with shoulder dystocia. Recommendations: the researchers recommend the need for appropriate documentation in the hospital, and the need for healthcare professionals to be mindful of the complications and the risk factors of BPBP so as to provide immediate and appropriate care. What is known about this topic Increase birth weight is a risk factor for BPBP; Birth weight above 4 kg is associated with an increased risk of shoulder dystocia and prolonged labour. Increase birth weight is a risk factor for BPBP; Birth weight above 4 kg is associated with an increased risk of shoulder dystocia and prolonged labour. What this study adds Increasing mothers' age was found to be directly related to the complications arising of the birth process such as prolonged labour and shoulder dystocia; Both nulliparous and multiparous mothers' are at risk for delivering babies who may suffer birth injuries because the average birth weight of children is the same for both groups of women; There is an associated increased risk of shoulder dystocia, high birth weight (macrosomia) and prolonged labour in mothers above 34 years. Increasing mothers' age was found to be directly related to the complications arising of the birth process such as prolonged labour and shoulder dystocia; Both nulliparous and multiparous mothers' are at risk for delivering babies who may suffer birth injuries because the average birth weight of children is the same for both groups of women; There is an associated increased risk of shoulder dystocia, high birth weight (macrosomia) and prolonged labour in mothers above 34 years.

Introduction: Brachial plexus birth palsy (BPBP) is a neurological condition that results from nerve injury to the brachial plexus; C5-T1, which supply the upper extremities [1, 2]. Al-Qattan et al. [3] classified BPBP into four main groups; group one shows injury to C5-C6 and results in paralysis of the shoulder and biceps muscles. Group two affects C5-C7, and typically results in paralysis of the shoulder, biceps and the forearm extensors [4]. Group three indicates injury to C5-T1 and results in complete paralysis of the affected upper limb. Group four affects C5-T1 and results in complete paralysis of the affected upper limb with Horner's syndrome. Zafeiriou and Psychogiou [5] further indicated that group one shows the least severity of injury whereas group four indicates the most severity of injury. Upper plexus injury is an injury to C5, C6 and sometimes C7 nerve of the cervical spine and lower plexus injury are injury to C8, and T1 nerves of the thoracic spine [6, 7]. The management of BPBP has many physical, psychological, financial, social and emotional implications on families and the country as a whole [8]. The prevalence of BPBP varies among countries. According to Foad et al. [9], the prevalence of BPBP in the United States is 1.5 in 1000 live births; shoulder dystocia had 100 times greater risk, an exceptionally large baby (>4.5 kg) had a fourteen times greater risk, and forceps delivery had a nine times greater risk for having a child with brachial plexus birth palsy. Having a twin or multiple birth mates and delivery by cesarean section had a protective effect against the occurrence of neonatal brachial plexus palsy. A study conducted by Evans-Jones and colleagues [10], reported the prevalence in the United Kingdom to be 0.42 per 1000 live births and the associated risk factors for BPBP was found to be shoulder dystocia, high birth weight and assisted delivery, but a considerably lower risk in infants delivered by caesarean section. The prevalence of children with BPBP in Nigeria over a ten year study period showed a persistent high prevalence, averaging 15.3% per year with associated problems such as birth asphyxia, humeral fracture, clavicular fracture and shoulder dislocation [4]. A study by Hamzat and colleagues [11] also reported the prevalence of BPBP in Accra, Ghana to be 27%, the results of the study further indicated that birth weight exceeding 4.0 kg, vertex presentation and vaginal delivery were the noticeable co-existing factors for BPBP in Accra. From the study done by Hamzat et al. [11] it was reported that only 55.2% of BPBP cases were referred for physiotherapy within one month after diagnosis and the treatment disposition for majority (88.1%) of the children were not documented and only 4.8% were formally discharged from physiotherapy. The prevalence of BPBP reported in Ghana focused on the Accra metropolis and there is no known data of BPBP in the Ho municipality, Volta region. The Ho municipality is a cosmopolitan urban city with health facilities serving the southern and central parts of the Volta region. BPBP is preventable when the risk factors and causes are known [12]. This work seeks to find out the prevalence and risk factors of BPBP in the Ho municipality of Ghana so as to advice policy makers on how to prevent the injury which will reduce the cost of treatment and increase productivity of the mothers. Managing of some children suffering from BPBP may need surgery and rehabilitation which poses financial issues to the family and loved ones [8]. Some children may have residual functional deficits and thereby affecting the way they function, as well psychological or emotional problems to the child [8]. This present study was undertaken to retrospectively investigate the prevalence of children who presented with BPBP and their predisposing factors in a regional hospital in Ghana. The clinical implication for this study is that there is not much information concerning the predisposing factors of BPBP in Ghana, apart from the study done by Hamzat et al. [11] that was done 10 years ago in Accra, an urban city in Ghana. Moreover the study by Hamzat et al. [11] did not provide information on the risk factors of BPBP in Ghana. As a result, there is paucity of information on the predisposing factors and the current prevalence of BPBP in a peri-urban city in Ghana. The aim of this study was to determine the prevalence and predisposing factors of BPBP over a five year period (January 2013-December 2017) at a regional hospital in Ghana. Methods: Study site: the study was conducted at the physiotherapy department of the Volta Regional Hospital (VRH), the main referral centre for the Volta region of Ghana. The hospital attends to over 177,281 inhabitants in the Ho municipality and its environs [13]. Study design: the study design employed was a retrospective quantitative study over a five-year period (January 2013 to December 2017). Ethical approval: the study was approved by the Research Ethics Committees of the University of Health and Allied Sciences and the VRH with protocol identification number UHAS-REC/A.5 (62) 17-18. Inclusion and exclusion criterion: all paediatric cases under 15 years who attended the physiotherapy clinic of the VRH over the study period were recruited. This age category is chosen because children within this age group are treated in the paediatric unit of the department. Paediatric cases above 15 years of age who reported for treatment over the study period were excluded. Also acute brachial plexus injuries not associated with delivery were also excluded. Procedure for data collection: records of all newly diagnosed clients' were retrieved to identify all paediatric conditions. Sampling of all children diagnosed with birth brachial injuries were done, and this provided a means of recording their folder numbers for further investigation. The folder numbers obtained aided in the retrieval of individual folders. The study involved all children 15 years and below who attended the physiotherapy clinic of the VRH within the study period. The second stage of data collection included children that have been diagnosed with brachial plexus birth palsy within the study year period. Demographic data and clinical profile of children and their mothers were recorded. Information such as mother's history of birth, age at delivery, maternal diabetic status, parity status as well as the mothers' occupation was recorded. Other information retrieved included the diagnosis, side of affectation, birth weight, place of delivery, complication at birth, mothers' occupation, antenatal care history, type of BPBP, presentation at birth and others were retrieved from the client folders. The clients were grouped in two categories of Erb's and Klumpke's palsy. Those who had good wrist control were in Erb's classification while those with minimal to less wrist control but good shoulder control were classified as Klumpke's palsy. Those with no control of shoulders and wrist were classified as having total BPBP. Each participant was assigned a code to ensure anonymity. The data was kept under lock and key, accessible only by the researchers. Data analysis: the data collected was entered into IBM SPSS version 20.0 analysed using descriptive statistics. Results: In all, the records of 47 participants who presented for BPBP were analysed. There were 28 (59.6%) female children with brachial plexus injuries, 17 (36.2%) were male children, and the gender for two (4.2%) of the cases were not documented. The average birth weight was 3.9 ± 0.5 kg. Table 1 presents the gender distribution for children diagnosed with BPBP in the five-year period with average birth weight and side of affectation. Table 2 shows the total number of paediatric conditions seen at the physiotherapy department of the Volta Regional Hospital during the study period. The weight category for children diagnosed with brachial plexus injuries was also recorded during the 5-year study period. Thirty-seven (37) of the children involved in the study had their birth weights recorded and their weights were categorised as shown in Table 2. The remaining 17.8% (n=8) of the children involved in the study had their birth weight not recorded. There was about half of brachial plexus injuries 48.7% (n=18) falling within the weight category (3.5 - 3.9 kg), 13.5% (n=5) falling within 3.0-3.4 kg and 24.3% (n=9) falling within the 4.5-5.0 kg weight category. Out of 320 paediatric cases reviewed during the study period, 47 (14.7%) of them were diagnosed with brachial plexus birth palsy. The rest of the paediatric cases constituted cerebral palsy (30.3%), musculoskeletal injuries (26.6%), acute flaccid paralysis (24%) and burns (4.4%). Brachial plexus birth palsy was the fourth (4th) leading cause for paediatric consultation and treatment at the physiotherapy department of the hospital. Gender distribution with their average birth weight and side of affectation Paediatric cases seen over the study period The clinical profile of the babies diagnosed with BPBP indicated that, the vast majority 44 (93%) had cephalic presentation at birth, 1 (2.1%) had breech presentation and 2 (4.2%) of the cases were not documented. Majority of the children 44 (93.6%) were delivered through spontaneous/normal vaginal delivery, 2 (4.3%) through caesarean section. However, the mode of delivery of one baby was not documented. 44 (93.6%) of the cases were diagnosed with Erb's palsy whereas 3 (6.4%) had Klumpke's palsy. Shoulder dystocia and prolonged labour were the most common complications documented; 13 (27.7%) and 20 (42.5%) respectively. 2 (4.3%) of the children had fractures of the clavicle and humerus. Unfortunately, there were no records on 12 (5.5%) of the children. In Table 3, the ages of the mothers were compared to the average birth weight of their children and complications such as shoulder dystocia and prolonged labour. Out of a total of nineteen (19) of mothers' whose age were documented, only thirteen (13) of their children's birth weight were documented. Within the age categories of 15-24, 25-34 and above 34 years of mothers' age, 71%, 75% and 80% of children's birth weight respectively were documented. Mother's whose ages were above 34 years recorded the highest average birth weight (4.07 ± 0.60 kg) for their children together with shoulder dystocia, n=6 and prolonged labour n=6. The number of children born with shoulder dystocia was directly proportional to the mothers' age. Mothers’ age category versus average birth weight, shoulder dystocia and prolonged labour Data is presented in mean ± SD Discussion: The aim of this study was to determine the prevalence of brachial plexus birth palsy (BPBP) over a five year period (January 2013-December 2017) and the predisposing factors of BPBP at a regional hospital in Ghana. The clinical profiles of 320 paediatric cases, within the study period were reviewed. The prevalence of BPBP was 14.7%, which is much lower than the prevalence (27.2%) recorded in a tertiary hospital in urban Ghana [11]. The difference in the prevalence could be attributed to the geographical difference (urban vs. rural). Majority (59.6%) of the cases recorded were females, with 80% of the cases affecting the right arm. This result is consistent with the findings of studies conducted by Toopchizadeh and colleagues [14] in Tabriz, Iran. Hamzat et al. [11] also found out, right arm palsy to be more prevalent than left arm palsy. In the Ghanaian, context, the high prevalence of right arm palsy could pose social problems for children because the Ghanaian culture tends to place much emphasis on the importance of the right hand in performing actions such as greeting, eating, handshaking and preparing meals [11]. The implication therefore is that children with right hand palsy may be seen as outcasts within the Ghanaian societal setup. Most (93.6%) of the study participants were cephalic spontaneous vaginal delivery with only two caesarean section (4.2%) and one having no documentation on the mode of delivery. About 95% of cephalic presentations were diagnosed with Erb's palsy, with 5% diagnosed with Klumpke's palsy; a breech presentation also resulted in Erb's palsy. These results confirm with a study conducted by Hale et al. [15], where they reported hyper abduction in breech presentation as a cause for Klumpke's palsy. For the current study, the high number of erb's palsy recorded for cephalic presentation will be due to the high birth weight of the children. Cephalic presentation with shoulder dystocia was recorded as the main mechanism putting a traction force on the upper brachial plexus [16]. The mechanism for Erb's palsy was reported to be due to, a traction force between the head and the shoulder, thereby putting a tension force on the nerve which might lead to tearing of some part or the entire upper brachial plexus [16]. This position supports the reason why majority of the children suffered from Erb's palsy in this study. According to Hehir and colleagues [16], cephalic presentation increase the risk for shoulder dystocia and causes a tension force on the upper brachial plexus, thereby leading to Erb's palsy in cases where the nerves are injured. The birth complications recorded in this study were shoulder dystocia (27.6%), prolonged labour (42.5%), clavicular fracture (2.1%) and humeral fracture (2.1%). These findings correspond to existing literature documenting shoulder dystocia and prolonged labour as the commonest complication for BPBP, with fractures recorded in some cases [17]. Similarly, Coroneos et al. [18] reported shoulder dystocia as the major cause for BPBP, but also found out that prolonged labour is the commonest risk factor resulting in BPBP. The recorded birth weight for the 47 BPBP participants used in this study were in the ranges of 3.0-5.0 kg with an average birth weight of 3.9 ± 0.5 kg. This finding shows that majority of the children's birth weight was within normal ranges (3.5-4.3 kg). This is consistent with existing literature which indicates that increase birth weight is a risk factor for BPBP [17]. The complications resulting in brachial plexus injury in this population is due to mothers not attending antenatal clinics, refusal to undergo caesarean section, delay in reporting to the hospital when labour sets in and lack of early referrals. Among the sampled population, majority of them suffered Erb's palsy (93.6%) with only 6.4% recording Klumpke's palsy. This shows that the number of children who suffered upper brachial injury was more than the lower plexus injuries. The clinical implication of this is that, clinicians and healthcare providers must be educated on the risk factors for brachial plexus injuries. Among the study population, only 19 mothers had their demographic details recorded. The average age and average birth weight of nulliparous mothers was 21 years and 3.8 kg respectively; whereas that for multiparous mothers was 34 years and 3.9 kg respectively. This shows that both nulliparous and multiparous mothers are at risk for delivering babies who may suffer birth injuries because the average birth weight is the same for both groups of women and therefore caregivers must provide the same level of care to prevent birth injuries. This study found out that age and parity status are risk free for brachial plexus injuries, since they are all at risk for the condition. This study contravenes other studies where mothers above 35 years and those below 16 years are considered as risk factors for brachial plexus injuries [19]. Birth weight above 4 kg was associated with an increased risk of shoulder dystocia and prolonged labour. Additionally, there was also an associated increased risk of shoulder dystocia, high birth weight (macrosomia) and prolonged labour in mothers above 34 years. This finding is consistent with the major risk factors reported by various studies [20, 21]. Limitations: this study was limited by the lack of complete and adequate maternal data. There was lack of documentation on birth weight for ten of the children and therefore conclusions about the study population cannot be representative. Conclusion: In conclusion, the results of this study showed a BPBP prevalence of 15% and most of the Erb's palsy cases had right arm affectation. The study also showed that majority of the babies had a birth weight >4.0 kg and this may accounts for the high number of shoulder dystocia cases. The presentations at birth were predominantly cephalic and this may pose an increased risk of upper plexus injuries among the macrosomic children. Majority of babies diagnosed with Erb's palsy were macrosomic with shoulder dystocia. Recommendations: the researchers recommend the need for appropriate documentation in the hospital, and the need for healthcare professionals to be mindful of the complications and the risk factors of BPBP so as to provide immediate and appropriate care. What is known about this topic Increase birth weight is a risk factor for BPBP; Birth weight above 4 kg is associated with an increased risk of shoulder dystocia and prolonged labour. Increase birth weight is a risk factor for BPBP; Birth weight above 4 kg is associated with an increased risk of shoulder dystocia and prolonged labour. What this study adds Increasing mothers' age was found to be directly related to the complications arising of the birth process such as prolonged labour and shoulder dystocia; Both nulliparous and multiparous mothers' are at risk for delivering babies who may suffer birth injuries because the average birth weight of children is the same for both groups of women; There is an associated increased risk of shoulder dystocia, high birth weight (macrosomia) and prolonged labour in mothers above 34 years. Increasing mothers' age was found to be directly related to the complications arising of the birth process such as prolonged labour and shoulder dystocia; Both nulliparous and multiparous mothers' are at risk for delivering babies who may suffer birth injuries because the average birth weight of children is the same for both groups of women; There is an associated increased risk of shoulder dystocia, high birth weight (macrosomia) and prolonged labour in mothers above 34 years. What is known about this topic: Increase birth weight is a risk factor for BPBP; Birth weight above 4 kg is associated with an increased risk of shoulder dystocia and prolonged labour. What this study adds: Increasing mothers' age was found to be directly related to the complications arising of the birth process such as prolonged labour and shoulder dystocia; Both nulliparous and multiparous mothers' are at risk for delivering babies who may suffer birth injuries because the average birth weight of children is the same for both groups of women; There is an associated increased risk of shoulder dystocia, high birth weight (macrosomia) and prolonged labour in mothers above 34 years. Competing interests: The authors declare no competing interests.

Background: Brachial plexus birth injury is one of the challenges associated with maternal delivery, with varying prevalence between countries. Brachial plexus birth injury poses negative health implications to children and also has socio-economic implications on families and the community as a whole. To treat brachial plexus birth injury, a multi-disciplinary treatment approach is recommended. Brachial plexus birth palsy (BPBP) is categorised into two-upper plexus injury (Erb's palsy) and lower plexus injury (Klumpke's palsy). These categories present with various degrees of injuries, with less severe injuries responding well to treatment and in most instances may resolve on their own, but serious and complicated injuries will require a multi-disciplinary treatment approach to treat and/or manage. Effective treatment and management depends on adequate knowledge of the disease condition. These include the risk factors and prevalence of brachial plexus birth palsy within a particular population at a specific period in time. The aim of this study was to determine the risk factors and the prevalence of a hospital based brachial plexus birth palsy within a five-year period (2013-2017). Methods: A five-year retrospective study design was used. The study involved selection of all clients' diagnosed with brachial plexus birth palsy, where their gender, birth weight, complications at birth, type of brachial plexus suffered, mothers' diabetes status, mother's age, birth attendant, side of affectation, presentation at birth and mode of delivery were recorded. Results: The prevalence rate of brachial plexus birth palsy was 14.7% out of a total of three hundred and twenty (320) cases reviewed over the study period in the Volta Regional Hospital. Erb's palsy was found to be the modal type of BPBP in this population (93.6%). Conclusions: There is the need to provide a nationwide education on the risk factors that predispose babies to brachial plexus birth palsy. There is also the need for frequent antenatal visit by pregnant women; this will help in the provision of best antenatal history, diagnostic investigation in determining the birth weight and safe mode of delivery.

Introduction: Brachial plexus birth palsy (BPBP) is a neurological condition that results from nerve injury to the brachial plexus; C5-T1, which supply the upper extremities [1, 2]. Al-Qattan et al. [3] classified BPBP into four main groups; group one shows injury to C5-C6 and results in paralysis of the shoulder and biceps muscles. Group two affects C5-C7, and typically results in paralysis of the shoulder, biceps and the forearm extensors [4]. Group three indicates injury to C5-T1 and results in complete paralysis of the affected upper limb. Group four affects C5-T1 and results in complete paralysis of the affected upper limb with Horner's syndrome. Zafeiriou and Psychogiou [5] further indicated that group one shows the least severity of injury whereas group four indicates the most severity of injury. Upper plexus injury is an injury to C5, C6 and sometimes C7 nerve of the cervical spine and lower plexus injury are injury to C8, and T1 nerves of the thoracic spine [6, 7]. The management of BPBP has many physical, psychological, financial, social and emotional implications on families and the country as a whole [8]. The prevalence of BPBP varies among countries. According to Foad et al. [9], the prevalence of BPBP in the United States is 1.5 in 1000 live births; shoulder dystocia had 100 times greater risk, an exceptionally large baby (>4.5 kg) had a fourteen times greater risk, and forceps delivery had a nine times greater risk for having a child with brachial plexus birth palsy. Having a twin or multiple birth mates and delivery by cesarean section had a protective effect against the occurrence of neonatal brachial plexus palsy. A study conducted by Evans-Jones and colleagues [10], reported the prevalence in the United Kingdom to be 0.42 per 1000 live births and the associated risk factors for BPBP was found to be shoulder dystocia, high birth weight and assisted delivery, but a considerably lower risk in infants delivered by caesarean section. The prevalence of children with BPBP in Nigeria over a ten year study period showed a persistent high prevalence, averaging 15.3% per year with associated problems such as birth asphyxia, humeral fracture, clavicular fracture and shoulder dislocation [4]. A study by Hamzat and colleagues [11] also reported the prevalence of BPBP in Accra, Ghana to be 27%, the results of the study further indicated that birth weight exceeding 4.0 kg, vertex presentation and vaginal delivery were the noticeable co-existing factors for BPBP in Accra. From the study done by Hamzat et al. [11] it was reported that only 55.2% of BPBP cases were referred for physiotherapy within one month after diagnosis and the treatment disposition for majority (88.1%) of the children were not documented and only 4.8% were formally discharged from physiotherapy. The prevalence of BPBP reported in Ghana focused on the Accra metropolis and there is no known data of BPBP in the Ho municipality, Volta region. The Ho municipality is a cosmopolitan urban city with health facilities serving the southern and central parts of the Volta region. BPBP is preventable when the risk factors and causes are known [12]. This work seeks to find out the prevalence and risk factors of BPBP in the Ho municipality of Ghana so as to advice policy makers on how to prevent the injury which will reduce the cost of treatment and increase productivity of the mothers. Managing of some children suffering from BPBP may need surgery and rehabilitation which poses financial issues to the family and loved ones [8]. Some children may have residual functional deficits and thereby affecting the way they function, as well psychological or emotional problems to the child [8]. This present study was undertaken to retrospectively investigate the prevalence of children who presented with BPBP and their predisposing factors in a regional hospital in Ghana. The clinical implication for this study is that there is not much information concerning the predisposing factors of BPBP in Ghana, apart from the study done by Hamzat et al. [11] that was done 10 years ago in Accra, an urban city in Ghana. Moreover the study by Hamzat et al. [11] did not provide information on the risk factors of BPBP in Ghana. As a result, there is paucity of information on the predisposing factors and the current prevalence of BPBP in a peri-urban city in Ghana. The aim of this study was to determine the prevalence and predisposing factors of BPBP over a five year period (January 2013-December 2017) at a regional hospital in Ghana. Conclusion: In conclusion, the results of this study showed a BPBP prevalence of 15% and most of the Erb's palsy cases had right arm affectation. The study also showed that majority of the babies had a birth weight >4.0 kg and this may accounts for the high number of shoulder dystocia cases. The presentations at birth were predominantly cephalic and this may pose an increased risk of upper plexus injuries among the macrosomic children. Majority of babies diagnosed with Erb's palsy were macrosomic with shoulder dystocia. Recommendations: the researchers recommend the need for appropriate documentation in the hospital, and the need for healthcare professionals to be mindful of the complications and the risk factors of BPBP so as to provide immediate and appropriate care. What is known about this topic Increase birth weight is a risk factor for BPBP; Birth weight above 4 kg is associated with an increased risk of shoulder dystocia and prolonged labour. Increase birth weight is a risk factor for BPBP; Birth weight above 4 kg is associated with an increased risk of shoulder dystocia and prolonged labour. What this study adds Increasing mothers' age was found to be directly related to the complications arising of the birth process such as prolonged labour and shoulder dystocia; Both nulliparous and multiparous mothers' are at risk for delivering babies who may suffer birth injuries because the average birth weight of children is the same for both groups of women; There is an associated increased risk of shoulder dystocia, high birth weight (macrosomia) and prolonged labour in mothers above 34 years. Increasing mothers' age was found to be directly related to the complications arising of the birth process such as prolonged labour and shoulder dystocia; Both nulliparous and multiparous mothers' are at risk for delivering babies who may suffer birth injuries because the average birth weight of children is the same for both groups of women; There is an associated increased risk of shoulder dystocia, high birth weight (macrosomia) and prolonged labour in mothers above 34 years.

Background: Brachial plexus birth injury is one of the challenges associated with maternal delivery, with varying prevalence between countries. Brachial plexus birth injury poses negative health implications to children and also has socio-economic implications on families and the community as a whole. To treat brachial plexus birth injury, a multi-disciplinary treatment approach is recommended. Brachial plexus birth palsy (BPBP) is categorised into two-upper plexus injury (Erb's palsy) and lower plexus injury (Klumpke's palsy). These categories present with various degrees of injuries, with less severe injuries responding well to treatment and in most instances may resolve on their own, but serious and complicated injuries will require a multi-disciplinary treatment approach to treat and/or manage. Effective treatment and management depends on adequate knowledge of the disease condition. These include the risk factors and prevalence of brachial plexus birth palsy within a particular population at a specific period in time. The aim of this study was to determine the risk factors and the prevalence of a hospital based brachial plexus birth palsy within a five-year period (2013-2017). Methods: A five-year retrospective study design was used. The study involved selection of all clients' diagnosed with brachial plexus birth palsy, where their gender, birth weight, complications at birth, type of brachial plexus suffered, mothers' diabetes status, mother's age, birth attendant, side of affectation, presentation at birth and mode of delivery were recorded. Results: The prevalence rate of brachial plexus birth palsy was 14.7% out of a total of three hundred and twenty (320) cases reviewed over the study period in the Volta Regional Hospital. Erb's palsy was found to be the modal type of BPBP in this population (93.6%). Conclusions: There is the need to provide a nationwide education on the risk factors that predispose babies to brachial plexus birth palsy. There is also the need for frequent antenatal visit by pregnant women; this will help in the provision of best antenatal history, diagnostic investigation in determining the birth weight and safe mode of delivery.

[CONTENT] Brachial plexus birth palsy | predisposing factors | erb’s palsy | Regional hospital | Ghana [SUMMARY]

[CONTENT] Brachial plexus birth palsy | predisposing factors | erb’s palsy | Regional hospital | Ghana [SUMMARY]

[CONTENT] Brachial plexus birth palsy | predisposing factors | erb’s palsy | Regional hospital | Ghana [SUMMARY]

[CONTENT] Brachial plexus birth palsy | predisposing factors | erb’s palsy | Regional hospital | Ghana [SUMMARY]

[CONTENT] Brachial plexus birth palsy | predisposing factors | erb’s palsy | Regional hospital | Ghana [SUMMARY]

[CONTENT] Brachial plexus birth palsy | predisposing factors | erb’s palsy | Regional hospital | Ghana [SUMMARY]

[CONTENT] Birth Weight | Delivery, Obstetric | Female | Ghana | Humans | Infant, Newborn | Male | Neonatal Brachial Plexus Palsy | Pregnancy | Prenatal Care | Prevalence | Retrospective Studies | Risk Factors | Severity of Illness Index [SUMMARY]

[CONTENT] Birth Weight | Delivery, Obstetric | Female | Ghana | Humans | Infant, Newborn | Male | Neonatal Brachial Plexus Palsy | Pregnancy | Prenatal Care | Prevalence | Retrospective Studies | Risk Factors | Severity of Illness Index [SUMMARY]

[CONTENT] Birth Weight | Delivery, Obstetric | Female | Ghana | Humans | Infant, Newborn | Male | Neonatal Brachial Plexus Palsy | Pregnancy | Prenatal Care | Prevalence | Retrospective Studies | Risk Factors | Severity of Illness Index [SUMMARY]

[CONTENT] Birth Weight | Delivery, Obstetric | Female | Ghana | Humans | Infant, Newborn | Male | Neonatal Brachial Plexus Palsy | Pregnancy | Prenatal Care | Prevalence | Retrospective Studies | Risk Factors | Severity of Illness Index [SUMMARY]

[CONTENT] Birth Weight | Delivery, Obstetric | Female | Ghana | Humans | Infant, Newborn | Male | Neonatal Brachial Plexus Palsy | Pregnancy | Prenatal Care | Prevalence | Retrospective Studies | Risk Factors | Severity of Illness Index [SUMMARY]

[CONTENT] Birth Weight | Delivery, Obstetric | Female | Ghana | Humans | Infant, Newborn | Male | Neonatal Brachial Plexus Palsy | Pregnancy | Prenatal Care | Prevalence | Retrospective Studies | Risk Factors | Severity of Illness Index [SUMMARY]

[CONTENT] child brachial plexus | birth palsy bpbp | brachial plexus injuries | palsy bpbp neurological | brachial plexus c5 [SUMMARY]

[CONTENT] child brachial plexus | birth palsy bpbp | brachial plexus injuries | palsy bpbp neurological | brachial plexus c5 [SUMMARY]

[CONTENT] child brachial plexus | birth palsy bpbp | brachial plexus injuries | palsy bpbp neurological | brachial plexus c5 [SUMMARY]

[CONTENT] child brachial plexus | birth palsy bpbp | brachial plexus injuries | palsy bpbp neurological | brachial plexus c5 [SUMMARY]

[CONTENT] child brachial plexus | birth palsy bpbp | brachial plexus injuries | palsy bpbp neurological | brachial plexus c5 [SUMMARY]

[CONTENT] child brachial plexus | birth palsy bpbp | brachial plexus injuries | palsy bpbp neurological | brachial plexus c5 [SUMMARY]

[CONTENT] birth | study | weight | bpbp | birth weight | children | risk | shoulder | palsy | mothers [SUMMARY]

[CONTENT] birth | study | weight | bpbp | birth weight | children | risk | shoulder | palsy | mothers [SUMMARY]

[CONTENT] birth | study | weight | bpbp | birth weight | children | risk | shoulder | palsy | mothers [SUMMARY]

[CONTENT] birth | study | weight | bpbp | birth weight | children | risk | shoulder | palsy | mothers [SUMMARY]

[CONTENT] birth | study | weight | bpbp | birth weight | children | risk | shoulder | palsy | mothers [SUMMARY]

[CONTENT] birth | study | weight | bpbp | birth weight | children | risk | shoulder | palsy | mothers [SUMMARY]

[CONTENT] bpbp | injury | prevalence | factors | ghana | c5 | study | group | risk | factors bpbp [SUMMARY]

[CONTENT] study | vrh | control | data | period | 15 years | wrist | retrieved | paediatric | birth [SUMMARY]

[CONTENT] children | documented | birth | weight | birth weight | table | average birth | average | average birth weight | paediatric [SUMMARY]

[CONTENT] birth | risk | weight | birth weight | shoulder dystocia | dystocia | shoulder | prolonged labour | prolonged | labour [SUMMARY]

[CONTENT] birth | risk | weight | birth weight | study | bpbp | dystocia | shoulder dystocia | shoulder | children [SUMMARY]

[CONTENT] birth | risk | weight | birth weight | study | bpbp | dystocia | shoulder dystocia | shoulder | children [SUMMARY]

[CONTENT] ||| ||| ||| two | Klumpke ||| ||| ||| ||| five-year | 2013-2017 [SUMMARY]

[CONTENT] five-year ||| [SUMMARY]

[CONTENT] 14.7% | three hundred and twenty | 320 | the Volta Regional Hospital ||| 93.6% [SUMMARY]

[CONTENT] ||| [SUMMARY]

[CONTENT] ||| ||| ||| two | Klumpke ||| ||| ||| ||| five-year | 2013-2017 ||| five-year ||| ||| ||| 14.7% | three hundred and twenty | 320 | the Volta Regional Hospital ||| 93.6% ||| ||| [SUMMARY]

[CONTENT] ||| ||| ||| two | Klumpke ||| ||| ||| ||| five-year | 2013-2017 ||| five-year ||| ||| ||| 14.7% | three hundred and twenty | 320 | the Volta Regional Hospital ||| 93.6% ||| ||| [SUMMARY]

Evolution of Estrogen Receptor Status from Primary Tumors to Metastasis and Serially Collected Circulating Tumor Cells.

32326116

The estrogen receptor (ER) can change expression between primary tumor (PT) and distant metastasis (DM) in breast cancer. A tissue biopsy reflects a momentary state at one location, whereas circulating tumor cells (CTCs) reflect real-time tumor progression. We evaluated ER-status during tumor progression from PT to DM and CTCs, and related the ER-status of CTCs to prognosis.

BACKGROUND

In a study of metastatic breast cancer, blood was collected at different timepoints. After CellSearch® enrichment, CTCs were captured on DropMount slides and evaluated for ER expression at baseline (BL) and after 1 and 3 months of therapy. Comparison of the ER-status of PT, DM, and CTCs at different timepoints was performed using the McNemar test. The primary endpoint was progression-free survival (PFS).

METHODS

Evidence of a shift from ER positivity to negativity between PT and DM was demonstrated (p = 0.019). We found strong evidence of similar shifts from PT to CTCs at different timepoints (p < 0.0001). ER-positive CTCs at 1 and 3 months were related to better prognosis.

RESULTS

A shift in ER-status from PT to DM/CTCs was demonstrated. ER-positive CTCs during systemic therapy might reflect the retention of a favorable phenotype that still responds to therapy.

CONCLUSIONS

7215368

1. Introduction

Breast cancer is the most common malignant disease in women, and although five-year survival is approaching 90%, 20–30% of women with an initially local disease develop metastatic disease within 10 years [1]. The median overall survival (OS) in women with metastatic breast cancer (MBC) is approximately two years, and only 25% survive beyond five years [2]. Evaluating the response of therapy in MBC can be difficult, especially in patients with non-measurable disease. Guidelines for therapy evaluation include diagnostic imaging, clinical assessment, and also evaluation of liquid-based tumor markers [2,3,4]. In breast cancer, the estrogen receptor (ER) and the human epidermal growth factor receptor 2 (HER2), which are predictive of the response to endocrine- and HER2-targeted therapy, respectively, are the most important treatment predictive markers. Since it was shown that the expression of ER and HER2 frequently changes between primary tumor (PT) and distant metastasis (DM) in breast cancer [5,6], it is important to re-evaluate the biomarker status in biopsies from metastatic lesions. Furthermore, there are indications that cancer cells lose their ER expression during endocrine treatment [7,8] as a mechanism of resistance. The choice of therapy in a metastatic setting should therefore preferably be based on the receptor status of the metastatic biopsy when available. Due to heterogeneity, biopsies from different sites should be evaluated. In clinical practice, however, it is not always feasible to biopsy multiple metastatic sites, and the metastasis might also be situated at a non-accessible site. Moreover, current diagnostic methods, i.e., needle biopsies, limit the amount of material available for further analyses. Liquid biopsies evaluating the presence of circulating tumor cells (CTCs) are non-invasive and easily accessible with a limited risk of complications in comparison to an invasive tissue biopsy. They can be repeated regularly, which allows for the serial monitoring of real-time tumor evolution/response to treatment. Additionally, a tissue biopsy reflects a momentary state of the tumor at one location in the body, whereas liquid-based markers reflect real-time tumor progression and are not dependent on repeated invasive tissue biopsies. CTCs can be detected in a liquid biopsy, and the CellSearch® system (©Menarini Silicon Biosystems, Inc 2020, USA) is considered the golden standard of CTC capturing and enumeration in MBC since it is the only Food and Drug Administration (FDA)-approved technology, so far. Multiple trials have shown CellSearch® enumerated CTCs as an independent prognostic factor in MBC [9,10,11,12,13,14,15]. CTCs have also been proposed to more accurately reflect tumor heterogeneity than biopsies of single metastases and are considered to originate from both the primary tumor and all metastatic sites [16]. Therefore, liquid biopsies enumerating CTCs might be a better proxy for the total metastatic burden. However, CTC enumeration alone is insufficient to predict the efficacy of therapy, and the addition of CTC tumor-marker expression might better predict the benefit of a therapy [17]. The primary aim of the present study was to characterize CTCs detected in blood samples from MBC patients and to evaluate ER-status during tumor progression from primary tumors to distant metastasis and CTCs, respectively. Our secondary aim was to relate the ER-status of CTCs to prognosis. We hypothesized that, by evaluating the ER-status of single CTCs, improved therapy guidance could be provided.

2. Results

For the study cohort with an available CTC status using CellSearch®, and with information on the ER-status in the PT (n = 147), the median age was 65 years (range 40–84 years; Figure 1, Table 1). In total, 63% (92/147) of the patients were given endocrine therapy as first-line systemic therapy, whereas 48% (70/147) had chemotherapy. Furthermore, 84% (123/147) of PTs and 86% (105/122) of DMs were ER-positive (Table 1). The median time from diagnosis of PT to diagnosis of MBC was five years (range 0–36 years). 2.1. CTC Status at Baseline and During Follow-Up At baseline (BL), 53% (76/144) of the patients were classified in the inferior prognostic group according to the previously approved and validated cut-off of ≥5 CTCs [9,10]. The corresponding percentages after 1 and 3 months of the first-line of therapy were 29% (37/128) and 19% (21/113), respectively. DropMount slides could be prepared for 113 cases (Figure 1). When the CTC-DropMount method was used for CTC characterization, the corresponding numbers for cases included in the inferior prognostic group were 54% (35/65), 32% (7/22), and 52% (12/23) at BL, after 1 month and after 3 months of therapy, respectively. At baseline (BL), 53% (76/144) of the patients were classified in the inferior prognostic group according to the previously approved and validated cut-off of ≥5 CTCs [9,10]. The corresponding percentages after 1 and 3 months of the first-line of therapy were 29% (37/128) and 19% (21/113), respectively. DropMount slides could be prepared for 113 cases (Figure 1). When the CTC-DropMount method was used for CTC characterization, the corresponding numbers for cases included in the inferior prognostic group were 54% (35/65), 32% (7/22), and 52% (12/23) at BL, after 1 month and after 3 months of therapy, respectively. 2.2. ER Status of CTCs By again applying the CTC-DropMount method for characterization, 26% (17/65) of CTCs were ER-positive at BL compared with 23% (5/22) after 1 month and 43% (10/23) after 3 months of therapy. The distribution of number of CTCs at BL and the ER-status are depicted in Figure 2a, b. For ER-positive CTC cases at BL (Figure 2a), the number of detected cells ranged from 1 to 47, where cases were considered ER-positive if at least one CTC was positive for ER. With a cut-off of ≥5CTCs, 13 cases would have been regarded as ER-positive at BL instead of 17. For ER-negative cases (Figure 2b), the number of evaluated CTCs varied from 1 to 1094. By again applying the CTC-DropMount method for characterization, 26% (17/65) of CTCs were ER-positive at BL compared with 23% (5/22) after 1 month and 43% (10/23) after 3 months of therapy. The distribution of number of CTCs at BL and the ER-status are depicted in Figure 2a, b. For ER-positive CTC cases at BL (Figure 2a), the number of detected cells ranged from 1 to 47, where cases were considered ER-positive if at least one CTC was positive for ER. With a cut-off of ≥5CTCs, 13 cases would have been regarded as ER-positive at BL instead of 17. For ER-negative cases (Figure 2b), the number of evaluated CTCs varied from 1 to 1094. 2.3. Shift in the ER Status Evidence of a shift in the ER-status from PT to DM was assessed with the McNemar test, showing a shift towards more ER-negative tumors in DM (Table 2; Figure 3a). In total, 74% (90/122) retained ER positivity from PT to DM, 12% (15/122) changed from ER-positive in PT to ER-negative in DM, and 3% (4/122) changed from ER-negative to ER-positive DM (Table 2, Figure 3a). Interestingly, when separately comparing the ER-status of the PT with the ER-status of CTCs at different timepoints, evidence of a shift from ER positivity to negativity was strong (Table 2; Figure 3b–d). Among 50 patients with ER-positive PTs, only 16 had ER-positive CTCs at BL (p < 0.0001; Figure 3b). McNemar analysis comparing the ER-status in DM and of CTCs at different timepoints showed evidence of a negative association at BL and after 1 month but not after 3 months (Table 2). Statistical power was low for a comparison of the ER-status of CTCs between follow-up timepoints (NBL = 17, N1 month = 19, and N3 months = 10; Table 2). Evidence of a shift in the ER-status from PT to DM was assessed with the McNemar test, showing a shift towards more ER-negative tumors in DM (Table 2; Figure 3a). In total, 74% (90/122) retained ER positivity from PT to DM, 12% (15/122) changed from ER-positive in PT to ER-negative in DM, and 3% (4/122) changed from ER-negative to ER-positive DM (Table 2, Figure 3a). Interestingly, when separately comparing the ER-status of the PT with the ER-status of CTCs at different timepoints, evidence of a shift from ER positivity to negativity was strong (Table 2; Figure 3b–d). Among 50 patients with ER-positive PTs, only 16 had ER-positive CTCs at BL (p < 0.0001; Figure 3b). McNemar analysis comparing the ER-status in DM and of CTCs at different timepoints showed evidence of a negative association at BL and after 1 month but not after 3 months (Table 2). Statistical power was low for a comparison of the ER-status of CTCs between follow-up timepoints (NBL = 17, N1 month = 19, and N3 months = 10; Table 2). 2.4. ER Status of CTCs and PFS Among all patients with positive CTCs (≥1 CTCs) at BL (n = 65), evidence of an association between the ER positivity of CTCs at BL, measured by the CTC-DropMount method, and improved PFS was weak (hazard ratio (HR): 0.57, 95% confidence interval (CI): 0.29–1.1, p = 0.10; Figure 4a). The effect, as well as the evidence, was stronger in landmark analyses using the ER-status of CTCs measured at 1 month (HR: 0.25, 95% CI: 0.055–1.1, p = 0.066; Figure 4b) and at 3 months (HR: 0.33, 95% CI: 0.11–1.01, p = 0.052; Figure 4c). Similar results were seen in patients with ≥5 CTCs at BL (n = 61). Thirteen cases were defined as ER-positive and 48 cases as ER-negative CTCs at BL, and ER positivity was associated with better PFS at BL (HR: 0.33, 95% CI: 0.13–0.83, p = 0.019), at 1 month (HR: 0.64, 95% CI: 0.071–5.8, p = 0.69) and at 3 months (HR: 0.32, 95% CI: 0.078–1.3, p = 0.12). Among all patients with positive CTCs (≥1 CTCs) at BL (n = 65), evidence of an association between the ER positivity of CTCs at BL, measured by the CTC-DropMount method, and improved PFS was weak (hazard ratio (HR): 0.57, 95% confidence interval (CI): 0.29–1.1, p = 0.10; Figure 4a). The effect, as well as the evidence, was stronger in landmark analyses using the ER-status of CTCs measured at 1 month (HR: 0.25, 95% CI: 0.055–1.1, p = 0.066; Figure 4b) and at 3 months (HR: 0.33, 95% CI: 0.11–1.01, p = 0.052; Figure 4c). Similar results were seen in patients with ≥5 CTCs at BL (n = 61). Thirteen cases were defined as ER-positive and 48 cases as ER-negative CTCs at BL, and ER positivity was associated with better PFS at BL (HR: 0.33, 95% CI: 0.13–0.83, p = 0.019), at 1 month (HR: 0.64, 95% CI: 0.071–5.8, p = 0.69) and at 3 months (HR: 0.32, 95% CI: 0.078–1.3, p = 0.12).

5. Conclusions

We demonstrated evidence of a shift from ER positivity to negativity in the PT compared to DM, as well as evidence of a similar shift in the ER-status of CTCs from PT and DM. Using the CTC-DropMount method, the ER positivity of CTCs at 1 and 3 months was related to better prognosis. The ER positivity of CTCs after the initiation of systemic therapy might reflect the retention of a favorable phenotype that still responds to therapy.

1. Introduction: Breast cancer is the most common malignant disease in women, and although five-year survival is approaching 90%, 20–30% of women with an initially local disease develop metastatic disease within 10 years [1]. The median overall survival (OS) in women with metastatic breast cancer (MBC) is approximately two years, and only 25% survive beyond five years [2]. Evaluating the response of therapy in MBC can be difficult, especially in patients with non-measurable disease. Guidelines for therapy evaluation include diagnostic imaging, clinical assessment, and also evaluation of liquid-based tumor markers [2,3,4]. In breast cancer, the estrogen receptor (ER) and the human epidermal growth factor receptor 2 (HER2), which are predictive of the response to endocrine- and HER2-targeted therapy, respectively, are the most important treatment predictive markers. Since it was shown that the expression of ER and HER2 frequently changes between primary tumor (PT) and distant metastasis (DM) in breast cancer [5,6], it is important to re-evaluate the biomarker status in biopsies from metastatic lesions. Furthermore, there are indications that cancer cells lose their ER expression during endocrine treatment [7,8] as a mechanism of resistance. The choice of therapy in a metastatic setting should therefore preferably be based on the receptor status of the metastatic biopsy when available. Due to heterogeneity, biopsies from different sites should be evaluated. In clinical practice, however, it is not always feasible to biopsy multiple metastatic sites, and the metastasis might also be situated at a non-accessible site. Moreover, current diagnostic methods, i.e., needle biopsies, limit the amount of material available for further analyses. Liquid biopsies evaluating the presence of circulating tumor cells (CTCs) are non-invasive and easily accessible with a limited risk of complications in comparison to an invasive tissue biopsy. They can be repeated regularly, which allows for the serial monitoring of real-time tumor evolution/response to treatment. Additionally, a tissue biopsy reflects a momentary state of the tumor at one location in the body, whereas liquid-based markers reflect real-time tumor progression and are not dependent on repeated invasive tissue biopsies. CTCs can be detected in a liquid biopsy, and the CellSearch® system (©Menarini Silicon Biosystems, Inc 2020, USA) is considered the golden standard of CTC capturing and enumeration in MBC since it is the only Food and Drug Administration (FDA)-approved technology, so far. Multiple trials have shown CellSearch® enumerated CTCs as an independent prognostic factor in MBC [9,10,11,12,13,14,15]. CTCs have also been proposed to more accurately reflect tumor heterogeneity than biopsies of single metastases and are considered to originate from both the primary tumor and all metastatic sites [16]. Therefore, liquid biopsies enumerating CTCs might be a better proxy for the total metastatic burden. However, CTC enumeration alone is insufficient to predict the efficacy of therapy, and the addition of CTC tumor-marker expression might better predict the benefit of a therapy [17]. The primary aim of the present study was to characterize CTCs detected in blood samples from MBC patients and to evaluate ER-status during tumor progression from primary tumors to distant metastasis and CTCs, respectively. Our secondary aim was to relate the ER-status of CTCs to prognosis. We hypothesized that, by evaluating the ER-status of single CTCs, improved therapy guidance could be provided. 2. Results: For the study cohort with an available CTC status using CellSearch®, and with information on the ER-status in the PT (n = 147), the median age was 65 years (range 40–84 years; Figure 1, Table 1). In total, 63% (92/147) of the patients were given endocrine therapy as first-line systemic therapy, whereas 48% (70/147) had chemotherapy. Furthermore, 84% (123/147) of PTs and 86% (105/122) of DMs were ER-positive (Table 1). The median time from diagnosis of PT to diagnosis of MBC was five years (range 0–36 years). 2.1. CTC Status at Baseline and During Follow-Up At baseline (BL), 53% (76/144) of the patients were classified in the inferior prognostic group according to the previously approved and validated cut-off of ≥5 CTCs [9,10]. The corresponding percentages after 1 and 3 months of the first-line of therapy were 29% (37/128) and 19% (21/113), respectively. DropMount slides could be prepared for 113 cases (Figure 1). When the CTC-DropMount method was used for CTC characterization, the corresponding numbers for cases included in the inferior prognostic group were 54% (35/65), 32% (7/22), and 52% (12/23) at BL, after 1 month and after 3 months of therapy, respectively. At baseline (BL), 53% (76/144) of the patients were classified in the inferior prognostic group according to the previously approved and validated cut-off of ≥5 CTCs [9,10]. The corresponding percentages after 1 and 3 months of the first-line of therapy were 29% (37/128) and 19% (21/113), respectively. DropMount slides could be prepared for 113 cases (Figure 1). When the CTC-DropMount method was used for CTC characterization, the corresponding numbers for cases included in the inferior prognostic group were 54% (35/65), 32% (7/22), and 52% (12/23) at BL, after 1 month and after 3 months of therapy, respectively. 2.2. ER Status of CTCs By again applying the CTC-DropMount method for characterization, 26% (17/65) of CTCs were ER-positive at BL compared with 23% (5/22) after 1 month and 43% (10/23) after 3 months of therapy. The distribution of number of CTCs at BL and the ER-status are depicted in Figure 2a, b. For ER-positive CTC cases at BL (Figure 2a), the number of detected cells ranged from 1 to 47, where cases were considered ER-positive if at least one CTC was positive for ER. With a cut-off of ≥5CTCs, 13 cases would have been regarded as ER-positive at BL instead of 17. For ER-negative cases (Figure 2b), the number of evaluated CTCs varied from 1 to 1094. By again applying the CTC-DropMount method for characterization, 26% (17/65) of CTCs were ER-positive at BL compared with 23% (5/22) after 1 month and 43% (10/23) after 3 months of therapy. The distribution of number of CTCs at BL and the ER-status are depicted in Figure 2a, b. For ER-positive CTC cases at BL (Figure 2a), the number of detected cells ranged from 1 to 47, where cases were considered ER-positive if at least one CTC was positive for ER. With a cut-off of ≥5CTCs, 13 cases would have been regarded as ER-positive at BL instead of 17. For ER-negative cases (Figure 2b), the number of evaluated CTCs varied from 1 to 1094. 2.3. Shift in the ER Status Evidence of a shift in the ER-status from PT to DM was assessed with the McNemar test, showing a shift towards more ER-negative tumors in DM (Table 2; Figure 3a). In total, 74% (90/122) retained ER positivity from PT to DM, 12% (15/122) changed from ER-positive in PT to ER-negative in DM, and 3% (4/122) changed from ER-negative to ER-positive DM (Table 2, Figure 3a). Interestingly, when separately comparing the ER-status of the PT with the ER-status of CTCs at different timepoints, evidence of a shift from ER positivity to negativity was strong (Table 2; Figure 3b–d). Among 50 patients with ER-positive PTs, only 16 had ER-positive CTCs at BL (p < 0.0001; Figure 3b). McNemar analysis comparing the ER-status in DM and of CTCs at different timepoints showed evidence of a negative association at BL and after 1 month but not after 3 months (Table 2). Statistical power was low for a comparison of the ER-status of CTCs between follow-up timepoints (NBL = 17, N1 month = 19, and N3 months = 10; Table 2). Evidence of a shift in the ER-status from PT to DM was assessed with the McNemar test, showing a shift towards more ER-negative tumors in DM (Table 2; Figure 3a). In total, 74% (90/122) retained ER positivity from PT to DM, 12% (15/122) changed from ER-positive in PT to ER-negative in DM, and 3% (4/122) changed from ER-negative to ER-positive DM (Table 2, Figure 3a). Interestingly, when separately comparing the ER-status of the PT with the ER-status of CTCs at different timepoints, evidence of a shift from ER positivity to negativity was strong (Table 2; Figure 3b–d). Among 50 patients with ER-positive PTs, only 16 had ER-positive CTCs at BL (p < 0.0001; Figure 3b). McNemar analysis comparing the ER-status in DM and of CTCs at different timepoints showed evidence of a negative association at BL and after 1 month but not after 3 months (Table 2). Statistical power was low for a comparison of the ER-status of CTCs between follow-up timepoints (NBL = 17, N1 month = 19, and N3 months = 10; Table 2). 2.4. ER Status of CTCs and PFS Among all patients with positive CTCs (≥1 CTCs) at BL (n = 65), evidence of an association between the ER positivity of CTCs at BL, measured by the CTC-DropMount method, and improved PFS was weak (hazard ratio (HR): 0.57, 95% confidence interval (CI): 0.29–1.1, p = 0.10; Figure 4a). The effect, as well as the evidence, was stronger in landmark analyses using the ER-status of CTCs measured at 1 month (HR: 0.25, 95% CI: 0.055–1.1, p = 0.066; Figure 4b) and at 3 months (HR: 0.33, 95% CI: 0.11–1.01, p = 0.052; Figure 4c). Similar results were seen in patients with ≥5 CTCs at BL (n = 61). Thirteen cases were defined as ER-positive and 48 cases as ER-negative CTCs at BL, and ER positivity was associated with better PFS at BL (HR: 0.33, 95% CI: 0.13–0.83, p = 0.019), at 1 month (HR: 0.64, 95% CI: 0.071–5.8, p = 0.69) and at 3 months (HR: 0.32, 95% CI: 0.078–1.3, p = 0.12). Among all patients with positive CTCs (≥1 CTCs) at BL (n = 65), evidence of an association between the ER positivity of CTCs at BL, measured by the CTC-DropMount method, and improved PFS was weak (hazard ratio (HR): 0.57, 95% confidence interval (CI): 0.29–1.1, p = 0.10; Figure 4a). The effect, as well as the evidence, was stronger in landmark analyses using the ER-status of CTCs measured at 1 month (HR: 0.25, 95% CI: 0.055–1.1, p = 0.066; Figure 4b) and at 3 months (HR: 0.33, 95% CI: 0.11–1.01, p = 0.052; Figure 4c). Similar results were seen in patients with ≥5 CTCs at BL (n = 61). Thirteen cases were defined as ER-positive and 48 cases as ER-negative CTCs at BL, and ER positivity was associated with better PFS at BL (HR: 0.33, 95% CI: 0.13–0.83, p = 0.019), at 1 month (HR: 0.64, 95% CI: 0.071–5.8, p = 0.69) and at 3 months (HR: 0.32, 95% CI: 0.078–1.3, p = 0.12). 2.1. CTC Status at Baseline and During Follow-Up: At baseline (BL), 53% (76/144) of the patients were classified in the inferior prognostic group according to the previously approved and validated cut-off of ≥5 CTCs [9,10]. The corresponding percentages after 1 and 3 months of the first-line of therapy were 29% (37/128) and 19% (21/113), respectively. DropMount slides could be prepared for 113 cases (Figure 1). When the CTC-DropMount method was used for CTC characterization, the corresponding numbers for cases included in the inferior prognostic group were 54% (35/65), 32% (7/22), and 52% (12/23) at BL, after 1 month and after 3 months of therapy, respectively. 2.2. ER Status of CTCs: By again applying the CTC-DropMount method for characterization, 26% (17/65) of CTCs were ER-positive at BL compared with 23% (5/22) after 1 month and 43% (10/23) after 3 months of therapy. The distribution of number of CTCs at BL and the ER-status are depicted in Figure 2a, b. For ER-positive CTC cases at BL (Figure 2a), the number of detected cells ranged from 1 to 47, where cases were considered ER-positive if at least one CTC was positive for ER. With a cut-off of ≥5CTCs, 13 cases would have been regarded as ER-positive at BL instead of 17. For ER-negative cases (Figure 2b), the number of evaluated CTCs varied from 1 to 1094. 2.3. Shift in the ER Status: Evidence of a shift in the ER-status from PT to DM was assessed with the McNemar test, showing a shift towards more ER-negative tumors in DM (Table 2; Figure 3a). In total, 74% (90/122) retained ER positivity from PT to DM, 12% (15/122) changed from ER-positive in PT to ER-negative in DM, and 3% (4/122) changed from ER-negative to ER-positive DM (Table 2, Figure 3a). Interestingly, when separately comparing the ER-status of the PT with the ER-status of CTCs at different timepoints, evidence of a shift from ER positivity to negativity was strong (Table 2; Figure 3b–d). Among 50 patients with ER-positive PTs, only 16 had ER-positive CTCs at BL (p < 0.0001; Figure 3b). McNemar analysis comparing the ER-status in DM and of CTCs at different timepoints showed evidence of a negative association at BL and after 1 month but not after 3 months (Table 2). Statistical power was low for a comparison of the ER-status of CTCs between follow-up timepoints (NBL = 17, N1 month = 19, and N3 months = 10; Table 2). 2.4. ER Status of CTCs and PFS: Among all patients with positive CTCs (≥1 CTCs) at BL (n = 65), evidence of an association between the ER positivity of CTCs at BL, measured by the CTC-DropMount method, and improved PFS was weak (hazard ratio (HR): 0.57, 95% confidence interval (CI): 0.29–1.1, p = 0.10; Figure 4a). The effect, as well as the evidence, was stronger in landmark analyses using the ER-status of CTCs measured at 1 month (HR: 0.25, 95% CI: 0.055–1.1, p = 0.066; Figure 4b) and at 3 months (HR: 0.33, 95% CI: 0.11–1.01, p = 0.052; Figure 4c). Similar results were seen in patients with ≥5 CTCs at BL (n = 61). Thirteen cases were defined as ER-positive and 48 cases as ER-negative CTCs at BL, and ER positivity was associated with better PFS at BL (HR: 0.33, 95% CI: 0.13–0.83, p = 0.019), at 1 month (HR: 0.64, 95% CI: 0.071–5.8, p = 0.69) and at 3 months (HR: 0.32, 95% CI: 0.078–1.3, p = 0.12). 3. Discussion: The tissue biopsy reflects a momentary state of a tumor at one location in the body, whereas liquid-based markers reflect real-time tumor progression and are not dependent on repeated invasive tissue biopsies. Enumeration and characterization of CTCs has potential as a less invasive diagnostic method to evaluate the progression of breast cancer. To our knowledge, this is one of a few studies evaluating the ER-status of single CTCs in metastatic breast cancer using an immunofluorescence method [18]. In the present study, we also evaluated the ER-status in PT and DM and identified a shift towards more ER-negative cases in DM. A shift in the ER-status from PT to CTCs was also detected. Out of 50 patients with ER-positive PT, only 16 had ER-positive CTCs at BL. In the present study, retained ER positivity of CTCs after initiation of systemic therapy was associated with better prognosis and thus seemed to reflect a favorable retained phenotype that still responded to therapy. Van de Ven et al. showed that, after neoadjuvant chemotherapy, there was a 2.5–17% change in the ER-status from positive to negative [19]; in metastatic breast cancer, 17% of the cases were found to have a different ER-status in metastasis compared to the primary tumor [20]. Hence, we hypothesized that, by evaluating the ER-status of single CTCs, improved therapy guidance could be provided. Bock et al. reported results similar to ours, with a shift towards more ER negativity in CTCs; 37% ER-positive in PT vs. 8% ER-positive in CTCs compared to our results, with 84% ER-positive in PT vs. 26% ER-positive in CTCs at BL [18]. Other studies showed ER-status discordance between CTCs and primary or metastatic biopsies similar to our results, ranging from 30% to 60% [21,22]. Lindström et al. showed an altered ER-status in one-third of patients during tumor progression from PT to DM, but CTCs were not assessed [23]. We identified a 16% (19/122) altered ER-status in PT vs. DM in the present study and four of them changed from ER-negative in PT to ER-positive in DM. Two of them had first-line endocrine therapy and had shorter time to progression than the median PFS in the study cohort. Unfortunately, the ER-status of the corresponding baseline CTCs was not available for these patients and consequently no conclusion on endocrine therapy resistance based on CTC data could be drawn. Furthermore, Lindström et al. stated that ER-positive recurrence had better prognosis than ER-negative recurrence, irrespective of the ER-status of the PT [23]. In our study, as treatment continued, the number of CTCs decreased, and this was confirmed by a decrease in the number of CTCs between different timepoints. However, we could not determine whether this decrease was due to technical issues with the CTC-DropMount method or a true biological difference. Fewer CTCs were found through immunofluorescence using the CTC-DropMount method than using the CellSearch® method. This might have been because cells were attached to the walls of the CellSearch® cartridge or were detached from the slides during the immunofluorescent-staining process. With fewer CTCs detected, the possibility of finding CTCs positive for ER consequently decreases. However, direct use of the CellSearch® system for the evaluation of ER-status, as Paoletti et al. presented [24], would demand larger amounts of blood at a higher cost, which was not possible in this study. Other studies acknowledged technical difficulties concerning CTC characterization using fluorescence in situ hybridization [25]. With few cells to evaluate, this had the largest impact on the evaluation of ER-negative cases. The question was whether cases were truly negative or classified as such only because ER-positive cells were not detected. According to the applied criteria for the definition of the ER-status of CTCs, only a few positive cells are needed for a sample to be considered ER-positive. Since CTCs disseminate from tumors, ER-negative cells could preferentially detach compared to ER-positive cells. This might explain the finding of a lower number of ER-positive CTCs than ER-positive PT. In line with our findings, Babayan et al. found 19% (9/16) ER-negative CTCs in patients who originally had ER-positive primary tumors [26]. As depicted in Figure 2, the number of detected cells ranged from 1 to 47 for samples classified as ER-positive. Regarding samples classified as ER-negative, on the other hand, the range was wider (1 to 1094). Ultimately, in cases where only one CTC was found, it is difficult to say whether this was or was not a true-negative CTC sample. The strengths of our study include a prospective study design, with a homogeneous cohort (all patients were included before they started the first-line of systemic therapy), with predefined timepoints for sampling; consequently, the ER-status of CTCs was measured at longitudinal timepoints after the initiation of therapy. Furthermore, information on the ER-status of both PT and DM was available for most cases, enabling a thorough comparison throughout the tumor evolution. Weaknesses are mainly attributed to applying an in-house technique for capturing CTCs from the cartridge, which could potentially cause a lower cell recovery rate. Consequently, we had few cases to compare between the different timepoints, limiting the statistical power. Furthermore, there are few similar publications in the field with which to compare our results. 4. Material and Methods: The study was previously described in detail [27,28]. In brief, 168 women with newly diagnosed MBC, previously untreated for metastatic disease, were included in a prospective observational study at Skåne University Hospital and Halmstad County Hospital (ClinicalTrials.gov NCT01322893). Ethical permission was granted by the Lund University Ethics Committee (LU 2010/135) and all patients signed informed consent forms. Briefly, inclusion criteria were MBC diagnosis, age ≥18 years, Eastern Cooperative Oncology Group (ECOG) performance status score 0–2, and predicted life expectancy of >2 months. Study protocol included blood sampling at baseline (BL) and during therapy at 1, 3, and 6 months. In the current study, DropMount slides from BL (n = 113) and after 1 (n = 60) and 3 months (n = 46) were used. In total, 63% (92/147) of patients received adjuvant endocrine treatment, and 48% (70/147) received adjuvant chemotherapy. The primary endpoint was progression-free survival (PFS). Patient and tumor characteristics (e.g., ER-status) were retrieved from clinical records and pathology reports. ER-status was routinely assessed by immunohistochemistry on formalin paraffin-embedded material (with commercially available antibodies) on surgical specimen from the primary tumors at the time of primary diagnosis and on diagnostic core biopsies from metastasis at the time of diagnosis of metastatic disease. All assessments were performed by board-certified pathologists and ≥10% were classified as ER-positive according to the European guidelines [2,29]. After CTC enumeration using CellSearch®, cells were fixated using the CTC-DropMount method as previously described [30]. In brief, isolated CTCs were separated using a magnetic stand, and all nonmagnetic fluid was removed. The remaining cells were resuspended in phosphate-buffered saline (PBS), dropped onto Superfrost™ slides (ThermoScientific™, Germany), dried at 37 °C for 30 min, and fixated in pure methanol for 5 min. The slides were stored at –20 °C until analysis (i.e., CTC-DropMount method). With at least five CTCs enumerated using CellSearch®, 113 cases were available as DropMount at BL. Using a previously described immunofluorescence method [30], cells were evaluated in the present study for ER expression at BL and after 1 and 3 months. A rabbit monoclonal anti-ERα antibody was used as a primary (Thermo Fisher Scientific, #9101S1210D) and an AlexaFluor488-labeled goat antirabbit as a secondary antibody (Life Technologies, #1423009). For CD45 staining, a rat monoclonal antibody was used as a primary (Thermo Fisher Scientific/Invitrogen, #MA5-17687) and an AlexaFluor647-labeled goat antirat as a secondary antibody (Thermo Fisher Scientific/Invitrogen, #A-21247). Cytokeratins 8, 9, and 19 were stained with phycoerythrin-labeled mouse monoclonal antibody provided as leftovers from the CellSearch®-kit (Menarini, #E491A). Blood spiked with cells from the MCF-7 cell line was used as a positive control for ER expression. Cells positive for 4’,6-diamidino-2-phenylindole (DAPI) and cytokeratins 8, 18, and 19 but negative for CD45 were considered to be CTCs. Samples were classified as ER-positive if at least one CTC was positive for ER. Assessment of the ER-status of CTCs was independently performed by EM and CF. The cells were scanned using a BX63 Upright Microscope (Olympus Corporation, LRI, Lund, Sweden) and CellSense Dimension software. The filters used were DAPI, Cy2/fluorescein isothiocyanate (FITC), Cy3/tetramethylrhodamine-5-isothiocyante (TRITC), and Cy5. Scans were performed at 20×, while objects of special interest were captured at 40×. A representative example is shown in Figure 5. Statistical Analysis The study followed the Reporting Recommendation for Tumor Marker (REMARK) criteria [31,32]. Kaplan–Meier plots and the log-rank test were used to compare PFS between patient groups. Survival analysis of ER expression measured at 1 and 3 months was performed with landmark analysis. Hazard ratios (HRs) were quantified by the univariable Cox proportional hazards-regression model, and proportional hazard assumptions were graphically checked. Comparison of the ER-status of PT, DM, and CTCs at different timepoints was performed using the McNemar test. Stata version 16.1 (StataCorp, College Station, TX, USA) was used for statistical analyses. The study followed the Reporting Recommendation for Tumor Marker (REMARK) criteria [31,32]. Kaplan–Meier plots and the log-rank test were used to compare PFS between patient groups. Survival analysis of ER expression measured at 1 and 3 months was performed with landmark analysis. Hazard ratios (HRs) were quantified by the univariable Cox proportional hazards-regression model, and proportional hazard assumptions were graphically checked. Comparison of the ER-status of PT, DM, and CTCs at different timepoints was performed using the McNemar test. Stata version 16.1 (StataCorp, College Station, TX, USA) was used for statistical analyses. Statistical Analysis: The study followed the Reporting Recommendation for Tumor Marker (REMARK) criteria [31,32]. Kaplan–Meier plots and the log-rank test were used to compare PFS between patient groups. Survival analysis of ER expression measured at 1 and 3 months was performed with landmark analysis. Hazard ratios (HRs) were quantified by the univariable Cox proportional hazards-regression model, and proportional hazard assumptions were graphically checked. Comparison of the ER-status of PT, DM, and CTCs at different timepoints was performed using the McNemar test. Stata version 16.1 (StataCorp, College Station, TX, USA) was used for statistical analyses. 5. Conclusions: We demonstrated evidence of a shift from ER positivity to negativity in the PT compared to DM, as well as evidence of a similar shift in the ER-status of CTCs from PT and DM. Using the CTC-DropMount method, the ER positivity of CTCs at 1 and 3 months was related to better prognosis. The ER positivity of CTCs after the initiation of systemic therapy might reflect the retention of a favorable phenotype that still responds to therapy.

Background: The estrogen receptor (ER) can change expression between primary tumor (PT) and distant metastasis (DM) in breast cancer. A tissue biopsy reflects a momentary state at one location, whereas circulating tumor cells (CTCs) reflect real-time tumor progression. We evaluated ER-status during tumor progression from PT to DM and CTCs, and related the ER-status of CTCs to prognosis. Methods: In a study of metastatic breast cancer, blood was collected at different timepoints. After CellSearch® enrichment, CTCs were captured on DropMount slides and evaluated for ER expression at baseline (BL) and after 1 and 3 months of therapy. Comparison of the ER-status of PT, DM, and CTCs at different timepoints was performed using the McNemar test. The primary endpoint was progression-free survival (PFS). Results: Evidence of a shift from ER positivity to negativity between PT and DM was demonstrated (p = 0.019). We found strong evidence of similar shifts from PT to CTCs at different timepoints (p < 0.0001). ER-positive CTCs at 1 and 3 months were related to better prognosis. Conclusions: A shift in ER-status from PT to DM/CTCs was demonstrated. ER-positive CTCs during systemic therapy might reflect the retention of a favorable phenotype that still responds to therapy.

1. Introduction: Breast cancer is the most common malignant disease in women, and although five-year survival is approaching 90%, 20–30% of women with an initially local disease develop metastatic disease within 10 years [1]. The median overall survival (OS) in women with metastatic breast cancer (MBC) is approximately two years, and only 25% survive beyond five years [2]. Evaluating the response of therapy in MBC can be difficult, especially in patients with non-measurable disease. Guidelines for therapy evaluation include diagnostic imaging, clinical assessment, and also evaluation of liquid-based tumor markers [2,3,4]. In breast cancer, the estrogen receptor (ER) and the human epidermal growth factor receptor 2 (HER2), which are predictive of the response to endocrine- and HER2-targeted therapy, respectively, are the most important treatment predictive markers. Since it was shown that the expression of ER and HER2 frequently changes between primary tumor (PT) and distant metastasis (DM) in breast cancer [5,6], it is important to re-evaluate the biomarker status in biopsies from metastatic lesions. Furthermore, there are indications that cancer cells lose their ER expression during endocrine treatment [7,8] as a mechanism of resistance. The choice of therapy in a metastatic setting should therefore preferably be based on the receptor status of the metastatic biopsy when available. Due to heterogeneity, biopsies from different sites should be evaluated. In clinical practice, however, it is not always feasible to biopsy multiple metastatic sites, and the metastasis might also be situated at a non-accessible site. Moreover, current diagnostic methods, i.e., needle biopsies, limit the amount of material available for further analyses. Liquid biopsies evaluating the presence of circulating tumor cells (CTCs) are non-invasive and easily accessible with a limited risk of complications in comparison to an invasive tissue biopsy. They can be repeated regularly, which allows for the serial monitoring of real-time tumor evolution/response to treatment. Additionally, a tissue biopsy reflects a momentary state of the tumor at one location in the body, whereas liquid-based markers reflect real-time tumor progression and are not dependent on repeated invasive tissue biopsies. CTCs can be detected in a liquid biopsy, and the CellSearch® system (©Menarini Silicon Biosystems, Inc 2020, USA) is considered the golden standard of CTC capturing and enumeration in MBC since it is the only Food and Drug Administration (FDA)-approved technology, so far. Multiple trials have shown CellSearch® enumerated CTCs as an independent prognostic factor in MBC [9,10,11,12,13,14,15]. CTCs have also been proposed to more accurately reflect tumor heterogeneity than biopsies of single metastases and are considered to originate from both the primary tumor and all metastatic sites [16]. Therefore, liquid biopsies enumerating CTCs might be a better proxy for the total metastatic burden. However, CTC enumeration alone is insufficient to predict the efficacy of therapy, and the addition of CTC tumor-marker expression might better predict the benefit of a therapy [17]. The primary aim of the present study was to characterize CTCs detected in blood samples from MBC patients and to evaluate ER-status during tumor progression from primary tumors to distant metastasis and CTCs, respectively. Our secondary aim was to relate the ER-status of CTCs to prognosis. We hypothesized that, by evaluating the ER-status of single CTCs, improved therapy guidance could be provided. 5. Conclusions: We demonstrated evidence of a shift from ER positivity to negativity in the PT compared to DM, as well as evidence of a similar shift in the ER-status of CTCs from PT and DM. Using the CTC-DropMount method, the ER positivity of CTCs at 1 and 3 months was related to better prognosis. The ER positivity of CTCs after the initiation of systemic therapy might reflect the retention of a favorable phenotype that still responds to therapy.

Background: The estrogen receptor (ER) can change expression between primary tumor (PT) and distant metastasis (DM) in breast cancer. A tissue biopsy reflects a momentary state at one location, whereas circulating tumor cells (CTCs) reflect real-time tumor progression. We evaluated ER-status during tumor progression from PT to DM and CTCs, and related the ER-status of CTCs to prognosis. Methods: In a study of metastatic breast cancer, blood was collected at different timepoints. After CellSearch® enrichment, CTCs were captured on DropMount slides and evaluated for ER expression at baseline (BL) and after 1 and 3 months of therapy. Comparison of the ER-status of PT, DM, and CTCs at different timepoints was performed using the McNemar test. The primary endpoint was progression-free survival (PFS). Results: Evidence of a shift from ER positivity to negativity between PT and DM was demonstrated (p = 0.019). We found strong evidence of similar shifts from PT to CTCs at different timepoints (p < 0.0001). ER-positive CTCs at 1 and 3 months were related to better prognosis. Conclusions: A shift in ER-status from PT to DM/CTCs was demonstrated. ER-positive CTCs during systemic therapy might reflect the retention of a favorable phenotype that still responds to therapy.

[CONTENT] circulating tumor cells | breast cancer | estrogen receptor: tumor progression | metastasis [SUMMARY]

[CONTENT] circulating tumor cells | breast cancer | estrogen receptor: tumor progression | metastasis [SUMMARY]

[CONTENT] circulating tumor cells | breast cancer | estrogen receptor: tumor progression | metastasis [SUMMARY]

[CONTENT] circulating tumor cells | breast cancer | estrogen receptor: tumor progression | metastasis [SUMMARY]

[CONTENT] circulating tumor cells | breast cancer | estrogen receptor: tumor progression | metastasis [SUMMARY]

[CONTENT] Adult | Aged | Aged, 80 and over | Biomarkers, Tumor | Breast Neoplasms | Disease Progression | Female | Follow-Up Studies | Humans | Middle Aged | Neoplasm Metastasis | Neoplasm Staging | Neoplasms | Neoplastic Cells, Circulating | Prognosis | Receptors, Estrogen | Survival Analysis [SUMMARY]

[CONTENT] Adult | Aged | Aged, 80 and over | Biomarkers, Tumor | Breast Neoplasms | Disease Progression | Female | Follow-Up Studies | Humans | Middle Aged | Neoplasm Metastasis | Neoplasm Staging | Neoplasms | Neoplastic Cells, Circulating | Prognosis | Receptors, Estrogen | Survival Analysis [SUMMARY]

[CONTENT] Adult | Aged | Aged, 80 and over | Biomarkers, Tumor | Breast Neoplasms | Disease Progression | Female | Follow-Up Studies | Humans | Middle Aged | Neoplasm Metastasis | Neoplasm Staging | Neoplasms | Neoplastic Cells, Circulating | Prognosis | Receptors, Estrogen | Survival Analysis [SUMMARY]

[CONTENT] Adult | Aged | Aged, 80 and over | Biomarkers, Tumor | Breast Neoplasms | Disease Progression | Female | Follow-Up Studies | Humans | Middle Aged | Neoplasm Metastasis | Neoplasm Staging | Neoplasms | Neoplastic Cells, Circulating | Prognosis | Receptors, Estrogen | Survival Analysis [SUMMARY]

[CONTENT] Adult | Aged | Aged, 80 and over | Biomarkers, Tumor | Breast Neoplasms | Disease Progression | Female | Follow-Up Studies | Humans | Middle Aged | Neoplasm Metastasis | Neoplasm Staging | Neoplasms | Neoplastic Cells, Circulating | Prognosis | Receptors, Estrogen | Survival Analysis [SUMMARY]

[CONTENT] women metastatic breast | metastasis dm breast | cancer estrogen receptor | progression breast cancer | markers breast cancer [SUMMARY]

[CONTENT] women metastatic breast | metastasis dm breast | cancer estrogen receptor | progression breast cancer | markers breast cancer [SUMMARY]

[CONTENT] women metastatic breast | metastasis dm breast | cancer estrogen receptor | progression breast cancer | markers breast cancer [SUMMARY]

[CONTENT] women metastatic breast | metastasis dm breast | cancer estrogen receptor | progression breast cancer | markers breast cancer [SUMMARY]

[CONTENT] women metastatic breast | metastasis dm breast | cancer estrogen receptor | progression breast cancer | markers breast cancer [SUMMARY]

[CONTENT] er | ctcs | status | positive | er status | bl | er positive | figure | ctc | cases [SUMMARY]

[CONTENT] er | ctcs | status | positive | er status | bl | er positive | figure | ctc | cases [SUMMARY]

[CONTENT] er | ctcs | status | positive | er status | bl | er positive | figure | ctc | cases [SUMMARY]

[CONTENT] er | ctcs | status | positive | er status | bl | er positive | figure | ctc | cases [SUMMARY]

[CONTENT] er | ctcs | status | positive | er status | bl | er positive | figure | ctc | cases [SUMMARY]

[CONTENT] tumor | metastatic | biopsies | liquid | cancer | biopsy | mbc | therapy | breast cancer | breast [SUMMARY]

[CONTENT] er | bl | positive | figure | ctcs | table | hr | 95 | ci | er positive [SUMMARY]

[CONTENT] er positivity | positivity | er | evidence | positivity ctcs | shift | shift er | er positivity ctcs | ctcs | compared dm [SUMMARY]

[CONTENT] er | ctcs | positive | bl | er positive | status | er status | cases | figure | dm [SUMMARY]

[CONTENT] er | ctcs | positive | bl | er positive | status | er status | cases | figure | dm [SUMMARY]

[CONTENT] estrogen | ER ||| one ||| ER | PT | ER [SUMMARY]

[CONTENT] ER | PT | DM | 0.019 ||| ||| ER | 1 and 3 months [SUMMARY]

[CONTENT] ER | PT ||| ER [SUMMARY]

[CONTENT] ER ||| one ||| ER | PT | ER ||| ||| CellSearch | DropMount | ER | 1 and 3 months ||| ER | PT | DM | McNemar ||| ||| ||| ER | PT | DM | 0.019 ||| ||| ER | 1 and 3 months ||| ER | PT ||| ER [SUMMARY]

[CONTENT] ER ||| one ||| ER | PT | ER ||| ||| CellSearch | DropMount | ER | 1 and 3 months ||| ER | PT | DM | McNemar ||| ||| ||| ER | PT | DM | 0.019 ||| ||| ER | 1 and 3 months ||| ER | PT ||| ER [SUMMARY]

Prognosis of ampullary cancer based on immunohistochemical type and expression of osteopontin.

21992455

Ampullary cancer (AC) was classified as pancreatobiliary, intestinal, or other subtype based on the expression of cytokeratin 7 (CK7) and cytokeratin 20 (CK20). We aimed to explore the association of AC subtype with patient prognosis.

BACKGROUND

The relationship of AC subtype and expression of Osteopontin (OPN) with the prognosis of 120 AC patients after pancreaticoduodenectomy was investigated.

METHODS

The patients had pancreatobiliary (CK7+/CK20-, n = 24, 20%), intestinal (CK7-/CK20+, n = 29, 24.2%) or other (CK7+/CK20+ or CK7-/CK20-, n = 67, 55.8%) subtypes of AC, and their median survival times were 23 ± 4.2, 38 ± 2.8 and 64 ± 16.8 months, respectively. The survival times of 64 OPN- patients (53.3%) and 56 OPN+ patients (46.7%) were 69 ± 18.4 and 36 ± 1.3 months, respectively. There was no significant effect of AC subtype on survival of OPN- patients. For OPN+ patients, those with pancreatobiliary AC had a shorter survival time (22 ± 6.6 months) than those with intestinal AC (37 ± 1.4 months, p = 0.041), and other AC subtype (36 ± 0.9 months, p = 0.010); intestinal and other AC subtypes had similar survival times.

RESULTS

The prognosis of AC patients can be estimated based on immunohistochemical classification and OPN status.

CONCLUSIONS

3213044

Background

Ampullary carcinoma (AC) is a relatively rare tumor of the hepatopancreatic ampulla that accounts for approximately 0.2% of gastrointestinal tract malignancies and 7% of periampullary carcinomas [1]. ACs have different anatomical origins. Kimura et al. initially classified AC as pancreatobiliary AC if it had papillary projections with scant fibrous cores and as intestinal AC if it resembled tubular adenocarcinoma of the stomach or colon [2]. Numerous studies have reported that intestinal AC is associated with a better prognosis than pancreatobiliary AC [1-5]. AC has also been classified based on immunohistochemical expression of cytokeratin 7 (CK7), Mucins and CDX2 [4,6-8] and HNF4α [9]. However, the clinical significance and survival rates of AC patients with these different immunohistochemical subtypes have not been definitely established. Histologic classification and immunohistochemical characterization by cytokeratins are in good agreement [5]. Fischer et al. reported that the histological subtypes of AC could be determined by the expression of CK7, CK20, and MUC2; pancreaticobiliary AC is CK7+/CK20-/MUC2-, and intestinal AC is CK7-/CK20+/MUC2+ [10]. Zhou et al. classified CK7-/CK20+ tumors as intestinal AC, CK7+/CK20- tumors as pancreatobiliary AC, and tumors that are CK7+/CK20+ or CK7-/CK20- as "other" [3]. However, there was no statistical difference in survival of patients with different CK7/CK20 subtypes [3] or with different CK20/MUC subtypes [11]. Osteopontin (OPN) is a secretory calcium-binding phosphorylated glycoprotein and plays an important role in bone metabolism. OPN is widely distributed in the urine, blood, gastrointestinal tract, pancreas, lungs and elsewhere. At the molecular level, OPN plays important roles in cellular adhesion and migration, tissue repair, and signal transduction and also in the invasion and metastasis of several cancers [12]. OPN is significantly associated with survival rate in several cancers and has value as a marker of clinical tumor progression [13,14]. In particular, low OPN levels were significantly associated with a favorable prognosis in patients with advanced non-small cell lung cancer [15], laryngeal and hypopharyngeal carcinomas [16], hepatocellular carcinoma [17], colorectal cancer [18], idiopathic pulmonary hypertension [19], upper urinary tract urothelial carcinoma [20], acute myeloid leukemia [21], oral squamous cell carcinoma [22], and endometrial cancer [23]. OPN may also be a suitable biomarker for overall survival and renal outcome of patients who are critically ill with acute kidney injury [24]. However, few studies have investigated the expression of OPN in patients with AC. Van Heek et al. reported higher OPN expression in the sera and tumors of AC patients than in the sera and duodenal samples of healthy controls [25]. Bloomston et al. reported that node-negative status and lack of OPN expression were associated with prolonged survival in patients with AC [26]. Hsu et al. reported that expression of OPN and the presence of tumor-associated macrophages in bulky AC were associated with tumor recurrence, and poorer disease-specific survival [27]. In the present study, we retrospectively analyzed the clinical data of 120 patients who were undergoing pancreaticoduodenectomy due to AC. We focused on the association of AC prognosis with the expression of CK7, CK20, and OPN.

Patients and Methods

Patients From January 1, 1994 to December 30, 2008, patients undergoing pancreaticoduodenectomy due to AC were recruited from the Department of Hepatobiliary Surgery of the General Hospital of the Peoples Liberation Army (Beijing, China). The exclusion criteria were: (i) duodenal cancer, cancer of the lower bile duct, or cancer of the pancreas or any of these cancers involving the ampulla or duodenal papilla, based on pathological examination; (ii) uncertain origin of the cancer; (iii) previous focal resection of duodenal papillary cancer or AC; (iv) metastasis to other organs; and (v) presence of concomitant heart disease, cerebrovascular disease, or pulmonary disease that made the patient ineligible for surgery. Follow-up examinations were performed at 3 months after surgery, once every 6 months for 3 years, and then once per year. These follow-up examinations included routine tests (liver and kidney function, blood electrolytes, routine blood test), tests for tumor markers, chest X-ray, and abdominal imaging by ultrasonography, CT, or MRI. The last follow-up was on January 31, 2010. Tumor stage and lymph node metastasis were evaluated according to Greene et al [28]. This study was approved by the hospital Institutional Review Board. From January 1, 1994 to December 30, 2008, patients undergoing pancreaticoduodenectomy due to AC were recruited from the Department of Hepatobiliary Surgery of the General Hospital of the Peoples Liberation Army (Beijing, China). The exclusion criteria were: (i) duodenal cancer, cancer of the lower bile duct, or cancer of the pancreas or any of these cancers involving the ampulla or duodenal papilla, based on pathological examination; (ii) uncertain origin of the cancer; (iii) previous focal resection of duodenal papillary cancer or AC; (iv) metastasis to other organs; and (v) presence of concomitant heart disease, cerebrovascular disease, or pulmonary disease that made the patient ineligible for surgery. Follow-up examinations were performed at 3 months after surgery, once every 6 months for 3 years, and then once per year. These follow-up examinations included routine tests (liver and kidney function, blood electrolytes, routine blood test), tests for tumor markers, chest X-ray, and abdominal imaging by ultrasonography, CT, or MRI. The last follow-up was on January 31, 2010. Tumor stage and lymph node metastasis were evaluated according to Greene et al [28]. This study was approved by the hospital Institutional Review Board. Immunohistochemistry Carcinoma specimens were embedded in paraffin, cut consecutively into sections (4 μm), and the streptavidin-peroxidase method was used for immunohistochemical visualization (UltraSensitive™ SP kit, Maximbio. Co. Ltd, Fuzhou, China). The primary antibodies were mouse anti-human CK7 or CK20 monoclonal antibodies and rabbit anti-human OPN polyclonal antibody (Lab Vision & NeoMarkers, USA). The normal serum from non-immunized goat was used as a negative control of the primary antibody, and CK7+/CK20+/OPN+ pancreatic carcinoma was used as a positive control. Details for the determination of positive staining were previously provided[3]. In brief, cells positive for CK7, CK20, or OPN had brown or yellow-brown granules, mainly in the cytoplasm. Sections were evaluated by two independent and blinded pathologists. No staining or staining in fewer than 10% of cells was considered negative, and staining of 10% or more of cells was considered positive. Carcinoma specimens were embedded in paraffin, cut consecutively into sections (4 μm), and the streptavidin-peroxidase method was used for immunohistochemical visualization (UltraSensitive™ SP kit, Maximbio. Co. Ltd, Fuzhou, China). The primary antibodies were mouse anti-human CK7 or CK20 monoclonal antibodies and rabbit anti-human OPN polyclonal antibody (Lab Vision & NeoMarkers, USA). The normal serum from non-immunized goat was used as a negative control of the primary antibody, and CK7+/CK20+/OPN+ pancreatic carcinoma was used as a positive control. Details for the determination of positive staining were previously provided[3]. In brief, cells positive for CK7, CK20, or OPN had brown or yellow-brown granules, mainly in the cytoplasm. Sections were evaluated by two independent and blinded pathologists. No staining or staining in fewer than 10% of cells was considered negative, and staining of 10% or more of cells was considered positive. Statistical analysis Results are expressed as means or medians with standard deviations, or counts and percentages. Survival analysis was analyzed by the Kaplan-Meier method and the log-rank test. Data were analyzed using SPSS 15.0 (SPSS, Inc., Chicago, IL, USA). All p-values were two-sided and were considered significant if p was less than 0.05. Results are expressed as means or medians with standard deviations, or counts and percentages. Survival analysis was analyzed by the Kaplan-Meier method and the log-rank test. Data were analyzed using SPSS 15.0 (SPSS, Inc., Chicago, IL, USA). All p-values were two-sided and were considered significant if p was less than 0.05.

Conclusions

XQZ carried out the study design, defined the intellectual content, participated in the literature research and manuscript preparation, analyzed data, and edited the manuscript. JHD do guarantor of integrity of the entire study, carried out the study concepts, and reviewed the manuscript. WZZ carried out the clinical studies, and acquired data. ZL carried out the experimental studies, and did statistical analysis. All authors read and approved the final manuscript.

Background: Ampullary carcinoma (AC) is a relatively rare tumor of the hepatopancreatic ampulla that accounts for approximately 0.2% of gastrointestinal tract malignancies and 7% of periampullary carcinomas [1]. ACs have different anatomical origins. Kimura et al. initially classified AC as pancreatobiliary AC if it had papillary projections with scant fibrous cores and as intestinal AC if it resembled tubular adenocarcinoma of the stomach or colon [2]. Numerous studies have reported that intestinal AC is associated with a better prognosis than pancreatobiliary AC [1-5]. AC has also been classified based on immunohistochemical expression of cytokeratin 7 (CK7), Mucins and CDX2 [4,6-8] and HNF4α [9]. However, the clinical significance and survival rates of AC patients with these different immunohistochemical subtypes have not been definitely established. Histologic classification and immunohistochemical characterization by cytokeratins are in good agreement [5]. Fischer et al. reported that the histological subtypes of AC could be determined by the expression of CK7, CK20, and MUC2; pancreaticobiliary AC is CK7+/CK20-/MUC2-, and intestinal AC is CK7-/CK20+/MUC2+ [10]. Zhou et al. classified CK7-/CK20+ tumors as intestinal AC, CK7+/CK20- tumors as pancreatobiliary AC, and tumors that are CK7+/CK20+ or CK7-/CK20- as "other" [3]. However, there was no statistical difference in survival of patients with different CK7/CK20 subtypes [3] or with different CK20/MUC subtypes [11]. Osteopontin (OPN) is a secretory calcium-binding phosphorylated glycoprotein and plays an important role in bone metabolism. OPN is widely distributed in the urine, blood, gastrointestinal tract, pancreas, lungs and elsewhere. At the molecular level, OPN plays important roles in cellular adhesion and migration, tissue repair, and signal transduction and also in the invasion and metastasis of several cancers [12]. OPN is significantly associated with survival rate in several cancers and has value as a marker of clinical tumor progression [13,14]. In particular, low OPN levels were significantly associated with a favorable prognosis in patients with advanced non-small cell lung cancer [15], laryngeal and hypopharyngeal carcinomas [16], hepatocellular carcinoma [17], colorectal cancer [18], idiopathic pulmonary hypertension [19], upper urinary tract urothelial carcinoma [20], acute myeloid leukemia [21], oral squamous cell carcinoma [22], and endometrial cancer [23]. OPN may also be a suitable biomarker for overall survival and renal outcome of patients who are critically ill with acute kidney injury [24]. However, few studies have investigated the expression of OPN in patients with AC. Van Heek et al. reported higher OPN expression in the sera and tumors of AC patients than in the sera and duodenal samples of healthy controls [25]. Bloomston et al. reported that node-negative status and lack of OPN expression were associated with prolonged survival in patients with AC [26]. Hsu et al. reported that expression of OPN and the presence of tumor-associated macrophages in bulky AC were associated with tumor recurrence, and poorer disease-specific survival [27]. In the present study, we retrospectively analyzed the clinical data of 120 patients who were undergoing pancreaticoduodenectomy due to AC. We focused on the association of AC prognosis with the expression of CK7, CK20, and OPN. Patients and Methods: Patients From January 1, 1994 to December 30, 2008, patients undergoing pancreaticoduodenectomy due to AC were recruited from the Department of Hepatobiliary Surgery of the General Hospital of the Peoples Liberation Army (Beijing, China). The exclusion criteria were: (i) duodenal cancer, cancer of the lower bile duct, or cancer of the pancreas or any of these cancers involving the ampulla or duodenal papilla, based on pathological examination; (ii) uncertain origin of the cancer; (iii) previous focal resection of duodenal papillary cancer or AC; (iv) metastasis to other organs; and (v) presence of concomitant heart disease, cerebrovascular disease, or pulmonary disease that made the patient ineligible for surgery. Follow-up examinations were performed at 3 months after surgery, once every 6 months for 3 years, and then once per year. These follow-up examinations included routine tests (liver and kidney function, blood electrolytes, routine blood test), tests for tumor markers, chest X-ray, and abdominal imaging by ultrasonography, CT, or MRI. The last follow-up was on January 31, 2010. Tumor stage and lymph node metastasis were evaluated according to Greene et al [28]. This study was approved by the hospital Institutional Review Board. From January 1, 1994 to December 30, 2008, patients undergoing pancreaticoduodenectomy due to AC were recruited from the Department of Hepatobiliary Surgery of the General Hospital of the Peoples Liberation Army (Beijing, China). The exclusion criteria were: (i) duodenal cancer, cancer of the lower bile duct, or cancer of the pancreas or any of these cancers involving the ampulla or duodenal papilla, based on pathological examination; (ii) uncertain origin of the cancer; (iii) previous focal resection of duodenal papillary cancer or AC; (iv) metastasis to other organs; and (v) presence of concomitant heart disease, cerebrovascular disease, or pulmonary disease that made the patient ineligible for surgery. Follow-up examinations were performed at 3 months after surgery, once every 6 months for 3 years, and then once per year. These follow-up examinations included routine tests (liver and kidney function, blood electrolytes, routine blood test), tests for tumor markers, chest X-ray, and abdominal imaging by ultrasonography, CT, or MRI. The last follow-up was on January 31, 2010. Tumor stage and lymph node metastasis were evaluated according to Greene et al [28]. This study was approved by the hospital Institutional Review Board. Immunohistochemistry Carcinoma specimens were embedded in paraffin, cut consecutively into sections (4 μm), and the streptavidin-peroxidase method was used for immunohistochemical visualization (UltraSensitive™ SP kit, Maximbio. Co. Ltd, Fuzhou, China). The primary antibodies were mouse anti-human CK7 or CK20 monoclonal antibodies and rabbit anti-human OPN polyclonal antibody (Lab Vision & NeoMarkers, USA). The normal serum from non-immunized goat was used as a negative control of the primary antibody, and CK7+/CK20+/OPN+ pancreatic carcinoma was used as a positive control. Details for the determination of positive staining were previously provided[3]. In brief, cells positive for CK7, CK20, or OPN had brown or yellow-brown granules, mainly in the cytoplasm. Sections were evaluated by two independent and blinded pathologists. No staining or staining in fewer than 10% of cells was considered negative, and staining of 10% or more of cells was considered positive. Carcinoma specimens were embedded in paraffin, cut consecutively into sections (4 μm), and the streptavidin-peroxidase method was used for immunohistochemical visualization (UltraSensitive™ SP kit, Maximbio. Co. Ltd, Fuzhou, China). The primary antibodies were mouse anti-human CK7 or CK20 monoclonal antibodies and rabbit anti-human OPN polyclonal antibody (Lab Vision & NeoMarkers, USA). The normal serum from non-immunized goat was used as a negative control of the primary antibody, and CK7+/CK20+/OPN+ pancreatic carcinoma was used as a positive control. Details for the determination of positive staining were previously provided[3]. In brief, cells positive for CK7, CK20, or OPN had brown or yellow-brown granules, mainly in the cytoplasm. Sections were evaluated by two independent and blinded pathologists. No staining or staining in fewer than 10% of cells was considered negative, and staining of 10% or more of cells was considered positive. Statistical analysis Results are expressed as means or medians with standard deviations, or counts and percentages. Survival analysis was analyzed by the Kaplan-Meier method and the log-rank test. Data were analyzed using SPSS 15.0 (SPSS, Inc., Chicago, IL, USA). All p-values were two-sided and were considered significant if p was less than 0.05. Results are expressed as means or medians with standard deviations, or counts and percentages. Survival analysis was analyzed by the Kaplan-Meier method and the log-rank test. Data were analyzed using SPSS 15.0 (SPSS, Inc., Chicago, IL, USA). All p-values were two-sided and were considered significant if p was less than 0.05. Patients: From January 1, 1994 to December 30, 2008, patients undergoing pancreaticoduodenectomy due to AC were recruited from the Department of Hepatobiliary Surgery of the General Hospital of the Peoples Liberation Army (Beijing, China). The exclusion criteria were: (i) duodenal cancer, cancer of the lower bile duct, or cancer of the pancreas or any of these cancers involving the ampulla or duodenal papilla, based on pathological examination; (ii) uncertain origin of the cancer; (iii) previous focal resection of duodenal papillary cancer or AC; (iv) metastasis to other organs; and (v) presence of concomitant heart disease, cerebrovascular disease, or pulmonary disease that made the patient ineligible for surgery. Follow-up examinations were performed at 3 months after surgery, once every 6 months for 3 years, and then once per year. These follow-up examinations included routine tests (liver and kidney function, blood electrolytes, routine blood test), tests for tumor markers, chest X-ray, and abdominal imaging by ultrasonography, CT, or MRI. The last follow-up was on January 31, 2010. Tumor stage and lymph node metastasis were evaluated according to Greene et al [28]. This study was approved by the hospital Institutional Review Board. Immunohistochemistry: Carcinoma specimens were embedded in paraffin, cut consecutively into sections (4 μm), and the streptavidin-peroxidase method was used for immunohistochemical visualization (UltraSensitive™ SP kit, Maximbio. Co. Ltd, Fuzhou, China). The primary antibodies were mouse anti-human CK7 or CK20 monoclonal antibodies and rabbit anti-human OPN polyclonal antibody (Lab Vision & NeoMarkers, USA). The normal serum from non-immunized goat was used as a negative control of the primary antibody, and CK7+/CK20+/OPN+ pancreatic carcinoma was used as a positive control. Details for the determination of positive staining were previously provided[3]. In brief, cells positive for CK7, CK20, or OPN had brown or yellow-brown granules, mainly in the cytoplasm. Sections were evaluated by two independent and blinded pathologists. No staining or staining in fewer than 10% of cells was considered negative, and staining of 10% or more of cells was considered positive. Statistical analysis: Results are expressed as means or medians with standard deviations, or counts and percentages. Survival analysis was analyzed by the Kaplan-Meier method and the log-rank test. Data were analyzed using SPSS 15.0 (SPSS, Inc., Chicago, IL, USA). All p-values were two-sided and were considered significant if p was less than 0.05. Results: Patient characteristics A total of 120 patients (84 males, 36 females) met our inclusion criteria and received follow-up examinations. The mean age was 55.1 ± 9.8 years and the mean tumor diameter was 2.4 ± 1.5 cm. The 1-, 3- and 5-year survival rates were 94.8%, 78.7%, and 68.0%, respectively. The median survival time was 38 ± 11.3 months and the mean survival time was 53.9 months. A total of 51 patients (42.5%) survived to the end of the follow-up period (January 31, 2010), 1 patient survived more than 10 years, and 2 patients survived more than 5 years. In addition to pancreaticoduodenectomy (given to all patients), 8 patients received chemotherapy. Among the 69 patients (57.5%) who died in the follow-up period, 21 died within 5 years after surgery and 1 died 11 years after surgery. A total of 120 patients (84 males, 36 females) met our inclusion criteria and received follow-up examinations. The mean age was 55.1 ± 9.8 years and the mean tumor diameter was 2.4 ± 1.5 cm. The 1-, 3- and 5-year survival rates were 94.8%, 78.7%, and 68.0%, respectively. The median survival time was 38 ± 11.3 months and the mean survival time was 53.9 months. A total of 51 patients (42.5%) survived to the end of the follow-up period (January 31, 2010), 1 patient survived more than 10 years, and 2 patients survived more than 5 years. In addition to pancreaticoduodenectomy (given to all patients), 8 patients received chemotherapy. Among the 69 patients (57.5%) who died in the follow-up period, 21 died within 5 years after surgery and 1 died 11 years after surgery. Survival of patients with different subtypes of AC Figure 1A-D shows representative immunohistochemical results of patients with intestinal AC (CK7-/CK20+), pancreatobiliary AC (CK7+/CK20-), and other subtype of AC (CK7+/CK20+ and CK7-/CK20-). Figure 1E shows representative positive and negative results for OPN staining. Representative immunohistochemical staining of ampullary cancer. A: intestinal type, CK7-/CK20+; B: pancreatobiliary type, CK7+/CK20-; C: other type, (C: CK7+/CK20+); D: other type, CK7-/CK20-); E: positive and negative immunohistochemical staining for osteopontin. Table 1 shows the survival times of AC patients stratified by immunohistochemical results, extent of tumor differentiation, amount of tumor invasion, and lymph node metastasis. None of the other associations were significant. OPN+ patients had marginally longer survival time than OPN- patients (p = 0.062; Figure 2A). The 1-, 3, and 5-year survival rates were 84.3%, 55.3%, and 0% for OPN+ patients and 91.8%, 68.4% and 7.2% for OPN- patients respectively. Characteristics of ampullary carcinoma and survival time of patients. Survival time is expressed as median ± SD and p-values were calculated by the log-rank test. Kaplan-Meier survival curves of patients with different immunohistochemical types of ampullary cancer and with positive or negative expression of osteopontin. *p < 0.05 by log-rank test. Figure 2B shows the survival times of patients with pancreatobiliary AC, intestinal AC, and other AC. Although pancreatobiliary AC was associated with slightly worse prognosis, a log-rank test indicated that this difference was only marginal (p = 0.052, Table 2). Figures 2C and 2D show the survival times of these three groups stratified by OPN expression. The results indicate that OPN- patients with different subtypes of AC had similar survival times (p = 0.75, Table 2). However, OPN+ patients with different subtypes of AC had significantly different survival times (p = 0.017, Table 2). Specifically, OPN+ patients with pancreatobiliary AC had worse prognosis than those with intestinal AC (p = 0.041) and other subtype of AC (p = 0.010); the survival time of the patients with intestinal and other types had no difference. In addition, the survival time of patients with OPN+ pancreatobiliary AC was shorter than those with OPN- intestinal AC (p = 0.043), and other subtype of AC (p = 0.002). Expression of OPN and survival time of patients with different subtypes of ampullary carcinoma (AC). Survival time is expressed as median ± SD. 1p-value for comparison of the 3 types; 2p-value for comparison of the OPN+ and OPN- groups. * p = 0.041 and p = 0.010 for comparison with the intestinal type and other type OPN+ AC, respectively. There was no difference between intestinal AC and other type of AC (p = 0.907). Figure 1A-D shows representative immunohistochemical results of patients with intestinal AC (CK7-/CK20+), pancreatobiliary AC (CK7+/CK20-), and other subtype of AC (CK7+/CK20+ and CK7-/CK20-). Figure 1E shows representative positive and negative results for OPN staining. Representative immunohistochemical staining of ampullary cancer. A: intestinal type, CK7-/CK20+; B: pancreatobiliary type, CK7+/CK20-; C: other type, (C: CK7+/CK20+); D: other type, CK7-/CK20-); E: positive and negative immunohistochemical staining for osteopontin. Table 1 shows the survival times of AC patients stratified by immunohistochemical results, extent of tumor differentiation, amount of tumor invasion, and lymph node metastasis. None of the other associations were significant. OPN+ patients had marginally longer survival time than OPN- patients (p = 0.062; Figure 2A). The 1-, 3, and 5-year survival rates were 84.3%, 55.3%, and 0% for OPN+ patients and 91.8%, 68.4% and 7.2% for OPN- patients respectively. Characteristics of ampullary carcinoma and survival time of patients. Survival time is expressed as median ± SD and p-values were calculated by the log-rank test. Kaplan-Meier survival curves of patients with different immunohistochemical types of ampullary cancer and with positive or negative expression of osteopontin. *p < 0.05 by log-rank test. Figure 2B shows the survival times of patients with pancreatobiliary AC, intestinal AC, and other AC. Although pancreatobiliary AC was associated with slightly worse prognosis, a log-rank test indicated that this difference was only marginal (p = 0.052, Table 2). Figures 2C and 2D show the survival times of these three groups stratified by OPN expression. The results indicate that OPN- patients with different subtypes of AC had similar survival times (p = 0.75, Table 2). However, OPN+ patients with different subtypes of AC had significantly different survival times (p = 0.017, Table 2). Specifically, OPN+ patients with pancreatobiliary AC had worse prognosis than those with intestinal AC (p = 0.041) and other subtype of AC (p = 0.010); the survival time of the patients with intestinal and other types had no difference. In addition, the survival time of patients with OPN+ pancreatobiliary AC was shorter than those with OPN- intestinal AC (p = 0.043), and other subtype of AC (p = 0.002). Expression of OPN and survival time of patients with different subtypes of ampullary carcinoma (AC). Survival time is expressed as median ± SD. 1p-value for comparison of the 3 types; 2p-value for comparison of the OPN+ and OPN- groups. * p = 0.041 and p = 0.010 for comparison with the intestinal type and other type OPN+ AC, respectively. There was no difference between intestinal AC and other type of AC (p = 0.907). Patient characteristics: A total of 120 patients (84 males, 36 females) met our inclusion criteria and received follow-up examinations. The mean age was 55.1 ± 9.8 years and the mean tumor diameter was 2.4 ± 1.5 cm. The 1-, 3- and 5-year survival rates were 94.8%, 78.7%, and 68.0%, respectively. The median survival time was 38 ± 11.3 months and the mean survival time was 53.9 months. A total of 51 patients (42.5%) survived to the end of the follow-up period (January 31, 2010), 1 patient survived more than 10 years, and 2 patients survived more than 5 years. In addition to pancreaticoduodenectomy (given to all patients), 8 patients received chemotherapy. Among the 69 patients (57.5%) who died in the follow-up period, 21 died within 5 years after surgery and 1 died 11 years after surgery. Survival of patients with different subtypes of AC: Figure 1A-D shows representative immunohistochemical results of patients with intestinal AC (CK7-/CK20+), pancreatobiliary AC (CK7+/CK20-), and other subtype of AC (CK7+/CK20+ and CK7-/CK20-). Figure 1E shows representative positive and negative results for OPN staining. Representative immunohistochemical staining of ampullary cancer. A: intestinal type, CK7-/CK20+; B: pancreatobiliary type, CK7+/CK20-; C: other type, (C: CK7+/CK20+); D: other type, CK7-/CK20-); E: positive and negative immunohistochemical staining for osteopontin. Table 1 shows the survival times of AC patients stratified by immunohistochemical results, extent of tumor differentiation, amount of tumor invasion, and lymph node metastasis. None of the other associations were significant. OPN+ patients had marginally longer survival time than OPN- patients (p = 0.062; Figure 2A). The 1-, 3, and 5-year survival rates were 84.3%, 55.3%, and 0% for OPN+ patients and 91.8%, 68.4% and 7.2% for OPN- patients respectively. Characteristics of ampullary carcinoma and survival time of patients. Survival time is expressed as median ± SD and p-values were calculated by the log-rank test. Kaplan-Meier survival curves of patients with different immunohistochemical types of ampullary cancer and with positive or negative expression of osteopontin. *p < 0.05 by log-rank test. Figure 2B shows the survival times of patients with pancreatobiliary AC, intestinal AC, and other AC. Although pancreatobiliary AC was associated with slightly worse prognosis, a log-rank test indicated that this difference was only marginal (p = 0.052, Table 2). Figures 2C and 2D show the survival times of these three groups stratified by OPN expression. The results indicate that OPN- patients with different subtypes of AC had similar survival times (p = 0.75, Table 2). However, OPN+ patients with different subtypes of AC had significantly different survival times (p = 0.017, Table 2). Specifically, OPN+ patients with pancreatobiliary AC had worse prognosis than those with intestinal AC (p = 0.041) and other subtype of AC (p = 0.010); the survival time of the patients with intestinal and other types had no difference. In addition, the survival time of patients with OPN+ pancreatobiliary AC was shorter than those with OPN- intestinal AC (p = 0.043), and other subtype of AC (p = 0.002). Expression of OPN and survival time of patients with different subtypes of ampullary carcinoma (AC). Survival time is expressed as median ± SD. 1p-value for comparison of the 3 types; 2p-value for comparison of the OPN+ and OPN- groups. * p = 0.041 and p = 0.010 for comparison with the intestinal type and other type OPN+ AC, respectively. There was no difference between intestinal AC and other type of AC (p = 0.907). Discussion: Many previous studies have examined the expression of histological markers in the carcinomas of patients with ACs. For example, de Paiva Haddad et al. reported that MUC1, MUC2, and CDX2 provided the best agreement with histomorphological classification of AC [4]. However, their multivariate analysis indicated that neither histological classification nor immunohistochemical results were statistically independent predictors of poor prognosis. Moriya et al. reported that MUC1 and MUC2 expression was useful for classification of pancreatobiliary and intestinal AC, and that pancreatobiliary AC was associated with worse prognosis than intestinal AC [8]. Ehehalt et al. reported that immunohistochemical determination of HNF4α expression distinguished different AC subtypes and that HNF4α protein expression was an independent predictor of favorable prognosis [9]. Westgaard et al. classified ACs by expression of CK7, MUC4, and CDX2 [6]. They found only moderate agreement of the immunohistochemical and histomorphological classifications, and that expression of MUC1 and/or MUC4 was independently associated with poor prognosis. Santini et al. reported that MUC2 and MUC5 expression was not associated with prognosis of the patients with radically resected AC [7]. Clearly, it is difficult to completely reconcile these diverse and occasionally contradictory results, which have examined many markers in diverse patient populations. Other researchers have reported immunohistochemical classification of AC based on expression of CK7 and CK20 [3,10]. Thus, in the present study we classified our AC patients as having pancreatobiliary AC (CK7+/CK20-, 20%), intestinal AC (CK7-/CK20+, 24.2%), or other subtype of AC (CK7+/CK20+ or CK7-/CK20-, 55.8%). Thus, the majority of our Chinese patients had other subtype of AC, in contrast to Kimura et al. [2], who studied Japanese patients, and Zhou et al. [3] who studied German patients and reported that pancreatobiliary AC was the most common subtype. This difference might be due to population differences and/or to the different reactivity thresholds used in the immunohistochemical classification [29]. Our survival analysis indicated no significant differences in survival among patients with the different subtypes of AC, consistent with the report of Zhou et al. [3]. Taken together, these findings suggest the immunohistologic subtype of AC alone has limited value in determination of the prognosis. Previous research has indicated that OPN expression is significantly associated with poor survival of patients with several forms of cancer [13,14]. It has also been reported that elevated OPN expression in AC patients predicts poor disease-specific survival [25-27]. However, when we pooled all AC subtypes, we found that survival time of OPN- and OPN+ patients had no significant difference. This is consistent with the results of Matsuzaki et al., who reported that OPN expression in AC patients was not associated with survival rate, although OPN expression in the carcinoma was higher than in adjacent normal tissues [30]. Our further analysis indicated that the subtype of AC (intestinal, pancreatobiliary, or other) had no significant effect on survival of patients with OPN- carcinomas. However, for patients with OPN+ carcinomas, those with intestinal AC or other subtype of AC had significantly better survival than those with pancreatobiliary AC. Thus, OPN expression appears to affect the biological behavior of AC, and this effect depends on the anatomical origin of the tumor. These results indicate that determination of the prognosis of patients with AC should consider OPN expression. Previous studies have reported interactions of OPN with other proteins. For example, in situ proximity ligation analysis indicated a molecular interaction of OPN and CD44 and that elevated expression of these proteins were associated with increased mitosis and significantly enhanced gastrointestinal stromal tumor cell proliferation in vitro [31]. Yang et al. reported that OPN combined with CD44 was a promising independent predictor of tumor recurrence and survival in patients with hepatocellular carcinoma [32]. OPN combined with CDX2 appears to predict survival of advanced gastric cancer patients, and CDX2 may be a transcription factor that modulates the expression of osteopontin [33]. Our results support previous reports which suggest that OPN has a role in the pathogenesis of AC [25-27]. However, a limitation of our study is that patients were enrolled retrospectively, and we did not include histomorphological classification of patients or immunochemical determination based on other factors including CDX2 and mucins. Clearly, the potential interaction of OPN, CK7, CDX2, mucins and other factors and the role of these in the pathogenesis of AC warrant further studies. Conclusions: In conclusion, our results indicate that it is difficult to determine prognosis of patients with AC based solely on immunohistochemical classification that considers CK7 and CK20 status. However, the additional consideration of OPN status allows determination of prognosis. Our results also suggest that OPN plays a role in the pathogenesis of AC, but its mechanisms and relationship with CK7 and CK20 warrant further studies.

Background: Ampullary cancer (AC) was classified as pancreatobiliary, intestinal, or other subtype based on the expression of cytokeratin 7 (CK7) and cytokeratin 20 (CK20). We aimed to explore the association of AC subtype with patient prognosis. Methods: The relationship of AC subtype and expression of Osteopontin (OPN) with the prognosis of 120 AC patients after pancreaticoduodenectomy was investigated. Results: The patients had pancreatobiliary (CK7+/CK20-, n = 24, 20%), intestinal (CK7-/CK20+, n = 29, 24.2%) or other (CK7+/CK20+ or CK7-/CK20-, n = 67, 55.8%) subtypes of AC, and their median survival times were 23 ± 4.2, 38 ± 2.8 and 64 ± 16.8 months, respectively. The survival times of 64 OPN- patients (53.3%) and 56 OPN+ patients (46.7%) were 69 ± 18.4 and 36 ± 1.3 months, respectively. There was no significant effect of AC subtype on survival of OPN- patients. For OPN+ patients, those with pancreatobiliary AC had a shorter survival time (22 ± 6.6 months) than those with intestinal AC (37 ± 1.4 months, p = 0.041), and other AC subtype (36 ± 0.9 months, p = 0.010); intestinal and other AC subtypes had similar survival times. Conclusions: The prognosis of AC patients can be estimated based on immunohistochemical classification and OPN status.

Background: Ampullary carcinoma (AC) is a relatively rare tumor of the hepatopancreatic ampulla that accounts for approximately 0.2% of gastrointestinal tract malignancies and 7% of periampullary carcinomas [1]. ACs have different anatomical origins. Kimura et al. initially classified AC as pancreatobiliary AC if it had papillary projections with scant fibrous cores and as intestinal AC if it resembled tubular adenocarcinoma of the stomach or colon [2]. Numerous studies have reported that intestinal AC is associated with a better prognosis than pancreatobiliary AC [1-5]. AC has also been classified based on immunohistochemical expression of cytokeratin 7 (CK7), Mucins and CDX2 [4,6-8] and HNF4α [9]. However, the clinical significance and survival rates of AC patients with these different immunohistochemical subtypes have not been definitely established. Histologic classification and immunohistochemical characterization by cytokeratins are in good agreement [5]. Fischer et al. reported that the histological subtypes of AC could be determined by the expression of CK7, CK20, and MUC2; pancreaticobiliary AC is CK7+/CK20-/MUC2-, and intestinal AC is CK7-/CK20+/MUC2+ [10]. Zhou et al. classified CK7-/CK20+ tumors as intestinal AC, CK7+/CK20- tumors as pancreatobiliary AC, and tumors that are CK7+/CK20+ or CK7-/CK20- as "other" [3]. However, there was no statistical difference in survival of patients with different CK7/CK20 subtypes [3] or with different CK20/MUC subtypes [11]. Osteopontin (OPN) is a secretory calcium-binding phosphorylated glycoprotein and plays an important role in bone metabolism. OPN is widely distributed in the urine, blood, gastrointestinal tract, pancreas, lungs and elsewhere. At the molecular level, OPN plays important roles in cellular adhesion and migration, tissue repair, and signal transduction and also in the invasion and metastasis of several cancers [12]. OPN is significantly associated with survival rate in several cancers and has value as a marker of clinical tumor progression [13,14]. In particular, low OPN levels were significantly associated with a favorable prognosis in patients with advanced non-small cell lung cancer [15], laryngeal and hypopharyngeal carcinomas [16], hepatocellular carcinoma [17], colorectal cancer [18], idiopathic pulmonary hypertension [19], upper urinary tract urothelial carcinoma [20], acute myeloid leukemia [21], oral squamous cell carcinoma [22], and endometrial cancer [23]. OPN may also be a suitable biomarker for overall survival and renal outcome of patients who are critically ill with acute kidney injury [24]. However, few studies have investigated the expression of OPN in patients with AC. Van Heek et al. reported higher OPN expression in the sera and tumors of AC patients than in the sera and duodenal samples of healthy controls [25]. Bloomston et al. reported that node-negative status and lack of OPN expression were associated with prolonged survival in patients with AC [26]. Hsu et al. reported that expression of OPN and the presence of tumor-associated macrophages in bulky AC were associated with tumor recurrence, and poorer disease-specific survival [27]. In the present study, we retrospectively analyzed the clinical data of 120 patients who were undergoing pancreaticoduodenectomy due to AC. We focused on the association of AC prognosis with the expression of CK7, CK20, and OPN. Conclusions: XQZ carried out the study design, defined the intellectual content, participated in the literature research and manuscript preparation, analyzed data, and edited the manuscript. JHD do guarantor of integrity of the entire study, carried out the study concepts, and reviewed the manuscript. WZZ carried out the clinical studies, and acquired data. ZL carried out the experimental studies, and did statistical analysis. All authors read and approved the final manuscript.

Background: Ampullary cancer (AC) was classified as pancreatobiliary, intestinal, or other subtype based on the expression of cytokeratin 7 (CK7) and cytokeratin 20 (CK20). We aimed to explore the association of AC subtype with patient prognosis. Methods: The relationship of AC subtype and expression of Osteopontin (OPN) with the prognosis of 120 AC patients after pancreaticoduodenectomy was investigated. Results: The patients had pancreatobiliary (CK7+/CK20-, n = 24, 20%), intestinal (CK7-/CK20+, n = 29, 24.2%) or other (CK7+/CK20+ or CK7-/CK20-, n = 67, 55.8%) subtypes of AC, and their median survival times were 23 ± 4.2, 38 ± 2.8 and 64 ± 16.8 months, respectively. The survival times of 64 OPN- patients (53.3%) and 56 OPN+ patients (46.7%) were 69 ± 18.4 and 36 ± 1.3 months, respectively. There was no significant effect of AC subtype on survival of OPN- patients. For OPN+ patients, those with pancreatobiliary AC had a shorter survival time (22 ± 6.6 months) than those with intestinal AC (37 ± 1.4 months, p = 0.041), and other AC subtype (36 ± 0.9 months, p = 0.010); intestinal and other AC subtypes had similar survival times. Conclusions: The prognosis of AC patients can be estimated based on immunohistochemical classification and OPN status.

[CONTENT] ampullary cancer | osteopontin | survival analysis | immunohistochemistry | classification | Cytokeratin 20 | Cytokeratin 7 [SUMMARY]

[CONTENT] ampullary cancer | osteopontin | survival analysis | immunohistochemistry | classification | Cytokeratin 20 | Cytokeratin 7 [SUMMARY]

[CONTENT] ampullary cancer | osteopontin | survival analysis | immunohistochemistry | classification | Cytokeratin 20 | Cytokeratin 7 [SUMMARY]

[CONTENT] ampullary cancer | osteopontin | survival analysis | immunohistochemistry | classification | Cytokeratin 20 | Cytokeratin 7 [SUMMARY]

[CONTENT] ampullary cancer | osteopontin | survival analysis | immunohistochemistry | classification | Cytokeratin 20 | Cytokeratin 7 [SUMMARY]

[CONTENT] Adenocarcinoma | Ampulla of Vater | Biomarkers, Tumor | Common Bile Duct Neoplasms | Female | Humans | Immunohistochemistry | Intestinal Neoplasms | Kaplan-Meier Estimate | Keratin-20 | Keratin-7 | Male | Middle Aged | Osteopontin | Prognosis [SUMMARY]

[CONTENT] Adenocarcinoma | Ampulla of Vater | Biomarkers, Tumor | Common Bile Duct Neoplasms | Female | Humans | Immunohistochemistry | Intestinal Neoplasms | Kaplan-Meier Estimate | Keratin-20 | Keratin-7 | Male | Middle Aged | Osteopontin | Prognosis [SUMMARY]

[CONTENT] Adenocarcinoma | Ampulla of Vater | Biomarkers, Tumor | Common Bile Duct Neoplasms | Female | Humans | Immunohistochemistry | Intestinal Neoplasms | Kaplan-Meier Estimate | Keratin-20 | Keratin-7 | Male | Middle Aged | Osteopontin | Prognosis [SUMMARY]

[CONTENT] Adenocarcinoma | Ampulla of Vater | Biomarkers, Tumor | Common Bile Duct Neoplasms | Female | Humans | Immunohistochemistry | Intestinal Neoplasms | Kaplan-Meier Estimate | Keratin-20 | Keratin-7 | Male | Middle Aged | Osteopontin | Prognosis [SUMMARY]

[CONTENT] Adenocarcinoma | Ampulla of Vater | Biomarkers, Tumor | Common Bile Duct Neoplasms | Female | Humans | Immunohistochemistry | Intestinal Neoplasms | Kaplan-Meier Estimate | Keratin-20 | Keratin-7 | Male | Middle Aged | Osteopontin | Prognosis [SUMMARY]

[CONTENT] ac intestinal pancreatobiliary | pancreatobiliary intestinal ac | ampullary cancer intestinal | ampullary carcinoma ac | pancreas cancers involving [SUMMARY]

[CONTENT] ac intestinal pancreatobiliary | pancreatobiliary intestinal ac | ampullary cancer intestinal | ampullary carcinoma ac | pancreas cancers involving [SUMMARY]

[CONTENT] ac intestinal pancreatobiliary | pancreatobiliary intestinal ac | ampullary cancer intestinal | ampullary carcinoma ac | pancreas cancers involving [SUMMARY]

[CONTENT] ac intestinal pancreatobiliary | pancreatobiliary intestinal ac | ampullary cancer intestinal | ampullary carcinoma ac | pancreas cancers involving [SUMMARY]

[CONTENT] ac intestinal pancreatobiliary | pancreatobiliary intestinal ac | ampullary cancer intestinal | ampullary carcinoma ac | pancreas cancers involving [SUMMARY]

[CONTENT] ac | patients | opn | survival | ck7 | ck20 | ck7 ck20 | intestinal | expression | survival time [SUMMARY]

[CONTENT] ac | patients | opn | survival | ck7 | ck20 | ck7 ck20 | intestinal | expression | survival time [SUMMARY]

[CONTENT] ac | patients | opn | survival | ck7 | ck20 | ck7 ck20 | intestinal | expression | survival time [SUMMARY]

[CONTENT] ac | patients | opn | survival | ck7 | ck20 | ck7 ck20 | intestinal | expression | survival time [SUMMARY]

[CONTENT] ac | patients | opn | survival | ck7 | ck20 | ck7 ck20 | intestinal | expression | survival time [SUMMARY]

[CONTENT] ac | opn | ck7 | ck20 | expression | ck7 ck20 | reported | tumors | associated | patients [SUMMARY]

[CONTENT] cancer | positive | staining | cells | duodenal | considered | surgery | disease | follow | spss [SUMMARY]

[CONTENT] status | prognosis | results | patients ac based | considers ck7 ck20 | considers ck7 ck20 status | mechanisms | mechanisms relationship | mechanisms relationship ck7 | mechanisms relationship ck7 ck20 [SUMMARY]

[CONTENT] ac | patients | opn | survival | ck7 | ck20 | ck7 ck20 | expression | intestinal | time [SUMMARY]

[CONTENT] ac | patients | opn | survival | ck7 | ck20 | ck7 ck20 | expression | intestinal | time [SUMMARY]

[CONTENT] AC | 7 | 20 ||| AC [SUMMARY]

[CONTENT] AC | Osteopontin | OPN | 120 [SUMMARY]

[CONTENT] AC | OPN [SUMMARY]

[CONTENT] AC | 7 | 20 ||| AC ||| AC | Osteopontin | OPN | 120 ||| ||| CK7+/CK20- | 24 | 20% | 29 | 24.2% | 67 | 55.8% | AC | 23 | 38 | 2.8 | 16.8 months ||| 64 | OPN- | 53.3% | 56 | 46.7% | 69 | 18.4 | 36 ± 1.3 months ||| AC | OPN- ||| OPN+ | AC | 6.6 months | AC | 37 | 1.4 months | 0.041 | AC | 36 ± 0.9 months | 0.010 | AC ||| AC | OPN [SUMMARY]

[CONTENT] AC | 7 | 20 ||| AC ||| AC | Osteopontin | OPN | 120 ||| ||| CK7+/CK20- | 24 | 20% | 29 | 24.2% | 67 | 55.8% | AC | 23 | 38 | 2.8 | 16.8 months ||| 64 | OPN- | 53.3% | 56 | 46.7% | 69 | 18.4 | 36 ± 1.3 months ||| AC | OPN- ||| OPN+ | AC | 6.6 months | AC | 37 | 1.4 months | 0.041 | AC | 36 ± 0.9 months | 0.010 | AC ||| AC | OPN [SUMMARY]

Identification of putative interactions between swine and human influenza A virus nucleoprotein and human host proteins.

25547032

Influenza A viruses (IAVs) are important pathogens that affect the health of humans and many additional animal species. IAVs are enveloped, negative single-stranded RNA viruses whose genome encodes at least ten proteins. The IAV nucleoprotein (NP) is a structural protein that associates with the viral RNA and is essential for virus replication. Understanding how IAVs interact with host proteins is essential for elucidating all of the required processes for viral replication, restrictions in species host range, and potential targets for antiviral therapies.

BACKGROUND

In this study, the NP from a swine IAV was cloned into a yeast two-hybrid "bait" vector for expression of a yeast Gal4 binding domain (BD)-NP fusion protein. This "bait" was used to screen a Y2H human HeLa cell "prey" library which consisted of human proteins fused to the Gal4 protein's activation domain (AD). The interaction of "bait" and "prey" proteins resulted in activation of reporter genes.

METHODS

Seventeen positive bait-prey interactions were isolated in yeast. All of the "prey" isolated also interact in yeast with a NP "bait" cloned from a human IAV strain. Isolation and sequence analysis of the cDNAs encoding the human prey proteins revealed ten different human proteins. These host proteins are involved in various host cell processes and structures, including purine biosynthesis (PAICS), metabolism (ACOT13), proteasome (PA28B), DNA-binding (MSANTD3), cytoskeleton (CKAP5), potassium channel formation (KCTD9), zinc transporter function (SLC30A9), Na+/K+ ATPase function (ATP1B1), and RNA splicing (TRA2B).

RESULTS

Ten human proteins were identified as interacting with IAV NP in a Y2H screen. Some of these human proteins were reported in previous screens aimed at elucidating host proteins relevant to specific viral life cycle processes such as replication. This study extends previous findings by suggesting a mechanism by which these host proteins associate with the IAV, i.e., physical interaction with NP. Furthermore, this study revealed novel host protein-NP interactions in yeast.

CONCLUSIONS

4297426

Background

Influenza A viruses (IAVs) are important pathogens that affect the health of humans and many additional animal species. In humans, seasonal IAV infection presents as a non-fatal, uncomplicated, acute infection characterized by the presence of upper respiratory symptoms as well as fever, headache, soreness and fatigue lasting 2–5 days [1]. However, deaths from seasonal IAV infections often arise when the normal flu symptoms are exacerbated by compromised immunity or age [1]. In addition to seasonal influenza A viruses, pandemic strains periodically appear causing increased mortality rates. For example, the 1918 Spanish flu resulted in approximately 50 million human deaths [2]. The most recent pandemic resulted from the emergence of the swine-origin H1N1 virus [3,4]. Due to influenza A’s potential for mortality, high mutation rates (resulting in genetic drift) and pandemic potential (resulting from genetic reassortment), it is critical to learn more about the virus, especially as it pertains to virulence, transmissibility, and identification of potential targets for development of therapeutics. Influenza A viruses exhibit a broad host range beyond humans [5]. Waterfowl serve as the central reservoir species. In addition to humans, various IAV subtypes circulate in pigs, poultry, horses, dogs as well as other species. Subtypes (H1N1 and H3N2) that circulate in the human population also circulate in the swine populations [6]. In addition to the concern for IAV in terms of swine health [7], IAV infection in swine is also important due to the potential for zoonotic infections as well as the potential for serving as a mixing vessel for the generation of human pandemic viruses [5]. The majority of inter-species transmission research has focused on amino acids in the hemagglutinin affecting viral attachment and entry [8]. Although known to be important, other than amino acids associated with temperature sensitivity (e.g., PB2 627 [9] or 701 [10]), the effects of differences in amino acid residues in proteins of the replication complex on species specificity is less well understood. Previous evidence suggests that there may be a role in some of these replication complex proteins in allowing for zoonotic infections [11]. The genome of the enveloped influenza A virion consists of eight segments of negative-sense single-stranded RNA that encode at least ten viral genes: Hemagglutinin (HA), Neuraminidase (NA), Matrix 2 protein (M2), Matrix 1 protein (M1), Non-structural Protein 1 (NS1), Non-structural Protein 2 (NS2), Nucleoprotein (NP), Polymerase Basic 1 (PB-1), Polymerase Basic 2 (PB-2), and Polymerase Acidic (PA) [1]. The viral envelope contains HA, NA, and M2 proteins, and M1 proteins form a layer inside the envelope. The IAV RNA genome located inside the virion is coated with NP and is bound by the replication complex consisting of PB1, PB2, and PA. The NP, the focus of this study, encoded by the fifth IAV RNA segment, binds with high affinity to viral RNA [12]. NP plays a role in viral RNA replication and transcription [12]. More recent data suggests that NP binds the polymerase as well as the newly replicated RNA and may act as a processivity factor that is necessary for replication of the viral RNA to be completed [13]. Phylogenic analysis of IAV NP shows distinct lineages of NP based on the host species [14,15]. Within the NP, there are amino acid signatures found within different host species [16,17]. These host-specific amino acid residues may result in differences in affinities for the various host proteins with which they interact (e.g., importin α1 [18], F- actin [19], nuclear factor 90 [20], cyclophilin E [21], exportin 1 [22], HMGB1 [23], HMGB2 [23], MxA/Mx1 [24,25], HSP40 [26], karyopherin alpha [27,28], clusterin [29], Raf-2p48/BAT1/UAP56 [30], Aly/REF [31], Tat-SF1 [32], TRIM22 [33], and alpha actinin 4 [34]) or they may result in differences in how the NP interacts with other viral proteins that have also made host-specific adaptations [11]. Given the suggestion that NP plays a role in determining host range [5,35-37], it is important to identify all host proteins that interact with NP. Replication of the IAV is dependent on the host cell machinery and interactions between host proteins and viral proteins. Therefore any inquiry attempting to investigate the life cycle of IAV must incorporate host-pathogen protein interactions in order to truly provide a mechanistic understanding of the process. Previous screens have identified important protein-protein interactions between IAV and host proteins, but often results are not consistent between screens. Therefore, corroborating evidence with additional screens is crucial to accumulating the best understanding of critical pathogen-host interactions in viral replication. As examples, several genome-wide screens have been performed using RNAi to systematically knockdown host genes and evaluate the effect on various stages of the IAV life cycle [38-41]. An integrated approach carried out by Shapira and colleagues involved transcription profiling and yeast two hybrid screens using specific viral proteins as baits [28]. Proteomics approaches have revealed host proteins found within viral particles [42,43]. The Random Homozygous Gene Perturbation strategy was used to identify host factors that prevent influenza-mediated killing of host cells [44]. In order to better understand the role of NP-host protein interactions in IAV replication, a Gal4-based yeast two-hybrid (Y2H) assay was used in this study. The Y2H system allows for the identification of unique human binding partners with NP. A “bait” plasmid encoding the binding domain (BD) of the Gal4 transcriptional activator fused to NP and a “prey” plasmid encoding Gal4’s activation domain (AD) fused to a protein encoded by a human cDNA are introduced into a yeast strain containing Gal4-responsive reporter genes. Interaction of the bait and prey brings together Gal4’s BD and AD, resulting in transcriptional activation of reporter genes [45]. The Y2H approach has successfully identified cellular factors (e.g., Raf-2p48/BAT1/UAP56, Hsp40, KPNA1, KNPA3, KPNA6, C16orf45, GMCL1, MAGED1, MLH1, USHBP1, ZBTB25, CLU, Aly/REF, and ACTN4) that interact with the IAV NP [26-31,34], (Table 1). In this study, the nucleoprotein from a classical swine H1N1 IAV (Sw/NC/44173/00) was used as the “bait” in a Y2H screen against a “prey” HeLa cDNA library. To investigate if the origin of NP affects the ability of the interactions to take place, the nucleoprotein from a contemporary human H3N2 IAV (A/Ca/07/04) was also used as the “bait” in a Y2H screen against the same prey HeLa cDNA library. By determination of the putative interaction partners between NP and human proteins, the possible functions of the newfound protein-protein interactions or verification and perhaps elucidation of previously identified host factors may be further investigated.Table 1 Summary of identified bait-prey interactions and previous reports of host protein association with influenza A virus Gene Protein description Complete ORF Number of prey clones Prey clones Previous reports with IAV PAICSPhosphoribosylaminoimidazole carboxylaseYes3R3, R9, R15Karlas et al. [40], Kumar and Nanduri [47], Kroeker et al. [48]MSANTD3Myb/SANT-like DNA-binding domain containing 3No3R10, R13, R27N/AFLJ30306PREDICTED: Homo sapiens uncharacterized LOC101059922, mRNAYes2R23, R33N/APA28BProteasome activator subunit 2Yes2R6, R19N/AKCTD9Potassium channel tetramerisation domain containing 9No2R34, R36N/AACOT13Acyl-CoA thioesterase 13Yes1R11N/ATRA2BTransformer 2 beta homolog (Drosophila)Yes1R12Zhu et al. [56]SLC30A9Solute carrier family 30 (zinc transporter), member 9No1R28Kumar and Nanduri [47], Sui et al. [44]ATP1B1ATPase, Na+/K+ transporting, beta 1 polypeptideNo1R7Mi et al. [51], Liu et al. [50]CKAP5Cytoskeleton associated protein 5No1R29N/A Host proteins identified as interacting with IAV nucleoprotein in previously published yeast two-hybrid screens BAT1/UAP56RNA-dependent ATPaseN/AN/AN/AMomose et al. [30]HSP40Heat shock protein 40N/AN/AN/ASharma et al. [26]NPI-1/ SRP1/KPNA1Karyopherin α1N/AN/AN/AO’Neill and Palese [27], Shapira et al. [28]KPNA3Karyopherin α3N/AN/AN/AShapira et al. [28]KPNA6Karyopherin α6N/AN/AN/AShapira et al. [28]C16orf45-N/AN/AN/AShapira et al. [28]GMCL1Germ cell-less, spermatogenesis associated 1N/AN/AN/AShapira et al. [28]MAGED1Melanoma antigen family D.1N/AN/AN/AShapira et al. [28]MLH1MutL homolog 1N/AN/AN/AShapira et al. [28]USHBP1Usher syndrome 1C binding protein 1N/AN/AN/AShapira et al. [28]ZBTB25Zinc finger and BTB domain containing 25N/AN/AN/AShapira et al. [28]CLUClusterinN/AN/AN/ATripathi et al. [29]ALY/REFRNA export adaptor proteinN/AN/AN/ABalasubramaniam et al. [31]ACTN4Alpha-actinin 4N/AN/AN/ASharma et al. [34] Summary of identified bait-prey interactions and previous reports of host protein association with influenza A virus Reported herein are ten potential human host cell proteins that interact with IAV NP in a Y2H screen. The relative strengths of these protein-protein interactions were characterized by a plating assay and beta-galactosidase assay.

Results

Constructing an NP bait and screening a human cDNA prey library The nucleoprotein open reading frame from the classical swine IAV H1N1 strain A/Sw/NC/44173/00 and contemporary human IAV H3N2 strain A/Ca/07/04 encode proteins with 89.1% identity and 94.8% similarity (Figure 1). Of the 49 amino acids that differ between the swine and human NPs, 29 are conservative differences. However, it is worth noting that several of these amino acid differences are signatures that differ between avian, swine and human isolates [16,17]. The two NP ORFs were amplified by PCR, and the resulting PCR products were separately inserted by recombination cloning into Y2H bait vector pGBKT7 as a C-terminal fusion with Gal4p’s DNA binding domain. Correct constructs were confirmed by DNA sequence analysis. Gal4(BD)-NP fusion proteins isolated from yeast cells were detected as 77 KDa bands on a Western blot probed with an anti-Myc antibody (data not shown). When introduced alone into yeast Y2HGold cells, the swine and human NP baits constructs were not toxic to yeast and did not autoactivate the Y2H reporter genes, indicating they were suitable for use in a Y2H screen (data not shown).Figure 1 Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Using a Y2H assay with Gal4 (BD)-swine NP as a bait and a human HeLa cDNA prey library, approximately 2 × 106 clones were screened resulting in 17 positive bait-prey interacting yeast strains. Sequence and bioinformatics analysis of the prey plasmids indicated that these 17 prey plasmids, listed by their initial positive interaction identification number (R3, R6, etc.), represent ten different human cDNAs (Table 1). An independent screen using Gal4 (BD)-human NP as a bait and a human HeLa cDNA prey library yielded a subset of these same host protein targets. Some of the identified clones contain partial cDNAs. Three independent prey plasmids isolated contained human cDNAs phosphoribosylaminoimidazole carboxylase, phosphoribosylaminoimidazole succinocarboxamide synthetase (PAICS) and another three contained Myb/SANT-like DNA-binding domain containing 3 (MSANTD3). Two independent prey plasmids were isolated for each of three human cDNAs: proteasome activator subunit 2 (PA28B), potassium channel tetramerisation domain (KCTD9), and an uncharacterized protein (FLJ30306). Each of the remaining five prey plasmids contained different human cDNAs. The nucleoprotein open reading frame from the classical swine IAV H1N1 strain A/Sw/NC/44173/00 and contemporary human IAV H3N2 strain A/Ca/07/04 encode proteins with 89.1% identity and 94.8% similarity (Figure 1). Of the 49 amino acids that differ between the swine and human NPs, 29 are conservative differences. However, it is worth noting that several of these amino acid differences are signatures that differ between avian, swine and human isolates [16,17]. The two NP ORFs were amplified by PCR, and the resulting PCR products were separately inserted by recombination cloning into Y2H bait vector pGBKT7 as a C-terminal fusion with Gal4p’s DNA binding domain. Correct constructs were confirmed by DNA sequence analysis. Gal4(BD)-NP fusion proteins isolated from yeast cells were detected as 77 KDa bands on a Western blot probed with an anti-Myc antibody (data not shown). When introduced alone into yeast Y2HGold cells, the swine and human NP baits constructs were not toxic to yeast and did not autoactivate the Y2H reporter genes, indicating they were suitable for use in a Y2H screen (data not shown).Figure 1 Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Using a Y2H assay with Gal4 (BD)-swine NP as a bait and a human HeLa cDNA prey library, approximately 2 × 106 clones were screened resulting in 17 positive bait-prey interacting yeast strains. Sequence and bioinformatics analysis of the prey plasmids indicated that these 17 prey plasmids, listed by their initial positive interaction identification number (R3, R6, etc.), represent ten different human cDNAs (Table 1). An independent screen using Gal4 (BD)-human NP as a bait and a human HeLa cDNA prey library yielded a subset of these same host protein targets. Some of the identified clones contain partial cDNAs. Three independent prey plasmids isolated contained human cDNAs phosphoribosylaminoimidazole carboxylase, phosphoribosylaminoimidazole succinocarboxamide synthetase (PAICS) and another three contained Myb/SANT-like DNA-binding domain containing 3 (MSANTD3). Two independent prey plasmids were isolated for each of three human cDNAs: proteasome activator subunit 2 (PA28B), potassium channel tetramerisation domain (KCTD9), and an uncharacterized protein (FLJ30306). Each of the remaining five prey plasmids contained different human cDNAs. Evaluation of the strength of bait-prey interactions The strength of interaction between each NP bait (swine or human) and human prey in yeast was measured using a qualitative growth assay and a quantitative β-galactosidase activity assay. First, yeast growth on selective media after cotransformation with swine or human NP and one of the identified prey was tested at both 30°C and 37°C as a qualitative metric of the protein-protein interaction’s strength (Figure 2, Table 2). All strains, including the positive and negative controls, grew on SD-Leu-Trp which selects for the presence of both bait and prey plasmids. Most strains showed decreased growth at 37°C relative to the more permissive temperature 30°C. Strains examined at 37°C which contained either swine or human NP bait grew best in the presence of prey CKAP5, MSANTD3, or TRA2B. In contrast, at 37°C strains containing swine NP grew poorly with prey ACOT13, whereas strains containing human NP grew poorly with prey PA28B.Figure 2 Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector).Table 2 Summary of affinity of bait-prey interactions Growth on SD-His agar plates* β-Gal activity** Gene Protein description Prey clone Human (30°C) Human (37°C) Swine (30°C) Swine (37°C) Human NP Swine NP PAICSPhosphoribosylaminoimidazole carboxylaseR15(++)(+)(++)(++)2.230.4MSANTD3Myb/SANT-like DNA-binding domain containing 3R10(++)(++)(++)(++)13.918.2FLJ30306PREDICTED: Homo sapiens uncharacterized LOC101059922, mRNAR23(+)(+)(++)(++)1.95.4PA28BProteasome (prosome, macropain) activator subunit 2R6(++)(−)(++)(+)6.733.3KCTD9Potassium channel tetramerisation domain containing 9R34(+)(+)(++)(++)5.016.4ACOT13Acyl-CoA thioesterase 13R11(++)(+)(++)(−)6.915.5TRA2BTransformer 2 beta homolog (Drosophila)R12(++)(++)(++)(++)1.90.85SLC30A9Solute carrier family 30 (zinc transporter), member 9R28(++)(+)(++)(+)82.9119.9ATP1B1ATPase, Na+/K+ transporting, beta 1 polypeptideR7(++)(+)(++)(++)3.44.3CKAP5Cytoskeleton associated protein 5R29(++)(++)(++)(++)70.190.3* Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2).** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector). Summary of affinity of bait-prey interactions * Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2). ** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). The strength of the interactions between IAV NP and host proteins identified by the library screen was also investigated by measuring the β-gal activity in liquid yeast cultures (Figure 3, Table 2). As observed with the qualitative growth assay, for all ten prey proteins, interactions were observed with both the swine and human NP. Of the ten interactions investigated, yeast strains containing swine or human NP and human prey SLC30A9 and CKAP3 showed the highest levels of β-gal activity compared to the other prey proteins examined. This is consistent with the growth of yeast strains containing human NP and SLC30A9 and CKAP3 on selective media at 37°C ( Figure 2, Table 2). Even strains with the lowest levels of β-gal activity observed (TRA2B and ATP1B1) have double the β-gal activity of the negative control. While the data suggest there may be differences in the strength of bait-prey interactions based on NP host origin, these differences should be interpreted carefully given the nature of the assay.Figure 3 β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. The strength of interaction between each NP bait (swine or human) and human prey in yeast was measured using a qualitative growth assay and a quantitative β-galactosidase activity assay. First, yeast growth on selective media after cotransformation with swine or human NP and one of the identified prey was tested at both 30°C and 37°C as a qualitative metric of the protein-protein interaction’s strength (Figure 2, Table 2). All strains, including the positive and negative controls, grew on SD-Leu-Trp which selects for the presence of both bait and prey plasmids. Most strains showed decreased growth at 37°C relative to the more permissive temperature 30°C. Strains examined at 37°C which contained either swine or human NP bait grew best in the presence of prey CKAP5, MSANTD3, or TRA2B. In contrast, at 37°C strains containing swine NP grew poorly with prey ACOT13, whereas strains containing human NP grew poorly with prey PA28B.Figure 2 Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector).Table 2 Summary of affinity of bait-prey interactions Growth on SD-His agar plates* β-Gal activity** Gene Protein description Prey clone Human (30°C) Human (37°C) Swine (30°C) Swine (37°C) Human NP Swine NP PAICSPhosphoribosylaminoimidazole carboxylaseR15(++)(+)(++)(++)2.230.4MSANTD3Myb/SANT-like DNA-binding domain containing 3R10(++)(++)(++)(++)13.918.2FLJ30306PREDICTED: Homo sapiens uncharacterized LOC101059922, mRNAR23(+)(+)(++)(++)1.95.4PA28BProteasome (prosome, macropain) activator subunit 2R6(++)(−)(++)(+)6.733.3KCTD9Potassium channel tetramerisation domain containing 9R34(+)(+)(++)(++)5.016.4ACOT13Acyl-CoA thioesterase 13R11(++)(+)(++)(−)6.915.5TRA2BTransformer 2 beta homolog (Drosophila)R12(++)(++)(++)(++)1.90.85SLC30A9Solute carrier family 30 (zinc transporter), member 9R28(++)(+)(++)(+)82.9119.9ATP1B1ATPase, Na+/K+ transporting, beta 1 polypeptideR7(++)(+)(++)(++)3.44.3CKAP5Cytoskeleton associated protein 5R29(++)(++)(++)(++)70.190.3* Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2).** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector). Summary of affinity of bait-prey interactions * Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2). ** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). The strength of the interactions between IAV NP and host proteins identified by the library screen was also investigated by measuring the β-gal activity in liquid yeast cultures (Figure 3, Table 2). As observed with the qualitative growth assay, for all ten prey proteins, interactions were observed with both the swine and human NP. Of the ten interactions investigated, yeast strains containing swine or human NP and human prey SLC30A9 and CKAP3 showed the highest levels of β-gal activity compared to the other prey proteins examined. This is consistent with the growth of yeast strains containing human NP and SLC30A9 and CKAP3 on selective media at 37°C ( Figure 2, Table 2). Even strains with the lowest levels of β-gal activity observed (TRA2B and ATP1B1) have double the β-gal activity of the negative control. While the data suggest there may be differences in the strength of bait-prey interactions based on NP host origin, these differences should be interpreted carefully given the nature of the assay.Figure 3 β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples.

Conclusions

A Y2H screen using a classical swine influenza A virus nucleoprotein as a bait resulted in isolation of ten different putative interacting human host proteins involved in a variety of cellular processes and structures: purine biosynthesis (PAICS), metabolism (ACOT13), proteasome (PA28B), DNA-binding (MSANTD3), cytoskeleton (CKAP5), potassium channel formation (KCTD9), zinc transporter function (SLC30A9), Na+/K+ ATPase function (ATP1B1), and RNA splicing (TRA2B). All of the identified human host proteins interacted in yeast with NP from both human and swine origin. Proteins PAICS, ATP1B1, TRA2B, and SLC30A9 have been previously identified in studies related to host response to the IAV. Host proteins SLC30A9 and CKAP5 displayed the strongest interactions with IAV NP in yeast in a quantitative β-gal assay.

Background: Influenza A viruses (IAVs) are important pathogens that affect the health of humans and many additional animal species. In humans, seasonal IAV infection presents as a non-fatal, uncomplicated, acute infection characterized by the presence of upper respiratory symptoms as well as fever, headache, soreness and fatigue lasting 2–5 days [1]. However, deaths from seasonal IAV infections often arise when the normal flu symptoms are exacerbated by compromised immunity or age [1]. In addition to seasonal influenza A viruses, pandemic strains periodically appear causing increased mortality rates. For example, the 1918 Spanish flu resulted in approximately 50 million human deaths [2]. The most recent pandemic resulted from the emergence of the swine-origin H1N1 virus [3,4]. Due to influenza A’s potential for mortality, high mutation rates (resulting in genetic drift) and pandemic potential (resulting from genetic reassortment), it is critical to learn more about the virus, especially as it pertains to virulence, transmissibility, and identification of potential targets for development of therapeutics. Influenza A viruses exhibit a broad host range beyond humans [5]. Waterfowl serve as the central reservoir species. In addition to humans, various IAV subtypes circulate in pigs, poultry, horses, dogs as well as other species. Subtypes (H1N1 and H3N2) that circulate in the human population also circulate in the swine populations [6]. In addition to the concern for IAV in terms of swine health [7], IAV infection in swine is also important due to the potential for zoonotic infections as well as the potential for serving as a mixing vessel for the generation of human pandemic viruses [5]. The majority of inter-species transmission research has focused on amino acids in the hemagglutinin affecting viral attachment and entry [8]. Although known to be important, other than amino acids associated with temperature sensitivity (e.g., PB2 627 [9] or 701 [10]), the effects of differences in amino acid residues in proteins of the replication complex on species specificity is less well understood. Previous evidence suggests that there may be a role in some of these replication complex proteins in allowing for zoonotic infections [11]. The genome of the enveloped influenza A virion consists of eight segments of negative-sense single-stranded RNA that encode at least ten viral genes: Hemagglutinin (HA), Neuraminidase (NA), Matrix 2 protein (M2), Matrix 1 protein (M1), Non-structural Protein 1 (NS1), Non-structural Protein 2 (NS2), Nucleoprotein (NP), Polymerase Basic 1 (PB-1), Polymerase Basic 2 (PB-2), and Polymerase Acidic (PA) [1]. The viral envelope contains HA, NA, and M2 proteins, and M1 proteins form a layer inside the envelope. The IAV RNA genome located inside the virion is coated with NP and is bound by the replication complex consisting of PB1, PB2, and PA. The NP, the focus of this study, encoded by the fifth IAV RNA segment, binds with high affinity to viral RNA [12]. NP plays a role in viral RNA replication and transcription [12]. More recent data suggests that NP binds the polymerase as well as the newly replicated RNA and may act as a processivity factor that is necessary for replication of the viral RNA to be completed [13]. Phylogenic analysis of IAV NP shows distinct lineages of NP based on the host species [14,15]. Within the NP, there are amino acid signatures found within different host species [16,17]. These host-specific amino acid residues may result in differences in affinities for the various host proteins with which they interact (e.g., importin α1 [18], F- actin [19], nuclear factor 90 [20], cyclophilin E [21], exportin 1 [22], HMGB1 [23], HMGB2 [23], MxA/Mx1 [24,25], HSP40 [26], karyopherin alpha [27,28], clusterin [29], Raf-2p48/BAT1/UAP56 [30], Aly/REF [31], Tat-SF1 [32], TRIM22 [33], and alpha actinin 4 [34]) or they may result in differences in how the NP interacts with other viral proteins that have also made host-specific adaptations [11]. Given the suggestion that NP plays a role in determining host range [5,35-37], it is important to identify all host proteins that interact with NP. Replication of the IAV is dependent on the host cell machinery and interactions between host proteins and viral proteins. Therefore any inquiry attempting to investigate the life cycle of IAV must incorporate host-pathogen protein interactions in order to truly provide a mechanistic understanding of the process. Previous screens have identified important protein-protein interactions between IAV and host proteins, but often results are not consistent between screens. Therefore, corroborating evidence with additional screens is crucial to accumulating the best understanding of critical pathogen-host interactions in viral replication. As examples, several genome-wide screens have been performed using RNAi to systematically knockdown host genes and evaluate the effect on various stages of the IAV life cycle [38-41]. An integrated approach carried out by Shapira and colleagues involved transcription profiling and yeast two hybrid screens using specific viral proteins as baits [28]. Proteomics approaches have revealed host proteins found within viral particles [42,43]. The Random Homozygous Gene Perturbation strategy was used to identify host factors that prevent influenza-mediated killing of host cells [44]. In order to better understand the role of NP-host protein interactions in IAV replication, a Gal4-based yeast two-hybrid (Y2H) assay was used in this study. The Y2H system allows for the identification of unique human binding partners with NP. A “bait” plasmid encoding the binding domain (BD) of the Gal4 transcriptional activator fused to NP and a “prey” plasmid encoding Gal4’s activation domain (AD) fused to a protein encoded by a human cDNA are introduced into a yeast strain containing Gal4-responsive reporter genes. Interaction of the bait and prey brings together Gal4’s BD and AD, resulting in transcriptional activation of reporter genes [45]. The Y2H approach has successfully identified cellular factors (e.g., Raf-2p48/BAT1/UAP56, Hsp40, KPNA1, KNPA3, KPNA6, C16orf45, GMCL1, MAGED1, MLH1, USHBP1, ZBTB25, CLU, Aly/REF, and ACTN4) that interact with the IAV NP [26-31,34], (Table 1). In this study, the nucleoprotein from a classical swine H1N1 IAV (Sw/NC/44173/00) was used as the “bait” in a Y2H screen against a “prey” HeLa cDNA library. To investigate if the origin of NP affects the ability of the interactions to take place, the nucleoprotein from a contemporary human H3N2 IAV (A/Ca/07/04) was also used as the “bait” in a Y2H screen against the same prey HeLa cDNA library. By determination of the putative interaction partners between NP and human proteins, the possible functions of the newfound protein-protein interactions or verification and perhaps elucidation of previously identified host factors may be further investigated.Table 1 Summary of identified bait-prey interactions and previous reports of host protein association with influenza A virus Gene Protein description Complete ORF Number of prey clones Prey clones Previous reports with IAV PAICSPhosphoribosylaminoimidazole carboxylaseYes3R3, R9, R15Karlas et al. [40], Kumar and Nanduri [47], Kroeker et al. [48]MSANTD3Myb/SANT-like DNA-binding domain containing 3No3R10, R13, R27N/AFLJ30306PREDICTED: Homo sapiens uncharacterized LOC101059922, mRNAYes2R23, R33N/APA28BProteasome activator subunit 2Yes2R6, R19N/AKCTD9Potassium channel tetramerisation domain containing 9No2R34, R36N/AACOT13Acyl-CoA thioesterase 13Yes1R11N/ATRA2BTransformer 2 beta homolog (Drosophila)Yes1R12Zhu et al. [56]SLC30A9Solute carrier family 30 (zinc transporter), member 9No1R28Kumar and Nanduri [47], Sui et al. [44]ATP1B1ATPase, Na+/K+ transporting, beta 1 polypeptideNo1R7Mi et al. [51], Liu et al. [50]CKAP5Cytoskeleton associated protein 5No1R29N/A Host proteins identified as interacting with IAV nucleoprotein in previously published yeast two-hybrid screens BAT1/UAP56RNA-dependent ATPaseN/AN/AN/AMomose et al. [30]HSP40Heat shock protein 40N/AN/AN/ASharma et al. [26]NPI-1/ SRP1/KPNA1Karyopherin α1N/AN/AN/AO’Neill and Palese [27], Shapira et al. [28]KPNA3Karyopherin α3N/AN/AN/AShapira et al. [28]KPNA6Karyopherin α6N/AN/AN/AShapira et al. [28]C16orf45-N/AN/AN/AShapira et al. [28]GMCL1Germ cell-less, spermatogenesis associated 1N/AN/AN/AShapira et al. [28]MAGED1Melanoma antigen family D.1N/AN/AN/AShapira et al. [28]MLH1MutL homolog 1N/AN/AN/AShapira et al. [28]USHBP1Usher syndrome 1C binding protein 1N/AN/AN/AShapira et al. [28]ZBTB25Zinc finger and BTB domain containing 25N/AN/AN/AShapira et al. [28]CLUClusterinN/AN/AN/ATripathi et al. [29]ALY/REFRNA export adaptor proteinN/AN/AN/ABalasubramaniam et al. [31]ACTN4Alpha-actinin 4N/AN/AN/ASharma et al. [34] Summary of identified bait-prey interactions and previous reports of host protein association with influenza A virus Reported herein are ten potential human host cell proteins that interact with IAV NP in a Y2H screen. The relative strengths of these protein-protein interactions were characterized by a plating assay and beta-galactosidase assay. Results: Constructing an NP bait and screening a human cDNA prey library The nucleoprotein open reading frame from the classical swine IAV H1N1 strain A/Sw/NC/44173/00 and contemporary human IAV H3N2 strain A/Ca/07/04 encode proteins with 89.1% identity and 94.8% similarity (Figure 1). Of the 49 amino acids that differ between the swine and human NPs, 29 are conservative differences. However, it is worth noting that several of these amino acid differences are signatures that differ between avian, swine and human isolates [16,17]. The two NP ORFs were amplified by PCR, and the resulting PCR products were separately inserted by recombination cloning into Y2H bait vector pGBKT7 as a C-terminal fusion with Gal4p’s DNA binding domain. Correct constructs were confirmed by DNA sequence analysis. Gal4(BD)-NP fusion proteins isolated from yeast cells were detected as 77 KDa bands on a Western blot probed with an anti-Myc antibody (data not shown). When introduced alone into yeast Y2HGold cells, the swine and human NP baits constructs were not toxic to yeast and did not autoactivate the Y2H reporter genes, indicating they were suitable for use in a Y2H screen (data not shown).Figure 1 Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Using a Y2H assay with Gal4 (BD)-swine NP as a bait and a human HeLa cDNA prey library, approximately 2 × 106 clones were screened resulting in 17 positive bait-prey interacting yeast strains. Sequence and bioinformatics analysis of the prey plasmids indicated that these 17 prey plasmids, listed by their initial positive interaction identification number (R3, R6, etc.), represent ten different human cDNAs (Table 1). An independent screen using Gal4 (BD)-human NP as a bait and a human HeLa cDNA prey library yielded a subset of these same host protein targets. Some of the identified clones contain partial cDNAs. Three independent prey plasmids isolated contained human cDNAs phosphoribosylaminoimidazole carboxylase, phosphoribosylaminoimidazole succinocarboxamide synthetase (PAICS) and another three contained Myb/SANT-like DNA-binding domain containing 3 (MSANTD3). Two independent prey plasmids were isolated for each of three human cDNAs: proteasome activator subunit 2 (PA28B), potassium channel tetramerisation domain (KCTD9), and an uncharacterized protein (FLJ30306). Each of the remaining five prey plasmids contained different human cDNAs. The nucleoprotein open reading frame from the classical swine IAV H1N1 strain A/Sw/NC/44173/00 and contemporary human IAV H3N2 strain A/Ca/07/04 encode proteins with 89.1% identity and 94.8% similarity (Figure 1). Of the 49 amino acids that differ between the swine and human NPs, 29 are conservative differences. However, it is worth noting that several of these amino acid differences are signatures that differ between avian, swine and human isolates [16,17]. The two NP ORFs were amplified by PCR, and the resulting PCR products were separately inserted by recombination cloning into Y2H bait vector pGBKT7 as a C-terminal fusion with Gal4p’s DNA binding domain. Correct constructs were confirmed by DNA sequence analysis. Gal4(BD)-NP fusion proteins isolated from yeast cells were detected as 77 KDa bands on a Western blot probed with an anti-Myc antibody (data not shown). When introduced alone into yeast Y2HGold cells, the swine and human NP baits constructs were not toxic to yeast and did not autoactivate the Y2H reporter genes, indicating they were suitable for use in a Y2H screen (data not shown).Figure 1 Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Using a Y2H assay with Gal4 (BD)-swine NP as a bait and a human HeLa cDNA prey library, approximately 2 × 106 clones were screened resulting in 17 positive bait-prey interacting yeast strains. Sequence and bioinformatics analysis of the prey plasmids indicated that these 17 prey plasmids, listed by their initial positive interaction identification number (R3, R6, etc.), represent ten different human cDNAs (Table 1). An independent screen using Gal4 (BD)-human NP as a bait and a human HeLa cDNA prey library yielded a subset of these same host protein targets. Some of the identified clones contain partial cDNAs. Three independent prey plasmids isolated contained human cDNAs phosphoribosylaminoimidazole carboxylase, phosphoribosylaminoimidazole succinocarboxamide synthetase (PAICS) and another three contained Myb/SANT-like DNA-binding domain containing 3 (MSANTD3). Two independent prey plasmids were isolated for each of three human cDNAs: proteasome activator subunit 2 (PA28B), potassium channel tetramerisation domain (KCTD9), and an uncharacterized protein (FLJ30306). Each of the remaining five prey plasmids contained different human cDNAs. Evaluation of the strength of bait-prey interactions The strength of interaction between each NP bait (swine or human) and human prey in yeast was measured using a qualitative growth assay and a quantitative β-galactosidase activity assay. First, yeast growth on selective media after cotransformation with swine or human NP and one of the identified prey was tested at both 30°C and 37°C as a qualitative metric of the protein-protein interaction’s strength (Figure 2, Table 2). All strains, including the positive and negative controls, grew on SD-Leu-Trp which selects for the presence of both bait and prey plasmids. Most strains showed decreased growth at 37°C relative to the more permissive temperature 30°C. Strains examined at 37°C which contained either swine or human NP bait grew best in the presence of prey CKAP5, MSANTD3, or TRA2B. In contrast, at 37°C strains containing swine NP grew poorly with prey ACOT13, whereas strains containing human NP grew poorly with prey PA28B.Figure 2 Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector).Table 2 Summary of affinity of bait-prey interactions Growth on SD-His agar plates* β-Gal activity** Gene Protein description Prey clone Human (30°C) Human (37°C) Swine (30°C) Swine (37°C) Human NP Swine NP PAICSPhosphoribosylaminoimidazole carboxylaseR15(++)(+)(++)(++)2.230.4MSANTD3Myb/SANT-like DNA-binding domain containing 3R10(++)(++)(++)(++)13.918.2FLJ30306PREDICTED: Homo sapiens uncharacterized LOC101059922, mRNAR23(+)(+)(++)(++)1.95.4PA28BProteasome (prosome, macropain) activator subunit 2R6(++)(−)(++)(+)6.733.3KCTD9Potassium channel tetramerisation domain containing 9R34(+)(+)(++)(++)5.016.4ACOT13Acyl-CoA thioesterase 13R11(++)(+)(++)(−)6.915.5TRA2BTransformer 2 beta homolog (Drosophila)R12(++)(++)(++)(++)1.90.85SLC30A9Solute carrier family 30 (zinc transporter), member 9R28(++)(+)(++)(+)82.9119.9ATP1B1ATPase, Na+/K+ transporting, beta 1 polypeptideR7(++)(+)(++)(++)3.44.3CKAP5Cytoskeleton associated protein 5R29(++)(++)(++)(++)70.190.3* Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2).** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector). Summary of affinity of bait-prey interactions * Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2). ** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). The strength of the interactions between IAV NP and host proteins identified by the library screen was also investigated by measuring the β-gal activity in liquid yeast cultures (Figure 3, Table 2). As observed with the qualitative growth assay, for all ten prey proteins, interactions were observed with both the swine and human NP. Of the ten interactions investigated, yeast strains containing swine or human NP and human prey SLC30A9 and CKAP3 showed the highest levels of β-gal activity compared to the other prey proteins examined. This is consistent with the growth of yeast strains containing human NP and SLC30A9 and CKAP3 on selective media at 37°C ( Figure 2, Table 2). Even strains with the lowest levels of β-gal activity observed (TRA2B and ATP1B1) have double the β-gal activity of the negative control. While the data suggest there may be differences in the strength of bait-prey interactions based on NP host origin, these differences should be interpreted carefully given the nature of the assay.Figure 3 β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. The strength of interaction between each NP bait (swine or human) and human prey in yeast was measured using a qualitative growth assay and a quantitative β-galactosidase activity assay. First, yeast growth on selective media after cotransformation with swine or human NP and one of the identified prey was tested at both 30°C and 37°C as a qualitative metric of the protein-protein interaction’s strength (Figure 2, Table 2). All strains, including the positive and negative controls, grew on SD-Leu-Trp which selects for the presence of both bait and prey plasmids. Most strains showed decreased growth at 37°C relative to the more permissive temperature 30°C. Strains examined at 37°C which contained either swine or human NP bait grew best in the presence of prey CKAP5, MSANTD3, or TRA2B. In contrast, at 37°C strains containing swine NP grew poorly with prey ACOT13, whereas strains containing human NP grew poorly with prey PA28B.Figure 2 Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector).Table 2 Summary of affinity of bait-prey interactions Growth on SD-His agar plates* β-Gal activity** Gene Protein description Prey clone Human (30°C) Human (37°C) Swine (30°C) Swine (37°C) Human NP Swine NP PAICSPhosphoribosylaminoimidazole carboxylaseR15(++)(+)(++)(++)2.230.4MSANTD3Myb/SANT-like DNA-binding domain containing 3R10(++)(++)(++)(++)13.918.2FLJ30306PREDICTED: Homo sapiens uncharacterized LOC101059922, mRNAR23(+)(+)(++)(++)1.95.4PA28BProteasome (prosome, macropain) activator subunit 2R6(++)(−)(++)(+)6.733.3KCTD9Potassium channel tetramerisation domain containing 9R34(+)(+)(++)(++)5.016.4ACOT13Acyl-CoA thioesterase 13R11(++)(+)(++)(−)6.915.5TRA2BTransformer 2 beta homolog (Drosophila)R12(++)(++)(++)(++)1.90.85SLC30A9Solute carrier family 30 (zinc transporter), member 9R28(++)(+)(++)(+)82.9119.9ATP1B1ATPase, Na+/K+ transporting, beta 1 polypeptideR7(++)(+)(++)(++)3.44.3CKAP5Cytoskeleton associated protein 5R29(++)(++)(++)(++)70.190.3* Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2).** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector). Summary of affinity of bait-prey interactions * Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2). ** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). The strength of the interactions between IAV NP and host proteins identified by the library screen was also investigated by measuring the β-gal activity in liquid yeast cultures (Figure 3, Table 2). As observed with the qualitative growth assay, for all ten prey proteins, interactions were observed with both the swine and human NP. Of the ten interactions investigated, yeast strains containing swine or human NP and human prey SLC30A9 and CKAP3 showed the highest levels of β-gal activity compared to the other prey proteins examined. This is consistent with the growth of yeast strains containing human NP and SLC30A9 and CKAP3 on selective media at 37°C ( Figure 2, Table 2). Even strains with the lowest levels of β-gal activity observed (TRA2B and ATP1B1) have double the β-gal activity of the negative control. While the data suggest there may be differences in the strength of bait-prey interactions based on NP host origin, these differences should be interpreted carefully given the nature of the assay.Figure 3 β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. Constructing an NP bait and screening a human cDNA prey library: The nucleoprotein open reading frame from the classical swine IAV H1N1 strain A/Sw/NC/44173/00 and contemporary human IAV H3N2 strain A/Ca/07/04 encode proteins with 89.1% identity and 94.8% similarity (Figure 1). Of the 49 amino acids that differ between the swine and human NPs, 29 are conservative differences. However, it is worth noting that several of these amino acid differences are signatures that differ between avian, swine and human isolates [16,17]. The two NP ORFs were amplified by PCR, and the resulting PCR products were separately inserted by recombination cloning into Y2H bait vector pGBKT7 as a C-terminal fusion with Gal4p’s DNA binding domain. Correct constructs were confirmed by DNA sequence analysis. Gal4(BD)-NP fusion proteins isolated from yeast cells were detected as 77 KDa bands on a Western blot probed with an anti-Myc antibody (data not shown). When introduced alone into yeast Y2HGold cells, the swine and human NP baits constructs were not toxic to yeast and did not autoactivate the Y2H reporter genes, indicating they were suitable for use in a Y2H screen (data not shown).Figure 1 Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Sequence alignment of nucleoproteins from swine and human influenza A virus strains. The swine and human NPs are 89.1% identical (yellow highlighting) and 94.8% similar (yellow and green highlighting). Nonconservative amino acids differences are represented in white. Amino acids with * represent signature amino acids associated with avian influenza A NP, and # and † represent signature amino acids associated with avian influenza A NP based on Chen et al. [16] and Pan et al. [17]. † Denotes the only deviation of A/Sw/NC 44173/00 from the consensus swine residues based on Pan et al. [17]. Using a Y2H assay with Gal4 (BD)-swine NP as a bait and a human HeLa cDNA prey library, approximately 2 × 106 clones were screened resulting in 17 positive bait-prey interacting yeast strains. Sequence and bioinformatics analysis of the prey plasmids indicated that these 17 prey plasmids, listed by their initial positive interaction identification number (R3, R6, etc.), represent ten different human cDNAs (Table 1). An independent screen using Gal4 (BD)-human NP as a bait and a human HeLa cDNA prey library yielded a subset of these same host protein targets. Some of the identified clones contain partial cDNAs. Three independent prey plasmids isolated contained human cDNAs phosphoribosylaminoimidazole carboxylase, phosphoribosylaminoimidazole succinocarboxamide synthetase (PAICS) and another three contained Myb/SANT-like DNA-binding domain containing 3 (MSANTD3). Two independent prey plasmids were isolated for each of three human cDNAs: proteasome activator subunit 2 (PA28B), potassium channel tetramerisation domain (KCTD9), and an uncharacterized protein (FLJ30306). Each of the remaining five prey plasmids contained different human cDNAs. Evaluation of the strength of bait-prey interactions: The strength of interaction between each NP bait (swine or human) and human prey in yeast was measured using a qualitative growth assay and a quantitative β-galactosidase activity assay. First, yeast growth on selective media after cotransformation with swine or human NP and one of the identified prey was tested at both 30°C and 37°C as a qualitative metric of the protein-protein interaction’s strength (Figure 2, Table 2). All strains, including the positive and negative controls, grew on SD-Leu-Trp which selects for the presence of both bait and prey plasmids. Most strains showed decreased growth at 37°C relative to the more permissive temperature 30°C. Strains examined at 37°C which contained either swine or human NP bait grew best in the presence of prey CKAP5, MSANTD3, or TRA2B. In contrast, at 37°C strains containing swine NP grew poorly with prey ACOT13, whereas strains containing human NP grew poorly with prey PA28B.Figure 2 Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector).Table 2 Summary of affinity of bait-prey interactions Growth on SD-His agar plates* β-Gal activity** Gene Protein description Prey clone Human (30°C) Human (37°C) Swine (30°C) Swine (37°C) Human NP Swine NP PAICSPhosphoribosylaminoimidazole carboxylaseR15(++)(+)(++)(++)2.230.4MSANTD3Myb/SANT-like DNA-binding domain containing 3R10(++)(++)(++)(++)13.918.2FLJ30306PREDICTED: Homo sapiens uncharacterized LOC101059922, mRNAR23(+)(+)(++)(++)1.95.4PA28BProteasome (prosome, macropain) activator subunit 2R6(++)(−)(++)(+)6.733.3KCTD9Potassium channel tetramerisation domain containing 9R34(+)(+)(++)(++)5.016.4ACOT13Acyl-CoA thioesterase 13R11(++)(+)(++)(−)6.915.5TRA2BTransformer 2 beta homolog (Drosophila)R12(++)(++)(++)(++)1.90.85SLC30A9Solute carrier family 30 (zinc transporter), member 9R28(++)(+)(++)(+)82.9119.9ATP1B1ATPase, Na+/K+ transporting, beta 1 polypeptideR7(++)(+)(++)(++)3.44.3CKAP5Cytoskeleton associated protein 5R29(++)(++)(++)(++)70.190.3* Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2).** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). Maintenance of protein-protein interactions at 30°C and 37°C. Yeast cells containing both a bait plasmid and prey plasmid were spotted in duplicate onto agar plates and incubated at 30°C or 37°C. SD-Leu-Trp media serves as a growth control, and SD-His is protein interaction reporter media. Growth on SD-His is represented as normal growth (++), reduced growth (+), and no growth (−) (see Table 2). NC and PC represent negative and positive controls, respectively. Additional yeast strain negative controls include h- (human NP bait and empty prey vector) and s- (swine NP bait and empty prey vector). Summary of affinity of bait-prey interactions * Results from growth on SD-His protein interaction assay media are presented as: (++) normal growth, (+) reduced growth, and (-) no growth (Figure 2). ** Mean β-galalactosidase activity from at least eight replicates reported in Miller units (Figure 3). The strength of the interactions between IAV NP and host proteins identified by the library screen was also investigated by measuring the β-gal activity in liquid yeast cultures (Figure 3, Table 2). As observed with the qualitative growth assay, for all ten prey proteins, interactions were observed with both the swine and human NP. Of the ten interactions investigated, yeast strains containing swine or human NP and human prey SLC30A9 and CKAP3 showed the highest levels of β-gal activity compared to the other prey proteins examined. This is consistent with the growth of yeast strains containing human NP and SLC30A9 and CKAP3 on selective media at 37°C ( Figure 2, Table 2). Even strains with the lowest levels of β-gal activity observed (TRA2B and ATP1B1) have double the β-gal activity of the negative control. While the data suggest there may be differences in the strength of bait-prey interactions based on NP host origin, these differences should be interpreted carefully given the nature of the assay.Figure 3 β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. β-galactosidase activity in liquid culture. Β-gal activity is a quantitative measure of the strength of interaction between bait swine or human NP and prey human host proteins identified in the Y2H screen. The β-gal activity for each interacting bait-prey pair is expressed in Miller units and represents the mean and standard error of the mean of at least eight independent samples. Discussion: This Y2H screen identified multiple candidate host proteins that were previously identified in genome-wide screens as playing a role in the IAV infection cycle (PAICS, ATP1B1, SLC30A9, and TRA2B), while also identifying new potential interactors with IAV nucleoprotein (ACOT13, CKAP5, KCTD9, PA28B, MSANDTD3, and FLJ30306). These results complement previous reports by suggesting the mechanism for involvement of the host proteins with the IAV (i.e. interaction with the NP). The yeast screen is internally validated by the observation that half of the prey proteins identified were independently isolated multiple times during the screen. All host prey proteins identified interact in yeast with IAV NP from both human and swine isolates. Although there can be variability in gene expression among yeast colonies containing the same combination of bait/prey, both a qualitative growth assay and a quantitative β-gal assay yielded similar results in terms of the strength of interaction between IAV NP bait and human prey proteins. Notably, CKAP5 displayed good growth at the more stringent 37°C in the qualitative assay and also had high β-gal activity. The putative interacting host proteins represent a variety of cellular processes, consistent with the diversity of processes required during the IAV infection cycle. Phosphoribosylaminoimidazole carboxylase, phosphoribosylaminoimidazole succinocarboxamide synthetase (PAICS) is an enzyme necessary to catalyze the sixth and seventh steps of the de novo purine biosynthesis pathway [46]. PAICS has been identified by two separate RNAi screens as being crucial to IAV replication [40,47]. Furthermore, proteins involved in purine biosynthesis, including PAICS, were found to be significantly up-regulated during an IAV infection [48]. The data presented herein suggest that the requirement for PAICS during IAV infection may be through the interaction with NP. The sodium/potassium-transporting ATPase subunit beta-1 (ATP1B1) represents another host protein that interacts in yeast with NP and that has been previously shown to be important for influenza virus replication in host cells. It belongs to the family of Na+/K + −ATPase beta chain proteins. Na+/K + −ATPase is necessary for the creation and maintenance of electrochemical gradients of sodium and potassium ions on either side of the plasma membrane [49]. This gradient is used by cells to permit functional nervous and muscular excitability, osmoregulation, and active transport of numerous molecules [49]. The beta subunit functions to regulate the quantity of sodium pumps transported to the plasma membrane [49]. Using a quantitative proteomics approach, ATP1B1 was one of 43 proteins found to be significantly up-regulated in IAV-infected primary human alveolar macrophages [50]. In a Y2H study, Mi and colleagues [51] isolated ATP1B1 as a protein that binds the cytoplasmic domain of IAV M2 and subsequently showed that knockdown of ATP1B1 in MDCK cells suppressed IAV replication [51]. These studies suggest that ATP1B1 may be interacting with multiple IAV proteins. The solute carrier family 30 (zinc transporter), member 9 (SLC30A9) is involved in nucleotide-excision repair of human DNA [52] and has an efflux motif characteristic of proteins in the SLC30 zinc efflux transporter family [53]. SLC30A9 protein found in the cytoplasm can bind nuclear receptors and/or nuclear receptor coactivators and translocate to the nucleus where it regulates gene transcription [54]. SLC30A9 was identified as a possible host target that confers upon the host cell resistance to IAV infection [44]. How SLC30A9 might interact with NP and be necessary for IAV replication is unclear. Another protein identified by this screen is TRA2B, a tissue-specific splicing factor [55]. TRA2B protein binds to AGAA and CAA RNA sequences affecting the inclusion of introns in the processed transcript [55]. With respect to influenza, expression of TRA2B protein was reported to be down-regulated in porcine alveolar macrophage cells infected by two swine IAVs [56]. Because influenza genes M2 and NS2 are spliced, it is possible that TRA2B is important for the splicing of these genes. However, how NP might be involved in TRA2B’s function in IAV replication is unknown. Another binding partner identified is Acyl CoA thioesterase 13 (ACOT13), a eukaryotic protein also known as thioesterase superfamily member 2 (Them2) [57,58]. ACOT13 catalyzes fatty acyl-CoA hydrolysis that exists primarily in association with mitochondria [57]. ACOT13/Them2 was identified in a Y2H screen focused on phosphatidylcholine transfer protein, suggestive of a role in fatty acid metabolism [57]. Although ACOT13 has not previously been reported to play a role in the IAV life cycle, another member of the ACOT family (ACOT9) which has two hot dog fold domains (ACOT13 has one) [59] has been shown to physically interact with the IAV PA protein [28]. How ACOT13 may be interacting with NP during IAV replication is unclear. Cytoskeletal associated protein 5 (CKAP5, also known as TOG) binds microtubules and is important for the process of centrosomal microtubule assembly especially during mitosis [60]. The microtubule network of the cell plays a role in movement of materials throughout the cell, including during infection [61]. TOG2 binds to a ribonucleoprotein and plays a role in RNA trafficking [62]. While a specific role for CKAP5 in the IAV life cycle is unknown, a genome-wide siRNA screen revealed CKAP5 as one of 96 human genes that supports replication of another RNA virus- the hepatitis C virus [63]. It’s possible that CKAP5 functions during the IAV infection cycle to organize microtubules for trafficking of viral components. Another interactor identified in this screen, KCTD9, has been previously shown to interact with a subunit of the Mediator complex, acting as a scaffold between regulatory proteins and the RNA polymerase II [64]. In addition, KCTD9 homolog FIP2 acts as a cytoskeletal rearrangement protein suspected to be involved in nuclear export in some plants through the organization of actin cables [64]. Perhaps KCTD9 aids in translocating NP to the nucleus for replication and transcription of the viral RNA. It has been shown that the translocation of NF-E2 related factor 2 (Nrf2) is mediated by cytoskeletal rearrangement in the oxidative stress response [65], and the level of Nrf2 expression is associated with IAV entry and replication in nasal epithelial cells [66]. Also identified by this screen was PA28B, a protein associated with the proteasome complex. In the typical proteasome, there is a 19S regulator, which is replaced by the 11S regulator in the modified immunoproteasome [67]. PA28B is part of the proteasome complex and acts as a subunit in the 11S alternate regulator. It is referred to as the beta subunit and is comprised of three beta and three alpha subunits arranged in a heterohexameric ring [68]. The proteasome, as a whole, is used for the cleaving of peptides and the 11S regulator, in particular, induces the degradation of short peptides, but not entire proteins [68]. Induction of 11S regulator expression is the responsibility of interferon gamma. PA28B is involved in cleavage of the peptides which bind to the major histocompatibility complex (MHC). It is possible that NP interacts with PA28B to prevent MHC I presentation of IAV antigens. Myb/SANT-like DNA-binding domain containing 3 (MSANTD3) is a member of the MSANTD3 family which contains DNA binding domains for Myb proteins and the SANT domain family. Myb proteins are proto-oncogenes which are necessary for hematopoiesis and possibly involved in tumorigenesis [69]. The SANT domain permits the interaction of chromatin remodeling proteins with histones and is found in chromatin-remodeling complexes as well as nuclear receptor co-repressors [70]. NP interactions with MSANTD3 could have an effect on gene transcription of the host through chromatin modification, presumably to inhibit antiviral genes or up-regulate genes beneficial to viral replication. The screen also identified FLJ30306 as a potential interactor with NP. FLJ30306 maintains a sequence similarity to retroviral elements, resulting in its classification as endogenous retrovirus group K3, member 1 [Homo sapiens] [71]. The endogenous retrovirus sequence was revealed by sequence analysis of the mRNA, not by direct protein sequencing. It is unclear whether the protein carries out a normal cellular effect, such as endogenous retrovirus 1’s (ERV-1) use in placental syncytia formation or if it is a result of mis-regulation due to influenza infection [71]. In Hodgkin’s lymphoma it has been shown that many endogenous retroviruses are reactivated and produce virus-like particles [71]. It may be that instead of directly utilizing FLJ30306 influenza may simply be up-regulating its expression due to its pathogenesis. Alternatively, FLJ30306 may contain useful enzymatic activity for influenza replication through an unknown mechanism. A characterization of normal FLJ30306 function would allow for appropriately testing either hypothesis. All human host proteins identified in this Y2H screen interact in yeast with NP from both swine and human IAV strains. While the data suggest there may be differences in the strength of bait-prey interactions in yeast based on NP host origin, these differences should be interpreted carefully given the nature of the assay. Given the suggestion that NP affects species specificity, additional screens using NP from other IAV strains as the bait and/or cDNA libraries from other susceptible species may shed light on amino acids within the NP and specific host factors that are responsible for host restrictions. As is the case with other published screens (e.g., [72]), it will be important to verify the putative interactions described here in infected cells by co-immunoprecipitation, co-localization, or RNA interference (e.g., [34]) and to investigate the role of the identified host proteins in the IAV life cycle. Conclusions: A Y2H screen using a classical swine influenza A virus nucleoprotein as a bait resulted in isolation of ten different putative interacting human host proteins involved in a variety of cellular processes and structures: purine biosynthesis (PAICS), metabolism (ACOT13), proteasome (PA28B), DNA-binding (MSANTD3), cytoskeleton (CKAP5), potassium channel formation (KCTD9), zinc transporter function (SLC30A9), Na+/K+ ATPase function (ATP1B1), and RNA splicing (TRA2B). All of the identified human host proteins interacted in yeast with NP from both human and swine origin. Proteins PAICS, ATP1B1, TRA2B, and SLC30A9 have been previously identified in studies related to host response to the IAV. Host proteins SLC30A9 and CKAP5 displayed the strongest interactions with IAV NP in yeast in a quantitative β-gal assay. Methods: Strains, plasmids, media, microbial growth conditions, and reagents Unless indicated otherwise, yeast strains (Table 3) were grown in liquid culture at 30°C at 180 rpm and on agar plates at 30°C. Bacterial strains were grown in liquid culture at 37°C at 200 rpm and on agar plates at 37°C. Selectable markers for the bait (pGBKT7) and prey (pGADT7) plasmids in the yeast Saccharomyces cerevisiae and bacterium E. coli are listed in Table 3. All media and Y2H reagents were purchased from Clontech (Mountain View, CA).Table 3 Strains and plasmids Yeast strain Genotype Reporter genes Y2HGold MATa, trp1-901, leu2-3, 112, ura3-52, his3-200, gal4Δ, gal80Δ, LYS2::GAL1UAS–Gal1TATA–His3, GAL2UAS–Gal2TATA–Ade2 URA3::MEL1UAS–Mel1TATA, AUR1-C MEL1 ADE2, HIS3, MEL1, AUR1-C Y187 ade2-101, trp1-901, leu2-3, 112, gal4Δ, gal80Δ, met–, URA3::GAL1UAS–Gal1TATA–LacZ, MEL1 MEL1, LacZ Plasmid Selectable markers (yeast, bacteria) Additional information Plasmid pGBKT7 (bait vector) TRP1, Kan GAL4(1–147)DNA-BD, TRP1, kanr, c-Myc epitope tagPlasmid pGADT7 (prey vector) LEU2, Amp GAL4(768–881)AD, LEU2, ampr, HA epitope tag Strains and plasmids Unless indicated otherwise, yeast strains (Table 3) were grown in liquid culture at 30°C at 180 rpm and on agar plates at 30°C. Bacterial strains were grown in liquid culture at 37°C at 200 rpm and on agar plates at 37°C. Selectable markers for the bait (pGBKT7) and prey (pGADT7) plasmids in the yeast Saccharomyces cerevisiae and bacterium E. coli are listed in Table 3. All media and Y2H reagents were purchased from Clontech (Mountain View, CA).Table 3 Strains and plasmids Yeast strain Genotype Reporter genes Y2HGold MATa, trp1-901, leu2-3, 112, ura3-52, his3-200, gal4Δ, gal80Δ, LYS2::GAL1UAS–Gal1TATA–His3, GAL2UAS–Gal2TATA–Ade2 URA3::MEL1UAS–Mel1TATA, AUR1-C MEL1 ADE2, HIS3, MEL1, AUR1-C Y187 ade2-101, trp1-901, leu2-3, 112, gal4Δ, gal80Δ, met–, URA3::GAL1UAS–Gal1TATA–LacZ, MEL1 MEL1, LacZ Plasmid Selectable markers (yeast, bacteria) Additional information Plasmid pGBKT7 (bait vector) TRP1, Kan GAL4(1–147)DNA-BD, TRP1, kanr, c-Myc epitope tagPlasmid pGADT7 (prey vector) LEU2, Amp GAL4(768–881)AD, LEU2, ampr, HA epitope tag Strains and plasmids Cloning Influenza A virus nucleoprotein into vector pGBKT7 The open reading frame (ORF) of the nucleoprotein gene from swine IAV H1N1 strain A/Sw/NC/44173/00 was amplified from the plasmid pScript A/Sw/NC/44173/00 NP (a gift from Christopher Olsen) by PCR using forward primer NP3F 5’-CATGGAGGCCGAATTCatggcgtctcaaggcaccaaacga-3’ and reverse primer NP2R 5’-GCAGGTCGACGGATCCattgtcatactcctctgcattgtctccgaaga-3’. The NP ORF from the human IAV H3N2 strain A/Ca/07/04 was amplified from plasmid pScript A/Ca/07/04 NP (a gift from Christopher Olsen) by PCR using forward primer NP2F 5’-CATGGAGGCCGAATTC atggcgtcccaaggcaccaaacg-3’ and reverse primer NP3R 5’-GCAGGTCGACGGATCC attgtcgtactcttctgcattgtctccgaaga-3’. For all primers the uppercase letters represent sequences homologous to the BamHI-EcoRI-linearized vector pGBKT7, and lower case letters represent sequences specific for the NP ORF. Reactions included 200 ng plasmid template DNA, 1X Advantage HD Buffer (Clontech, Mountain View, CA), 0.2 mM dNTPs, 0.25 μM forward and reverse primers, and 0.625 Units Advantage HD Polymerase (Clontech, Mountain View, CA). Thermocycler conditions were as follows: 95°C for 3 minutes; 30 cycles of [95°C for 15 seconds, 55°C for 5 seconds, 72°C for 100 seconds], 72°C for 10 minutes. InFusion cloning (Clontech, Mountain View, CA) was used to insert the NP ORFs into BamHI-EcoRI-linearized pGBKT7 3’ and in-frame with DNA encoding Gal4’s DNA binding domain (BD). Conditions were as follows: 60 ng gel-purified BamHI-EcoRI-linearized pGBKT7, 50 ng gel-purified NP PCR product, and 1X InFusion HD Enzyme Premix in a 5 μl reaction were incubated at 50°C for 15 min. An aliquot of the cloning reaction was transformed into Stellar competent cells per the manufacturer’s directions (Clontech, Mountain View, CA). All plasmid sequences were verified by DNA sequencing. The open reading frame (ORF) of the nucleoprotein gene from swine IAV H1N1 strain A/Sw/NC/44173/00 was amplified from the plasmid pScript A/Sw/NC/44173/00 NP (a gift from Christopher Olsen) by PCR using forward primer NP3F 5’-CATGGAGGCCGAATTCatggcgtctcaaggcaccaaacga-3’ and reverse primer NP2R 5’-GCAGGTCGACGGATCCattgtcatactcctctgcattgtctccgaaga-3’. The NP ORF from the human IAV H3N2 strain A/Ca/07/04 was amplified from plasmid pScript A/Ca/07/04 NP (a gift from Christopher Olsen) by PCR using forward primer NP2F 5’-CATGGAGGCCGAATTC atggcgtcccaaggcaccaaacg-3’ and reverse primer NP3R 5’-GCAGGTCGACGGATCC attgtcgtactcttctgcattgtctccgaaga-3’. For all primers the uppercase letters represent sequences homologous to the BamHI-EcoRI-linearized vector pGBKT7, and lower case letters represent sequences specific for the NP ORF. Reactions included 200 ng plasmid template DNA, 1X Advantage HD Buffer (Clontech, Mountain View, CA), 0.2 mM dNTPs, 0.25 μM forward and reverse primers, and 0.625 Units Advantage HD Polymerase (Clontech, Mountain View, CA). Thermocycler conditions were as follows: 95°C for 3 minutes; 30 cycles of [95°C for 15 seconds, 55°C for 5 seconds, 72°C for 100 seconds], 72°C for 10 minutes. InFusion cloning (Clontech, Mountain View, CA) was used to insert the NP ORFs into BamHI-EcoRI-linearized pGBKT7 3’ and in-frame with DNA encoding Gal4’s DNA binding domain (BD). Conditions were as follows: 60 ng gel-purified BamHI-EcoRI-linearized pGBKT7, 50 ng gel-purified NP PCR product, and 1X InFusion HD Enzyme Premix in a 5 μl reaction were incubated at 50°C for 15 min. An aliquot of the cloning reaction was transformed into Stellar competent cells per the manufacturer’s directions (Clontech, Mountain View, CA). All plasmid sequences were verified by DNA sequencing. Yeast two-hybrid screen The Yeast Two-Hybrid MatchMaker Gold system (Clontech, Mountain View, CA) was used to isolate prey plasmids that interact with bait pGBKT7-NP. A 50 ml SD-TRP liquid media culture inoculated with Y2HGold cells containing the bait plasmid pGBKT7-NP was grown in a 200 rpm shaker at 30 degrees to an OD600 of 0.8 (approximately 18 hours). Bait cells were pelleted by centrifugation, resuspended in 4 ml SD-TRP, and mixed with 45 ml YPDA in a 2 L flask with 2 × 107 Y187 cells containing a commercially-available normalized HeLa S3 Mate and Plate cDNA prey library (Clontech, Mountain View, CA). Bait and prey cells were mated by slow shaking (40 rpm) for 24 hours at 30°C. After 20 hours, mating cells were observed microscopically for the presence of zygotes. Mated cells were centrifuged 1,000 × g for 10 min, and the pellet was resuspended in 5 ml 0.5X YPDA containing 50 μg/ml kanamycin. Mated cells (0.2 ml) were spread on each of 55 SD-LEU-TRP/X-a-Gal/125 ng/ml AbA agar plates, and plates were incubated 30°C for 5–8 days. This concentration of AbA is considered high stringency and could exclude isolation of low-affinity interactors. The mating efficiency was determined by plating 1:1,000 and 1:10,000 dilutions on SD-LEU, SD-TRP, and SD-LEU-TRP agar plates. Blue colonies from the Y2H screen SD-LEU-TRP/X-a-Gal/AbA plates were picked as small patches to an SD-LEU-TRP plate, grown 2 days at 30°C, and replica plated to agar media (SD-ADE, SD-HIS, SD-LEU-TRP/X-a-Gal, SD-LEU-TRP/AbA) to test for activation of four reporter genes: ADE2, HIS3, MEL1C, and AUR1-C, respectively. Replica plates were incubated at 30°C and observed daily for 3 days. Cells from potential positive interactors (growing on all types of reporter media and blue/light blue on SD-LEU-TRP/X-α-Gal) were single-colony purified on SD-LEU-TRP/X-a-Gal and incubated at 30°C for 4 days. A single blue colony from each potential positive interaction strain was single-colony purified a second time on SD-LEU-TRP/X-α-Gal. Following two rounds of streaking, the prey plasmid from each potential positive interaction yeast strain was isolated using the “Easy Yeast Plasmid Isolation Kit” (Clontech, Mountain View, CA). The resulting yeast plasmid DNAs were transformed into and subsequently re-isolated from E. coli cells. The Yeast Two-Hybrid MatchMaker Gold system (Clontech, Mountain View, CA) was used to isolate prey plasmids that interact with bait pGBKT7-NP. A 50 ml SD-TRP liquid media culture inoculated with Y2HGold cells containing the bait plasmid pGBKT7-NP was grown in a 200 rpm shaker at 30 degrees to an OD600 of 0.8 (approximately 18 hours). Bait cells were pelleted by centrifugation, resuspended in 4 ml SD-TRP, and mixed with 45 ml YPDA in a 2 L flask with 2 × 107 Y187 cells containing a commercially-available normalized HeLa S3 Mate and Plate cDNA prey library (Clontech, Mountain View, CA). Bait and prey cells were mated by slow shaking (40 rpm) for 24 hours at 30°C. After 20 hours, mating cells were observed microscopically for the presence of zygotes. Mated cells were centrifuged 1,000 × g for 10 min, and the pellet was resuspended in 5 ml 0.5X YPDA containing 50 μg/ml kanamycin. Mated cells (0.2 ml) were spread on each of 55 SD-LEU-TRP/X-a-Gal/125 ng/ml AbA agar plates, and plates were incubated 30°C for 5–8 days. This concentration of AbA is considered high stringency and could exclude isolation of low-affinity interactors. The mating efficiency was determined by plating 1:1,000 and 1:10,000 dilutions on SD-LEU, SD-TRP, and SD-LEU-TRP agar plates. Blue colonies from the Y2H screen SD-LEU-TRP/X-a-Gal/AbA plates were picked as small patches to an SD-LEU-TRP plate, grown 2 days at 30°C, and replica plated to agar media (SD-ADE, SD-HIS, SD-LEU-TRP/X-a-Gal, SD-LEU-TRP/AbA) to test for activation of four reporter genes: ADE2, HIS3, MEL1C, and AUR1-C, respectively. Replica plates were incubated at 30°C and observed daily for 3 days. Cells from potential positive interactors (growing on all types of reporter media and blue/light blue on SD-LEU-TRP/X-α-Gal) were single-colony purified on SD-LEU-TRP/X-a-Gal and incubated at 30°C for 4 days. A single blue colony from each potential positive interaction strain was single-colony purified a second time on SD-LEU-TRP/X-α-Gal. Following two rounds of streaking, the prey plasmid from each potential positive interaction yeast strain was isolated using the “Easy Yeast Plasmid Isolation Kit” (Clontech, Mountain View, CA). The resulting yeast plasmid DNAs were transformed into and subsequently re-isolated from E. coli cells. Testing potential positive prey plasmids for toxicity to yeast and autoactivation of the Y2H reporter genes Isolated prey plasmid DNAs were re-introduced into Y2HGold cells using a yeast transformation system (Geno Technology). Y2HGold cells containing prey plasmids were single-colony purified on SD-Leu agar plates alongside a control (Y2HGold containing the empty prey vector pGADT7), and cell growth was observed after 3 days at 30C to test for toxicity. To test for false positives, patches of Y2HGold cells containing prey plasmids were replica plated to assess autoactivation of reporter genes on reporter media (SD-Ade, SD-His, SD-Leu-Trp + X-α-Gal, SD-Leu-Trp + AbA). Prey strains determined to be false positives in yeast were excluded from the pool of identified interacting host proteins. Isolated prey plasmid DNAs were re-introduced into Y2HGold cells using a yeast transformation system (Geno Technology). Y2HGold cells containing prey plasmids were single-colony purified on SD-Leu agar plates alongside a control (Y2HGold containing the empty prey vector pGADT7), and cell growth was observed after 3 days at 30C to test for toxicity. To test for false positives, patches of Y2HGold cells containing prey plasmids were replica plated to assess autoactivation of reporter genes on reporter media (SD-Ade, SD-His, SD-Leu-Trp + X-α-Gal, SD-Leu-Trp + AbA). Prey strains determined to be false positives in yeast were excluded from the pool of identified interacting host proteins. Bioinformatics The 5’ and 3’ ends of human cDNA within each confirmed positive prey plasmid were determined by sequencing at the Iowa State University (ISU) DNA Facility using primers T7-1 (AATACGACTCACTATAG) and 3’AD (AGATGGTGCACGATGCACAG) or 3’ poly-TA (TTTTTTTTTTTTTTTTTTTTTTTTTA), respectively. The sequence data was used in a BLAST to query the NCBI human genome and human transcript databases. The 5’ and 3’ ends of human cDNA within each confirmed positive prey plasmid were determined by sequencing at the Iowa State University (ISU) DNA Facility using primers T7-1 (AATACGACTCACTATAG) and 3’AD (AGATGGTGCACGATGCACAG) or 3’ poly-TA (TTTTTTTTTTTTTTTTTTTTTTTTTA), respectively. The sequence data was used in a BLAST to query the NCBI human genome and human transcript databases. Maintenance of protein-protein interactions at 30°C and 37°C Y2HGold yeast cells were separately co-transformed with a bait plasmid (either pGBKT7 bait vector containing NP from human influenza A strain H3N2, A/Ca/07/04 or NP from swine influenza A strain H1N1, A/Sw/NC/44173/00) and a prey plasmid (pGADT7 containing one of the human cDNAs isolated as a positive interactor with human NP bait). Y2HGold cells containing known bait/prey interactors pGBKT7-53 and pGADT7-T and known non-interactors pGBKT7-Lam and pGADT7-T were used as positive and negative controls, respectively. Additional negative interaction controls included Y2HGold cells containing human or swine NP (described above) and an empty prey vector pGADT7. All transformants were grown in YPD media in overnight cultures and were diluted to a concentration of 5 × 106 cells/ml (OD = 0.417). 200 ml of culture was placed into a 96 well plate. A 48 sample pinner was used to spot yeast onto 2 plates each of SD-Leu-Trp (control), and SD-His. One plate of each type was incubated at 30°C, the other was incubated at 37°C. Growth was assessed and images taken 4 days after the initial plating procedure. At least two replicates of each strain containing the bait (swine NP or human NP) and prey were tested on SD-Leu-Trp and SD-His at both 30°C and 37°C. Y2HGold yeast cells were separately co-transformed with a bait plasmid (either pGBKT7 bait vector containing NP from human influenza A strain H3N2, A/Ca/07/04 or NP from swine influenza A strain H1N1, A/Sw/NC/44173/00) and a prey plasmid (pGADT7 containing one of the human cDNAs isolated as a positive interactor with human NP bait). Y2HGold cells containing known bait/prey interactors pGBKT7-53 and pGADT7-T and known non-interactors pGBKT7-Lam and pGADT7-T were used as positive and negative controls, respectively. Additional negative interaction controls included Y2HGold cells containing human or swine NP (described above) and an empty prey vector pGADT7. All transformants were grown in YPD media in overnight cultures and were diluted to a concentration of 5 × 106 cells/ml (OD = 0.417). 200 ml of culture was placed into a 96 well plate. A 48 sample pinner was used to spot yeast onto 2 plates each of SD-Leu-Trp (control), and SD-His. One plate of each type was incubated at 30°C, the other was incubated at 37°C. Growth was assessed and images taken 4 days after the initial plating procedure. At least two replicates of each strain containing the bait (swine NP or human NP) and prey were tested on SD-Leu-Trp and SD-His at both 30°C and 37°C. ß-galactosidase assay Diploid cells containing bait and prey were prepared by mating Y187 cells containing each of the prey plasmids with Y2HGold cells containing each of the bait plasmids (swine NP or human NP) on YPD agar (MatchMaker Gold Y2H protocol, Clontech, Mountainview, CA). Eight replicate samples of diploids were selected by subsequent growth on SD-Leu-Trp agar for each protein interaction. Diploid cells containing bait/prey were inoculated in 2 mL SD-Leu/Trp liquid media and grown in a 200 rpm shaker at 30°C overnight. 2 mL of the overnight cultures were diluted to 4 mL with YPD media and grown in a 200 rpm shaker at 30°C until they reached mid-log phase (OD600 of 0.5-0.8). Cultures were separated into 1.5 mL aliquots, pelleted, washed once, and then resuspended in 300 μL of Z buffer (16.1 g/L Na2HPO4⋅7H2O, 5.5 g/L NaH2PO4⋅H2O, 0.75 g/L KCl, 0.246 g/L MgSO4⋅7H2O, pH7.0). Samples of 100 μL were lysed by three freeze/thaw cycles with liquid nitrogen and a 37°C water bath. After the cells were lysed, the sample was split into two 150 μL aliquots. A blank tube, containing 100uL of Z buffer, was also set up. To analyze the enzyme activity, 350 μL of Z buffer with 0.27% β-mercaptoethanol and 4 mg/ml of ONPG as substrate were added. The yellow color was then allowed to develop at 30°C, and the reaction was stopped by the addition of 200 μL of 1 M Na2CO3. To remove cell debris, the samples were centrifuged for 10 minutes at 16,000 RCF, and OD420 values were recorded. In order to express the enzyme activity, Miller Units were calculated using the formula: 1000*OD420/(t*V*OD600) where t refers to the elapsed time, and V refers to the dilution factor (0.25 in this case). Diploid cells containing bait and prey were prepared by mating Y187 cells containing each of the prey plasmids with Y2HGold cells containing each of the bait plasmids (swine NP or human NP) on YPD agar (MatchMaker Gold Y2H protocol, Clontech, Mountainview, CA). Eight replicate samples of diploids were selected by subsequent growth on SD-Leu-Trp agar for each protein interaction. Diploid cells containing bait/prey were inoculated in 2 mL SD-Leu/Trp liquid media and grown in a 200 rpm shaker at 30°C overnight. 2 mL of the overnight cultures were diluted to 4 mL with YPD media and grown in a 200 rpm shaker at 30°C until they reached mid-log phase (OD600 of 0.5-0.8). Cultures were separated into 1.5 mL aliquots, pelleted, washed once, and then resuspended in 300 μL of Z buffer (16.1 g/L Na2HPO4⋅7H2O, 5.5 g/L NaH2PO4⋅H2O, 0.75 g/L KCl, 0.246 g/L MgSO4⋅7H2O, pH7.0). Samples of 100 μL were lysed by three freeze/thaw cycles with liquid nitrogen and a 37°C water bath. After the cells were lysed, the sample was split into two 150 μL aliquots. A blank tube, containing 100uL of Z buffer, was also set up. To analyze the enzyme activity, 350 μL of Z buffer with 0.27% β-mercaptoethanol and 4 mg/ml of ONPG as substrate were added. The yellow color was then allowed to develop at 30°C, and the reaction was stopped by the addition of 200 μL of 1 M Na2CO3. To remove cell debris, the samples were centrifuged for 10 minutes at 16,000 RCF, and OD420 values were recorded. In order to express the enzyme activity, Miller Units were calculated using the formula: 1000*OD420/(t*V*OD600) where t refers to the elapsed time, and V refers to the dilution factor (0.25 in this case). Strains, plasmids, media, microbial growth conditions, and reagents: Unless indicated otherwise, yeast strains (Table 3) were grown in liquid culture at 30°C at 180 rpm and on agar plates at 30°C. Bacterial strains were grown in liquid culture at 37°C at 200 rpm and on agar plates at 37°C. Selectable markers for the bait (pGBKT7) and prey (pGADT7) plasmids in the yeast Saccharomyces cerevisiae and bacterium E. coli are listed in Table 3. All media and Y2H reagents were purchased from Clontech (Mountain View, CA).Table 3 Strains and plasmids Yeast strain Genotype Reporter genes Y2HGold MATa, trp1-901, leu2-3, 112, ura3-52, his3-200, gal4Δ, gal80Δ, LYS2::GAL1UAS–Gal1TATA–His3, GAL2UAS–Gal2TATA–Ade2 URA3::MEL1UAS–Mel1TATA, AUR1-C MEL1 ADE2, HIS3, MEL1, AUR1-C Y187 ade2-101, trp1-901, leu2-3, 112, gal4Δ, gal80Δ, met–, URA3::GAL1UAS–Gal1TATA–LacZ, MEL1 MEL1, LacZ Plasmid Selectable markers (yeast, bacteria) Additional information Plasmid pGBKT7 (bait vector) TRP1, Kan GAL4(1–147)DNA-BD, TRP1, kanr, c-Myc epitope tagPlasmid pGADT7 (prey vector) LEU2, Amp GAL4(768–881)AD, LEU2, ampr, HA epitope tag Strains and plasmids Cloning Influenza A virus nucleoprotein into vector pGBKT7: The open reading frame (ORF) of the nucleoprotein gene from swine IAV H1N1 strain A/Sw/NC/44173/00 was amplified from the plasmid pScript A/Sw/NC/44173/00 NP (a gift from Christopher Olsen) by PCR using forward primer NP3F 5’-CATGGAGGCCGAATTCatggcgtctcaaggcaccaaacga-3’ and reverse primer NP2R 5’-GCAGGTCGACGGATCCattgtcatactcctctgcattgtctccgaaga-3’. The NP ORF from the human IAV H3N2 strain A/Ca/07/04 was amplified from plasmid pScript A/Ca/07/04 NP (a gift from Christopher Olsen) by PCR using forward primer NP2F 5’-CATGGAGGCCGAATTC atggcgtcccaaggcaccaaacg-3’ and reverse primer NP3R 5’-GCAGGTCGACGGATCC attgtcgtactcttctgcattgtctccgaaga-3’. For all primers the uppercase letters represent sequences homologous to the BamHI-EcoRI-linearized vector pGBKT7, and lower case letters represent sequences specific for the NP ORF. Reactions included 200 ng plasmid template DNA, 1X Advantage HD Buffer (Clontech, Mountain View, CA), 0.2 mM dNTPs, 0.25 μM forward and reverse primers, and 0.625 Units Advantage HD Polymerase (Clontech, Mountain View, CA). Thermocycler conditions were as follows: 95°C for 3 minutes; 30 cycles of [95°C for 15 seconds, 55°C for 5 seconds, 72°C for 100 seconds], 72°C for 10 minutes. InFusion cloning (Clontech, Mountain View, CA) was used to insert the NP ORFs into BamHI-EcoRI-linearized pGBKT7 3’ and in-frame with DNA encoding Gal4’s DNA binding domain (BD). Conditions were as follows: 60 ng gel-purified BamHI-EcoRI-linearized pGBKT7, 50 ng gel-purified NP PCR product, and 1X InFusion HD Enzyme Premix in a 5 μl reaction were incubated at 50°C for 15 min. An aliquot of the cloning reaction was transformed into Stellar competent cells per the manufacturer’s directions (Clontech, Mountain View, CA). All plasmid sequences were verified by DNA sequencing. Yeast two-hybrid screen: The Yeast Two-Hybrid MatchMaker Gold system (Clontech, Mountain View, CA) was used to isolate prey plasmids that interact with bait pGBKT7-NP. A 50 ml SD-TRP liquid media culture inoculated with Y2HGold cells containing the bait plasmid pGBKT7-NP was grown in a 200 rpm shaker at 30 degrees to an OD600 of 0.8 (approximately 18 hours). Bait cells were pelleted by centrifugation, resuspended in 4 ml SD-TRP, and mixed with 45 ml YPDA in a 2 L flask with 2 × 107 Y187 cells containing a commercially-available normalized HeLa S3 Mate and Plate cDNA prey library (Clontech, Mountain View, CA). Bait and prey cells were mated by slow shaking (40 rpm) for 24 hours at 30°C. After 20 hours, mating cells were observed microscopically for the presence of zygotes. Mated cells were centrifuged 1,000 × g for 10 min, and the pellet was resuspended in 5 ml 0.5X YPDA containing 50 μg/ml kanamycin. Mated cells (0.2 ml) were spread on each of 55 SD-LEU-TRP/X-a-Gal/125 ng/ml AbA agar plates, and plates were incubated 30°C for 5–8 days. This concentration of AbA is considered high stringency and could exclude isolation of low-affinity interactors. The mating efficiency was determined by plating 1:1,000 and 1:10,000 dilutions on SD-LEU, SD-TRP, and SD-LEU-TRP agar plates. Blue colonies from the Y2H screen SD-LEU-TRP/X-a-Gal/AbA plates were picked as small patches to an SD-LEU-TRP plate, grown 2 days at 30°C, and replica plated to agar media (SD-ADE, SD-HIS, SD-LEU-TRP/X-a-Gal, SD-LEU-TRP/AbA) to test for activation of four reporter genes: ADE2, HIS3, MEL1C, and AUR1-C, respectively. Replica plates were incubated at 30°C and observed daily for 3 days. Cells from potential positive interactors (growing on all types of reporter media and blue/light blue on SD-LEU-TRP/X-α-Gal) were single-colony purified on SD-LEU-TRP/X-a-Gal and incubated at 30°C for 4 days. A single blue colony from each potential positive interaction strain was single-colony purified a second time on SD-LEU-TRP/X-α-Gal. Following two rounds of streaking, the prey plasmid from each potential positive interaction yeast strain was isolated using the “Easy Yeast Plasmid Isolation Kit” (Clontech, Mountain View, CA). The resulting yeast plasmid DNAs were transformed into and subsequently re-isolated from E. coli cells. Testing potential positive prey plasmids for toxicity to yeast and autoactivation of the Y2H reporter genes: Isolated prey plasmid DNAs were re-introduced into Y2HGold cells using a yeast transformation system (Geno Technology). Y2HGold cells containing prey plasmids were single-colony purified on SD-Leu agar plates alongside a control (Y2HGold containing the empty prey vector pGADT7), and cell growth was observed after 3 days at 30C to test for toxicity. To test for false positives, patches of Y2HGold cells containing prey plasmids were replica plated to assess autoactivation of reporter genes on reporter media (SD-Ade, SD-His, SD-Leu-Trp + X-α-Gal, SD-Leu-Trp + AbA). Prey strains determined to be false positives in yeast were excluded from the pool of identified interacting host proteins. Bioinformatics: The 5’ and 3’ ends of human cDNA within each confirmed positive prey plasmid were determined by sequencing at the Iowa State University (ISU) DNA Facility using primers T7-1 (AATACGACTCACTATAG) and 3’AD (AGATGGTGCACGATGCACAG) or 3’ poly-TA (TTTTTTTTTTTTTTTTTTTTTTTTTA), respectively. The sequence data was used in a BLAST to query the NCBI human genome and human transcript databases. Maintenance of protein-protein interactions at 30°C and 37°C: Y2HGold yeast cells were separately co-transformed with a bait plasmid (either pGBKT7 bait vector containing NP from human influenza A strain H3N2, A/Ca/07/04 or NP from swine influenza A strain H1N1, A/Sw/NC/44173/00) and a prey plasmid (pGADT7 containing one of the human cDNAs isolated as a positive interactor with human NP bait). Y2HGold cells containing known bait/prey interactors pGBKT7-53 and pGADT7-T and known non-interactors pGBKT7-Lam and pGADT7-T were used as positive and negative controls, respectively. Additional negative interaction controls included Y2HGold cells containing human or swine NP (described above) and an empty prey vector pGADT7. All transformants were grown in YPD media in overnight cultures and were diluted to a concentration of 5 × 106 cells/ml (OD = 0.417). 200 ml of culture was placed into a 96 well plate. A 48 sample pinner was used to spot yeast onto 2 plates each of SD-Leu-Trp (control), and SD-His. One plate of each type was incubated at 30°C, the other was incubated at 37°C. Growth was assessed and images taken 4 days after the initial plating procedure. At least two replicates of each strain containing the bait (swine NP or human NP) and prey were tested on SD-Leu-Trp and SD-His at both 30°C and 37°C. ß-galactosidase assay: Diploid cells containing bait and prey were prepared by mating Y187 cells containing each of the prey plasmids with Y2HGold cells containing each of the bait plasmids (swine NP or human NP) on YPD agar (MatchMaker Gold Y2H protocol, Clontech, Mountainview, CA). Eight replicate samples of diploids were selected by subsequent growth on SD-Leu-Trp agar for each protein interaction. Diploid cells containing bait/prey were inoculated in 2 mL SD-Leu/Trp liquid media and grown in a 200 rpm shaker at 30°C overnight. 2 mL of the overnight cultures were diluted to 4 mL with YPD media and grown in a 200 rpm shaker at 30°C until they reached mid-log phase (OD600 of 0.5-0.8). Cultures were separated into 1.5 mL aliquots, pelleted, washed once, and then resuspended in 300 μL of Z buffer (16.1 g/L Na2HPO4⋅7H2O, 5.5 g/L NaH2PO4⋅H2O, 0.75 g/L KCl, 0.246 g/L MgSO4⋅7H2O, pH7.0). Samples of 100 μL were lysed by three freeze/thaw cycles with liquid nitrogen and a 37°C water bath. After the cells were lysed, the sample was split into two 150 μL aliquots. A blank tube, containing 100uL of Z buffer, was also set up. To analyze the enzyme activity, 350 μL of Z buffer with 0.27% β-mercaptoethanol and 4 mg/ml of ONPG as substrate were added. The yellow color was then allowed to develop at 30°C, and the reaction was stopped by the addition of 200 μL of 1 M Na2CO3. To remove cell debris, the samples were centrifuged for 10 minutes at 16,000 RCF, and OD420 values were recorded. In order to express the enzyme activity, Miller Units were calculated using the formula: 1000*OD420/(t*V*OD600) where t refers to the elapsed time, and V refers to the dilution factor (0.25 in this case).

Background: Influenza A viruses (IAVs) are important pathogens that affect the health of humans and many additional animal species. IAVs are enveloped, negative single-stranded RNA viruses whose genome encodes at least ten proteins. The IAV nucleoprotein (NP) is a structural protein that associates with the viral RNA and is essential for virus replication. Understanding how IAVs interact with host proteins is essential for elucidating all of the required processes for viral replication, restrictions in species host range, and potential targets for antiviral therapies. Methods: In this study, the NP from a swine IAV was cloned into a yeast two-hybrid "bait" vector for expression of a yeast Gal4 binding domain (BD)-NP fusion protein. This "bait" was used to screen a Y2H human HeLa cell "prey" library which consisted of human proteins fused to the Gal4 protein's activation domain (AD). The interaction of "bait" and "prey" proteins resulted in activation of reporter genes. Results: Seventeen positive bait-prey interactions were isolated in yeast. All of the "prey" isolated also interact in yeast with a NP "bait" cloned from a human IAV strain. Isolation and sequence analysis of the cDNAs encoding the human prey proteins revealed ten different human proteins. These host proteins are involved in various host cell processes and structures, including purine biosynthesis (PAICS), metabolism (ACOT13), proteasome (PA28B), DNA-binding (MSANTD3), cytoskeleton (CKAP5), potassium channel formation (KCTD9), zinc transporter function (SLC30A9), Na+/K+ ATPase function (ATP1B1), and RNA splicing (TRA2B). Conclusions: Ten human proteins were identified as interacting with IAV NP in a Y2H screen. Some of these human proteins were reported in previous screens aimed at elucidating host proteins relevant to specific viral life cycle processes such as replication. This study extends previous findings by suggesting a mechanism by which these host proteins associate with the IAV, i.e., physical interaction with NP. Furthermore, this study revealed novel host protein-NP interactions in yeast.

Background: Influenza A viruses (IAVs) are important pathogens that affect the health of humans and many additional animal species. In humans, seasonal IAV infection presents as a non-fatal, uncomplicated, acute infection characterized by the presence of upper respiratory symptoms as well as fever, headache, soreness and fatigue lasting 2–5 days [1]. However, deaths from seasonal IAV infections often arise when the normal flu symptoms are exacerbated by compromised immunity or age [1]. In addition to seasonal influenza A viruses, pandemic strains periodically appear causing increased mortality rates. For example, the 1918 Spanish flu resulted in approximately 50 million human deaths [2]. The most recent pandemic resulted from the emergence of the swine-origin H1N1 virus [3,4]. Due to influenza A’s potential for mortality, high mutation rates (resulting in genetic drift) and pandemic potential (resulting from genetic reassortment), it is critical to learn more about the virus, especially as it pertains to virulence, transmissibility, and identification of potential targets for development of therapeutics. Influenza A viruses exhibit a broad host range beyond humans [5]. Waterfowl serve as the central reservoir species. In addition to humans, various IAV subtypes circulate in pigs, poultry, horses, dogs as well as other species. Subtypes (H1N1 and H3N2) that circulate in the human population also circulate in the swine populations [6]. In addition to the concern for IAV in terms of swine health [7], IAV infection in swine is also important due to the potential for zoonotic infections as well as the potential for serving as a mixing vessel for the generation of human pandemic viruses [5]. The majority of inter-species transmission research has focused on amino acids in the hemagglutinin affecting viral attachment and entry [8]. Although known to be important, other than amino acids associated with temperature sensitivity (e.g., PB2 627 [9] or 701 [10]), the effects of differences in amino acid residues in proteins of the replication complex on species specificity is less well understood. Previous evidence suggests that there may be a role in some of these replication complex proteins in allowing for zoonotic infections [11]. The genome of the enveloped influenza A virion consists of eight segments of negative-sense single-stranded RNA that encode at least ten viral genes: Hemagglutinin (HA), Neuraminidase (NA), Matrix 2 protein (M2), Matrix 1 protein (M1), Non-structural Protein 1 (NS1), Non-structural Protein 2 (NS2), Nucleoprotein (NP), Polymerase Basic 1 (PB-1), Polymerase Basic 2 (PB-2), and Polymerase Acidic (PA) [1]. The viral envelope contains HA, NA, and M2 proteins, and M1 proteins form a layer inside the envelope. The IAV RNA genome located inside the virion is coated with NP and is bound by the replication complex consisting of PB1, PB2, and PA. The NP, the focus of this study, encoded by the fifth IAV RNA segment, binds with high affinity to viral RNA [12]. NP plays a role in viral RNA replication and transcription [12]. More recent data suggests that NP binds the polymerase as well as the newly replicated RNA and may act as a processivity factor that is necessary for replication of the viral RNA to be completed [13]. Phylogenic analysis of IAV NP shows distinct lineages of NP based on the host species [14,15]. Within the NP, there are amino acid signatures found within different host species [16,17]. These host-specific amino acid residues may result in differences in affinities for the various host proteins with which they interact (e.g., importin α1 [18], F- actin [19], nuclear factor 90 [20], cyclophilin E [21], exportin 1 [22], HMGB1 [23], HMGB2 [23], MxA/Mx1 [24,25], HSP40 [26], karyopherin alpha [27,28], clusterin [29], Raf-2p48/BAT1/UAP56 [30], Aly/REF [31], Tat-SF1 [32], TRIM22 [33], and alpha actinin 4 [34]) or they may result in differences in how the NP interacts with other viral proteins that have also made host-specific adaptations [11]. Given the suggestion that NP plays a role in determining host range [5,35-37], it is important to identify all host proteins that interact with NP. Replication of the IAV is dependent on the host cell machinery and interactions between host proteins and viral proteins. Therefore any inquiry attempting to investigate the life cycle of IAV must incorporate host-pathogen protein interactions in order to truly provide a mechanistic understanding of the process. Previous screens have identified important protein-protein interactions between IAV and host proteins, but often results are not consistent between screens. Therefore, corroborating evidence with additional screens is crucial to accumulating the best understanding of critical pathogen-host interactions in viral replication. As examples, several genome-wide screens have been performed using RNAi to systematically knockdown host genes and evaluate the effect on various stages of the IAV life cycle [38-41]. An integrated approach carried out by Shapira and colleagues involved transcription profiling and yeast two hybrid screens using specific viral proteins as baits [28]. Proteomics approaches have revealed host proteins found within viral particles [42,43]. The Random Homozygous Gene Perturbation strategy was used to identify host factors that prevent influenza-mediated killing of host cells [44]. In order to better understand the role of NP-host protein interactions in IAV replication, a Gal4-based yeast two-hybrid (Y2H) assay was used in this study. The Y2H system allows for the identification of unique human binding partners with NP. A “bait” plasmid encoding the binding domain (BD) of the Gal4 transcriptional activator fused to NP and a “prey” plasmid encoding Gal4’s activation domain (AD) fused to a protein encoded by a human cDNA are introduced into a yeast strain containing Gal4-responsive reporter genes. Interaction of the bait and prey brings together Gal4’s BD and AD, resulting in transcriptional activation of reporter genes [45]. The Y2H approach has successfully identified cellular factors (e.g., Raf-2p48/BAT1/UAP56, Hsp40, KPNA1, KNPA3, KPNA6, C16orf45, GMCL1, MAGED1, MLH1, USHBP1, ZBTB25, CLU, Aly/REF, and ACTN4) that interact with the IAV NP [26-31,34], (Table 1). In this study, the nucleoprotein from a classical swine H1N1 IAV (Sw/NC/44173/00) was used as the “bait” in a Y2H screen against a “prey” HeLa cDNA library. To investigate if the origin of NP affects the ability of the interactions to take place, the nucleoprotein from a contemporary human H3N2 IAV (A/Ca/07/04) was also used as the “bait” in a Y2H screen against the same prey HeLa cDNA library. By determination of the putative interaction partners between NP and human proteins, the possible functions of the newfound protein-protein interactions or verification and perhaps elucidation of previously identified host factors may be further investigated.Table 1 Summary of identified bait-prey interactions and previous reports of host protein association with influenza A virus Gene Protein description Complete ORF Number of prey clones Prey clones Previous reports with IAV PAICSPhosphoribosylaminoimidazole carboxylaseYes3R3, R9, R15Karlas et al. [40], Kumar and Nanduri [47], Kroeker et al. [48]MSANTD3Myb/SANT-like DNA-binding domain containing 3No3R10, R13, R27N/AFLJ30306PREDICTED: Homo sapiens uncharacterized LOC101059922, mRNAYes2R23, R33N/APA28BProteasome activator subunit 2Yes2R6, R19N/AKCTD9Potassium channel tetramerisation domain containing 9No2R34, R36N/AACOT13Acyl-CoA thioesterase 13Yes1R11N/ATRA2BTransformer 2 beta homolog (Drosophila)Yes1R12Zhu et al. [56]SLC30A9Solute carrier family 30 (zinc transporter), member 9No1R28Kumar and Nanduri [47], Sui et al. [44]ATP1B1ATPase, Na+/K+ transporting, beta 1 polypeptideNo1R7Mi et al. [51], Liu et al. [50]CKAP5Cytoskeleton associated protein 5No1R29N/A Host proteins identified as interacting with IAV nucleoprotein in previously published yeast two-hybrid screens BAT1/UAP56RNA-dependent ATPaseN/AN/AN/AMomose et al. [30]HSP40Heat shock protein 40N/AN/AN/ASharma et al. [26]NPI-1/ SRP1/KPNA1Karyopherin α1N/AN/AN/AO’Neill and Palese [27], Shapira et al. [28]KPNA3Karyopherin α3N/AN/AN/AShapira et al. [28]KPNA6Karyopherin α6N/AN/AN/AShapira et al. [28]C16orf45-N/AN/AN/AShapira et al. [28]GMCL1Germ cell-less, spermatogenesis associated 1N/AN/AN/AShapira et al. [28]MAGED1Melanoma antigen family D.1N/AN/AN/AShapira et al. [28]MLH1MutL homolog 1N/AN/AN/AShapira et al. [28]USHBP1Usher syndrome 1C binding protein 1N/AN/AN/AShapira et al. [28]ZBTB25Zinc finger and BTB domain containing 25N/AN/AN/AShapira et al. [28]CLUClusterinN/AN/AN/ATripathi et al. [29]ALY/REFRNA export adaptor proteinN/AN/AN/ABalasubramaniam et al. [31]ACTN4Alpha-actinin 4N/AN/AN/ASharma et al. [34] Summary of identified bait-prey interactions and previous reports of host protein association with influenza A virus Reported herein are ten potential human host cell proteins that interact with IAV NP in a Y2H screen. The relative strengths of these protein-protein interactions were characterized by a plating assay and beta-galactosidase assay. Conclusions: A Y2H screen using a classical swine influenza A virus nucleoprotein as a bait resulted in isolation of ten different putative interacting human host proteins involved in a variety of cellular processes and structures: purine biosynthesis (PAICS), metabolism (ACOT13), proteasome (PA28B), DNA-binding (MSANTD3), cytoskeleton (CKAP5), potassium channel formation (KCTD9), zinc transporter function (SLC30A9), Na+/K+ ATPase function (ATP1B1), and RNA splicing (TRA2B). All of the identified human host proteins interacted in yeast with NP from both human and swine origin. Proteins PAICS, ATP1B1, TRA2B, and SLC30A9 have been previously identified in studies related to host response to the IAV. Host proteins SLC30A9 and CKAP5 displayed the strongest interactions with IAV NP in yeast in a quantitative β-gal assay.

Background: Influenza A viruses (IAVs) are important pathogens that affect the health of humans and many additional animal species. IAVs are enveloped, negative single-stranded RNA viruses whose genome encodes at least ten proteins. The IAV nucleoprotein (NP) is a structural protein that associates with the viral RNA and is essential for virus replication. Understanding how IAVs interact with host proteins is essential for elucidating all of the required processes for viral replication, restrictions in species host range, and potential targets for antiviral therapies. Methods: In this study, the NP from a swine IAV was cloned into a yeast two-hybrid "bait" vector for expression of a yeast Gal4 binding domain (BD)-NP fusion protein. This "bait" was used to screen a Y2H human HeLa cell "prey" library which consisted of human proteins fused to the Gal4 protein's activation domain (AD). The interaction of "bait" and "prey" proteins resulted in activation of reporter genes. Results: Seventeen positive bait-prey interactions were isolated in yeast. All of the "prey" isolated also interact in yeast with a NP "bait" cloned from a human IAV strain. Isolation and sequence analysis of the cDNAs encoding the human prey proteins revealed ten different human proteins. These host proteins are involved in various host cell processes and structures, including purine biosynthesis (PAICS), metabolism (ACOT13), proteasome (PA28B), DNA-binding (MSANTD3), cytoskeleton (CKAP5), potassium channel formation (KCTD9), zinc transporter function (SLC30A9), Na+/K+ ATPase function (ATP1B1), and RNA splicing (TRA2B). Conclusions: Ten human proteins were identified as interacting with IAV NP in a Y2H screen. Some of these human proteins were reported in previous screens aimed at elucidating host proteins relevant to specific viral life cycle processes such as replication. This study extends previous findings by suggesting a mechanism by which these host proteins associate with the IAV, i.e., physical interaction with NP. Furthermore, this study revealed novel host protein-NP interactions in yeast.

[CONTENT] Influenza A virus | Nucleoprotein | Yeast two-hybrid | Host restriction [SUMMARY]

[CONTENT] Influenza A virus | Nucleoprotein | Yeast two-hybrid | Host restriction [SUMMARY]

[CONTENT] Influenza A virus | Nucleoprotein | Yeast two-hybrid | Host restriction [SUMMARY]

[CONTENT] Influenza A virus | Nucleoprotein | Yeast two-hybrid | Host restriction [SUMMARY]

[CONTENT] Influenza A virus | Nucleoprotein | Yeast two-hybrid | Host restriction [SUMMARY]

[CONTENT] Animals | Genes, Reporter | HeLa Cells | Host-Pathogen Interactions | Humans | Influenzavirus A | Nucleocapsid Proteins | Protein Interaction Mapping | RNA-Binding Proteins | Swine | Two-Hybrid System Techniques | Viral Core Proteins [SUMMARY]

[CONTENT] Animals | Genes, Reporter | HeLa Cells | Host-Pathogen Interactions | Humans | Influenzavirus A | Nucleocapsid Proteins | Protein Interaction Mapping | RNA-Binding Proteins | Swine | Two-Hybrid System Techniques | Viral Core Proteins [SUMMARY]

[CONTENT] Animals | Genes, Reporter | HeLa Cells | Host-Pathogen Interactions | Humans | Influenzavirus A | Nucleocapsid Proteins | Protein Interaction Mapping | RNA-Binding Proteins | Swine | Two-Hybrid System Techniques | Viral Core Proteins [SUMMARY]

[CONTENT] Animals | Genes, Reporter | HeLa Cells | Host-Pathogen Interactions | Humans | Influenzavirus A | Nucleocapsid Proteins | Protein Interaction Mapping | RNA-Binding Proteins | Swine | Two-Hybrid System Techniques | Viral Core Proteins [SUMMARY]

[CONTENT] Animals | Genes, Reporter | HeLa Cells | Host-Pathogen Interactions | Humans | Influenzavirus A | Nucleocapsid Proteins | Protein Interaction Mapping | RNA-Binding Proteins | Swine | Two-Hybrid System Techniques | Viral Core Proteins [SUMMARY]

[CONTENT] seasonal influenza viruses | human influenza | influenza potential mortality | influenza genes | virus influenza potential [SUMMARY]

[CONTENT] seasonal influenza viruses | human influenza | influenza potential mortality | influenza genes | virus influenza potential [SUMMARY]

[CONTENT] seasonal influenza viruses | human influenza | influenza potential mortality | influenza genes | virus influenza potential [SUMMARY]

[CONTENT] seasonal influenza viruses | human influenza | influenza potential mortality | influenza genes | virus influenza potential [SUMMARY]

[CONTENT] seasonal influenza viruses | human influenza | influenza potential mortality | influenza genes | virus influenza potential [SUMMARY]

[CONTENT] prey | np | human | bait | sd | swine | yeast | growth | cells | protein [SUMMARY]

[CONTENT] prey | np | human | bait | sd | swine | yeast | growth | cells | protein [SUMMARY]

[CONTENT] prey | np | human | bait | sd | swine | yeast | growth | cells | protein [SUMMARY]

[CONTENT] prey | np | human | bait | sd | swine | yeast | growth | cells | protein [SUMMARY]

[CONTENT] prey | np | human | bait | sd | swine | yeast | growth | cells | protein [SUMMARY]

[CONTENT] host | protein | iav | 28 | viral | ashapira | ashapira 28 | proteins | np | replication [SUMMARY]

[CONTENT] human | growth | prey | np | swine | swine human | activity | bait | figure | human np [SUMMARY]

[CONTENT] slc30a9 | host | proteins | function | host proteins | ckap5 | paics | tra2b | human host proteins | atp1b1 [SUMMARY]

[CONTENT] np | sd | human | prey | bait | cells | swine | containing | yeast | trp [SUMMARY]

[CONTENT] np | sd | human | prey | bait | cells | swine | containing | yeast | trp [SUMMARY]

[CONTENT] Influenza A ||| RNA | at least ten ||| NP | RNA ||| [SUMMARY]

[CONTENT] Seventeen ||| NP ||| ten ||| PAICS | MSANTD3 | cytoskeleton | CKAP5 | KCTD9 | ATPase | RNA [SUMMARY]

[CONTENT] Ten | NP ||| ||| IAV | NP ||| NP [SUMMARY]

[CONTENT] ||| RNA | at least ten ||| NP | RNA ||| ||| NP | IAV | two ||| HeLa ||| ||| ||| Seventeen ||| NP ||| ten ||| PAICS | MSANTD3 | cytoskeleton | CKAP5 | KCTD9 | ATPase | RNA ||| Ten | NP ||| ||| IAV | NP ||| NP [SUMMARY]

[CONTENT] ||| RNA | at least ten ||| NP | RNA ||| ||| NP | IAV | two ||| HeLa ||| ||| ||| Seventeen ||| NP ||| ten ||| PAICS | MSANTD3 | cytoskeleton | CKAP5 | KCTD9 | ATPase | RNA ||| Ten | NP ||| ||| IAV | NP ||| NP [SUMMARY]

Antibiotic resistance in Helicobacter pylori strains and its effect on H. pylori eradication rates in a single center in Korea.

24205490

Clarithromycin, amoxicillin, metronidazole, tetracycline, and levofloxacin have been commonly used for the eradication of Helicobacter pylori. We compared the change in antibiotic resistance of H. pylori strains during two separate periods and investigated the effect of antibiotic resistance on H. pylori eradication.

BACKGROUND

H. pylori strains were isolated from 71 patients between 2009 and 2010 and from 94 patients between 2011 and 2012. The distribution of minimal inhibitory concentration (MIC) of 5 antibiotics was assessed using the agar dilution method, and H. pylori eradication based on the antimicrobial susceptibility of the isolates was investigated retrospectively.

METHODS

Antibiotic resistance rate against clarithromycin, amoxicillin, tetracycline, metronidazole, and levofloxacin for the 2009-2010 isolates were 7.0% (5/71), 2.8% (2/71), 0% (0/71), 45.1% (32/71), and 26.8% (19/71), respectively, and for the 2011-2012 isolates were 16.0% (15/94), 2.1% (2/94), 0% (0/94), 56.3% (53/94), and 22.3% (21/94), respectively. Multi-drug resistance for 2 or more antibiotics increased slightly from 16.9% (12/71) in the 2009-2010 isolates to 23.4% (22/94) in the 2011-2012 isolates. In follow-up testing of 66 patients, first-line treatment successfully eradicated H. pylori in 50 patients (75.8%) and failed in 4 of 7 patients (57.1%) in a clarithromycin-resistant and amoxicillin-susceptible group.

RESULTS

We observed an increase in resistance to clarithromycin and an overall increase in multi-drug resistance during the 2 study periods. The effectiveness of the eradication regimen was low with combinations of clarithromycin and amoxicillin, particularly in the clarithromycin-resistant group. Thus, eradication of H. pylori depends upon periodic monitoring of antimicrobial susceptibility.

CONCLUSIONS

3819440

INTRODUCTION

Eradication of gastric colonies of Helicobacter pylori helps heal gastritis and peptic ulcer disease and has beneficial effects on the regression of atrophic gastritis and the prevention of distal gastric cancer [1, 2]. Triple therapy using a proton pump inhibitor (PPI) with clarithromycin and amoxicillin or metronidazole is recommended as the first-line treatment regimen for H. pylori eradication. If it fails, bismuth-containing quadruple therapy, which involves inclusion of additional antibiotics to the first-line treatment regimen is used [3, 4]. The increase in clarithromycin resistance in Korea is considered to be closely related to the decrease of eradication rate in first-line therapy. According to recent data, clarithromycin resistance sharply increased from 16.7% to 38.5% from 2003 through 2009, and eradication rates have decreased by 77-87% since 2003 [4-6]; these rates are inclusive of regional and institutional differences. Although regular antibiotic resistance monitoring is important in the clinical setting, the labor-intensive and time-consuming nature of H. pylori isolation from clinical samples complicates comparative antibiotic susceptibility testing. In this study, we investigated H. pylori antibiotic resistance and its effect on eradication rates in a single center in Korea between 2009-2010 and 2011-2012.

METHODS

1. Patients H. pylori strains were isolated from 71 patients with H. pylori infections from July 2009 to December 2010 and from 94 patients from June 2011 through December 2012 at the Yongin Severance Hospital of Yonsei University, Korea. Of these patients, 66 (clinical characteristics listed in Table 1) had previously undergone eradication treatments, including week-long first-line treatment with PPI (pantoprazole or esomeprazole 30 mg, bid), amoxicillin (2,250 mg, tid), clarithromycin (1,000 mg, bid). First-line therapy failed in 16 patients, and they were subjected to second-line treatment with PPI (30 mg, bid), bismuth (300 mg, bid), metronidazole (2,250 mg, bid), and tetracycline (1,000 mg, qid). Eradication of H. pylori was verified by a negative result in a 13C-urea breath test (Isotechnika, Alberta, Canada) after at least 4 weeks of drug administration. This study was conducted retrospectively to follow up the results of eradication of H. pylori on the basis of antimicrobial susceptibility of the isolates, and it did not interfere with patient management decisions. The study was approved by the Institutional Review Board of Yonsei University College of Medicine (No. 4-2011-0508). Written informed consent was provided by all patients at the time of their first visit to the hospital. H. pylori strains were isolated from 71 patients with H. pylori infections from July 2009 to December 2010 and from 94 patients from June 2011 through December 2012 at the Yongin Severance Hospital of Yonsei University, Korea. Of these patients, 66 (clinical characteristics listed in Table 1) had previously undergone eradication treatments, including week-long first-line treatment with PPI (pantoprazole or esomeprazole 30 mg, bid), amoxicillin (2,250 mg, tid), clarithromycin (1,000 mg, bid). First-line therapy failed in 16 patients, and they were subjected to second-line treatment with PPI (30 mg, bid), bismuth (300 mg, bid), metronidazole (2,250 mg, bid), and tetracycline (1,000 mg, qid). Eradication of H. pylori was verified by a negative result in a 13C-urea breath test (Isotechnika, Alberta, Canada) after at least 4 weeks of drug administration. This study was conducted retrospectively to follow up the results of eradication of H. pylori on the basis of antimicrobial susceptibility of the isolates, and it did not interfere with patient management decisions. The study was approved by the Institutional Review Board of Yonsei University College of Medicine (No. 4-2011-0508). Written informed consent was provided by all patients at the time of their first visit to the hospital. 2. H. pylori culture The culture medium used in this study was composed of Brucella broth (BBL, Sparks, MD, USA) containing 1.2% agar, 10% bovine serum, and selected antibiotics (Oxoid Limited, Hampshire, England) (10 µg/mL vancomycin, 5 µg/mL trimethoprim, 5 µg/mL cefsulodin, and 5 µg/mL amphotericin B). Completely minced gastric biopsy specimens were incubated under 10% CO2, 5% O2, and 100% humidity at 37℃ for 3-5 days. Strains were identified as H. pylori by Gram staining; colony morphology analysis; and oxidase, catalase, and urease tests. The H. pylori ATCC 43504 strain was cultured as a standard using the same methods described above for quality control assessment. The culture medium used in this study was composed of Brucella broth (BBL, Sparks, MD, USA) containing 1.2% agar, 10% bovine serum, and selected antibiotics (Oxoid Limited, Hampshire, England) (10 µg/mL vancomycin, 5 µg/mL trimethoprim, 5 µg/mL cefsulodin, and 5 µg/mL amphotericin B). Completely minced gastric biopsy specimens were incubated under 10% CO2, 5% O2, and 100% humidity at 37℃ for 3-5 days. Strains were identified as H. pylori by Gram staining; colony morphology analysis; and oxidase, catalase, and urease tests. The H. pylori ATCC 43504 strain was cultured as a standard using the same methods described above for quality control assessment. 3. Susceptibility tests The minimal inhibitory concentrations (MICs) for clarithromycin (Sigma-Aldrich Co., St. Louis, MO, USA), amoxicillin (Sigma-Aldrich), tetracycline (Sigma-Aldrich), metronidazole (Sigma-Aldrich), and levofloxacin (Sigma-Aldrich) were determined using a slightly modified agar dilution method (using Brucella broth base containing 1.2% agar). Clarithromycin resistance was defined according to the CLSI-approved breakpoint (≥1 µg/mL) [7]. Isolates were defined as resistant to amoxicillin, tetracycline, metronidazole, and levofloxacin, when MICs were ≥1, ≥4, ≥8, and ≥1 µg/mL, respectively [8-10]. For H. pylori ATCC 43504, the MIC ranges for clarithromycin, amoxicillin, metronidazole, tetracycline, and levofloxacin were 0.016-0.125 µg/mL, 0.016-0.125 µg/mL, 64-256 µg/mL, 0.125-1 µg/mL, and 0.064-0.5 µg/mL, respectively. The minimal inhibitory concentrations (MICs) for clarithromycin (Sigma-Aldrich Co., St. Louis, MO, USA), amoxicillin (Sigma-Aldrich), tetracycline (Sigma-Aldrich), metronidazole (Sigma-Aldrich), and levofloxacin (Sigma-Aldrich) were determined using a slightly modified agar dilution method (using Brucella broth base containing 1.2% agar). Clarithromycin resistance was defined according to the CLSI-approved breakpoint (≥1 µg/mL) [7]. Isolates were defined as resistant to amoxicillin, tetracycline, metronidazole, and levofloxacin, when MICs were ≥1, ≥4, ≥8, and ≥1 µg/mL, respectively [8-10]. For H. pylori ATCC 43504, the MIC ranges for clarithromycin, amoxicillin, metronidazole, tetracycline, and levofloxacin were 0.016-0.125 µg/mL, 0.016-0.125 µg/mL, 64-256 µg/mL, 0.125-1 µg/mL, and 0.064-0.5 µg/mL, respectively. 4. Statistical analysis Statistical analysis was performed using SPSS (Statistical Package for the Social Sciences version 18.0; SPSS Ins., Chicago, IL, USA). Data of antibiotic resistance were analyzed using the student t test and Chi-square test. P<0.05 was considered statistically significant. Statistical analysis was performed using SPSS (Statistical Package for the Social Sciences version 18.0; SPSS Ins., Chicago, IL, USA). Data of antibiotic resistance were analyzed using the student t test and Chi-square test. P<0.05 was considered statistically significant.

RESULTS

1. Antibiotic resistance of H. pylori The antibiotic resistance rates for the isolates from the 2009-2010 group against clarithromycin, amoxicillin, tetracycline, metronidazole, and levofloxacin were 7.0% (5/71), 2.8% (2/71), 0% (0/71), 45.1% (32/71), and 26.8% (19/71), respectively, and those for the isolates from the 2011-2012 group were 16.0% (15/94), 2.1% (2/94), 0% (0/94), 56.3% (53/94), and 22.3% (21/94), respectively. The rate of H. pylori resistance to clarithromycin and metronidazole increased from 7.0% to 16.0% and from 45.1% to 56.3% for the 2 periods, respectively (Fig. 1), although the increase was not statistically significant. When the MIC distribution profiles for the 2 study periods were compared, the MICs for clarithromycin showed notable differences between the susceptible and resistant strains (Fig. 2). The MIC of the clarithromycin-susceptible strains was less than 0.125 µg/mL. Resistance to tetracycline was not detected in any strain (based on a cut-off of ≥ 4 µg/mL). The MIC range of tetracycline was 0.031-2 µg/mL. The MIC of metronidazole varied widely (8-256 µg/mL). Multi-drug resistance for 2 or more antibiotics was more frequent in the isolates from 2011-2012 (23.4%, 22/94) than in the isolates from 2009-2010 (16.9%, 12/71), but there was no statistical significance (P<0.082). Only 1 strain exhibited multi-drug resistance to clarithromycin, amoxicillin, metronidazole, and levofloxacin (Table 2). The antibiotic resistance rates for the isolates from the 2009-2010 group against clarithromycin, amoxicillin, tetracycline, metronidazole, and levofloxacin were 7.0% (5/71), 2.8% (2/71), 0% (0/71), 45.1% (32/71), and 26.8% (19/71), respectively, and those for the isolates from the 2011-2012 group were 16.0% (15/94), 2.1% (2/94), 0% (0/94), 56.3% (53/94), and 22.3% (21/94), respectively. The rate of H. pylori resistance to clarithromycin and metronidazole increased from 7.0% to 16.0% and from 45.1% to 56.3% for the 2 periods, respectively (Fig. 1), although the increase was not statistically significant. When the MIC distribution profiles for the 2 study periods were compared, the MICs for clarithromycin showed notable differences between the susceptible and resistant strains (Fig. 2). The MIC of the clarithromycin-susceptible strains was less than 0.125 µg/mL. Resistance to tetracycline was not detected in any strain (based on a cut-off of ≥ 4 µg/mL). The MIC range of tetracycline was 0.031-2 µg/mL. The MIC of metronidazole varied widely (8-256 µg/mL). Multi-drug resistance for 2 or more antibiotics was more frequent in the isolates from 2011-2012 (23.4%, 22/94) than in the isolates from 2009-2010 (16.9%, 12/71), but there was no statistical significance (P<0.082). Only 1 strain exhibited multi-drug resistance to clarithromycin, amoxicillin, metronidazole, and levofloxacin (Table 2). 2. Effect of antibiotic resistance on H. pylori eradication rates Of the 165 patients studied during the 2009-2012 period, 66 patients who were subjected to the first-line therapy were followed up for H. pylori eradication after treatment. Among these 66 patients, no significant differences were found with respect to sex, age, and endoscopic diagnosis. Eradication of H. pylori was successful in 50 of these 66 patients (75.8%). The effects of antibiotic resistance on H. pylori eradication rates are shown in Table 3. Eradication rates were 79.3% (46/58) for the clarithromycin-susceptible and amoxicillin-susceptible strains, and 100% (1/1) for the clarithromycin-susceptible and amoxicillin-resistant strains. A significant difference was observed between the eradication rates for the clarithromycin-resistant (42.9%, 3/7) and the clarithromycin-sensitive (79.7%, 47/59) strains (P<0.001) . Second-line therapy was prescribed for the 16 patients in whom first-line therapy failed. The eradication rates for the tetracycline-susceptible and metronidazole-susceptible strains and the tetracycline-susceptible and metronidazole-resistant strains were 50.0% (4/8) and 25.0% (2/8), respectively (P<0.32). Of the 165 patients studied during the 2009-2012 period, 66 patients who were subjected to the first-line therapy were followed up for H. pylori eradication after treatment. Among these 66 patients, no significant differences were found with respect to sex, age, and endoscopic diagnosis. Eradication of H. pylori was successful in 50 of these 66 patients (75.8%). The effects of antibiotic resistance on H. pylori eradication rates are shown in Table 3. Eradication rates were 79.3% (46/58) for the clarithromycin-susceptible and amoxicillin-susceptible strains, and 100% (1/1) for the clarithromycin-susceptible and amoxicillin-resistant strains. A significant difference was observed between the eradication rates for the clarithromycin-resistant (42.9%, 3/7) and the clarithromycin-sensitive (79.7%, 47/59) strains (P<0.001) . Second-line therapy was prescribed for the 16 patients in whom first-line therapy failed. The eradication rates for the tetracycline-susceptible and metronidazole-susceptible strains and the tetracycline-susceptible and metronidazole-resistant strains were 50.0% (4/8) and 25.0% (2/8), respectively (P<0.32).

INTRODUCTION: Eradication of gastric colonies of Helicobacter pylori helps heal gastritis and peptic ulcer disease and has beneficial effects on the regression of atrophic gastritis and the prevention of distal gastric cancer [1, 2]. Triple therapy using a proton pump inhibitor (PPI) with clarithromycin and amoxicillin or metronidazole is recommended as the first-line treatment regimen for H. pylori eradication. If it fails, bismuth-containing quadruple therapy, which involves inclusion of additional antibiotics to the first-line treatment regimen is used [3, 4]. The increase in clarithromycin resistance in Korea is considered to be closely related to the decrease of eradication rate in first-line therapy. According to recent data, clarithromycin resistance sharply increased from 16.7% to 38.5% from 2003 through 2009, and eradication rates have decreased by 77-87% since 2003 [4-6]; these rates are inclusive of regional and institutional differences. Although regular antibiotic resistance monitoring is important in the clinical setting, the labor-intensive and time-consuming nature of H. pylori isolation from clinical samples complicates comparative antibiotic susceptibility testing. In this study, we investigated H. pylori antibiotic resistance and its effect on eradication rates in a single center in Korea between 2009-2010 and 2011-2012. METHODS: 1. Patients H. pylori strains were isolated from 71 patients with H. pylori infections from July 2009 to December 2010 and from 94 patients from June 2011 through December 2012 at the Yongin Severance Hospital of Yonsei University, Korea. Of these patients, 66 (clinical characteristics listed in Table 1) had previously undergone eradication treatments, including week-long first-line treatment with PPI (pantoprazole or esomeprazole 30 mg, bid), amoxicillin (2,250 mg, tid), clarithromycin (1,000 mg, bid). First-line therapy failed in 16 patients, and they were subjected to second-line treatment with PPI (30 mg, bid), bismuth (300 mg, bid), metronidazole (2,250 mg, bid), and tetracycline (1,000 mg, qid). Eradication of H. pylori was verified by a negative result in a 13C-urea breath test (Isotechnika, Alberta, Canada) after at least 4 weeks of drug administration. This study was conducted retrospectively to follow up the results of eradication of H. pylori on the basis of antimicrobial susceptibility of the isolates, and it did not interfere with patient management decisions. The study was approved by the Institutional Review Board of Yonsei University College of Medicine (No. 4-2011-0508). Written informed consent was provided by all patients at the time of their first visit to the hospital. H. pylori strains were isolated from 71 patients with H. pylori infections from July 2009 to December 2010 and from 94 patients from June 2011 through December 2012 at the Yongin Severance Hospital of Yonsei University, Korea. Of these patients, 66 (clinical characteristics listed in Table 1) had previously undergone eradication treatments, including week-long first-line treatment with PPI (pantoprazole or esomeprazole 30 mg, bid), amoxicillin (2,250 mg, tid), clarithromycin (1,000 mg, bid). First-line therapy failed in 16 patients, and they were subjected to second-line treatment with PPI (30 mg, bid), bismuth (300 mg, bid), metronidazole (2,250 mg, bid), and tetracycline (1,000 mg, qid). Eradication of H. pylori was verified by a negative result in a 13C-urea breath test (Isotechnika, Alberta, Canada) after at least 4 weeks of drug administration. This study was conducted retrospectively to follow up the results of eradication of H. pylori on the basis of antimicrobial susceptibility of the isolates, and it did not interfere with patient management decisions. The study was approved by the Institutional Review Board of Yonsei University College of Medicine (No. 4-2011-0508). Written informed consent was provided by all patients at the time of their first visit to the hospital. 2. H. pylori culture The culture medium used in this study was composed of Brucella broth (BBL, Sparks, MD, USA) containing 1.2% agar, 10% bovine serum, and selected antibiotics (Oxoid Limited, Hampshire, England) (10 µg/mL vancomycin, 5 µg/mL trimethoprim, 5 µg/mL cefsulodin, and 5 µg/mL amphotericin B). Completely minced gastric biopsy specimens were incubated under 10% CO2, 5% O2, and 100% humidity at 37℃ for 3-5 days. Strains were identified as H. pylori by Gram staining; colony morphology analysis; and oxidase, catalase, and urease tests. The H. pylori ATCC 43504 strain was cultured as a standard using the same methods described above for quality control assessment. The culture medium used in this study was composed of Brucella broth (BBL, Sparks, MD, USA) containing 1.2% agar, 10% bovine serum, and selected antibiotics (Oxoid Limited, Hampshire, England) (10 µg/mL vancomycin, 5 µg/mL trimethoprim, 5 µg/mL cefsulodin, and 5 µg/mL amphotericin B). Completely minced gastric biopsy specimens were incubated under 10% CO2, 5% O2, and 100% humidity at 37℃ for 3-5 days. Strains were identified as H. pylori by Gram staining; colony morphology analysis; and oxidase, catalase, and urease tests. The H. pylori ATCC 43504 strain was cultured as a standard using the same methods described above for quality control assessment. 3. Susceptibility tests The minimal inhibitory concentrations (MICs) for clarithromycin (Sigma-Aldrich Co., St. Louis, MO, USA), amoxicillin (Sigma-Aldrich), tetracycline (Sigma-Aldrich), metronidazole (Sigma-Aldrich), and levofloxacin (Sigma-Aldrich) were determined using a slightly modified agar dilution method (using Brucella broth base containing 1.2% agar). Clarithromycin resistance was defined according to the CLSI-approved breakpoint (≥1 µg/mL) [7]. Isolates were defined as resistant to amoxicillin, tetracycline, metronidazole, and levofloxacin, when MICs were ≥1, ≥4, ≥8, and ≥1 µg/mL, respectively [8-10]. For H. pylori ATCC 43504, the MIC ranges for clarithromycin, amoxicillin, metronidazole, tetracycline, and levofloxacin were 0.016-0.125 µg/mL, 0.016-0.125 µg/mL, 64-256 µg/mL, 0.125-1 µg/mL, and 0.064-0.5 µg/mL, respectively. The minimal inhibitory concentrations (MICs) for clarithromycin (Sigma-Aldrich Co., St. Louis, MO, USA), amoxicillin (Sigma-Aldrich), tetracycline (Sigma-Aldrich), metronidazole (Sigma-Aldrich), and levofloxacin (Sigma-Aldrich) were determined using a slightly modified agar dilution method (using Brucella broth base containing 1.2% agar). Clarithromycin resistance was defined according to the CLSI-approved breakpoint (≥1 µg/mL) [7]. Isolates were defined as resistant to amoxicillin, tetracycline, metronidazole, and levofloxacin, when MICs were ≥1, ≥4, ≥8, and ≥1 µg/mL, respectively [8-10]. For H. pylori ATCC 43504, the MIC ranges for clarithromycin, amoxicillin, metronidazole, tetracycline, and levofloxacin were 0.016-0.125 µg/mL, 0.016-0.125 µg/mL, 64-256 µg/mL, 0.125-1 µg/mL, and 0.064-0.5 µg/mL, respectively. 4. Statistical analysis Statistical analysis was performed using SPSS (Statistical Package for the Social Sciences version 18.0; SPSS Ins., Chicago, IL, USA). Data of antibiotic resistance were analyzed using the student t test and Chi-square test. P<0.05 was considered statistically significant. Statistical analysis was performed using SPSS (Statistical Package for the Social Sciences version 18.0; SPSS Ins., Chicago, IL, USA). Data of antibiotic resistance were analyzed using the student t test and Chi-square test. P<0.05 was considered statistically significant. 1. Patients: H. pylori strains were isolated from 71 patients with H. pylori infections from July 2009 to December 2010 and from 94 patients from June 2011 through December 2012 at the Yongin Severance Hospital of Yonsei University, Korea. Of these patients, 66 (clinical characteristics listed in Table 1) had previously undergone eradication treatments, including week-long first-line treatment with PPI (pantoprazole or esomeprazole 30 mg, bid), amoxicillin (2,250 mg, tid), clarithromycin (1,000 mg, bid). First-line therapy failed in 16 patients, and they were subjected to second-line treatment with PPI (30 mg, bid), bismuth (300 mg, bid), metronidazole (2,250 mg, bid), and tetracycline (1,000 mg, qid). Eradication of H. pylori was verified by a negative result in a 13C-urea breath test (Isotechnika, Alberta, Canada) after at least 4 weeks of drug administration. This study was conducted retrospectively to follow up the results of eradication of H. pylori on the basis of antimicrobial susceptibility of the isolates, and it did not interfere with patient management decisions. The study was approved by the Institutional Review Board of Yonsei University College of Medicine (No. 4-2011-0508). Written informed consent was provided by all patients at the time of their first visit to the hospital. 2. H. pylori culture: The culture medium used in this study was composed of Brucella broth (BBL, Sparks, MD, USA) containing 1.2% agar, 10% bovine serum, and selected antibiotics (Oxoid Limited, Hampshire, England) (10 µg/mL vancomycin, 5 µg/mL trimethoprim, 5 µg/mL cefsulodin, and 5 µg/mL amphotericin B). Completely minced gastric biopsy specimens were incubated under 10% CO2, 5% O2, and 100% humidity at 37℃ for 3-5 days. Strains were identified as H. pylori by Gram staining; colony morphology analysis; and oxidase, catalase, and urease tests. The H. pylori ATCC 43504 strain was cultured as a standard using the same methods described above for quality control assessment. 3. Susceptibility tests: The minimal inhibitory concentrations (MICs) for clarithromycin (Sigma-Aldrich Co., St. Louis, MO, USA), amoxicillin (Sigma-Aldrich), tetracycline (Sigma-Aldrich), metronidazole (Sigma-Aldrich), and levofloxacin (Sigma-Aldrich) were determined using a slightly modified agar dilution method (using Brucella broth base containing 1.2% agar). Clarithromycin resistance was defined according to the CLSI-approved breakpoint (≥1 µg/mL) [7]. Isolates were defined as resistant to amoxicillin, tetracycline, metronidazole, and levofloxacin, when MICs were ≥1, ≥4, ≥8, and ≥1 µg/mL, respectively [8-10]. For H. pylori ATCC 43504, the MIC ranges for clarithromycin, amoxicillin, metronidazole, tetracycline, and levofloxacin were 0.016-0.125 µg/mL, 0.016-0.125 µg/mL, 64-256 µg/mL, 0.125-1 µg/mL, and 0.064-0.5 µg/mL, respectively. 4. Statistical analysis: Statistical analysis was performed using SPSS (Statistical Package for the Social Sciences version 18.0; SPSS Ins., Chicago, IL, USA). Data of antibiotic resistance were analyzed using the student t test and Chi-square test. P<0.05 was considered statistically significant. RESULTS: 1. Antibiotic resistance of H. pylori The antibiotic resistance rates for the isolates from the 2009-2010 group against clarithromycin, amoxicillin, tetracycline, metronidazole, and levofloxacin were 7.0% (5/71), 2.8% (2/71), 0% (0/71), 45.1% (32/71), and 26.8% (19/71), respectively, and those for the isolates from the 2011-2012 group were 16.0% (15/94), 2.1% (2/94), 0% (0/94), 56.3% (53/94), and 22.3% (21/94), respectively. The rate of H. pylori resistance to clarithromycin and metronidazole increased from 7.0% to 16.0% and from 45.1% to 56.3% for the 2 periods, respectively (Fig. 1), although the increase was not statistically significant. When the MIC distribution profiles for the 2 study periods were compared, the MICs for clarithromycin showed notable differences between the susceptible and resistant strains (Fig. 2). The MIC of the clarithromycin-susceptible strains was less than 0.125 µg/mL. Resistance to tetracycline was not detected in any strain (based on a cut-off of ≥ 4 µg/mL). The MIC range of tetracycline was 0.031-2 µg/mL. The MIC of metronidazole varied widely (8-256 µg/mL). Multi-drug resistance for 2 or more antibiotics was more frequent in the isolates from 2011-2012 (23.4%, 22/94) than in the isolates from 2009-2010 (16.9%, 12/71), but there was no statistical significance (P<0.082). Only 1 strain exhibited multi-drug resistance to clarithromycin, amoxicillin, metronidazole, and levofloxacin (Table 2). The antibiotic resistance rates for the isolates from the 2009-2010 group against clarithromycin, amoxicillin, tetracycline, metronidazole, and levofloxacin were 7.0% (5/71), 2.8% (2/71), 0% (0/71), 45.1% (32/71), and 26.8% (19/71), respectively, and those for the isolates from the 2011-2012 group were 16.0% (15/94), 2.1% (2/94), 0% (0/94), 56.3% (53/94), and 22.3% (21/94), respectively. The rate of H. pylori resistance to clarithromycin and metronidazole increased from 7.0% to 16.0% and from 45.1% to 56.3% for the 2 periods, respectively (Fig. 1), although the increase was not statistically significant. When the MIC distribution profiles for the 2 study periods were compared, the MICs for clarithromycin showed notable differences between the susceptible and resistant strains (Fig. 2). The MIC of the clarithromycin-susceptible strains was less than 0.125 µg/mL. Resistance to tetracycline was not detected in any strain (based on a cut-off of ≥ 4 µg/mL). The MIC range of tetracycline was 0.031-2 µg/mL. The MIC of metronidazole varied widely (8-256 µg/mL). Multi-drug resistance for 2 or more antibiotics was more frequent in the isolates from 2011-2012 (23.4%, 22/94) than in the isolates from 2009-2010 (16.9%, 12/71), but there was no statistical significance (P<0.082). Only 1 strain exhibited multi-drug resistance to clarithromycin, amoxicillin, metronidazole, and levofloxacin (Table 2). 2. Effect of antibiotic resistance on H. pylori eradication rates Of the 165 patients studied during the 2009-2012 period, 66 patients who were subjected to the first-line therapy were followed up for H. pylori eradication after treatment. Among these 66 patients, no significant differences were found with respect to sex, age, and endoscopic diagnosis. Eradication of H. pylori was successful in 50 of these 66 patients (75.8%). The effects of antibiotic resistance on H. pylori eradication rates are shown in Table 3. Eradication rates were 79.3% (46/58) for the clarithromycin-susceptible and amoxicillin-susceptible strains, and 100% (1/1) for the clarithromycin-susceptible and amoxicillin-resistant strains. A significant difference was observed between the eradication rates for the clarithromycin-resistant (42.9%, 3/7) and the clarithromycin-sensitive (79.7%, 47/59) strains (P<0.001) . Second-line therapy was prescribed for the 16 patients in whom first-line therapy failed. The eradication rates for the tetracycline-susceptible and metronidazole-susceptible strains and the tetracycline-susceptible and metronidazole-resistant strains were 50.0% (4/8) and 25.0% (2/8), respectively (P<0.32). Of the 165 patients studied during the 2009-2012 period, 66 patients who were subjected to the first-line therapy were followed up for H. pylori eradication after treatment. Among these 66 patients, no significant differences were found with respect to sex, age, and endoscopic diagnosis. Eradication of H. pylori was successful in 50 of these 66 patients (75.8%). The effects of antibiotic resistance on H. pylori eradication rates are shown in Table 3. Eradication rates were 79.3% (46/58) for the clarithromycin-susceptible and amoxicillin-susceptible strains, and 100% (1/1) for the clarithromycin-susceptible and amoxicillin-resistant strains. A significant difference was observed between the eradication rates for the clarithromycin-resistant (42.9%, 3/7) and the clarithromycin-sensitive (79.7%, 47/59) strains (P<0.001) . Second-line therapy was prescribed for the 16 patients in whom first-line therapy failed. The eradication rates for the tetracycline-susceptible and metronidazole-susceptible strains and the tetracycline-susceptible and metronidazole-resistant strains were 50.0% (4/8) and 25.0% (2/8), respectively (P<0.32). 1. Antibiotic resistance of H. pylori: The antibiotic resistance rates for the isolates from the 2009-2010 group against clarithromycin, amoxicillin, tetracycline, metronidazole, and levofloxacin were 7.0% (5/71), 2.8% (2/71), 0% (0/71), 45.1% (32/71), and 26.8% (19/71), respectively, and those for the isolates from the 2011-2012 group were 16.0% (15/94), 2.1% (2/94), 0% (0/94), 56.3% (53/94), and 22.3% (21/94), respectively. The rate of H. pylori resistance to clarithromycin and metronidazole increased from 7.0% to 16.0% and from 45.1% to 56.3% for the 2 periods, respectively (Fig. 1), although the increase was not statistically significant. When the MIC distribution profiles for the 2 study periods were compared, the MICs for clarithromycin showed notable differences between the susceptible and resistant strains (Fig. 2). The MIC of the clarithromycin-susceptible strains was less than 0.125 µg/mL. Resistance to tetracycline was not detected in any strain (based on a cut-off of ≥ 4 µg/mL). The MIC range of tetracycline was 0.031-2 µg/mL. The MIC of metronidazole varied widely (8-256 µg/mL). Multi-drug resistance for 2 or more antibiotics was more frequent in the isolates from 2011-2012 (23.4%, 22/94) than in the isolates from 2009-2010 (16.9%, 12/71), but there was no statistical significance (P<0.082). Only 1 strain exhibited multi-drug resistance to clarithromycin, amoxicillin, metronidazole, and levofloxacin (Table 2). 2. Effect of antibiotic resistance on H. pylori eradication rates: Of the 165 patients studied during the 2009-2012 period, 66 patients who were subjected to the first-line therapy were followed up for H. pylori eradication after treatment. Among these 66 patients, no significant differences were found with respect to sex, age, and endoscopic diagnosis. Eradication of H. pylori was successful in 50 of these 66 patients (75.8%). The effects of antibiotic resistance on H. pylori eradication rates are shown in Table 3. Eradication rates were 79.3% (46/58) for the clarithromycin-susceptible and amoxicillin-susceptible strains, and 100% (1/1) for the clarithromycin-susceptible and amoxicillin-resistant strains. A significant difference was observed between the eradication rates for the clarithromycin-resistant (42.9%, 3/7) and the clarithromycin-sensitive (79.7%, 47/59) strains (P<0.001) . Second-line therapy was prescribed for the 16 patients in whom first-line therapy failed. The eradication rates for the tetracycline-susceptible and metronidazole-susceptible strains and the tetracycline-susceptible and metronidazole-resistant strains were 50.0% (4/8) and 25.0% (2/8), respectively (P<0.32). DISCUSSION: Recently, H. pylori eradication rates of 70-95% have been reported [4-6]. Failure of eradication may be attributed to increase in antibiotic resistance associated with problems in patient compliance, such as difficulties in taking drugs, or side effects [11-13]. In this study, the antimicrobial susceptibility test was conducted for H. pylori strains isolated from a single center over 2 periods, followed by examination of the factors affecting failure. Clarithromycin resistance rates increased from 7.0% in the 2009-2010 patient group to 16.0% in the 2011-2012 patient group. These rates are slightly lower than that reported in a previous study, which showed that the overall frequency of clarithromycin-resistant H. pylori in 2008 was 21.6% [14]. This discrepancy is conceivably attributable to regional differences in the location of the studies. The primary factor influencing clarithromycin resistance is known to be the A2142-4 point mutation in the 23S rRNA [14-17]. Amoxicillin resistance rates decreased slightly from 2.8% (2009-2011) to 2.1% (2011-2012). Resistance to tetracycline was not detected in any strain when the cut-off was set at ≥ 4 µg/mL, and the MICs were as low as 0.031-2 µg/mL. Recently, tetracycline resistance rates of 0-36% have been reported. However, as with clarithromycin resistance, the differences could be due to regional differences [18, 19]. Metronidazole resistance rates were higher than those for all other antibiotics, ranging from 45.1% in the 2009-2010 group to 56.3% in the 2011-2012 group, and the MIC of metronidazole was the highest among all the studied antibiotics (8-256 µg/mL). Levofloxacin resistance rates decreased slightly from 26.8% in 2009-2010 to 23.3% in 2011-2012. This finding is consistent with that reported in a previous domestic study (resistance rates decreased from 26.3% to 22.5%) [18, 19]. The continuous increase in levofloxacin resistance warrants the use of rescue therapy based on the results of antimicrobial susceptibility tests. Although differences in resistance rates to the 5 antibiotics in the 2 study periods failed to reach statistical significance, increases in the resistance to clarithromycin and metronidazole were identified. Moreover, multi-drug resistance for 2 or more antibiotics increased from 16.9% (12/71) in 2009-2010 group to 23.4% (22/94) in the 2011-2012 group, but there was no statistical significance (P<0.082). The overall eradication rate in patients who received first-line therapy with clarithromycin and amoxicillin was 75.8% (50/66), ranging from 78.1% in 2009-2010 to 73.5% in 2011-2012 (data not shown). The eradication rate for clarithromycin-resistant strains (42.9%, 3/7) was significantly lower than that for the clarithromycin-susceptible strains (79.7%, 47/59) (P<0.001). These results indicate that resistance to clarithromycin is a critical factor in the effectiveness of eradication with the first-line regimen. Complete eradication rate was only 79.3% in strains susceptible to both clarithromycin and amoxicillin with first-line therapy (Table 3). Thus failure rate of 20.7% may be attributable to problems with patient compliance; however, a more extensive follow-up survey is needed to confirm it. Eradication was successful in 6 of 16 patients, who received second-line therapy including tetracycline and metronidazole; second-line therapy failed in 4 patients, and data were unavailable for the remaining 6 patients. Previous treatment histories of the 16 patients in the second-line treatment group were as follows: clarithromycin treatment for eradication of H. pylori, 2 patients; treatment for liver cirrhosis, 1 patient; and poor compliance, 1 patient. Specific histories were unavailable for the remaining 12 patients. Although failure of eradication is generally linked to antibiotic resistance, increase in antibiotic resistance does not always correlate with decrease in eradication rates; therefore, further studies are required to identify other factors affecting eradication rates. In conclusion, the effectiveness of eradication using first-line therapy with clarithromycin and amoxicillin decreased, especially in the clarithromycin-resistant group, and clarithromycin resistance was considered crucial for the eradication of H. pylori. This result suggested that eradication of H. pylori is greatly dependent on periodic monitoring of antimicrobial susceptibility, which is necessary for selection of an appropriate antibiotic regimen.

Background: Clarithromycin, amoxicillin, metronidazole, tetracycline, and levofloxacin have been commonly used for the eradication of Helicobacter pylori. We compared the change in antibiotic resistance of H. pylori strains during two separate periods and investigated the effect of antibiotic resistance on H. pylori eradication. Methods: H. pylori strains were isolated from 71 patients between 2009 and 2010 and from 94 patients between 2011 and 2012. The distribution of minimal inhibitory concentration (MIC) of 5 antibiotics was assessed using the agar dilution method, and H. pylori eradication based on the antimicrobial susceptibility of the isolates was investigated retrospectively. Results: Antibiotic resistance rate against clarithromycin, amoxicillin, tetracycline, metronidazole, and levofloxacin for the 2009-2010 isolates were 7.0% (5/71), 2.8% (2/71), 0% (0/71), 45.1% (32/71), and 26.8% (19/71), respectively, and for the 2011-2012 isolates were 16.0% (15/94), 2.1% (2/94), 0% (0/94), 56.3% (53/94), and 22.3% (21/94), respectively. Multi-drug resistance for 2 or more antibiotics increased slightly from 16.9% (12/71) in the 2009-2010 isolates to 23.4% (22/94) in the 2011-2012 isolates. In follow-up testing of 66 patients, first-line treatment successfully eradicated H. pylori in 50 patients (75.8%) and failed in 4 of 7 patients (57.1%) in a clarithromycin-resistant and amoxicillin-susceptible group. Conclusions: We observed an increase in resistance to clarithromycin and an overall increase in multi-drug resistance during the 2 study periods. The effectiveness of the eradication regimen was low with combinations of clarithromycin and amoxicillin, particularly in the clarithromycin-resistant group. Thus, eradication of H. pylori depends upon periodic monitoring of antimicrobial susceptibility.

[CONTENT] Helicobacter pylori | Antibiotic resistance | Eradication [SUMMARY]

[CONTENT] Helicobacter pylori | Antibiotic resistance | Eradication [SUMMARY]

[CONTENT] Helicobacter pylori | Antibiotic resistance | Eradication [SUMMARY]

[CONTENT] Helicobacter pylori | Antibiotic resistance | Eradication [SUMMARY]

[CONTENT] Adult | Aged | Anti-Bacterial Agents | Drug Resistance, Multiple, Bacterial | Female | Helicobacter Infections | Helicobacter pylori | Humans | Male | Microbial Sensitivity Tests | Middle Aged | Peptic Ulcer | Republic of Korea | Retrospective Studies | Treatment Outcome [SUMMARY]

[CONTENT] Adult | Aged | Anti-Bacterial Agents | Drug Resistance, Multiple, Bacterial | Female | Helicobacter Infections | Helicobacter pylori | Humans | Male | Microbial Sensitivity Tests | Middle Aged | Peptic Ulcer | Republic of Korea | Retrospective Studies | Treatment Outcome [SUMMARY]

[CONTENT] Adult | Aged | Anti-Bacterial Agents | Drug Resistance, Multiple, Bacterial | Female | Helicobacter Infections | Helicobacter pylori | Humans | Male | Microbial Sensitivity Tests | Middle Aged | Peptic Ulcer | Republic of Korea | Retrospective Studies | Treatment Outcome [SUMMARY]

[CONTENT] Adult | Aged | Anti-Bacterial Agents | Drug Resistance, Multiple, Bacterial | Female | Helicobacter Infections | Helicobacter pylori | Humans | Male | Microbial Sensitivity Tests | Middle Aged | Peptic Ulcer | Republic of Korea | Retrospective Studies | Treatment Outcome [SUMMARY]

[CONTENT] pylori eradication fails | pylori patients treatment | followed pylori eradication | pylori resistance clarithromycin | clarithromycin resistant pylori [SUMMARY]

[CONTENT] pylori eradication fails | pylori patients treatment | followed pylori eradication | pylori resistance clarithromycin | clarithromycin resistant pylori [SUMMARY]

[CONTENT] pylori eradication fails | pylori patients treatment | followed pylori eradication | pylori resistance clarithromycin | clarithromycin resistant pylori [SUMMARY]

[CONTENT] pylori eradication fails | pylori patients treatment | followed pylori eradication | pylori resistance clarithromycin | clarithromycin resistant pylori [SUMMARY]

[CONTENT] clarithromycin | resistance | pylori | ml | eradication | µg | µg ml | patients | metronidazole | strains [SUMMARY]

[CONTENT] clarithromycin | resistance | pylori | ml | eradication | µg | µg ml | patients | metronidazole | strains [SUMMARY]

[CONTENT] clarithromycin | resistance | pylori | ml | eradication | µg | µg ml | patients | metronidazole | strains [SUMMARY]

[CONTENT] clarithromycin | resistance | pylori | ml | eradication | µg | µg ml | patients | metronidazole | strains [SUMMARY]

[CONTENT] eradication | line treatment regimen | 2003 | treatment regimen | gastritis | rates | resistance | regimen | pylori | line [SUMMARY]

[CONTENT] mg | µg | µg ml | ml | aldrich | bid | mg bid | sigma | sigma aldrich | patients [SUMMARY]

[CONTENT] susceptible | clarithromycin | 71 | 94 | eradication | strains | rates | resistance | eradication rates | patients [SUMMARY]

[CONTENT] ml | µg | µg ml | eradication | clarithromycin | resistance | patients | pylori | rates | susceptible [SUMMARY]

[CONTENT] Clarithromycin ||| two [SUMMARY]

[CONTENT] 71 | between 2009 and 2010 | 94 | between 2011 and 2012 ||| MIC | 5 [SUMMARY]

[CONTENT] 2009-2010 | 7.0% | 5/71 | 2.8% | 2/71 | 0% | 0/71 | 45.1% | 32/71 | 26.8% | 19/71 | 2011-2012 | 16.0% | 15/94 | 2.1% | 2/94 | 0% | 0/94 | 56.3% | 22.3% | 21/94 ||| 2 | 16.9% | 12/71 | 2009-2010 | 23.4% | 22/94 | 2011-2012 ||| 66 | first | 50 | 75.8% | 4 | 7 | 57.1% [SUMMARY]

[CONTENT] Clarithromycin ||| two ||| 71 | between 2009 and 2010 | 94 | between 2011 and 2012 ||| MIC | 5 ||| ||| 2009-2010 | 7.0% | 5/71 | 2.8% | 2/71 | 0% | 0/71 | 45.1% | 32/71 | 26.8% | 19/71 | 2011-2012 | 16.0% | 15/94 | 2.1% | 2/94 | 0% | 0/94 | 56.3% | 22.3% | 21/94 ||| 2 | 16.9% | 12/71 | 2009-2010 | 23.4% | 22/94 | 2011-2012 ||| 66 | first | 50 | 75.8% | 4 | 7 | 57.1% ||| 2 ||| ||| [SUMMARY]

LPS preconditioning redirects TLR signaling following stroke: TRIF-IRF3 plays a seminal role in mediating tolerance to ischemic injury.

21999375

Toll-like receptor 4 (TLR4) is activated in response to cerebral ischemia leading to substantial brain damage. In contrast, mild activation of TLR4 by preconditioning with low dose exposure to lipopolysaccharide (LPS) prior to cerebral ischemia dramatically improves outcome by reprogramming the signaling response to injury. This suggests that TLR4 signaling can be altered to induce an endogenously neuroprotective phenotype. However, the TLR4 signaling events involved in this neuroprotective response are poorly understood. Here we define several molecular mediators of the primary signaling cascades induced by LPS preconditioning that give rise to the reprogrammed response to cerebral ischemia and confer the neuroprotective phenotype.

BACKGROUND

C57BL6 mice were preconditioned with low dose LPS prior to transient middle cerebral artery occlusion (MCAO). Cortical tissue and blood were collected following MCAO. Microarray and qtPCR were performed to analyze gene expression associated with TLR4 signaling. EMSA and DNA binding ELISA were used to evaluate NFκB and IRF3 activity. Protein expression was determined using Western blot or ELISA. MyD88-/- and TRIF-/- mice were utilized to evaluate signaling in LPS preconditioning-induced neuroprotection.

METHODS

Gene expression analyses revealed that LPS preconditioning resulted in a marked upregulation of anti-inflammatory/type I IFN-associated genes following ischemia while pro-inflammatory genes induced following ischemia were present but not differentially modulated by LPS. Interestingly, although expression of pro-inflammatory genes was observed, there was decreased activity of NFκB p65 and increased presence of NFκB inhibitors, including Ship1, Tollip, and p105, in LPS-preconditioned mice following stroke. In contrast, IRF3 activity was enhanced in LPS-preconditioned mice following stroke. TRIF and MyD88 deficient mice revealed that neuroprotection induced by LPS depends on TLR4 signaling via TRIF, which activates IRF3, but does not depend on MyD88 signaling.

RESULTS

Our results characterize several critical mediators of the TLR4 signaling events associated with neuroprotection. LPS preconditioning redirects TLR4 signaling in response to stroke through suppression of NFκB activity, enhanced IRF3 activity, and increased anti-inflammatory/type I IFN gene expression. Interestingly, this protective phenotype does not require the suppression of pro-inflammatory mediators. Furthermore, our results highlight a critical role for TRIF-IRF3 signaling as the governing mechanism in the neuroprotective response to stroke.

CONCLUSION

3217906

Background

Stroke is one of the leading causes of death and the leading cause of morbidity in the United States [1]. The inflammatory response to stroke substantially exacerbates ischemic damage. The acute activation of the NFκB transcription factor has been linked to the inflammatory response to stroke [2] and suppression of NFκB activity following stroke has been found to reduce damage [3]. NFκB activation can lead to the dramatic upregulation of inflammatory molecules and cytokines including TNFα, IL6, IL1β, and COX2 [2]. The source of these inflammatory molecules in the acute response to stroke appears to stem from the cells of the central nervous system (CNS), including neurons and glial cells [2]. The cells in the CNS play a particularly dominant role early in the response to ischemia because infiltrating leukocytes do not appear in substantial numbers in the brain until 24 hr following injury [4]. Stroke also induces an acute inflammatory response in the circulating blood. Inflammatory cytokine and chemokine levels, including IL6, IL1β, MCP-1 and TNFα are elevated in the circulation following stroke [5]. This suggests there is an intimate relationship between responses in the brain and blood following stroke--responses that result in increased inflammation. Toll-like receptors (TLRs), traditionally considered innate immune receptors, signal through the adaptor proteins MyD88 and TRIF to activate NFκB and interferon regulatory factors (IRFs). It has been shown recently that TLRs become activated in response to endogenous ligands, known as damage associated molecular patterns (DAMPs), released during injury. Interestingly, animals deficient in TLR2 or TLR4 have significantly reduced infarct sizes in several models of stroke [6-11]. This suggests that TLR2 and TLR4 activation in response to ischemic injury exacerbates damage. In addition, a recent investigation in humans showed that the inflammatory responses to stroke in the blood were linked to increased TLR2 and TLR4 expression on hematopoetic cells and associated with worse outcome in stroke [12]. The detrimental effect of TLR signaling is associated with the pathways that lead to NFκB activation and pro-inflammatory responses. In contrast, TLR signaling pathways that activate IRFs can induce anti-inflammatory mediators and type I IFNs that have been associated with neuroprotection [13,14]. Thus, in TLR signaling there is a fine balance between pathways leading to injury or protection. TLR ligands have been a major source of interest as preconditioning agents for prophylactic therapy against ischemic injury. Such therapies would target a population of patients that are at risk of ischemia in the setting of surgery [15-18]. Preconditioning with low doses of ligands for TLR2, TLR4, and TLR9 all successfully reduce infarct size in experimental models of stroke [19-21], including a recent study showing that a TLR9 ligand is neuroprotective in a nonhuman primate model of stroke [22]. Emerging evidence suggests that TLR-induced neuroprotection occurs by reprogramming the genomic response to the DAMPs, which are produced in response to ischemic injury. In this reprogrammed state, the resultant pathway activation of TLR4 signaling preferentially leads to IRF-mediated gene expression [13,14]. However, whether TLR preconditioning affects NFκB activity and pro-inflammatory signaling is unknown. As yet, a complete analysis of the characteristic TLR signaling responses to stroke following preconditioning has not been reported. The objective of this study is to utilize LPS preconditioning followed by transient middle cerebral artery occlusion (MCAO) to elucidate the reprogrammed TLR response to stroke and to determine the major pathways involved in producing the neuroprotective phenotype. Here we show that preconditioning against ischemia using LPS leads to suppressed NFκB activity--although pro-inflammatory gene expression does not appear to be attenuated. We also demonstrate that LPS-preconditioned mice have enhanced IRF3 activity and anti-inflammatory/type I IFN gene expression in the ischemic brain. This expression pattern was recapitulated in the blood where plasma levels of pro-inflammatory cytokine proteins were comparable in LPS-preconditioned and control mice while IRF-associated proteins were enhanced in LPS preconditioned mice. To our knowledge, we provide the first evidence that protection due to LPS preconditioning stems from TRIF signaling, the cascade that is associated with IRF3 activation, and is independent of MyD88 signaling. These molecular features suggest that, following stroke, signaling is directed away from NFκB activity and towards a dominant TRIF-IRF3 response. Understanding the endogenous signaling events that promote protection against ischemic injury is integral to the identification and development of novel stroke therapeutics. In particular, the evidence presented here further highlights a key role for IRF3 activity in the protective response to stroke.

Methods

Animals C57Bl/6J mice (male, 8-12 weeks) were purchased from Jackson Laboratories (West Sacramento, CA). C57Bl/6J-Ticam1LPS2/J (TRIF-/-) mice were also obtained from Jackson Laboratories. MyD88-/- mice were a kind gift of Dr. Shizuo Akira (Osaka University, Osaka Japan) and were bred in our facility. All mice were housed in an American Association for Laboratory Animal Care-approved facility. Procedures were conducted according to Oregon Health & Science University, Institutional Animal Care and Use Committee, and National Institutes of Health guidelines. C57Bl/6J mice (male, 8-12 weeks) were purchased from Jackson Laboratories (West Sacramento, CA). C57Bl/6J-Ticam1LPS2/J (TRIF-/-) mice were also obtained from Jackson Laboratories. MyD88-/- mice were a kind gift of Dr. Shizuo Akira (Osaka University, Osaka Japan) and were bred in our facility. All mice were housed in an American Association for Laboratory Animal Care-approved facility. Procedures were conducted according to Oregon Health & Science University, Institutional Animal Care and Use Committee, and National Institutes of Health guidelines. LPS treatment Mice were preconditioned with LPS (0.2 or 0.8 mg/kg, Escherichia coli serotype 0111:B4; Sigma) or saline by one subcutaneous injection, unless otherwise indicated, 72 hr prior to MCAO. Each new lot of LPS was titrated for the optimal dose that confers neuroprotection. No differences were observed in the genomic responses to LPS for each dose used and route of administration (subcutaneous or intraperitoneal, data not shown). Mice were preconditioned with LPS (0.2 or 0.8 mg/kg, Escherichia coli serotype 0111:B4; Sigma) or saline by one subcutaneous injection, unless otherwise indicated, 72 hr prior to MCAO. Each new lot of LPS was titrated for the optimal dose that confers neuroprotection. No differences were observed in the genomic responses to LPS for each dose used and route of administration (subcutaneous or intraperitoneal, data not shown). Middle Cerebral Artery Occlusion (MCAO) Mice were anesthetized with isoflurane (1.5-2%) and subjected to MCAO using the monofilament suture method described previously [23]. Briefly, a silicone-coated 7-0 monofilament nylon surgical suture was threaded through the external carotid artery to the internal carotid artery to block the middle cerebral artery, and maintained intraluminally for 40 to 60 min. The suture was then removed to restore blood flow. The duration of occlusion was optimized based on the specific surgeon who performed the MCAO to yield comparable infarct sizes in the saline treated control animals (~35-40%). The selected duration of MCAO was held constant within experiments. Cerebral blood flow (CBF) was monitored throughout surgery by laser doppler flowmetry. Any mouse that did not maintain a CBF during occlusion of <25% of baseline was excluded from the study. The reduction of CBF was comparable in LPS and saline preconditioned mice in response to MCAO. Body temperature was monitored and maintained at 37°C with a thermostat-controlled heating pad. Infarct measurements were made using triphenyltetrazolium chloride (TTC) staining of 1 mm coronal brain sections. Mice were anesthetized with isoflurane (1.5-2%) and subjected to MCAO using the monofilament suture method described previously [23]. Briefly, a silicone-coated 7-0 monofilament nylon surgical suture was threaded through the external carotid artery to the internal carotid artery to block the middle cerebral artery, and maintained intraluminally for 40 to 60 min. The suture was then removed to restore blood flow. The duration of occlusion was optimized based on the specific surgeon who performed the MCAO to yield comparable infarct sizes in the saline treated control animals (~35-40%). The selected duration of MCAO was held constant within experiments. Cerebral blood flow (CBF) was monitored throughout surgery by laser doppler flowmetry. Any mouse that did not maintain a CBF during occlusion of <25% of baseline was excluded from the study. The reduction of CBF was comparable in LPS and saline preconditioned mice in response to MCAO. Body temperature was monitored and maintained at 37°C with a thermostat-controlled heating pad. Infarct measurements were made using triphenyltetrazolium chloride (TTC) staining of 1 mm coronal brain sections. Tissue collection Under deep isoflurane anesthesia, approximately ~0.5-1.0 ml of blood was collected via cardiac puncture in a heparinized syringe. Subsequently, the mice were perfused with heparinized (2 U/ml) saline followed by rapid removal of the brain. The olfactory bulbs were removed and the first 4 mm of tissue was collected beginning at the rostral end. The striatum was dissected and removed and the remaining cortex was utilized for RNA isolation or protein extraction. The collected blood was centrifuged at 5000 × g for 20 min to obtain plasma that was stored at -80°C. Under deep isoflurane anesthesia, approximately ~0.5-1.0 ml of blood was collected via cardiac puncture in a heparinized syringe. Subsequently, the mice were perfused with heparinized (2 U/ml) saline followed by rapid removal of the brain. The olfactory bulbs were removed and the first 4 mm of tissue was collected beginning at the rostral end. The striatum was dissected and removed and the remaining cortex was utilized for RNA isolation or protein extraction. The collected blood was centrifuged at 5000 × g for 20 min to obtain plasma that was stored at -80°C. Genomic profiling of TLR associated mediators For the genes displayed in Figure 1, the transcript expression levels were determined as previously described from our microarray experiments examining the brain cortical response to stroke and 3 different preconditioning stimuli [14]. In brief, total RNA was isolated from the ipsilateral cortex (n = 4 mice/treatment/timepoint), using the Qiagen Rneasy Lipid Mini Kit (Qiagen). Microarray assays were performed in the Affymetrix Microarray Core of the Oregon Health & Science University Gene Microarray Shared Resource. RNA samples were labeled using the NuGEN Ovation Biotin RNA Amplification and Labeling System_V1. Hybridization was performed as described in the Affymetrix technical manual (Affymetrix) with modification as recommended for the Ovation labeling protocol (NuGEN Technologies). Labeled cRNA target was quality-checked based on yield and size distribution. Quality-tested samples were hybridized to the MOE430 2.0 array. The array image was processed with Affymetrix GeneChip Operating Software (GCOS). Affymetrix CEL files were then uploaded into GeneSifter (http://www.genesifter.net) and normalized using RMA. Microarray analysis of anti-inflammatory/type I IFN and pro-inflammatory gene expression. Microarray analysis revealed enhanced anti-inflammatory/type I IFN and comparable pro-inflammatory gene expression profiles in the brain of LPS-preconditioned (0.2 mg/kg, intraperitoneal injection) mice following 45 min MCAO. Heatmap representing level of gene expression immediately prior to (0 hr) MCAO and 3 and 24 hr post MCAO; n = 4/treatment/timepoint. Lt. Select anti-inflammatory/type I IFN genes. Rt. Select pro-inflammatory genes. Color scale from green to red represents relative decreased or increased gene expression levels, respectively. For the genes displayed in Figure 1, the transcript expression levels were determined as previously described from our microarray experiments examining the brain cortical response to stroke and 3 different preconditioning stimuli [14]. In brief, total RNA was isolated from the ipsilateral cortex (n = 4 mice/treatment/timepoint), using the Qiagen Rneasy Lipid Mini Kit (Qiagen). Microarray assays were performed in the Affymetrix Microarray Core of the Oregon Health & Science University Gene Microarray Shared Resource. RNA samples were labeled using the NuGEN Ovation Biotin RNA Amplification and Labeling System_V1. Hybridization was performed as described in the Affymetrix technical manual (Affymetrix) with modification as recommended for the Ovation labeling protocol (NuGEN Technologies). Labeled cRNA target was quality-checked based on yield and size distribution. Quality-tested samples were hybridized to the MOE430 2.0 array. The array image was processed with Affymetrix GeneChip Operating Software (GCOS). Affymetrix CEL files were then uploaded into GeneSifter (http://www.genesifter.net) and normalized using RMA. Microarray analysis of anti-inflammatory/type I IFN and pro-inflammatory gene expression. Microarray analysis revealed enhanced anti-inflammatory/type I IFN and comparable pro-inflammatory gene expression profiles in the brain of LPS-preconditioned (0.2 mg/kg, intraperitoneal injection) mice following 45 min MCAO. Heatmap representing level of gene expression immediately prior to (0 hr) MCAO and 3 and 24 hr post MCAO; n = 4/treatment/timepoint. Lt. Select anti-inflammatory/type I IFN genes. Rt. Select pro-inflammatory genes. Color scale from green to red represents relative decreased or increased gene expression levels, respectively. RNA isolation, Reverse Transcription, and qtPCR RNA was isolated from cortical tissue 72 hr post injection or from ipsilateral cortical tissue at 3 or 24 hr following MCAO (n ≥ 4 mice/treatment/timepoint) using a Lipid Mini RNA isolation kit (Qiagen). Reverse transcription was performed on 2 μg of RNA using Omniscript (Qiagen). Quantitative PCR was performed using Taqman Gene Expression Assays (Applied Biosystems) for each gene of interest on an ABI Prism 7700. Results were normalized to β-Actin expression and analyzed relative to their saline preconditioned counterparts. The relative quantification of the gene of interest was determined using the comparative CT method (2-DDCt). RNA was isolated from cortical tissue 72 hr post injection or from ipsilateral cortical tissue at 3 or 24 hr following MCAO (n ≥ 4 mice/treatment/timepoint) using a Lipid Mini RNA isolation kit (Qiagen). Reverse transcription was performed on 2 μg of RNA using Omniscript (Qiagen). Quantitative PCR was performed using Taqman Gene Expression Assays (Applied Biosystems) for each gene of interest on an ABI Prism 7700. Results were normalized to β-Actin expression and analyzed relative to their saline preconditioned counterparts. The relative quantification of the gene of interest was determined using the comparative CT method (2-DDCt). Western Blot Protein extraction was performed as described previously [24] with some modifications. Briefly, tissue samples (n ≥ 4 mice/treatment/timepoint) were dissected from the ipsilateral cortex and lysed in a buffer containing a protease inhibitor cocktail (Roche). Protein concentrations were determined using the BCA method (Pierce-Endogen). Protein samples (50 μg) were denatured in a gel-loading buffer (Bio-Rad Laboratories) at 100°C for 5 min and then loaded onto 12% Bis-Tris polyacrylamide gels (Bio-Rad Laboratories). Following electrophoresis, proteins were transferred to polyvinylodene difluoride membranes (Bio-Rad Laboratories) and incubated with primary antibodies for Ship-1 (Santa Cruz, sc8425), Tollip (AbCam, Ab37155), p105 (Santa Cruz, sc7178), or β-Actin (Santa Cruz, sc1616R) at 4°C overnight. Membranes were then incubated with horseradish peroxidase conjugated anti-rabbit, anti-goat, or anti-mouse antibody (Santa Cruz Biotechnology) and detected by chemiluminescence (NEN Life Science Products) and exposed to Kodak film (Biomax). Images were captured using an Epson scanner and the densitometry of the gel bands, including β-Actin loading control, was analyzed using ImageJ (NIH). Protein extraction was performed as described previously [24] with some modifications. Briefly, tissue samples (n ≥ 4 mice/treatment/timepoint) were dissected from the ipsilateral cortex and lysed in a buffer containing a protease inhibitor cocktail (Roche). Protein concentrations were determined using the BCA method (Pierce-Endogen). Protein samples (50 μg) were denatured in a gel-loading buffer (Bio-Rad Laboratories) at 100°C for 5 min and then loaded onto 12% Bis-Tris polyacrylamide gels (Bio-Rad Laboratories). Following electrophoresis, proteins were transferred to polyvinylodene difluoride membranes (Bio-Rad Laboratories) and incubated with primary antibodies for Ship-1 (Santa Cruz, sc8425), Tollip (AbCam, Ab37155), p105 (Santa Cruz, sc7178), or β-Actin (Santa Cruz, sc1616R) at 4°C overnight. Membranes were then incubated with horseradish peroxidase conjugated anti-rabbit, anti-goat, or anti-mouse antibody (Santa Cruz Biotechnology) and detected by chemiluminescence (NEN Life Science Products) and exposed to Kodak film (Biomax). Images were captured using an Epson scanner and the densitometry of the gel bands, including β-Actin loading control, was analyzed using ImageJ (NIH). Electrophoretic Mobility Shift Assay (EMSA) Nuclear protein extracts (n = 4 mice/treatment/timepoint) were prepared from tissue dissected from the ipsilateral cortex. Homogenized tissue was incubated in Buffer A (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 1 mM DTT, 1 mM PMSF) for 5 min on ice and centrifuged at 3000 rpm for 5 min at 4°C. The pellets were washed in Buffer B (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 0.5% NP-40, 1 mM DTT, 1 mM PMSF), resuspended in Buffer C (250 mM Tris pH7.8, 60 mM KCl, 1 mM DTT, 1 mM PMSF), and freeze-thawed 3 times in liquid nitrogen. All buffers contained a protease inhibitor cocktail (Roche). After centrifuging at 10,000 rpm for 10 min at 4°C, the supernatant was collected and stored as nuclear extract at -80°C. Nuclear protein concentrations were determined using the BCA method (Pierce-Endogen). EMSAs were performed using the Promega Gel Shift Assay System according to the manufacturer's instructions. Briefly, 15 μg of nuclear protein was incubated with 32P-labeled NFκB consensus oligonucleotide (Promega), either with or without unlabeled competitor oligonucleotide, unlabeled noncompetitor oligonucleotide, or anti-p65 antibody (Santa Cruz). Samples were electrophoresed on a 4% acrylamide gel, dried and exposed to phosphorimager overnight. The densitometry of the gel bands was analyzed using scanning integrated optical density software (ImageJ). IRF3 Activity Assay Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Nuclear protein extracts (n = 4 mice/treatment/timepoint) were prepared from tissue dissected from the ipsilateral cortex. Homogenized tissue was incubated in Buffer A (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 1 mM DTT, 1 mM PMSF) for 5 min on ice and centrifuged at 3000 rpm for 5 min at 4°C. The pellets were washed in Buffer B (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 0.5% NP-40, 1 mM DTT, 1 mM PMSF), resuspended in Buffer C (250 mM Tris pH7.8, 60 mM KCl, 1 mM DTT, 1 mM PMSF), and freeze-thawed 3 times in liquid nitrogen. All buffers contained a protease inhibitor cocktail (Roche). After centrifuging at 10,000 rpm for 10 min at 4°C, the supernatant was collected and stored as nuclear extract at -80°C. Nuclear protein concentrations were determined using the BCA method (Pierce-Endogen). EMSAs were performed using the Promega Gel Shift Assay System according to the manufacturer's instructions. Briefly, 15 μg of nuclear protein was incubated with 32P-labeled NFκB consensus oligonucleotide (Promega), either with or without unlabeled competitor oligonucleotide, unlabeled noncompetitor oligonucleotide, or anti-p65 antibody (Santa Cruz). Samples were electrophoresed on a 4% acrylamide gel, dried and exposed to phosphorimager overnight. The densitometry of the gel bands was analyzed using scanning integrated optical density software (ImageJ). IRF3 Activity Assay Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Cytokine Analysis Cytokine/chemokine analysis for IL1β, IL1α, MIP-1α, MCP-1, RANTES, and IL10 was performed on plasma samples (n ≥ 3 mice/treatment/timepoint) using a multiplex ELISA (Quansys). An IFNβ ELISA (PBL Interferon Source) was used to measure plasma levels of IFNβ. Cytokine/chemokine analysis for IL1β, IL1α, MIP-1α, MCP-1, RANTES, and IL10 was performed on plasma samples (n ≥ 3 mice/treatment/timepoint) using a multiplex ELISA (Quansys). An IFNβ ELISA (PBL Interferon Source) was used to measure plasma levels of IFNβ. Statistical Analysis Data is represented as mean ± SEM. The n for each experiment is greater than or equal to 3, as specified in each figure. Statistical analysis was performed using GraphPad Prism5 software. Two-way ANOVA with Bonferroni Post Hoc test and Student's t-test were utilized as specified. Significance was determined as p < 0.05. Data is represented as mean ± SEM. The n for each experiment is greater than or equal to 3, as specified in each figure. Statistical analysis was performed using GraphPad Prism5 software. Two-way ANOVA with Bonferroni Post Hoc test and Student's t-test were utilized as specified. Significance was determined as p < 0.05.

Conclusions

KBV performed experiments, collected data, conceived of the idea for the paper, and wrote the manuscript. SLS worked on the microarray, provided guidance in the production of data, and edited the paper. BJM performed experiments and contributed to the writing of the Methods section. RWK performed experiments. NL performed the MCAO surgeries. MSP provided critical guidance and worked on the manuscript. All authors approved of the final manuscript.

Background: Stroke is one of the leading causes of death and the leading cause of morbidity in the United States [1]. The inflammatory response to stroke substantially exacerbates ischemic damage. The acute activation of the NFκB transcription factor has been linked to the inflammatory response to stroke [2] and suppression of NFκB activity following stroke has been found to reduce damage [3]. NFκB activation can lead to the dramatic upregulation of inflammatory molecules and cytokines including TNFα, IL6, IL1β, and COX2 [2]. The source of these inflammatory molecules in the acute response to stroke appears to stem from the cells of the central nervous system (CNS), including neurons and glial cells [2]. The cells in the CNS play a particularly dominant role early in the response to ischemia because infiltrating leukocytes do not appear in substantial numbers in the brain until 24 hr following injury [4]. Stroke also induces an acute inflammatory response in the circulating blood. Inflammatory cytokine and chemokine levels, including IL6, IL1β, MCP-1 and TNFα are elevated in the circulation following stroke [5]. This suggests there is an intimate relationship between responses in the brain and blood following stroke--responses that result in increased inflammation. Toll-like receptors (TLRs), traditionally considered innate immune receptors, signal through the adaptor proteins MyD88 and TRIF to activate NFκB and interferon regulatory factors (IRFs). It has been shown recently that TLRs become activated in response to endogenous ligands, known as damage associated molecular patterns (DAMPs), released during injury. Interestingly, animals deficient in TLR2 or TLR4 have significantly reduced infarct sizes in several models of stroke [6-11]. This suggests that TLR2 and TLR4 activation in response to ischemic injury exacerbates damage. In addition, a recent investigation in humans showed that the inflammatory responses to stroke in the blood were linked to increased TLR2 and TLR4 expression on hematopoetic cells and associated with worse outcome in stroke [12]. The detrimental effect of TLR signaling is associated with the pathways that lead to NFκB activation and pro-inflammatory responses. In contrast, TLR signaling pathways that activate IRFs can induce anti-inflammatory mediators and type I IFNs that have been associated with neuroprotection [13,14]. Thus, in TLR signaling there is a fine balance between pathways leading to injury or protection. TLR ligands have been a major source of interest as preconditioning agents for prophylactic therapy against ischemic injury. Such therapies would target a population of patients that are at risk of ischemia in the setting of surgery [15-18]. Preconditioning with low doses of ligands for TLR2, TLR4, and TLR9 all successfully reduce infarct size in experimental models of stroke [19-21], including a recent study showing that a TLR9 ligand is neuroprotective in a nonhuman primate model of stroke [22]. Emerging evidence suggests that TLR-induced neuroprotection occurs by reprogramming the genomic response to the DAMPs, which are produced in response to ischemic injury. In this reprogrammed state, the resultant pathway activation of TLR4 signaling preferentially leads to IRF-mediated gene expression [13,14]. However, whether TLR preconditioning affects NFκB activity and pro-inflammatory signaling is unknown. As yet, a complete analysis of the characteristic TLR signaling responses to stroke following preconditioning has not been reported. The objective of this study is to utilize LPS preconditioning followed by transient middle cerebral artery occlusion (MCAO) to elucidate the reprogrammed TLR response to stroke and to determine the major pathways involved in producing the neuroprotective phenotype. Here we show that preconditioning against ischemia using LPS leads to suppressed NFκB activity--although pro-inflammatory gene expression does not appear to be attenuated. We also demonstrate that LPS-preconditioned mice have enhanced IRF3 activity and anti-inflammatory/type I IFN gene expression in the ischemic brain. This expression pattern was recapitulated in the blood where plasma levels of pro-inflammatory cytokine proteins were comparable in LPS-preconditioned and control mice while IRF-associated proteins were enhanced in LPS preconditioned mice. To our knowledge, we provide the first evidence that protection due to LPS preconditioning stems from TRIF signaling, the cascade that is associated with IRF3 activation, and is independent of MyD88 signaling. These molecular features suggest that, following stroke, signaling is directed away from NFκB activity and towards a dominant TRIF-IRF3 response. Understanding the endogenous signaling events that promote protection against ischemic injury is integral to the identification and development of novel stroke therapeutics. In particular, the evidence presented here further highlights a key role for IRF3 activity in the protective response to stroke. Methods: Animals C57Bl/6J mice (male, 8-12 weeks) were purchased from Jackson Laboratories (West Sacramento, CA). C57Bl/6J-Ticam1LPS2/J (TRIF-/-) mice were also obtained from Jackson Laboratories. MyD88-/- mice were a kind gift of Dr. Shizuo Akira (Osaka University, Osaka Japan) and were bred in our facility. All mice were housed in an American Association for Laboratory Animal Care-approved facility. Procedures were conducted according to Oregon Health & Science University, Institutional Animal Care and Use Committee, and National Institutes of Health guidelines. C57Bl/6J mice (male, 8-12 weeks) were purchased from Jackson Laboratories (West Sacramento, CA). C57Bl/6J-Ticam1LPS2/J (TRIF-/-) mice were also obtained from Jackson Laboratories. MyD88-/- mice were a kind gift of Dr. Shizuo Akira (Osaka University, Osaka Japan) and were bred in our facility. All mice were housed in an American Association for Laboratory Animal Care-approved facility. Procedures were conducted according to Oregon Health & Science University, Institutional Animal Care and Use Committee, and National Institutes of Health guidelines. LPS treatment Mice were preconditioned with LPS (0.2 or 0.8 mg/kg, Escherichia coli serotype 0111:B4; Sigma) or saline by one subcutaneous injection, unless otherwise indicated, 72 hr prior to MCAO. Each new lot of LPS was titrated for the optimal dose that confers neuroprotection. No differences were observed in the genomic responses to LPS for each dose used and route of administration (subcutaneous or intraperitoneal, data not shown). Mice were preconditioned with LPS (0.2 or 0.8 mg/kg, Escherichia coli serotype 0111:B4; Sigma) or saline by one subcutaneous injection, unless otherwise indicated, 72 hr prior to MCAO. Each new lot of LPS was titrated for the optimal dose that confers neuroprotection. No differences were observed in the genomic responses to LPS for each dose used and route of administration (subcutaneous or intraperitoneal, data not shown). Middle Cerebral Artery Occlusion (MCAO) Mice were anesthetized with isoflurane (1.5-2%) and subjected to MCAO using the monofilament suture method described previously [23]. Briefly, a silicone-coated 7-0 monofilament nylon surgical suture was threaded through the external carotid artery to the internal carotid artery to block the middle cerebral artery, and maintained intraluminally for 40 to 60 min. The suture was then removed to restore blood flow. The duration of occlusion was optimized based on the specific surgeon who performed the MCAO to yield comparable infarct sizes in the saline treated control animals (~35-40%). The selected duration of MCAO was held constant within experiments. Cerebral blood flow (CBF) was monitored throughout surgery by laser doppler flowmetry. Any mouse that did not maintain a CBF during occlusion of <25% of baseline was excluded from the study. The reduction of CBF was comparable in LPS and saline preconditioned mice in response to MCAO. Body temperature was monitored and maintained at 37°C with a thermostat-controlled heating pad. Infarct measurements were made using triphenyltetrazolium chloride (TTC) staining of 1 mm coronal brain sections. Mice were anesthetized with isoflurane (1.5-2%) and subjected to MCAO using the monofilament suture method described previously [23]. Briefly, a silicone-coated 7-0 monofilament nylon surgical suture was threaded through the external carotid artery to the internal carotid artery to block the middle cerebral artery, and maintained intraluminally for 40 to 60 min. The suture was then removed to restore blood flow. The duration of occlusion was optimized based on the specific surgeon who performed the MCAO to yield comparable infarct sizes in the saline treated control animals (~35-40%). The selected duration of MCAO was held constant within experiments. Cerebral blood flow (CBF) was monitored throughout surgery by laser doppler flowmetry. Any mouse that did not maintain a CBF during occlusion of <25% of baseline was excluded from the study. The reduction of CBF was comparable in LPS and saline preconditioned mice in response to MCAO. Body temperature was monitored and maintained at 37°C with a thermostat-controlled heating pad. Infarct measurements were made using triphenyltetrazolium chloride (TTC) staining of 1 mm coronal brain sections. Tissue collection Under deep isoflurane anesthesia, approximately ~0.5-1.0 ml of blood was collected via cardiac puncture in a heparinized syringe. Subsequently, the mice were perfused with heparinized (2 U/ml) saline followed by rapid removal of the brain. The olfactory bulbs were removed and the first 4 mm of tissue was collected beginning at the rostral end. The striatum was dissected and removed and the remaining cortex was utilized for RNA isolation or protein extraction. The collected blood was centrifuged at 5000 × g for 20 min to obtain plasma that was stored at -80°C. Under deep isoflurane anesthesia, approximately ~0.5-1.0 ml of blood was collected via cardiac puncture in a heparinized syringe. Subsequently, the mice were perfused with heparinized (2 U/ml) saline followed by rapid removal of the brain. The olfactory bulbs were removed and the first 4 mm of tissue was collected beginning at the rostral end. The striatum was dissected and removed and the remaining cortex was utilized for RNA isolation or protein extraction. The collected blood was centrifuged at 5000 × g for 20 min to obtain plasma that was stored at -80°C. Genomic profiling of TLR associated mediators For the genes displayed in Figure 1, the transcript expression levels were determined as previously described from our microarray experiments examining the brain cortical response to stroke and 3 different preconditioning stimuli [14]. In brief, total RNA was isolated from the ipsilateral cortex (n = 4 mice/treatment/timepoint), using the Qiagen Rneasy Lipid Mini Kit (Qiagen). Microarray assays were performed in the Affymetrix Microarray Core of the Oregon Health & Science University Gene Microarray Shared Resource. RNA samples were labeled using the NuGEN Ovation Biotin RNA Amplification and Labeling System_V1. Hybridization was performed as described in the Affymetrix technical manual (Affymetrix) with modification as recommended for the Ovation labeling protocol (NuGEN Technologies). Labeled cRNA target was quality-checked based on yield and size distribution. Quality-tested samples were hybridized to the MOE430 2.0 array. The array image was processed with Affymetrix GeneChip Operating Software (GCOS). Affymetrix CEL files were then uploaded into GeneSifter (http://www.genesifter.net) and normalized using RMA. Microarray analysis of anti-inflammatory/type I IFN and pro-inflammatory gene expression. Microarray analysis revealed enhanced anti-inflammatory/type I IFN and comparable pro-inflammatory gene expression profiles in the brain of LPS-preconditioned (0.2 mg/kg, intraperitoneal injection) mice following 45 min MCAO. Heatmap representing level of gene expression immediately prior to (0 hr) MCAO and 3 and 24 hr post MCAO; n = 4/treatment/timepoint. Lt. Select anti-inflammatory/type I IFN genes. Rt. Select pro-inflammatory genes. Color scale from green to red represents relative decreased or increased gene expression levels, respectively. For the genes displayed in Figure 1, the transcript expression levels were determined as previously described from our microarray experiments examining the brain cortical response to stroke and 3 different preconditioning stimuli [14]. In brief, total RNA was isolated from the ipsilateral cortex (n = 4 mice/treatment/timepoint), using the Qiagen Rneasy Lipid Mini Kit (Qiagen). Microarray assays were performed in the Affymetrix Microarray Core of the Oregon Health & Science University Gene Microarray Shared Resource. RNA samples were labeled using the NuGEN Ovation Biotin RNA Amplification and Labeling System_V1. Hybridization was performed as described in the Affymetrix technical manual (Affymetrix) with modification as recommended for the Ovation labeling protocol (NuGEN Technologies). Labeled cRNA target was quality-checked based on yield and size distribution. Quality-tested samples were hybridized to the MOE430 2.0 array. The array image was processed with Affymetrix GeneChip Operating Software (GCOS). Affymetrix CEL files were then uploaded into GeneSifter (http://www.genesifter.net) and normalized using RMA. Microarray analysis of anti-inflammatory/type I IFN and pro-inflammatory gene expression. Microarray analysis revealed enhanced anti-inflammatory/type I IFN and comparable pro-inflammatory gene expression profiles in the brain of LPS-preconditioned (0.2 mg/kg, intraperitoneal injection) mice following 45 min MCAO. Heatmap representing level of gene expression immediately prior to (0 hr) MCAO and 3 and 24 hr post MCAO; n = 4/treatment/timepoint. Lt. Select anti-inflammatory/type I IFN genes. Rt. Select pro-inflammatory genes. Color scale from green to red represents relative decreased or increased gene expression levels, respectively. RNA isolation, Reverse Transcription, and qtPCR RNA was isolated from cortical tissue 72 hr post injection or from ipsilateral cortical tissue at 3 or 24 hr following MCAO (n ≥ 4 mice/treatment/timepoint) using a Lipid Mini RNA isolation kit (Qiagen). Reverse transcription was performed on 2 μg of RNA using Omniscript (Qiagen). Quantitative PCR was performed using Taqman Gene Expression Assays (Applied Biosystems) for each gene of interest on an ABI Prism 7700. Results were normalized to β-Actin expression and analyzed relative to their saline preconditioned counterparts. The relative quantification of the gene of interest was determined using the comparative CT method (2-DDCt). RNA was isolated from cortical tissue 72 hr post injection or from ipsilateral cortical tissue at 3 or 24 hr following MCAO (n ≥ 4 mice/treatment/timepoint) using a Lipid Mini RNA isolation kit (Qiagen). Reverse transcription was performed on 2 μg of RNA using Omniscript (Qiagen). Quantitative PCR was performed using Taqman Gene Expression Assays (Applied Biosystems) for each gene of interest on an ABI Prism 7700. Results were normalized to β-Actin expression and analyzed relative to their saline preconditioned counterparts. The relative quantification of the gene of interest was determined using the comparative CT method (2-DDCt). Western Blot Protein extraction was performed as described previously [24] with some modifications. Briefly, tissue samples (n ≥ 4 mice/treatment/timepoint) were dissected from the ipsilateral cortex and lysed in a buffer containing a protease inhibitor cocktail (Roche). Protein concentrations were determined using the BCA method (Pierce-Endogen). Protein samples (50 μg) were denatured in a gel-loading buffer (Bio-Rad Laboratories) at 100°C for 5 min and then loaded onto 12% Bis-Tris polyacrylamide gels (Bio-Rad Laboratories). Following electrophoresis, proteins were transferred to polyvinylodene difluoride membranes (Bio-Rad Laboratories) and incubated with primary antibodies for Ship-1 (Santa Cruz, sc8425), Tollip (AbCam, Ab37155), p105 (Santa Cruz, sc7178), or β-Actin (Santa Cruz, sc1616R) at 4°C overnight. Membranes were then incubated with horseradish peroxidase conjugated anti-rabbit, anti-goat, or anti-mouse antibody (Santa Cruz Biotechnology) and detected by chemiluminescence (NEN Life Science Products) and exposed to Kodak film (Biomax). Images were captured using an Epson scanner and the densitometry of the gel bands, including β-Actin loading control, was analyzed using ImageJ (NIH). Protein extraction was performed as described previously [24] with some modifications. Briefly, tissue samples (n ≥ 4 mice/treatment/timepoint) were dissected from the ipsilateral cortex and lysed in a buffer containing a protease inhibitor cocktail (Roche). Protein concentrations were determined using the BCA method (Pierce-Endogen). Protein samples (50 μg) were denatured in a gel-loading buffer (Bio-Rad Laboratories) at 100°C for 5 min and then loaded onto 12% Bis-Tris polyacrylamide gels (Bio-Rad Laboratories). Following electrophoresis, proteins were transferred to polyvinylodene difluoride membranes (Bio-Rad Laboratories) and incubated with primary antibodies for Ship-1 (Santa Cruz, sc8425), Tollip (AbCam, Ab37155), p105 (Santa Cruz, sc7178), or β-Actin (Santa Cruz, sc1616R) at 4°C overnight. Membranes were then incubated with horseradish peroxidase conjugated anti-rabbit, anti-goat, or anti-mouse antibody (Santa Cruz Biotechnology) and detected by chemiluminescence (NEN Life Science Products) and exposed to Kodak film (Biomax). Images were captured using an Epson scanner and the densitometry of the gel bands, including β-Actin loading control, was analyzed using ImageJ (NIH). Electrophoretic Mobility Shift Assay (EMSA) Nuclear protein extracts (n = 4 mice/treatment/timepoint) were prepared from tissue dissected from the ipsilateral cortex. Homogenized tissue was incubated in Buffer A (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 1 mM DTT, 1 mM PMSF) for 5 min on ice and centrifuged at 3000 rpm for 5 min at 4°C. The pellets were washed in Buffer B (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 0.5% NP-40, 1 mM DTT, 1 mM PMSF), resuspended in Buffer C (250 mM Tris pH7.8, 60 mM KCl, 1 mM DTT, 1 mM PMSF), and freeze-thawed 3 times in liquid nitrogen. All buffers contained a protease inhibitor cocktail (Roche). After centrifuging at 10,000 rpm for 10 min at 4°C, the supernatant was collected and stored as nuclear extract at -80°C. Nuclear protein concentrations were determined using the BCA method (Pierce-Endogen). EMSAs were performed using the Promega Gel Shift Assay System according to the manufacturer's instructions. Briefly, 15 μg of nuclear protein was incubated with 32P-labeled NFκB consensus oligonucleotide (Promega), either with or without unlabeled competitor oligonucleotide, unlabeled noncompetitor oligonucleotide, or anti-p65 antibody (Santa Cruz). Samples were electrophoresed on a 4% acrylamide gel, dried and exposed to phosphorimager overnight. The densitometry of the gel bands was analyzed using scanning integrated optical density software (ImageJ). IRF3 Activity Assay Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Nuclear protein extracts (n = 4 mice/treatment/timepoint) were prepared from tissue dissected from the ipsilateral cortex. Homogenized tissue was incubated in Buffer A (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 1 mM DTT, 1 mM PMSF) for 5 min on ice and centrifuged at 3000 rpm for 5 min at 4°C. The pellets were washed in Buffer B (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 0.5% NP-40, 1 mM DTT, 1 mM PMSF), resuspended in Buffer C (250 mM Tris pH7.8, 60 mM KCl, 1 mM DTT, 1 mM PMSF), and freeze-thawed 3 times in liquid nitrogen. All buffers contained a protease inhibitor cocktail (Roche). After centrifuging at 10,000 rpm for 10 min at 4°C, the supernatant was collected and stored as nuclear extract at -80°C. Nuclear protein concentrations were determined using the BCA method (Pierce-Endogen). EMSAs were performed using the Promega Gel Shift Assay System according to the manufacturer's instructions. Briefly, 15 μg of nuclear protein was incubated with 32P-labeled NFκB consensus oligonucleotide (Promega), either with or without unlabeled competitor oligonucleotide, unlabeled noncompetitor oligonucleotide, or anti-p65 antibody (Santa Cruz). Samples were electrophoresed on a 4% acrylamide gel, dried and exposed to phosphorimager overnight. The densitometry of the gel bands was analyzed using scanning integrated optical density software (ImageJ). IRF3 Activity Assay Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Cytokine Analysis Cytokine/chemokine analysis for IL1β, IL1α, MIP-1α, MCP-1, RANTES, and IL10 was performed on plasma samples (n ≥ 3 mice/treatment/timepoint) using a multiplex ELISA (Quansys). An IFNβ ELISA (PBL Interferon Source) was used to measure plasma levels of IFNβ. Cytokine/chemokine analysis for IL1β, IL1α, MIP-1α, MCP-1, RANTES, and IL10 was performed on plasma samples (n ≥ 3 mice/treatment/timepoint) using a multiplex ELISA (Quansys). An IFNβ ELISA (PBL Interferon Source) was used to measure plasma levels of IFNβ. Statistical Analysis Data is represented as mean ± SEM. The n for each experiment is greater than or equal to 3, as specified in each figure. Statistical analysis was performed using GraphPad Prism5 software. Two-way ANOVA with Bonferroni Post Hoc test and Student's t-test were utilized as specified. Significance was determined as p < 0.05. Data is represented as mean ± SEM. The n for each experiment is greater than or equal to 3, as specified in each figure. Statistical analysis was performed using GraphPad Prism5 software. Two-way ANOVA with Bonferroni Post Hoc test and Student's t-test were utilized as specified. Significance was determined as p < 0.05. Animals: C57Bl/6J mice (male, 8-12 weeks) were purchased from Jackson Laboratories (West Sacramento, CA). C57Bl/6J-Ticam1LPS2/J (TRIF-/-) mice were also obtained from Jackson Laboratories. MyD88-/- mice were a kind gift of Dr. Shizuo Akira (Osaka University, Osaka Japan) and were bred in our facility. All mice were housed in an American Association for Laboratory Animal Care-approved facility. Procedures were conducted according to Oregon Health & Science University, Institutional Animal Care and Use Committee, and National Institutes of Health guidelines. LPS treatment: Mice were preconditioned with LPS (0.2 or 0.8 mg/kg, Escherichia coli serotype 0111:B4; Sigma) or saline by one subcutaneous injection, unless otherwise indicated, 72 hr prior to MCAO. Each new lot of LPS was titrated for the optimal dose that confers neuroprotection. No differences were observed in the genomic responses to LPS for each dose used and route of administration (subcutaneous or intraperitoneal, data not shown). Middle Cerebral Artery Occlusion (MCAO): Mice were anesthetized with isoflurane (1.5-2%) and subjected to MCAO using the monofilament suture method described previously [23]. Briefly, a silicone-coated 7-0 monofilament nylon surgical suture was threaded through the external carotid artery to the internal carotid artery to block the middle cerebral artery, and maintained intraluminally for 40 to 60 min. The suture was then removed to restore blood flow. The duration of occlusion was optimized based on the specific surgeon who performed the MCAO to yield comparable infarct sizes in the saline treated control animals (~35-40%). The selected duration of MCAO was held constant within experiments. Cerebral blood flow (CBF) was monitored throughout surgery by laser doppler flowmetry. Any mouse that did not maintain a CBF during occlusion of <25% of baseline was excluded from the study. The reduction of CBF was comparable in LPS and saline preconditioned mice in response to MCAO. Body temperature was monitored and maintained at 37°C with a thermostat-controlled heating pad. Infarct measurements were made using triphenyltetrazolium chloride (TTC) staining of 1 mm coronal brain sections. Tissue collection: Under deep isoflurane anesthesia, approximately ~0.5-1.0 ml of blood was collected via cardiac puncture in a heparinized syringe. Subsequently, the mice were perfused with heparinized (2 U/ml) saline followed by rapid removal of the brain. The olfactory bulbs were removed and the first 4 mm of tissue was collected beginning at the rostral end. The striatum was dissected and removed and the remaining cortex was utilized for RNA isolation or protein extraction. The collected blood was centrifuged at 5000 × g for 20 min to obtain plasma that was stored at -80°C. Genomic profiling of TLR associated mediators: For the genes displayed in Figure 1, the transcript expression levels were determined as previously described from our microarray experiments examining the brain cortical response to stroke and 3 different preconditioning stimuli [14]. In brief, total RNA was isolated from the ipsilateral cortex (n = 4 mice/treatment/timepoint), using the Qiagen Rneasy Lipid Mini Kit (Qiagen). Microarray assays were performed in the Affymetrix Microarray Core of the Oregon Health & Science University Gene Microarray Shared Resource. RNA samples were labeled using the NuGEN Ovation Biotin RNA Amplification and Labeling System_V1. Hybridization was performed as described in the Affymetrix technical manual (Affymetrix) with modification as recommended for the Ovation labeling protocol (NuGEN Technologies). Labeled cRNA target was quality-checked based on yield and size distribution. Quality-tested samples were hybridized to the MOE430 2.0 array. The array image was processed with Affymetrix GeneChip Operating Software (GCOS). Affymetrix CEL files were then uploaded into GeneSifter (http://www.genesifter.net) and normalized using RMA. Microarray analysis of anti-inflammatory/type I IFN and pro-inflammatory gene expression. Microarray analysis revealed enhanced anti-inflammatory/type I IFN and comparable pro-inflammatory gene expression profiles in the brain of LPS-preconditioned (0.2 mg/kg, intraperitoneal injection) mice following 45 min MCAO. Heatmap representing level of gene expression immediately prior to (0 hr) MCAO and 3 and 24 hr post MCAO; n = 4/treatment/timepoint. Lt. Select anti-inflammatory/type I IFN genes. Rt. Select pro-inflammatory genes. Color scale from green to red represents relative decreased or increased gene expression levels, respectively. RNA isolation, Reverse Transcription, and qtPCR: RNA was isolated from cortical tissue 72 hr post injection or from ipsilateral cortical tissue at 3 or 24 hr following MCAO (n ≥ 4 mice/treatment/timepoint) using a Lipid Mini RNA isolation kit (Qiagen). Reverse transcription was performed on 2 μg of RNA using Omniscript (Qiagen). Quantitative PCR was performed using Taqman Gene Expression Assays (Applied Biosystems) for each gene of interest on an ABI Prism 7700. Results were normalized to β-Actin expression and analyzed relative to their saline preconditioned counterparts. The relative quantification of the gene of interest was determined using the comparative CT method (2-DDCt). Western Blot: Protein extraction was performed as described previously [24] with some modifications. Briefly, tissue samples (n ≥ 4 mice/treatment/timepoint) were dissected from the ipsilateral cortex and lysed in a buffer containing a protease inhibitor cocktail (Roche). Protein concentrations were determined using the BCA method (Pierce-Endogen). Protein samples (50 μg) were denatured in a gel-loading buffer (Bio-Rad Laboratories) at 100°C for 5 min and then loaded onto 12% Bis-Tris polyacrylamide gels (Bio-Rad Laboratories). Following electrophoresis, proteins were transferred to polyvinylodene difluoride membranes (Bio-Rad Laboratories) and incubated with primary antibodies for Ship-1 (Santa Cruz, sc8425), Tollip (AbCam, Ab37155), p105 (Santa Cruz, sc7178), or β-Actin (Santa Cruz, sc1616R) at 4°C overnight. Membranes were then incubated with horseradish peroxidase conjugated anti-rabbit, anti-goat, or anti-mouse antibody (Santa Cruz Biotechnology) and detected by chemiluminescence (NEN Life Science Products) and exposed to Kodak film (Biomax). Images were captured using an Epson scanner and the densitometry of the gel bands, including β-Actin loading control, was analyzed using ImageJ (NIH). Electrophoretic Mobility Shift Assay (EMSA): Nuclear protein extracts (n = 4 mice/treatment/timepoint) were prepared from tissue dissected from the ipsilateral cortex. Homogenized tissue was incubated in Buffer A (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 1 mM DTT, 1 mM PMSF) for 5 min on ice and centrifuged at 3000 rpm for 5 min at 4°C. The pellets were washed in Buffer B (10 mM Hepes-KOH pH7.9, 60 mM KCl, 1 mM EDTA, 0.5% NP-40, 1 mM DTT, 1 mM PMSF), resuspended in Buffer C (250 mM Tris pH7.8, 60 mM KCl, 1 mM DTT, 1 mM PMSF), and freeze-thawed 3 times in liquid nitrogen. All buffers contained a protease inhibitor cocktail (Roche). After centrifuging at 10,000 rpm for 10 min at 4°C, the supernatant was collected and stored as nuclear extract at -80°C. Nuclear protein concentrations were determined using the BCA method (Pierce-Endogen). EMSAs were performed using the Promega Gel Shift Assay System according to the manufacturer's instructions. Briefly, 15 μg of nuclear protein was incubated with 32P-labeled NFκB consensus oligonucleotide (Promega), either with or without unlabeled competitor oligonucleotide, unlabeled noncompetitor oligonucleotide, or anti-p65 antibody (Santa Cruz). Samples were electrophoresed on a 4% acrylamide gel, dried and exposed to phosphorimager overnight. The densitometry of the gel bands was analyzed using scanning integrated optical density software (ImageJ). IRF3 Activity Assay Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. IRF3 Activity Assay: Nuclear protein (n ≥ 4 mice/treatment/timepoint) was isolated from fresh cortical tissue at 72 hr post injection and from ipsilateral cortices at 3 or 24 hr following MCAO using a Nuclear Extraction Kit (Active Motif, Inc.). IRF3 activity was measured using 10 μg of nuclear protein in an IRF3 activity ELISA (Active Motif, Inc), that utilizes colorimetric detection of active IRF3 bound to immobilized oligonucleotides. Cytokine Analysis: Cytokine/chemokine analysis for IL1β, IL1α, MIP-1α, MCP-1, RANTES, and IL10 was performed on plasma samples (n ≥ 3 mice/treatment/timepoint) using a multiplex ELISA (Quansys). An IFNβ ELISA (PBL Interferon Source) was used to measure plasma levels of IFNβ. Statistical Analysis: Data is represented as mean ± SEM. The n for each experiment is greater than or equal to 3, as specified in each figure. Statistical analysis was performed using GraphPad Prism5 software. Two-way ANOVA with Bonferroni Post Hoc test and Student's t-test were utilized as specified. Significance was determined as p < 0.05. Results: LPS preconditioning does not affect inflammatory gene expression in the brain following stroke We used gene microarray analysis to elucidate the pattern of inflammatory or anti-inflammatory/type I IFN gene expression in the brain following stroke. In the setting of stroke, LPS preconditioned animals exhibited regulation of a number of genes typically found downstream of TLR signaling. The inflammatory profile reveals that the gene regulation is similar at each timepoint following stroke in LPS or saline preconditioned animals (Figure 1, Rt.). There is no evidence of inflammatory gene expression present immediately prior to stroke (Figure 1, Rt. 0 hr). At 3 hr post MCAO, several inflammatory genes are upregulated including IL6, IL1β, Ptgs2/COX2, and CCL2/MCP-1 (Figure 1, Rt.) and this upregulation is sustained at the 24 hr timepoint following MCAO (Figure 1, Rt.). TNFα, which is commonly shown to be upregulated following MCAO [25,26], only shows marginal levels of upregulation in LPS or saline preconditioned mice (Figure 1, Rt.). To confirm the microarray results, a subset of selected inflammatory genes including IL6, IL1β, COX2, and TNFα, were analyzed using qtPCR. Each of these genes were upregulated following MCAO in LPS and saline preconditioned mice, but there were no significant differences based on treatment at 3 hr (data not shown) and 24 hr (Figure 2) following MCAO. Enhanced anti-inflammatory/type I IFN gene expression but comparable pro-inflammatory gene expression in LPS-preconditioned mice post MCAO. Gene regulation 24 hr post MCAO measured by qtPCR reveals that anti-inflammatory/type I IFN-associated genes TGFβ, RANTES, and IFIT1 are significantly upregulated in LPS preconditioned mice compared to saline. Pro-inflammatory genes IL6, IL1β, COX2, and TNFα show similar regulation in LPS and saline preconditioned mice. These results confirm the gene microarray data. Samples from mice receiving a 45 (LPS: 0.2 mg/kg) or 60 min (LPS: 0.8 mg/kg) MCAO were combined due to comparable gene regulation (see methods). ND = not detected. Student's t-test, LPS vs. saline 24 hr post MCAO, **p < 0.01, n ≥ 4 per treatment. We used gene microarray analysis to elucidate the pattern of inflammatory or anti-inflammatory/type I IFN gene expression in the brain following stroke. In the setting of stroke, LPS preconditioned animals exhibited regulation of a number of genes typically found downstream of TLR signaling. The inflammatory profile reveals that the gene regulation is similar at each timepoint following stroke in LPS or saline preconditioned animals (Figure 1, Rt.). There is no evidence of inflammatory gene expression present immediately prior to stroke (Figure 1, Rt. 0 hr). At 3 hr post MCAO, several inflammatory genes are upregulated including IL6, IL1β, Ptgs2/COX2, and CCL2/MCP-1 (Figure 1, Rt.) and this upregulation is sustained at the 24 hr timepoint following MCAO (Figure 1, Rt.). TNFα, which is commonly shown to be upregulated following MCAO [25,26], only shows marginal levels of upregulation in LPS or saline preconditioned mice (Figure 1, Rt.). To confirm the microarray results, a subset of selected inflammatory genes including IL6, IL1β, COX2, and TNFα, were analyzed using qtPCR. Each of these genes were upregulated following MCAO in LPS and saline preconditioned mice, but there were no significant differences based on treatment at 3 hr (data not shown) and 24 hr (Figure 2) following MCAO. Enhanced anti-inflammatory/type I IFN gene expression but comparable pro-inflammatory gene expression in LPS-preconditioned mice post MCAO. Gene regulation 24 hr post MCAO measured by qtPCR reveals that anti-inflammatory/type I IFN-associated genes TGFβ, RANTES, and IFIT1 are significantly upregulated in LPS preconditioned mice compared to saline. Pro-inflammatory genes IL6, IL1β, COX2, and TNFα show similar regulation in LPS and saline preconditioned mice. These results confirm the gene microarray data. Samples from mice receiving a 45 (LPS: 0.2 mg/kg) or 60 min (LPS: 0.8 mg/kg) MCAO were combined due to comparable gene regulation (see methods). ND = not detected. Student's t-test, LPS vs. saline 24 hr post MCAO, **p < 0.01, n ≥ 4 per treatment. LPS preconditioning upregulates anti-inflammatory/type I IFN gene expression in the brain following MCAO Although pro-inflammatory gene expression was not differentially modulated in preconditioned animals, microarray results revealed that the majority of the anti-inflammatory/type I IFN genes, such as TGFβ, IL1 receptor antagonist (IL1rn), RANTES, and IRF7, were upregulated following stroke in the brains of LPS versus saline preconditioned mice (Figure 1, Lt.). IL10 gene expression was not detected at any timepoint (Figure 1, Lt.). TGFβ, IL10, RANTES, and IFIT1 were selected for qtPCR analysis. TGFβ, RANTES, and IFIT1 were significantly upregulated in the LPS-preconditioned brain compared to saline 24 hr following stroke (Figure 2). RANTES was also significantly upregulated at 3 hr following stroke in LPS-preconditioned mice compared to saline (data not shown). IL10 expression remained undetectable by qtPCR analysis (Figure 2), suggesting that IL10 mRNA is not present at these timepoints in the brain following stroke. These qtPCR results confirm the gene expression profile observed on the microarray. Taken together, these data indicate an enhanced anti-inflammatory/type I IFN gene expression profile in the brain of LPS-preconditioned animals following MCAO while the inflammatory gene expression is unaffected. Although pro-inflammatory gene expression was not differentially modulated in preconditioned animals, microarray results revealed that the majority of the anti-inflammatory/type I IFN genes, such as TGFβ, IL1 receptor antagonist (IL1rn), RANTES, and IRF7, were upregulated following stroke in the brains of LPS versus saline preconditioned mice (Figure 1, Lt.). IL10 gene expression was not detected at any timepoint (Figure 1, Lt.). TGFβ, IL10, RANTES, and IFIT1 were selected for qtPCR analysis. TGFβ, RANTES, and IFIT1 were significantly upregulated in the LPS-preconditioned brain compared to saline 24 hr following stroke (Figure 2). RANTES was also significantly upregulated at 3 hr following stroke in LPS-preconditioned mice compared to saline (data not shown). IL10 expression remained undetectable by qtPCR analysis (Figure 2), suggesting that IL10 mRNA is not present at these timepoints in the brain following stroke. These qtPCR results confirm the gene expression profile observed on the microarray. Taken together, these data indicate an enhanced anti-inflammatory/type I IFN gene expression profile in the brain of LPS-preconditioned animals following MCAO while the inflammatory gene expression is unaffected. NFκB activity is suppressed in the brain of LPS-preconditioned animals 24 hr post MCAO NFκB activity is associated with damage and inflammation in the brain that occurs in response to stroke. We used EMSAs to evaluate the activity of the NFκB subunit p65 in the brain following stroke. The results indicated that LPS and saline preconditioned mice have comparable NFκB activity at 3 hr post MCAO (Figure 3A). However, at 24 hr post MCAO, LPS-preconditioned animals have significantly suppressed NFκB activity compared to saline preconditioned mice (Figure 3A). NFκB is suppressed 24 hr post MCAO in LPS-preconditioned mice. (A) Nuclear protein obtained from ipsilateral cortices was used to measure p65 activity by EMSA analysis. EMSA gel of pooled samples (n = 4) following 45 min MCAO for saline and LPS preconditioned (0.8 mg/kg) mice (Lt.). Quantification of band intensity of individual mice following MCAO (Rt.). NFκB is significantly decreased in LPS-preconditioned mice 24 hr post MCAO compared to saline. Supershift assay confirmed specificity for p65 oligos (data not shown). (B) Ship1 and Tollip mRNA are significantly upregulated 24 hr post 60 minute MCAO in LPS-preconditioned (0.8 mg/kg) mice compared to saline, n ≥ 4 per treatment/timepoint. (C) Western blot for Ship1 and Tollip and relative band quantification showing significant upregulation of Ship1 and Tollip protein 24 hr post 45 min MCAO in LPS-preconditioned mice (0.8 mg/kg), n ≥ 3 per treatment/timepoint. (D) Western blot and relative band quantification for p105 at 24 hr post 45 minute MCAO showing significant upregulation in LPS-preconditioned (0.8 mg/kg) mice, n ≥ 3 per treatment/timepoint. (A) Two-Way ANOVA, Bonferroni Post Hoc, *p < 0.05. (B-D) Student's t-test, LPS vs. saline, **p < 0.01. Ship1 and Tollip are cytosolic molecules that inhibit TLR signaling, which leads to the suppression of NFκB activity. We found that Ship1 and Tollip mRNA are upregulated in the brain 72 hr post injection versus saline controls (2.06 ± 0.27 and 2.31 ± 0.35, respectively) but not at 3 hr post stroke (1.09 ± 0.10 and 1.05 ± 0.09, respectively). However, by 24 hr post MCAO, Ship1 and Tollip mRNA are significantly enhanced in the brain of LPS-preconditioned mice compared to saline controls (2.62 ± 0.84 and 4.01 ± 1.06, respectively, Figure 3B). Ship1 protein is not upregulated at 72 hr post injection (Fold change vs. saline: 1.01 ± 0.32), but becomes significantly enhanced in LPS-preconditioned mice at 3 hr (Fold change vs saline: 1.83 ± 0.13) and at 24 hr (Fold change vs. saline: 8.81 ± 1.54, Figure 3C) post MCAO. Tollip protein is not affected by LPS preconditioning at 72 hr post injection or 3 hr post MCAO (Fold change vs. saline: 1.42 ± 0.10 and 0.83 ± 0.10, respectively), but it is significantly enhanced in LPS-preconditioned mice compared to saline controls at 24 hr post MCAO (Fold change vs. saline: 2.42 ± 0.20, Figure 3C). Additionally, the p50 precursor protein p105, which inhibits NFκB activity by acting like an IκB molecule by sequestering NFκB in the cytosol [27,28], was significantly upregulated 24 hr post stroke in LPS-preconditioned mice compared to saline (Figure 3D). Thus, despite the upregulation of inflammatory genes, the activity of NFκB is suppressed in the late-phase of the neuroprotective response of LPS-preconditioned mice. NFκB activity is associated with damage and inflammation in the brain that occurs in response to stroke. We used EMSAs to evaluate the activity of the NFκB subunit p65 in the brain following stroke. The results indicated that LPS and saline preconditioned mice have comparable NFκB activity at 3 hr post MCAO (Figure 3A). However, at 24 hr post MCAO, LPS-preconditioned animals have significantly suppressed NFκB activity compared to saline preconditioned mice (Figure 3A). NFκB is suppressed 24 hr post MCAO in LPS-preconditioned mice. (A) Nuclear protein obtained from ipsilateral cortices was used to measure p65 activity by EMSA analysis. EMSA gel of pooled samples (n = 4) following 45 min MCAO for saline and LPS preconditioned (0.8 mg/kg) mice (Lt.). Quantification of band intensity of individual mice following MCAO (Rt.). NFκB is significantly decreased in LPS-preconditioned mice 24 hr post MCAO compared to saline. Supershift assay confirmed specificity for p65 oligos (data not shown). (B) Ship1 and Tollip mRNA are significantly upregulated 24 hr post 60 minute MCAO in LPS-preconditioned (0.8 mg/kg) mice compared to saline, n ≥ 4 per treatment/timepoint. (C) Western blot for Ship1 and Tollip and relative band quantification showing significant upregulation of Ship1 and Tollip protein 24 hr post 45 min MCAO in LPS-preconditioned mice (0.8 mg/kg), n ≥ 3 per treatment/timepoint. (D) Western blot and relative band quantification for p105 at 24 hr post 45 minute MCAO showing significant upregulation in LPS-preconditioned (0.8 mg/kg) mice, n ≥ 3 per treatment/timepoint. (A) Two-Way ANOVA, Bonferroni Post Hoc, *p < 0.05. (B-D) Student's t-test, LPS vs. saline, **p < 0.01. Ship1 and Tollip are cytosolic molecules that inhibit TLR signaling, which leads to the suppression of NFκB activity. We found that Ship1 and Tollip mRNA are upregulated in the brain 72 hr post injection versus saline controls (2.06 ± 0.27 and 2.31 ± 0.35, respectively) but not at 3 hr post stroke (1.09 ± 0.10 and 1.05 ± 0.09, respectively). However, by 24 hr post MCAO, Ship1 and Tollip mRNA are significantly enhanced in the brain of LPS-preconditioned mice compared to saline controls (2.62 ± 0.84 and 4.01 ± 1.06, respectively, Figure 3B). Ship1 protein is not upregulated at 72 hr post injection (Fold change vs. saline: 1.01 ± 0.32), but becomes significantly enhanced in LPS-preconditioned mice at 3 hr (Fold change vs saline: 1.83 ± 0.13) and at 24 hr (Fold change vs. saline: 8.81 ± 1.54, Figure 3C) post MCAO. Tollip protein is not affected by LPS preconditioning at 72 hr post injection or 3 hr post MCAO (Fold change vs. saline: 1.42 ± 0.10 and 0.83 ± 0.10, respectively), but it is significantly enhanced in LPS-preconditioned mice compared to saline controls at 24 hr post MCAO (Fold change vs. saline: 2.42 ± 0.20, Figure 3C). Additionally, the p50 precursor protein p105, which inhibits NFκB activity by acting like an IκB molecule by sequestering NFκB in the cytosol [27,28], was significantly upregulated 24 hr post stroke in LPS-preconditioned mice compared to saline (Figure 3D). Thus, despite the upregulation of inflammatory genes, the activity of NFκB is suppressed in the late-phase of the neuroprotective response of LPS-preconditioned mice. IRF3 activity in the brain is enhanced following MCAO in LPS-preconditioned mice IRF3 activation downstream of TLR4 is associated with anti-inflammatory/type I IFN responses. Using an IRF3 activity ELISA, we determined that IRF3 activity is comparable immediately prior to stroke (data not shown) and subsequently enhanced in the brains of LPS-preconditioned mice following MCAO (Figure 4). The trend for increased IRF3 activity is present at 3 hr post MCAO and is significantly increased at 24 hr in LPS-preconditioned mice (Figure 4). Saline treated animals showed no evidence of increased IRF3 activity following stroke (Figure 4). This indicates that LPS preconditioning alters the response to ischemic injury by activating IRF3--a finding that is consistent with the enhanced anti-inflammatory/type I IFN gene expression. IRF3 activity is enhanced following MCAO in LPS-preconditioned mice. Nuclear protein obtained from ipsilateral cortex post 60 min MCAO analyzed using an IRF3 activity ELISA (Active Motif, Inc.) revealed a significant increase in IRF3 activity in LPS-preconditioned (0.8 mg/kg) mice. Two-way ANOVA, Bonferroni Post Hoc, LPS vs. saline, *p < 0.05, n ≥ 4 per treatment. IRF3 activation downstream of TLR4 is associated with anti-inflammatory/type I IFN responses. Using an IRF3 activity ELISA, we determined that IRF3 activity is comparable immediately prior to stroke (data not shown) and subsequently enhanced in the brains of LPS-preconditioned mice following MCAO (Figure 4). The trend for increased IRF3 activity is present at 3 hr post MCAO and is significantly increased at 24 hr in LPS-preconditioned mice (Figure 4). Saline treated animals showed no evidence of increased IRF3 activity following stroke (Figure 4). This indicates that LPS preconditioning alters the response to ischemic injury by activating IRF3--a finding that is consistent with the enhanced anti-inflammatory/type I IFN gene expression. IRF3 activity is enhanced following MCAO in LPS-preconditioned mice. Nuclear protein obtained from ipsilateral cortex post 60 min MCAO analyzed using an IRF3 activity ELISA (Active Motif, Inc.) revealed a significant increase in IRF3 activity in LPS-preconditioned (0.8 mg/kg) mice. Two-way ANOVA, Bonferroni Post Hoc, LPS vs. saline, *p < 0.05, n ≥ 4 per treatment. Blood cytokine/chemokine levels parallel the expression in the brain Evidence indicates that stroke alters the cytokine profile in the plasma of circulating blood [5,29]. To determine whether LPS preconditioning changes the balance of pro- and anti-inflammatory cytokines and chemokines in the plasma we examined the levels of seven molecules using ELISAs. The results indicate that the level of pro-inflammatory cytokines, such as IL6, IL1β, and MCP-1, are increased in both LPS and saline preconditioned mice (Figure 5). The pro-inflammatory cytokines MIP-1α and IL1α were not detected in the serum (data not shown). The anti-inflammatory cytokine IL10 was significantly increased only in the plasma of LPS-preconditioned mice compared to saline preconditioned mice following stroke (Figure 5). RANTES, which is a chemokine associated with IRF3 and IRF7 activity [30], was present in the blood of LPS-preconditioned mice at significantly greater levels than saline preconditioned mice (Figure 5). IFNβ was not detectable in the blood of LPS or saline preconditioned animals following stroke (data not shown). Overall, this suggests that the pro-inflammatory and anti-inflammatory/type I IFN-associated response in the blood parallels the response in the brain following stroke. Blood cytokine/chemokine levels show alterations in gene expression patterns comparable to the brain. Plasma collected from saline or LPS-preconditioned (0.8 mg/kg) mice at the time of or following 60 min MCAO was examined using a multikine ELISA (Quansys). Results indicated that pro-inflammatory cytokines IL1β, IL6 and MCP-1 are similar in saline and LPS-preconditioned mice. In contrast, LPS-preconditioned mice have significantly enhanced levels of the anti-inflammatory/type I IFN-associated cytokine and chemokine IL10 and RANTES compared to saline following MCAO. Two-way ANOVA, LPS vs. saline, *p < 0.05, n ≥ 3 per treatment. Evidence indicates that stroke alters the cytokine profile in the plasma of circulating blood [5,29]. To determine whether LPS preconditioning changes the balance of pro- and anti-inflammatory cytokines and chemokines in the plasma we examined the levels of seven molecules using ELISAs. The results indicate that the level of pro-inflammatory cytokines, such as IL6, IL1β, and MCP-1, are increased in both LPS and saline preconditioned mice (Figure 5). The pro-inflammatory cytokines MIP-1α and IL1α were not detected in the serum (data not shown). The anti-inflammatory cytokine IL10 was significantly increased only in the plasma of LPS-preconditioned mice compared to saline preconditioned mice following stroke (Figure 5). RANTES, which is a chemokine associated with IRF3 and IRF7 activity [30], was present in the blood of LPS-preconditioned mice at significantly greater levels than saline preconditioned mice (Figure 5). IFNβ was not detectable in the blood of LPS or saline preconditioned animals following stroke (data not shown). Overall, this suggests that the pro-inflammatory and anti-inflammatory/type I IFN-associated response in the blood parallels the response in the brain following stroke. Blood cytokine/chemokine levels show alterations in gene expression patterns comparable to the brain. Plasma collected from saline or LPS-preconditioned (0.8 mg/kg) mice at the time of or following 60 min MCAO was examined using a multikine ELISA (Quansys). Results indicated that pro-inflammatory cytokines IL1β, IL6 and MCP-1 are similar in saline and LPS-preconditioned mice. In contrast, LPS-preconditioned mice have significantly enhanced levels of the anti-inflammatory/type I IFN-associated cytokine and chemokine IL10 and RANTES compared to saline following MCAO. Two-way ANOVA, LPS vs. saline, *p < 0.05, n ≥ 3 per treatment. TRIF dependent LPS preconditioning induced neuroprotection Evidence presented here and previously suggests that signaling following stroke is redirected towards IRF3 [13,14]. TLR4 signaling, which activates IRF3, is initiated by the adaptor molecule TRIF, while TLR4 signaling that activates NFκB is initiated by the adaptor molecule MyD88. The individual roles of these adaptor molecules in neuroprotection induced by LPS preconditioning are unknown. To test whether either of these key molecular adaptors were important in mediating the neuroprotective effects of LPS, we exposed MyD88-/- and TRIF-/- mice to LPS preconditioning (n = 4-10 mice/treatment). We found that MyD88-/- mice preconditioned with LPS had significantly reduced infarct sizes in response to MCAO compared to saline controls (Figure 6), indicating that LPS preconditioning is able to induce neuroprotection in mice lacking MyD88. In contrast, TRIF-/- mice preconditioned with LPS or saline had comparable infarct sizes (Figure 6), indicating that LPS preconditioning is not able to induce neuroprotection in mice lacking TRIF. Importantly, the TRIF adaptor is responsible for activation of IRF3, thus, our finding that TRIF is required for LPS preconditioning provides further support for a protective role of IRF3 activity in neuroprotection. LPS preconditioning requires TLR signaling through TRIF to promote neuroprotection. WT, MyD88-/-, and TRIF-/- mice were preconditioned with LPS (0.8 mg/kg) 3 days prior to 40 min MCAO. MyD88-/- mice were protected by LPS preconditioning resulting in smaller infarct sizes. TRIF-/- mice did not have reduced infarct sizes, demonstrating that TRIF deficient mice are not protected by LPS preconditioning. Thus, TRIF is required for LPS preconditioning induced neuroprotection. Student's t-test, LPS vs. saline, *p < 0.05, n = 4-10 per treatment. Evidence presented here and previously suggests that signaling following stroke is redirected towards IRF3 [13,14]. TLR4 signaling, which activates IRF3, is initiated by the adaptor molecule TRIF, while TLR4 signaling that activates NFκB is initiated by the adaptor molecule MyD88. The individual roles of these adaptor molecules in neuroprotection induced by LPS preconditioning are unknown. To test whether either of these key molecular adaptors were important in mediating the neuroprotective effects of LPS, we exposed MyD88-/- and TRIF-/- mice to LPS preconditioning (n = 4-10 mice/treatment). We found that MyD88-/- mice preconditioned with LPS had significantly reduced infarct sizes in response to MCAO compared to saline controls (Figure 6), indicating that LPS preconditioning is able to induce neuroprotection in mice lacking MyD88. In contrast, TRIF-/- mice preconditioned with LPS or saline had comparable infarct sizes (Figure 6), indicating that LPS preconditioning is not able to induce neuroprotection in mice lacking TRIF. Importantly, the TRIF adaptor is responsible for activation of IRF3, thus, our finding that TRIF is required for LPS preconditioning provides further support for a protective role of IRF3 activity in neuroprotection. LPS preconditioning requires TLR signaling through TRIF to promote neuroprotection. WT, MyD88-/-, and TRIF-/- mice were preconditioned with LPS (0.8 mg/kg) 3 days prior to 40 min MCAO. MyD88-/- mice were protected by LPS preconditioning resulting in smaller infarct sizes. TRIF-/- mice did not have reduced infarct sizes, demonstrating that TRIF deficient mice are not protected by LPS preconditioning. Thus, TRIF is required for LPS preconditioning induced neuroprotection. Student's t-test, LPS vs. saline, *p < 0.05, n = 4-10 per treatment. LPS preconditioning does not affect inflammatory gene expression in the brain following stroke: We used gene microarray analysis to elucidate the pattern of inflammatory or anti-inflammatory/type I IFN gene expression in the brain following stroke. In the setting of stroke, LPS preconditioned animals exhibited regulation of a number of genes typically found downstream of TLR signaling. The inflammatory profile reveals that the gene regulation is similar at each timepoint following stroke in LPS or saline preconditioned animals (Figure 1, Rt.). There is no evidence of inflammatory gene expression present immediately prior to stroke (Figure 1, Rt. 0 hr). At 3 hr post MCAO, several inflammatory genes are upregulated including IL6, IL1β, Ptgs2/COX2, and CCL2/MCP-1 (Figure 1, Rt.) and this upregulation is sustained at the 24 hr timepoint following MCAO (Figure 1, Rt.). TNFα, which is commonly shown to be upregulated following MCAO [25,26], only shows marginal levels of upregulation in LPS or saline preconditioned mice (Figure 1, Rt.). To confirm the microarray results, a subset of selected inflammatory genes including IL6, IL1β, COX2, and TNFα, were analyzed using qtPCR. Each of these genes were upregulated following MCAO in LPS and saline preconditioned mice, but there were no significant differences based on treatment at 3 hr (data not shown) and 24 hr (Figure 2) following MCAO. Enhanced anti-inflammatory/type I IFN gene expression but comparable pro-inflammatory gene expression in LPS-preconditioned mice post MCAO. Gene regulation 24 hr post MCAO measured by qtPCR reveals that anti-inflammatory/type I IFN-associated genes TGFβ, RANTES, and IFIT1 are significantly upregulated in LPS preconditioned mice compared to saline. Pro-inflammatory genes IL6, IL1β, COX2, and TNFα show similar regulation in LPS and saline preconditioned mice. These results confirm the gene microarray data. Samples from mice receiving a 45 (LPS: 0.2 mg/kg) or 60 min (LPS: 0.8 mg/kg) MCAO were combined due to comparable gene regulation (see methods). ND = not detected. Student's t-test, LPS vs. saline 24 hr post MCAO, **p < 0.01, n ≥ 4 per treatment. LPS preconditioning upregulates anti-inflammatory/type I IFN gene expression in the brain following MCAO: Although pro-inflammatory gene expression was not differentially modulated in preconditioned animals, microarray results revealed that the majority of the anti-inflammatory/type I IFN genes, such as TGFβ, IL1 receptor antagonist (IL1rn), RANTES, and IRF7, were upregulated following stroke in the brains of LPS versus saline preconditioned mice (Figure 1, Lt.). IL10 gene expression was not detected at any timepoint (Figure 1, Lt.). TGFβ, IL10, RANTES, and IFIT1 were selected for qtPCR analysis. TGFβ, RANTES, and IFIT1 were significantly upregulated in the LPS-preconditioned brain compared to saline 24 hr following stroke (Figure 2). RANTES was also significantly upregulated at 3 hr following stroke in LPS-preconditioned mice compared to saline (data not shown). IL10 expression remained undetectable by qtPCR analysis (Figure 2), suggesting that IL10 mRNA is not present at these timepoints in the brain following stroke. These qtPCR results confirm the gene expression profile observed on the microarray. Taken together, these data indicate an enhanced anti-inflammatory/type I IFN gene expression profile in the brain of LPS-preconditioned animals following MCAO while the inflammatory gene expression is unaffected. NFκB activity is suppressed in the brain of LPS-preconditioned animals 24 hr post MCAO: NFκB activity is associated with damage and inflammation in the brain that occurs in response to stroke. We used EMSAs to evaluate the activity of the NFκB subunit p65 in the brain following stroke. The results indicated that LPS and saline preconditioned mice have comparable NFκB activity at 3 hr post MCAO (Figure 3A). However, at 24 hr post MCAO, LPS-preconditioned animals have significantly suppressed NFκB activity compared to saline preconditioned mice (Figure 3A). NFκB is suppressed 24 hr post MCAO in LPS-preconditioned mice. (A) Nuclear protein obtained from ipsilateral cortices was used to measure p65 activity by EMSA analysis. EMSA gel of pooled samples (n = 4) following 45 min MCAO for saline and LPS preconditioned (0.8 mg/kg) mice (Lt.). Quantification of band intensity of individual mice following MCAO (Rt.). NFκB is significantly decreased in LPS-preconditioned mice 24 hr post MCAO compared to saline. Supershift assay confirmed specificity for p65 oligos (data not shown). (B) Ship1 and Tollip mRNA are significantly upregulated 24 hr post 60 minute MCAO in LPS-preconditioned (0.8 mg/kg) mice compared to saline, n ≥ 4 per treatment/timepoint. (C) Western blot for Ship1 and Tollip and relative band quantification showing significant upregulation of Ship1 and Tollip protein 24 hr post 45 min MCAO in LPS-preconditioned mice (0.8 mg/kg), n ≥ 3 per treatment/timepoint. (D) Western blot and relative band quantification for p105 at 24 hr post 45 minute MCAO showing significant upregulation in LPS-preconditioned (0.8 mg/kg) mice, n ≥ 3 per treatment/timepoint. (A) Two-Way ANOVA, Bonferroni Post Hoc, *p < 0.05. (B-D) Student's t-test, LPS vs. saline, **p < 0.01. Ship1 and Tollip are cytosolic molecules that inhibit TLR signaling, which leads to the suppression of NFκB activity. We found that Ship1 and Tollip mRNA are upregulated in the brain 72 hr post injection versus saline controls (2.06 ± 0.27 and 2.31 ± 0.35, respectively) but not at 3 hr post stroke (1.09 ± 0.10 and 1.05 ± 0.09, respectively). However, by 24 hr post MCAO, Ship1 and Tollip mRNA are significantly enhanced in the brain of LPS-preconditioned mice compared to saline controls (2.62 ± 0.84 and 4.01 ± 1.06, respectively, Figure 3B). Ship1 protein is not upregulated at 72 hr post injection (Fold change vs. saline: 1.01 ± 0.32), but becomes significantly enhanced in LPS-preconditioned mice at 3 hr (Fold change vs saline: 1.83 ± 0.13) and at 24 hr (Fold change vs. saline: 8.81 ± 1.54, Figure 3C) post MCAO. Tollip protein is not affected by LPS preconditioning at 72 hr post injection or 3 hr post MCAO (Fold change vs. saline: 1.42 ± 0.10 and 0.83 ± 0.10, respectively), but it is significantly enhanced in LPS-preconditioned mice compared to saline controls at 24 hr post MCAO (Fold change vs. saline: 2.42 ± 0.20, Figure 3C). Additionally, the p50 precursor protein p105, which inhibits NFκB activity by acting like an IκB molecule by sequestering NFκB in the cytosol [27,28], was significantly upregulated 24 hr post stroke in LPS-preconditioned mice compared to saline (Figure 3D). Thus, despite the upregulation of inflammatory genes, the activity of NFκB is suppressed in the late-phase of the neuroprotective response of LPS-preconditioned mice. IRF3 activity in the brain is enhanced following MCAO in LPS-preconditioned mice: IRF3 activation downstream of TLR4 is associated with anti-inflammatory/type I IFN responses. Using an IRF3 activity ELISA, we determined that IRF3 activity is comparable immediately prior to stroke (data not shown) and subsequently enhanced in the brains of LPS-preconditioned mice following MCAO (Figure 4). The trend for increased IRF3 activity is present at 3 hr post MCAO and is significantly increased at 24 hr in LPS-preconditioned mice (Figure 4). Saline treated animals showed no evidence of increased IRF3 activity following stroke (Figure 4). This indicates that LPS preconditioning alters the response to ischemic injury by activating IRF3--a finding that is consistent with the enhanced anti-inflammatory/type I IFN gene expression. IRF3 activity is enhanced following MCAO in LPS-preconditioned mice. Nuclear protein obtained from ipsilateral cortex post 60 min MCAO analyzed using an IRF3 activity ELISA (Active Motif, Inc.) revealed a significant increase in IRF3 activity in LPS-preconditioned (0.8 mg/kg) mice. Two-way ANOVA, Bonferroni Post Hoc, LPS vs. saline, *p < 0.05, n ≥ 4 per treatment. Blood cytokine/chemokine levels parallel the expression in the brain: Evidence indicates that stroke alters the cytokine profile in the plasma of circulating blood [5,29]. To determine whether LPS preconditioning changes the balance of pro- and anti-inflammatory cytokines and chemokines in the plasma we examined the levels of seven molecules using ELISAs. The results indicate that the level of pro-inflammatory cytokines, such as IL6, IL1β, and MCP-1, are increased in both LPS and saline preconditioned mice (Figure 5). The pro-inflammatory cytokines MIP-1α and IL1α were not detected in the serum (data not shown). The anti-inflammatory cytokine IL10 was significantly increased only in the plasma of LPS-preconditioned mice compared to saline preconditioned mice following stroke (Figure 5). RANTES, which is a chemokine associated with IRF3 and IRF7 activity [30], was present in the blood of LPS-preconditioned mice at significantly greater levels than saline preconditioned mice (Figure 5). IFNβ was not detectable in the blood of LPS or saline preconditioned animals following stroke (data not shown). Overall, this suggests that the pro-inflammatory and anti-inflammatory/type I IFN-associated response in the blood parallels the response in the brain following stroke. Blood cytokine/chemokine levels show alterations in gene expression patterns comparable to the brain. Plasma collected from saline or LPS-preconditioned (0.8 mg/kg) mice at the time of or following 60 min MCAO was examined using a multikine ELISA (Quansys). Results indicated that pro-inflammatory cytokines IL1β, IL6 and MCP-1 are similar in saline and LPS-preconditioned mice. In contrast, LPS-preconditioned mice have significantly enhanced levels of the anti-inflammatory/type I IFN-associated cytokine and chemokine IL10 and RANTES compared to saline following MCAO. Two-way ANOVA, LPS vs. saline, *p < 0.05, n ≥ 3 per treatment. TRIF dependent LPS preconditioning induced neuroprotection: Evidence presented here and previously suggests that signaling following stroke is redirected towards IRF3 [13,14]. TLR4 signaling, which activates IRF3, is initiated by the adaptor molecule TRIF, while TLR4 signaling that activates NFκB is initiated by the adaptor molecule MyD88. The individual roles of these adaptor molecules in neuroprotection induced by LPS preconditioning are unknown. To test whether either of these key molecular adaptors were important in mediating the neuroprotective effects of LPS, we exposed MyD88-/- and TRIF-/- mice to LPS preconditioning (n = 4-10 mice/treatment). We found that MyD88-/- mice preconditioned with LPS had significantly reduced infarct sizes in response to MCAO compared to saline controls (Figure 6), indicating that LPS preconditioning is able to induce neuroprotection in mice lacking MyD88. In contrast, TRIF-/- mice preconditioned with LPS or saline had comparable infarct sizes (Figure 6), indicating that LPS preconditioning is not able to induce neuroprotection in mice lacking TRIF. Importantly, the TRIF adaptor is responsible for activation of IRF3, thus, our finding that TRIF is required for LPS preconditioning provides further support for a protective role of IRF3 activity in neuroprotection. LPS preconditioning requires TLR signaling through TRIF to promote neuroprotection. WT, MyD88-/-, and TRIF-/- mice were preconditioned with LPS (0.8 mg/kg) 3 days prior to 40 min MCAO. MyD88-/- mice were protected by LPS preconditioning resulting in smaller infarct sizes. TRIF-/- mice did not have reduced infarct sizes, demonstrating that TRIF deficient mice are not protected by LPS preconditioning. Thus, TRIF is required for LPS preconditioning induced neuroprotection. Student's t-test, LPS vs. saline, *p < 0.05, n = 4-10 per treatment. Discussion: Here we sought to describe the LPS-induced reprogrammed response to stroke and to determine the important signaling events involved in neuroprotection against ischemic injury. Our results demonstrated that NFκB activity was suppressed and that the cytosolic inhibitors of NFκB, Ship1, Tollip, and p105, were present 24 hr post MCAO although pro-inflammatory gene expression was unaffected (diagrammed in Figure 7). Interestingly, there is evidence that suppression of NFκB can promote protection against cerebral ischemia without influencing pro-inflammatory cytokine production [3,31]. In particular, administration of the NFκB inhibitor Tat-NEMO Binding Domain provided protection against hypoxia-ischemia in neonatal rats without affecting TNFα or IL1β production [3]. Furthermore, TLR4 deficient mice have smaller infarcts in response to MCAO, yet the production of TNFα and IL1β was unaffected [6]. This suggests that reduced ischemic injury can be achieved by suppressing NFκB activity without suppressing pro-inflammatory cytokines and that TLR4 signaling and NFκB activation is not the sole source of these pro-inflammatory cytokines in response to ischemic injury, implicating other signaling cascades and transcription factors in the inflammatory response. Thus, consistent with our result, reprogramming the TLR4 response would not alter inflammatory gene expression in the brain. Schematic of TLR4 signaling and gene expression following stroke. (Top) TLR4 signaling cascades following stroke. In the absence of LPS preconditioning, stroke leads to NFκB activation without IRF3 activation. LPS preconditioning prior to stroke leads to robust activation of IRF3 and suppressed NFκB activity compared to stroke alone. (Bottom) Gene expression 24 hr post stroke. Stroke alone dramatically upregulates pro-inflammatory genes. LPS preconditioning prior to stroke dramatically upregulates anti-inflammatory/Type I IFN genes, many of which are associated with IRF3, while still maintaining a pro-inflammatory response. NFκB is known to be induced acutely in response to ischemic injury; however, investigation into the role of NFκB activity has revealed conflicting results [2]. For instance, NFκB is constitutively active in neurons, a requirement for their survival, while the surrounding glial cells have inducible NFκB activity [32]. In response to ischemic challenge, NFκB activity in astrocytes is responsible for detrimental inflammation [33]. This concept of pleotropic roles also applies to many of the inflammatory genes expressed in the brain in the setting of stroke [34,35]. For example, intracerebroventricular injection of recombinant IL6 significantly decreased the infarct size in rats 24 hr post MCAO [36]. IL1β is a potent inducer of IL1 receptor antagonist (IL-rn), which significantly reduces damage in response to stroke [37] and, notably, is upregulated in our microarray (Figure 1, Lt.). TNFα is considered to play multiple roles in stroke injury mediating many neuroprotective and injurious effects [34]. Furthermore, in response to viral challenge, the simultaneous presence of inflammatory cytokines, such as TNFα, and type I IFNs can alter their effects and synergize to promote a more protective state [38]. Thus, alterations in the environment in which NFκB is activated and inflammatory genes are present may affect the roles pro-inflammatory mediators play in injury and may even contribute to the protective phenotype. IRF3 activity induces the expression of anti-inflammatory and type I IFN-associated genes. Interestingly, mice deficient in IRF3 are not protected against cerebral ischemia by LPS preconditioning [13]. We have further established the importance of IRF3 in neuroprotection by identifying that multiple preconditioning paradigms including LPS, CpG (TLR9 agonist) and brief ischemia induce a common set of IRF-mediated genes in the neuroprotective environment following MCAO [14]. Here we demonstrate that IRF3 activity is upregulated in the brain of LPS-preconditioned mice in response to MCAO and that several IRF3-mediated genes are also upregulated, including RANTES and IFIT1 (diagrammed in Figure 7), which may mitigate the damaging effects of ischemia. Many of the upregulated anti-inflammatory/type I IFN genes in the brain following stroke have several identified neuroprotective functions. TGFβ has been shown to protect neurons from apoptosis, promote angiogenesis, decrease microglial activation, and reduce edema [34,39]. RANTES, which is induced by IRF3 and IRF7 [30], has been shown to protect neurons from cell death in response to HIV-1 glycoprotein gp120 [40]. In the setting of brain ischemia, mice deficient in the RANTES receptor, CCR5, have larger infarcts, suggesting a neuroprotective role for CCR5 activation [41]. Notably, the expression of CCR5 is upregulated in our microarray data (Figure 1, Lt). IFIT1 is commonly associated with IRF3 signaling in response to IFN treatment and viral infection [42]. Little is known about a role for IFIT1 in ischemic injury; however, it is inducible in microglia and neurons and has been shown to affect NFκB and IRF3 activation [42-45]. Additional anti-inflammatory/type I IFN genes shown to be upregulated in our microarray studies have potential roles in neuroprotection including IL-receptor antagonist (IL-rn), which is associated with reduced infarct size in response to stroke [34,46]. A recombinant form of IL-rn is being tested in Phase II clinical trials as an acute stroke therapy [47,48]. Although not detected in our gene microarray studies here, perhaps due to assay sensitivity for IFNβ transcript on the microarray, we have previously published that IFNβ mRNA, a type I IFN known to have neuroprotective properties, is upregulated following stroke in the brain of LPS-preconditioned mice using qtPCR [13]. The protective functions of these genes may be of considerable importance to the neuroprotective phenotype following MCAO induced by LPS preconditioning. Research strongly suggests that cerebral ischemia dramatically alters the protein and gene expression profile in the peripheral blood [5,29,49,50]. Our results demonstrate that the cytokine and chemokine response in the blood paralleled the pattern of gene expression in the brain. Overall, inflammatory cytokine protein levels were similarly induced in LPS and saline preconditioned mice following stroke. However, we have previously published that TNFα is significantly reduced in the plasma of LPS-preconditioned mice following MCAO [51]. The anti-inflammatory and type I IFN-induced cytokines and chemokines measured in the blood were enhanced in LPS-preconditioned mice compared to saline. In particular, IL10 was significantly upregulated in the blood following MCAO of LPS-preconditioned mice. Importantly, in humans, upregulation of IL10 in the blood has been correlated with improved outcome in stroke [52]. While IL10 mRNA was not detectable in the brain, IL10 can be induced by IRF3 activity and therefore is indicative of the same redirected response seen in the brain. IFNβ was not detected in the blood 24 hr post MCAO. This may be due to the kinetics of IFNβ expression. Further investigation into the time course of IFNβ induction in the blood is necessary to fully understand the role of IFNβ in this system. The redirected signaling observed in the blood may stem from the brain's response to injury by leaking proteins into the peripheral circulation; however, this is not considered a major source of plasma cytokines at these early timepoints following stroke [29]. Alternately, because LPS administration occurs by a systemic route, target cells in the periphery may become tolerant to activation by the secondary stimuli resulting from ischemic injury. Although our data does not distinguish between these possibilities, it is clear that LPS preconditioning alters the response to injury in the brain and the blood in a manner that promotes a protective phenotype. TLR4 signals through the adaptor molecules MyD88 and TRIF. MyD88 signaling culminates in NFκB activation. TRIF signaling can activate both IRF3 and NFκB, although IRF3 activation often is more rapid and robust, while activation of NFκB is a secondary effect that occurs as part of late-phase TLR signaling [53]. The data presented in this paper and Marsh et al., 2009 [13] suggests a dominant role for IRF3 signaling in LPS-induced neuroprotection, which implicates the TRIF adaptor as a key player in the reprogrammed TLR4 response to stroke. Support for this lies in our finding that mice deficient in TRIF are not protected by LPS preconditioning. In contrast, MyD88 deficient mice preconditioned with LPS are still protected against MCAO. Taken together, these data strongly support a protective role for TRIF-mediated IRF3 activation in the neuroprotective phenotype induced by LPS preconditioning. TLRs have the ability to self regulate in a manner that redirects their signaling. The classic example is endotoxin tolerance, whereby a low dose of the TLR4 ligand LPS reprograms TLR4 signaling in response to a subsequent toxic dose of LPS, leading to a protective phenotype [54]. This reprogrammed response comes in two major forms: (1.) suppressed pro-inflammatory signaling and enhanced anti-inflammatory/type I IFN signaling, or (2.) enhanced anti-inflammatory/type I IFN signaling in the absence of suppressed pro-inflammatory signaling. Thus, the suppressed NFκB activity, the enhanced IRF3 activity, and the upregulated anti-inflammatory/type I IFN associated genes seen in the LPS-preconditioned brain following stroke is reminiscent of endotoxin tolerance--a phenomenon that has been best described in macrophages in vitro, but more recently in animals. Many other key features of endotoxin tolerance are seen in the reprogrammed response to stroke produced by LPS preconditioning. For example, Tollip and Ship1 are known to be induced in endotoxin tolerance and lead to suppressed NFκB activity. TGFβ has been shown to play an important role in endotoxin tolerance, whereby TGFβ-mediated induction of SMAD4 is required to promote complete endotoxin tolerance and to induce the NFκB inhibitor, Ship1 [55]. Interestingly, in our system the upregulation of TGFβ corresponds to Ship1 upregulation 24 hr post MCAO in LPS-preconditioned mice compared to saline. Furthermore, cells deficient in TRIF or IRF3 are unable to develop tolerance to endotoxin [56]. This is similar to TRIF deficient or IRF3 deficient mice not being protected by LPS preconditioning against cerebral ischemia. Taken together, this suggests that the cellular phenomenon of endotoxin tolerance is potentially the same response observed in LPS preconditioning wherein LPS exposure leads to a reprogrammed TLR signaling response in the brain following stroke to produce protection. Conclusions: The findings reported here provide an important characterization of the LPS-induced neuroprotective response following stroke. We show that LPS preconditioning induces a reprogrammed response to stroke, whereby NFκB activity is suppressed, IRF3 activity is enhanced, and anti-inflammatory/type-I IFN genes are upregulated (diagrammed in Figure 7). Interestingly, the suppression of pro-inflammatory genes is not a necessary part of the neuroprotective response induced by LPS preconditioning. Further evaluation into the TLR4 signaling cascades revealed a seminal role for the TRIF cascade in producing the neuroprotection initiated by LPS preconditioning. As TRIF signaling culminates in IRF3 activation, this finding provides further evidence for the importance of IRF3 in the neuroprotective response to stroke.

Background: Toll-like receptor 4 (TLR4) is activated in response to cerebral ischemia leading to substantial brain damage. In contrast, mild activation of TLR4 by preconditioning with low dose exposure to lipopolysaccharide (LPS) prior to cerebral ischemia dramatically improves outcome by reprogramming the signaling response to injury. This suggests that TLR4 signaling can be altered to induce an endogenously neuroprotective phenotype. However, the TLR4 signaling events involved in this neuroprotective response are poorly understood. Here we define several molecular mediators of the primary signaling cascades induced by LPS preconditioning that give rise to the reprogrammed response to cerebral ischemia and confer the neuroprotective phenotype. Methods: C57BL6 mice were preconditioned with low dose LPS prior to transient middle cerebral artery occlusion (MCAO). Cortical tissue and blood were collected following MCAO. Microarray and qtPCR were performed to analyze gene expression associated with TLR4 signaling. EMSA and DNA binding ELISA were used to evaluate NFκB and IRF3 activity. Protein expression was determined using Western blot or ELISA. MyD88-/- and TRIF-/- mice were utilized to evaluate signaling in LPS preconditioning-induced neuroprotection. Results: Gene expression analyses revealed that LPS preconditioning resulted in a marked upregulation of anti-inflammatory/type I IFN-associated genes following ischemia while pro-inflammatory genes induced following ischemia were present but not differentially modulated by LPS. Interestingly, although expression of pro-inflammatory genes was observed, there was decreased activity of NFκB p65 and increased presence of NFκB inhibitors, including Ship1, Tollip, and p105, in LPS-preconditioned mice following stroke. In contrast, IRF3 activity was enhanced in LPS-preconditioned mice following stroke. TRIF and MyD88 deficient mice revealed that neuroprotection induced by LPS depends on TLR4 signaling via TRIF, which activates IRF3, but does not depend on MyD88 signaling. Conclusions: Our results characterize several critical mediators of the TLR4 signaling events associated with neuroprotection. LPS preconditioning redirects TLR4 signaling in response to stroke through suppression of NFκB activity, enhanced IRF3 activity, and increased anti-inflammatory/type I IFN gene expression. Interestingly, this protective phenotype does not require the suppression of pro-inflammatory mediators. Furthermore, our results highlight a critical role for TRIF-IRF3 signaling as the governing mechanism in the neuroprotective response to stroke.

Background: Stroke is one of the leading causes of death and the leading cause of morbidity in the United States [1]. The inflammatory response to stroke substantially exacerbates ischemic damage. The acute activation of the NFκB transcription factor has been linked to the inflammatory response to stroke [2] and suppression of NFκB activity following stroke has been found to reduce damage [3]. NFκB activation can lead to the dramatic upregulation of inflammatory molecules and cytokines including TNFα, IL6, IL1β, and COX2 [2]. The source of these inflammatory molecules in the acute response to stroke appears to stem from the cells of the central nervous system (CNS), including neurons and glial cells [2]. The cells in the CNS play a particularly dominant role early in the response to ischemia because infiltrating leukocytes do not appear in substantial numbers in the brain until 24 hr following injury [4]. Stroke also induces an acute inflammatory response in the circulating blood. Inflammatory cytokine and chemokine levels, including IL6, IL1β, MCP-1 and TNFα are elevated in the circulation following stroke [5]. This suggests there is an intimate relationship between responses in the brain and blood following stroke--responses that result in increased inflammation. Toll-like receptors (TLRs), traditionally considered innate immune receptors, signal through the adaptor proteins MyD88 and TRIF to activate NFκB and interferon regulatory factors (IRFs). It has been shown recently that TLRs become activated in response to endogenous ligands, known as damage associated molecular patterns (DAMPs), released during injury. Interestingly, animals deficient in TLR2 or TLR4 have significantly reduced infarct sizes in several models of stroke [6-11]. This suggests that TLR2 and TLR4 activation in response to ischemic injury exacerbates damage. In addition, a recent investigation in humans showed that the inflammatory responses to stroke in the blood were linked to increased TLR2 and TLR4 expression on hematopoetic cells and associated with worse outcome in stroke [12]. The detrimental effect of TLR signaling is associated with the pathways that lead to NFκB activation and pro-inflammatory responses. In contrast, TLR signaling pathways that activate IRFs can induce anti-inflammatory mediators and type I IFNs that have been associated with neuroprotection [13,14]. Thus, in TLR signaling there is a fine balance between pathways leading to injury or protection. TLR ligands have been a major source of interest as preconditioning agents for prophylactic therapy against ischemic injury. Such therapies would target a population of patients that are at risk of ischemia in the setting of surgery [15-18]. Preconditioning with low doses of ligands for TLR2, TLR4, and TLR9 all successfully reduce infarct size in experimental models of stroke [19-21], including a recent study showing that a TLR9 ligand is neuroprotective in a nonhuman primate model of stroke [22]. Emerging evidence suggests that TLR-induced neuroprotection occurs by reprogramming the genomic response to the DAMPs, which are produced in response to ischemic injury. In this reprogrammed state, the resultant pathway activation of TLR4 signaling preferentially leads to IRF-mediated gene expression [13,14]. However, whether TLR preconditioning affects NFκB activity and pro-inflammatory signaling is unknown. As yet, a complete analysis of the characteristic TLR signaling responses to stroke following preconditioning has not been reported. The objective of this study is to utilize LPS preconditioning followed by transient middle cerebral artery occlusion (MCAO) to elucidate the reprogrammed TLR response to stroke and to determine the major pathways involved in producing the neuroprotective phenotype. Here we show that preconditioning against ischemia using LPS leads to suppressed NFκB activity--although pro-inflammatory gene expression does not appear to be attenuated. We also demonstrate that LPS-preconditioned mice have enhanced IRF3 activity and anti-inflammatory/type I IFN gene expression in the ischemic brain. This expression pattern was recapitulated in the blood where plasma levels of pro-inflammatory cytokine proteins were comparable in LPS-preconditioned and control mice while IRF-associated proteins were enhanced in LPS preconditioned mice. To our knowledge, we provide the first evidence that protection due to LPS preconditioning stems from TRIF signaling, the cascade that is associated with IRF3 activation, and is independent of MyD88 signaling. These molecular features suggest that, following stroke, signaling is directed away from NFκB activity and towards a dominant TRIF-IRF3 response. Understanding the endogenous signaling events that promote protection against ischemic injury is integral to the identification and development of novel stroke therapeutics. In particular, the evidence presented here further highlights a key role for IRF3 activity in the protective response to stroke. Conclusions: KBV performed experiments, collected data, conceived of the idea for the paper, and wrote the manuscript. SLS worked on the microarray, provided guidance in the production of data, and edited the paper. BJM performed experiments and contributed to the writing of the Methods section. RWK performed experiments. NL performed the MCAO surgeries. MSP provided critical guidance and worked on the manuscript. All authors approved of the final manuscript.

Background: Toll-like receptor 4 (TLR4) is activated in response to cerebral ischemia leading to substantial brain damage. In contrast, mild activation of TLR4 by preconditioning with low dose exposure to lipopolysaccharide (LPS) prior to cerebral ischemia dramatically improves outcome by reprogramming the signaling response to injury. This suggests that TLR4 signaling can be altered to induce an endogenously neuroprotective phenotype. However, the TLR4 signaling events involved in this neuroprotective response are poorly understood. Here we define several molecular mediators of the primary signaling cascades induced by LPS preconditioning that give rise to the reprogrammed response to cerebral ischemia and confer the neuroprotective phenotype. Methods: C57BL6 mice were preconditioned with low dose LPS prior to transient middle cerebral artery occlusion (MCAO). Cortical tissue and blood were collected following MCAO. Microarray and qtPCR were performed to analyze gene expression associated with TLR4 signaling. EMSA and DNA binding ELISA were used to evaluate NFκB and IRF3 activity. Protein expression was determined using Western blot or ELISA. MyD88-/- and TRIF-/- mice were utilized to evaluate signaling in LPS preconditioning-induced neuroprotection. Results: Gene expression analyses revealed that LPS preconditioning resulted in a marked upregulation of anti-inflammatory/type I IFN-associated genes following ischemia while pro-inflammatory genes induced following ischemia were present but not differentially modulated by LPS. Interestingly, although expression of pro-inflammatory genes was observed, there was decreased activity of NFκB p65 and increased presence of NFκB inhibitors, including Ship1, Tollip, and p105, in LPS-preconditioned mice following stroke. In contrast, IRF3 activity was enhanced in LPS-preconditioned mice following stroke. TRIF and MyD88 deficient mice revealed that neuroprotection induced by LPS depends on TLR4 signaling via TRIF, which activates IRF3, but does not depend on MyD88 signaling. Conclusions: Our results characterize several critical mediators of the TLR4 signaling events associated with neuroprotection. LPS preconditioning redirects TLR4 signaling in response to stroke through suppression of NFκB activity, enhanced IRF3 activity, and increased anti-inflammatory/type I IFN gene expression. Interestingly, this protective phenotype does not require the suppression of pro-inflammatory mediators. Furthermore, our results highlight a critical role for TRIF-IRF3 signaling as the governing mechanism in the neuroprotective response to stroke.

[CONTENT] Toll-like receptors | stroke | NFκB | inflammation | preconditioning | neuroprotection [SUMMARY]

[CONTENT] Toll-like receptors | stroke | NFκB | inflammation | preconditioning | neuroprotection [SUMMARY]

[CONTENT] Toll-like receptors | stroke | NFκB | inflammation | preconditioning | neuroprotection [SUMMARY]

[CONTENT] Toll-like receptors | stroke | NFκB | inflammation | preconditioning | neuroprotection [SUMMARY]

[CONTENT] Toll-like receptors | stroke | NFκB | inflammation | preconditioning | neuroprotection [SUMMARY]

[CONTENT] Adaptor Proteins, Vesicular Transport | Animals | Brain Ischemia | Chemokines | Cytokines | Gene Expression Profiling | Humans | Infarction, Middle Cerebral Artery | Interferon Regulatory Factor-3 | Ischemic Preconditioning | Lipopolysaccharides | Male | Mice | Mice, Inbred C57BL | Mice, Knockout | Microarray Analysis | NF-kappa B | Signal Transduction | Stroke | Toll-Like Receptor 4 [SUMMARY]

[CONTENT] Adaptor Proteins, Vesicular Transport | Animals | Brain Ischemia | Chemokines | Cytokines | Gene Expression Profiling | Humans | Infarction, Middle Cerebral Artery | Interferon Regulatory Factor-3 | Ischemic Preconditioning | Lipopolysaccharides | Male | Mice | Mice, Inbred C57BL | Mice, Knockout | Microarray Analysis | NF-kappa B | Signal Transduction | Stroke | Toll-Like Receptor 4 [SUMMARY]

[CONTENT] Adaptor Proteins, Vesicular Transport | Animals | Brain Ischemia | Chemokines | Cytokines | Gene Expression Profiling | Humans | Infarction, Middle Cerebral Artery | Interferon Regulatory Factor-3 | Ischemic Preconditioning | Lipopolysaccharides | Male | Mice | Mice, Inbred C57BL | Mice, Knockout | Microarray Analysis | NF-kappa B | Signal Transduction | Stroke | Toll-Like Receptor 4 [SUMMARY]

[CONTENT] Adaptor Proteins, Vesicular Transport | Animals | Brain Ischemia | Chemokines | Cytokines | Gene Expression Profiling | Humans | Infarction, Middle Cerebral Artery | Interferon Regulatory Factor-3 | Ischemic Preconditioning | Lipopolysaccharides | Male | Mice | Mice, Inbred C57BL | Mice, Knockout | Microarray Analysis | NF-kappa B | Signal Transduction | Stroke | Toll-Like Receptor 4 [SUMMARY]

[CONTENT] Adaptor Proteins, Vesicular Transport | Animals | Brain Ischemia | Chemokines | Cytokines | Gene Expression Profiling | Humans | Infarction, Middle Cerebral Artery | Interferon Regulatory Factor-3 | Ischemic Preconditioning | Lipopolysaccharides | Male | Mice | Mice, Inbred C57BL | Mice, Knockout | Microarray Analysis | NF-kappa B | Signal Transduction | Stroke | Toll-Like Receptor 4 [SUMMARY]

[CONTENT] stroke leads nfκb | inflammatory response nfκb | nfκb activated inflammatory | inflammatory response stroke | stroke blood cytokine [SUMMARY]

[CONTENT] stroke leads nfκb | inflammatory response nfκb | nfκb activated inflammatory | inflammatory response stroke | stroke blood cytokine [SUMMARY]

[CONTENT] stroke leads nfκb | inflammatory response nfκb | nfκb activated inflammatory | inflammatory response stroke | stroke blood cytokine [SUMMARY]

[CONTENT] stroke leads nfκb | inflammatory response nfκb | nfκb activated inflammatory | inflammatory response stroke | stroke blood cytokine [SUMMARY]

[CONTENT] stroke leads nfκb | inflammatory response nfκb | nfκb activated inflammatory | inflammatory response stroke | stroke blood cytokine [SUMMARY]

[CONTENT] lps | mice | preconditioned | mcao | inflammatory | saline | hr | following | stroke | lps preconditioned [SUMMARY]

[CONTENT] lps | mice | preconditioned | mcao | inflammatory | saline | hr | following | stroke | lps preconditioned [SUMMARY]

[CONTENT] lps | mice | preconditioned | mcao | inflammatory | saline | hr | following | stroke | lps preconditioned [SUMMARY]

[CONTENT] lps | mice | preconditioned | mcao | inflammatory | saline | hr | following | stroke | lps preconditioned [SUMMARY]

[CONTENT] lps | mice | preconditioned | mcao | inflammatory | saline | hr | following | stroke | lps preconditioned [SUMMARY]

[CONTENT] stroke | inflammatory | signaling | response | injury | tlr | nfκb | ischemic | tlr2 | tlr2 tlr4 [SUMMARY]

[CONTENT] mm | nuclear | protein | rna | tissue | mice | performed | nuclear protein | affymetrix | mcao [SUMMARY]

[CONTENT] neuroprotective response | response | neuroprotective | lps preconditioning | lps | preconditioning | irf3 | stroke | induced | signaling [SUMMARY]

[CONTENT] lps | mice | inflammatory | preconditioned | irf3 | saline | mcao | hr | stroke | activity [SUMMARY]

[CONTENT] lps | mice | inflammatory | preconditioned | irf3 | saline | mcao | hr | stroke | activity [SUMMARY]

[CONTENT] 4 ||| LPS ||| ||| ||| LPS [SUMMARY]

[CONTENT] C57BL6 | LPS | MCAO ||| MCAO ||| Microarray ||| EMSA | ELISA | NFκB ||| ELISA ||| MyD88-/- | TRIF-/- | LPS [SUMMARY]

[CONTENT] ||| LPS | NFκB | IFN ||| ||| TRIF [SUMMARY]

[CONTENT] 4 ||| LPS ||| ||| ||| LPS ||| C57BL6 | LPS | MCAO ||| MCAO ||| Microarray ||| EMSA | ELISA | NFκB ||| ELISA ||| MyD88-/- | TRIF-/- | LPS ||| ||| LPS | IFN | LPS ||| NFκB | NFκB | LPS ||| LPS ||| TRIF | MyD88 | LPS | TRIF | MyD88 ||| ||| LPS | NFκB | IFN ||| ||| TRIF [SUMMARY]

[CONTENT] 4 ||| LPS ||| ||| ||| LPS ||| C57BL6 | LPS | MCAO ||| MCAO ||| Microarray ||| EMSA | ELISA | NFκB ||| ELISA ||| MyD88-/- | TRIF-/- | LPS ||| ||| LPS | IFN | LPS ||| NFκB | NFκB | LPS ||| LPS ||| TRIF | MyD88 | LPS | TRIF | MyD88 ||| ||| LPS | NFκB | IFN ||| ||| TRIF [SUMMARY]

Conditional survival of cancer patients: an Australian perspective.

23043308

Estimated conditional survival for cancer patients diagnosed at different ages and disease stage provides important information for cancer patients and clinicians in planning follow-up, surveillance and ongoing management.

BACKGROUND

Using population-based cancer registry data for New South Wales Australia, we estimated conditional 5-year relative survival for 11 major cancers diagnosed 1972-2006 by time since diagnosis and age and stage at diagnosis.

METHODS

193,182 cases were included, with the most common cancers being prostate (39,851), female breast (36,585) and colorectal (35,455). Five-year relative survival tended to increase with increasing years already survived and improvement was greatest for cancers with poor prognosis at diagnosis (lung or pancreas) and for those with advanced stage or older age at diagnosis. After surviving 10 years, conditional 5-year survival was over 95% for 6 localised, 6 regional, 3 distant and 3 unknown stage cancers. For the remaining patient groups, conditional 5-year survival ranged from 74% (for distant stage bladder cancer) to 94% (for 4 cancers at different stages), indicating that they continue to have excess mortality 10-15 years after diagnosis.

RESULTS

These data provide important information for cancer patients, based on age and stage at diagnosis, as they continue on their cancer journey. This information may also be used by clinicians as a tool to make more evidence-based decisions regarding follow-up, surveillance, or ongoing management according to patients' changing survival expectations over time.

CONCLUSION

3519618

Background

Survival estimates for cancer patients are traditionally reported from the time of diagnosis such as five-year survival. It is useful for answering questions that many people ask about their prognosis when first diagnosed with cancer. For cancer patients who have already survived a number of years, survival expectations at diagnosis are too pessimistic because they include people who have already died. An ongoing question among these surviving patients is “now that I have survived for x years, what is the probability that I will survive another y years”. Over the past decade, the concept of conditional survival (CS) has emerged to directly address this question, because it provides cancer patients with survival expectations based on people who have reached a similar point in their cancer journey. However, numerous previously published CS estimates have focused on one or a few cancer types, including cancer of the head and neck [1], stomach [2], colon [3-5], rectum [6], lung [7,8], breast [9] and melanoma of the skin [10-12]. Only a few published studies provided estimates for many cancer sites [13-18], and an even smaller number have included stratification by age group and stage at diagnosis [15-17]. Ellison et al. [14] acknowledged that a stratification of conditional survival estimates by age group at diagnosis provides more relevant clinical information for clinicians and cancer patients. Similarly other studies have acknowledged the limitation of excluding information about stage at diagnosis [13,15]. This has been shown to be an important prognostic factor for survival outcomes [19]. While it has been suggested that the impact of stage reduces and can disappear for long term conditional survival [16], there are currently no published Australian data describing conditional survival outcomes according to the stage at diagnosis. This paper provides conditional survival estimates from New South Wales (NSW), Australia stratified by age group and stage at diagnosis for 11 major cancers.

Methods

Study population New South Wales is the most populous state in Australia with a population of 7.2 million, approximately one-third of the Australian population. Age-standardised mortality rates from cancer in NSW are almost identical to the national rates (187.8 per 100,000 vs 187.1 per 100,00) [20]. The de-identified records of people diagnosed with one of 11 major cancers in NSW (Table 1) were obtained from the NSW Central Cancer Registry. The Registry maintains a record of all cases of cancer diagnosed in NSW residents since 1972, with notifications from multiple sources and linkage to death certificates. We included cases diagnosed in 1972–2006 and aged 15–89 years at diagnosis. Cases reported to the Registry through death certificate only or first identified at post-mortem were excluded. Conditional 5-year relative survival estimates, by type of cancer and number of years since diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998-2006 The NSW Central Cancer Registry is the only population-based cancer registry in Australia that routinely collects information on spread of disease at diagnosis which had been used as an indicator of disease stage at diagnosis in this study. Medical coders from the Registry categorise stage based on information from statutory notification forms and pathology reports using a modified summary classification similar to the Surveillance, Epidemiology, and End Results (SEER) summary stage [21]. Categories are localised (stage I confined to tissue or organ of origin), regional (stage II spread to adjacent organs or tissues or stage III spread to regional lymph nodes), distant (stage IV with metastases to distant organs), or unknown stage (insufficient information available) [22]. Survival status was obtained through record linkage of the cancer cases in the Registry with the death records from the NSW Register of Births, Deaths, and Marriages and the National Death Index. All eligible cases were followed up to 31 December 2006 to determine survival status. This passive approach to follow-up may fail to ascertain all deaths and may incorrectly link some incidence and death records. A previous study investigating its completeness and accuracy found loss to follow-up to be uniform from 1980 to 1993 and estimated the resulting overestimation of relative survival to be a maximum of 2% [23]. This study was approved by the NSW Population and Health Service Research Ethics Committee (reference number: 2011/04/317). New South Wales is the most populous state in Australia with a population of 7.2 million, approximately one-third of the Australian population. Age-standardised mortality rates from cancer in NSW are almost identical to the national rates (187.8 per 100,000 vs 187.1 per 100,00) [20]. The de-identified records of people diagnosed with one of 11 major cancers in NSW (Table 1) were obtained from the NSW Central Cancer Registry. The Registry maintains a record of all cases of cancer diagnosed in NSW residents since 1972, with notifications from multiple sources and linkage to death certificates. We included cases diagnosed in 1972–2006 and aged 15–89 years at diagnosis. Cases reported to the Registry through death certificate only or first identified at post-mortem were excluded. Conditional 5-year relative survival estimates, by type of cancer and number of years since diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998-2006 The NSW Central Cancer Registry is the only population-based cancer registry in Australia that routinely collects information on spread of disease at diagnosis which had been used as an indicator of disease stage at diagnosis in this study. Medical coders from the Registry categorise stage based on information from statutory notification forms and pathology reports using a modified summary classification similar to the Surveillance, Epidemiology, and End Results (SEER) summary stage [21]. Categories are localised (stage I confined to tissue or organ of origin), regional (stage II spread to adjacent organs or tissues or stage III spread to regional lymph nodes), distant (stage IV with metastases to distant organs), or unknown stage (insufficient information available) [22]. Survival status was obtained through record linkage of the cancer cases in the Registry with the death records from the NSW Register of Births, Deaths, and Marriages and the National Death Index. All eligible cases were followed up to 31 December 2006 to determine survival status. This passive approach to follow-up may fail to ascertain all deaths and may incorrectly link some incidence and death records. A previous study investigating its completeness and accuracy found loss to follow-up to be uniform from 1980 to 1993 and estimated the resulting overestimation of relative survival to be a maximum of 2% [23]. This study was approved by the NSW Population and Health Service Research Ethics Committee (reference number: 2011/04/317). Statistical methods Estimation of relative survival overcomes the possibility that cause of death on death certificates may be inaccurate [24]. Relative survival is the ratio of the observed proportion surviving in a group of patients to the expected proportion that would have survived in an age- and sex-comparable group of people from the general population [25]. We calculated relative survival using the period approach [26], with cancer patients under observation between 1 January 1998 and 31 December 2006. In period analysis survival times can be left-truncated at the beginning of the period of interest in addition to being right-censored at its end. Expected survival was estimated using the Ederer and Heise (Ederer II) method [27]. Observed survival was measured from the month of diagnosis to the date of death or censoring (31 December 2006) whichever occurred first. Survival estimates were stratified by age group (15–49, 50–69 and 70–89) and stage at diagnosis separately. Stata 11 (College Station, TX: StataCorp) was used for all analyses together with publically available commands for estimating relative survival from Dickman et al. [28]. Estimation of relative survival overcomes the possibility that cause of death on death certificates may be inaccurate [24]. Relative survival is the ratio of the observed proportion surviving in a group of patients to the expected proportion that would have survived in an age- and sex-comparable group of people from the general population [25]. We calculated relative survival using the period approach [26], with cancer patients under observation between 1 January 1998 and 31 December 2006. In period analysis survival times can be left-truncated at the beginning of the period of interest in addition to being right-censored at its end. Expected survival was estimated using the Ederer and Heise (Ederer II) method [27]. Observed survival was measured from the month of diagnosis to the date of death or censoring (31 December 2006) whichever occurred first. Survival estimates were stratified by age group (15–49, 50–69 and 70–89) and stage at diagnosis separately. Stata 11 (College Station, TX: StataCorp) was used for all analyses together with publically available commands for estimating relative survival from Dickman et al. [28]. Conditional survival Conditional survival is defined as the probability of surviving an additional y years on the condition that the patient has survived x years. It is calculated by dividing the relative survival at (x + y) years after diagnosis by the relative survival at x years after diagnosis [8]. For each type of cancer, 5-year conditional survival is estimated at 1, 3, 5 and 10 years after diagnosis. We calculated the 95% confidence intervals assuming that CS follows a normal distribution and using Paul Dickman’s method for period analysis, the details of which can be found on his website [29]. Conditional survival is defined as the probability of surviving an additional y years on the condition that the patient has survived x years. It is calculated by dividing the relative survival at (x + y) years after diagnosis by the relative survival at x years after diagnosis [8]. For each type of cancer, 5-year conditional survival is estimated at 1, 3, 5 and 10 years after diagnosis. We calculated the 95% confidence intervals assuming that CS follows a normal distribution and using Paul Dickman’s method for period analysis, the details of which can be found on his website [29].

Results

A total of 193,182 cases were included in this study, with the most common cancers being prostate (39,851), female breast (36,585) and colorectum (35,455) (Table 1). Table 1 shows the 5-year relative survival estimates at diagnosis for each of the 11 selected cancer types, along with 5-year CS estimates for patients who have survived 1, 3, 5 and 10 years after diagnosis. Overall, 5-year relative survival tended to increase when conditional on increasing years after diagnosis and the greatest changes in CS occurred for cancers with poor prognosis at diagnosis for example, patients with aggressive cancers or those with advanced stage or at older age. For example, people diagnosed with lung cancer had an initial 5-year relative survival of 14%. However, their conditional 5-year relative survival increased to 33% after they survived one-year after diagnosis, and reached 85% if they survived 10 years after diagnosis. In contrast, 5-year relative survival was initially very high for men with prostate cancer (90%), with no change after surviving 10 years since diagnosis (90%). Table 2 shows the 5-year relative survival estimates stratified by stage at diagnosis for each of the 11 selected cancers and conditional on having survived 1, 3, 5 and 10 years after diagnosis. The improvement in 5-year relative survival was greatest for cases with distant metastases when conditional on increasing years already survived whereas the impact on early stage cancers was much smaller. Conditional 5-year relative survival estimates, by type of cancer, disease stage and number of years since diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998-2006 † Thickness of the lesion was also used to categorise disease stage at diagnosis for melanoma [T3 (thickness = 2.01-4.0 mm) and T4 (thickness > 4.0 mm) stages were grouped with ‘spread to regional lymph nodes’ as ”regional”]. Since thickness data prior to 1983 were considered unreliable, we used melanoma data from 1983 onward for the stage-specific survival estimates. Age-specific and stage-specific conditional 5-year relative survival at 0, 1, 3, 5 and 10 years after diagnosis for each selected cancer are also presented graphically in Figures 1 and 2. For most cancers, the age or stage differential in survival at diagnosis generally decreased over time except for cancers of the lung and pancreas. Age-specific conditional 5-year relative survival at 0, 1, 3, 5, 10 years after diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998–2006. Stage-specific conditional 5-year relative survival at 0, 1, 3, 5, 10 years after diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998–2006.

Discussion and conclusion

The pre-publication history for this paper can be accessed here: http://www.biomedcentral.com/1471-2407/12/460/prepub

Background: Survival estimates for cancer patients are traditionally reported from the time of diagnosis such as five-year survival. It is useful for answering questions that many people ask about their prognosis when first diagnosed with cancer. For cancer patients who have already survived a number of years, survival expectations at diagnosis are too pessimistic because they include people who have already died. An ongoing question among these surviving patients is “now that I have survived for x years, what is the probability that I will survive another y years”. Over the past decade, the concept of conditional survival (CS) has emerged to directly address this question, because it provides cancer patients with survival expectations based on people who have reached a similar point in their cancer journey. However, numerous previously published CS estimates have focused on one or a few cancer types, including cancer of the head and neck [1], stomach [2], colon [3-5], rectum [6], lung [7,8], breast [9] and melanoma of the skin [10-12]. Only a few published studies provided estimates for many cancer sites [13-18], and an even smaller number have included stratification by age group and stage at diagnosis [15-17]. Ellison et al. [14] acknowledged that a stratification of conditional survival estimates by age group at diagnosis provides more relevant clinical information for clinicians and cancer patients. Similarly other studies have acknowledged the limitation of excluding information about stage at diagnosis [13,15]. This has been shown to be an important prognostic factor for survival outcomes [19]. While it has been suggested that the impact of stage reduces and can disappear for long term conditional survival [16], there are currently no published Australian data describing conditional survival outcomes according to the stage at diagnosis. This paper provides conditional survival estimates from New South Wales (NSW), Australia stratified by age group and stage at diagnosis for 11 major cancers. Methods: Study population New South Wales is the most populous state in Australia with a population of 7.2 million, approximately one-third of the Australian population. Age-standardised mortality rates from cancer in NSW are almost identical to the national rates (187.8 per 100,000 vs 187.1 per 100,00) [20]. The de-identified records of people diagnosed with one of 11 major cancers in NSW (Table 1) were obtained from the NSW Central Cancer Registry. The Registry maintains a record of all cases of cancer diagnosed in NSW residents since 1972, with notifications from multiple sources and linkage to death certificates. We included cases diagnosed in 1972–2006 and aged 15–89 years at diagnosis. Cases reported to the Registry through death certificate only or first identified at post-mortem were excluded. Conditional 5-year relative survival estimates, by type of cancer and number of years since diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998-2006 The NSW Central Cancer Registry is the only population-based cancer registry in Australia that routinely collects information on spread of disease at diagnosis which had been used as an indicator of disease stage at diagnosis in this study. Medical coders from the Registry categorise stage based on information from statutory notification forms and pathology reports using a modified summary classification similar to the Surveillance, Epidemiology, and End Results (SEER) summary stage [21]. Categories are localised (stage I confined to tissue or organ of origin), regional (stage II spread to adjacent organs or tissues or stage III spread to regional lymph nodes), distant (stage IV with metastases to distant organs), or unknown stage (insufficient information available) [22]. Survival status was obtained through record linkage of the cancer cases in the Registry with the death records from the NSW Register of Births, Deaths, and Marriages and the National Death Index. All eligible cases were followed up to 31 December 2006 to determine survival status. This passive approach to follow-up may fail to ascertain all deaths and may incorrectly link some incidence and death records. A previous study investigating its completeness and accuracy found loss to follow-up to be uniform from 1980 to 1993 and estimated the resulting overestimation of relative survival to be a maximum of 2% [23]. This study was approved by the NSW Population and Health Service Research Ethics Committee (reference number: 2011/04/317). New South Wales is the most populous state in Australia with a population of 7.2 million, approximately one-third of the Australian population. Age-standardised mortality rates from cancer in NSW are almost identical to the national rates (187.8 per 100,000 vs 187.1 per 100,00) [20]. The de-identified records of people diagnosed with one of 11 major cancers in NSW (Table 1) were obtained from the NSW Central Cancer Registry. The Registry maintains a record of all cases of cancer diagnosed in NSW residents since 1972, with notifications from multiple sources and linkage to death certificates. We included cases diagnosed in 1972–2006 and aged 15–89 years at diagnosis. Cases reported to the Registry through death certificate only or first identified at post-mortem were excluded. Conditional 5-year relative survival estimates, by type of cancer and number of years since diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998-2006 The NSW Central Cancer Registry is the only population-based cancer registry in Australia that routinely collects information on spread of disease at diagnosis which had been used as an indicator of disease stage at diagnosis in this study. Medical coders from the Registry categorise stage based on information from statutory notification forms and pathology reports using a modified summary classification similar to the Surveillance, Epidemiology, and End Results (SEER) summary stage [21]. Categories are localised (stage I confined to tissue or organ of origin), regional (stage II spread to adjacent organs or tissues or stage III spread to regional lymph nodes), distant (stage IV with metastases to distant organs), or unknown stage (insufficient information available) [22]. Survival status was obtained through record linkage of the cancer cases in the Registry with the death records from the NSW Register of Births, Deaths, and Marriages and the National Death Index. All eligible cases were followed up to 31 December 2006 to determine survival status. This passive approach to follow-up may fail to ascertain all deaths and may incorrectly link some incidence and death records. A previous study investigating its completeness and accuracy found loss to follow-up to be uniform from 1980 to 1993 and estimated the resulting overestimation of relative survival to be a maximum of 2% [23]. This study was approved by the NSW Population and Health Service Research Ethics Committee (reference number: 2011/04/317). Statistical methods Estimation of relative survival overcomes the possibility that cause of death on death certificates may be inaccurate [24]. Relative survival is the ratio of the observed proportion surviving in a group of patients to the expected proportion that would have survived in an age- and sex-comparable group of people from the general population [25]. We calculated relative survival using the period approach [26], with cancer patients under observation between 1 January 1998 and 31 December 2006. In period analysis survival times can be left-truncated at the beginning of the period of interest in addition to being right-censored at its end. Expected survival was estimated using the Ederer and Heise (Ederer II) method [27]. Observed survival was measured from the month of diagnosis to the date of death or censoring (31 December 2006) whichever occurred first. Survival estimates were stratified by age group (15–49, 50–69 and 70–89) and stage at diagnosis separately. Stata 11 (College Station, TX: StataCorp) was used for all analyses together with publically available commands for estimating relative survival from Dickman et al. [28]. Estimation of relative survival overcomes the possibility that cause of death on death certificates may be inaccurate [24]. Relative survival is the ratio of the observed proportion surviving in a group of patients to the expected proportion that would have survived in an age- and sex-comparable group of people from the general population [25]. We calculated relative survival using the period approach [26], with cancer patients under observation between 1 January 1998 and 31 December 2006. In period analysis survival times can be left-truncated at the beginning of the period of interest in addition to being right-censored at its end. Expected survival was estimated using the Ederer and Heise (Ederer II) method [27]. Observed survival was measured from the month of diagnosis to the date of death or censoring (31 December 2006) whichever occurred first. Survival estimates were stratified by age group (15–49, 50–69 and 70–89) and stage at diagnosis separately. Stata 11 (College Station, TX: StataCorp) was used for all analyses together with publically available commands for estimating relative survival from Dickman et al. [28]. Conditional survival Conditional survival is defined as the probability of surviving an additional y years on the condition that the patient has survived x years. It is calculated by dividing the relative survival at (x + y) years after diagnosis by the relative survival at x years after diagnosis [8]. For each type of cancer, 5-year conditional survival is estimated at 1, 3, 5 and 10 years after diagnosis. We calculated the 95% confidence intervals assuming that CS follows a normal distribution and using Paul Dickman’s method for period analysis, the details of which can be found on his website [29]. Conditional survival is defined as the probability of surviving an additional y years on the condition that the patient has survived x years. It is calculated by dividing the relative survival at (x + y) years after diagnosis by the relative survival at x years after diagnosis [8]. For each type of cancer, 5-year conditional survival is estimated at 1, 3, 5 and 10 years after diagnosis. We calculated the 95% confidence intervals assuming that CS follows a normal distribution and using Paul Dickman’s method for period analysis, the details of which can be found on his website [29]. Study population: New South Wales is the most populous state in Australia with a population of 7.2 million, approximately one-third of the Australian population. Age-standardised mortality rates from cancer in NSW are almost identical to the national rates (187.8 per 100,000 vs 187.1 per 100,00) [20]. The de-identified records of people diagnosed with one of 11 major cancers in NSW (Table 1) were obtained from the NSW Central Cancer Registry. The Registry maintains a record of all cases of cancer diagnosed in NSW residents since 1972, with notifications from multiple sources and linkage to death certificates. We included cases diagnosed in 1972–2006 and aged 15–89 years at diagnosis. Cases reported to the Registry through death certificate only or first identified at post-mortem were excluded. Conditional 5-year relative survival estimates, by type of cancer and number of years since diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998-2006 The NSW Central Cancer Registry is the only population-based cancer registry in Australia that routinely collects information on spread of disease at diagnosis which had been used as an indicator of disease stage at diagnosis in this study. Medical coders from the Registry categorise stage based on information from statutory notification forms and pathology reports using a modified summary classification similar to the Surveillance, Epidemiology, and End Results (SEER) summary stage [21]. Categories are localised (stage I confined to tissue or organ of origin), regional (stage II spread to adjacent organs or tissues or stage III spread to regional lymph nodes), distant (stage IV with metastases to distant organs), or unknown stage (insufficient information available) [22]. Survival status was obtained through record linkage of the cancer cases in the Registry with the death records from the NSW Register of Births, Deaths, and Marriages and the National Death Index. All eligible cases were followed up to 31 December 2006 to determine survival status. This passive approach to follow-up may fail to ascertain all deaths and may incorrectly link some incidence and death records. A previous study investigating its completeness and accuracy found loss to follow-up to be uniform from 1980 to 1993 and estimated the resulting overestimation of relative survival to be a maximum of 2% [23]. This study was approved by the NSW Population and Health Service Research Ethics Committee (reference number: 2011/04/317). Statistical methods: Estimation of relative survival overcomes the possibility that cause of death on death certificates may be inaccurate [24]. Relative survival is the ratio of the observed proportion surviving in a group of patients to the expected proportion that would have survived in an age- and sex-comparable group of people from the general population [25]. We calculated relative survival using the period approach [26], with cancer patients under observation between 1 January 1998 and 31 December 2006. In period analysis survival times can be left-truncated at the beginning of the period of interest in addition to being right-censored at its end. Expected survival was estimated using the Ederer and Heise (Ederer II) method [27]. Observed survival was measured from the month of diagnosis to the date of death or censoring (31 December 2006) whichever occurred first. Survival estimates were stratified by age group (15–49, 50–69 and 70–89) and stage at diagnosis separately. Stata 11 (College Station, TX: StataCorp) was used for all analyses together with publically available commands for estimating relative survival from Dickman et al. [28]. Conditional survival: Conditional survival is defined as the probability of surviving an additional y years on the condition that the patient has survived x years. It is calculated by dividing the relative survival at (x + y) years after diagnosis by the relative survival at x years after diagnosis [8]. For each type of cancer, 5-year conditional survival is estimated at 1, 3, 5 and 10 years after diagnosis. We calculated the 95% confidence intervals assuming that CS follows a normal distribution and using Paul Dickman’s method for period analysis, the details of which can be found on his website [29]. Results: A total of 193,182 cases were included in this study, with the most common cancers being prostate (39,851), female breast (36,585) and colorectum (35,455) (Table 1). Table 1 shows the 5-year relative survival estimates at diagnosis for each of the 11 selected cancer types, along with 5-year CS estimates for patients who have survived 1, 3, 5 and 10 years after diagnosis. Overall, 5-year relative survival tended to increase when conditional on increasing years after diagnosis and the greatest changes in CS occurred for cancers with poor prognosis at diagnosis for example, patients with aggressive cancers or those with advanced stage or at older age. For example, people diagnosed with lung cancer had an initial 5-year relative survival of 14%. However, their conditional 5-year relative survival increased to 33% after they survived one-year after diagnosis, and reached 85% if they survived 10 years after diagnosis. In contrast, 5-year relative survival was initially very high for men with prostate cancer (90%), with no change after surviving 10 years since diagnosis (90%). Table 2 shows the 5-year relative survival estimates stratified by stage at diagnosis for each of the 11 selected cancers and conditional on having survived 1, 3, 5 and 10 years after diagnosis. The improvement in 5-year relative survival was greatest for cases with distant metastases when conditional on increasing years already survived whereas the impact on early stage cancers was much smaller. Conditional 5-year relative survival estimates, by type of cancer, disease stage and number of years since diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998-2006 † Thickness of the lesion was also used to categorise disease stage at diagnosis for melanoma [T3 (thickness = 2.01-4.0 mm) and T4 (thickness > 4.0 mm) stages were grouped with ‘spread to regional lymph nodes’ as ”regional”]. Since thickness data prior to 1983 were considered unreliable, we used melanoma data from 1983 onward for the stage-specific survival estimates. Age-specific and stage-specific conditional 5-year relative survival at 0, 1, 3, 5 and 10 years after diagnosis for each selected cancer are also presented graphically in Figures 1 and 2. For most cancers, the age or stage differential in survival at diagnosis generally decreased over time except for cancers of the lung and pancreas. Age-specific conditional 5-year relative survival at 0, 1, 3, 5, 10 years after diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998–2006. Stage-specific conditional 5-year relative survival at 0, 1, 3, 5, 10 years after diagnosis, for patients aged 15–89 years at diagnosis, NSW Australia 1998–2006. Discussion and conclusion: This study provides quantitative evidence that Australian cancer patients who are still alive ten years after their cancer diagnosis, even those diagnosed with advanced stage disease or at older age, have, at that moment, a much better survival outlook over the next five years than they did at diagnosis. The information is important for cancer patients as they face important life decisions in trying to plan their remaining life, and to provide evidence-based optimism as they continue living after their initial cancer diagnosis. When 5-year relative survival exceeds 95%, the excess mortality is minimal, and so the survival for this group is considered similar to the general population with the same age structure [15,16], although this does not necessarily indicate cure of cancer. In NSW, we found that patients who were diagnosed with cancer of the colorectum, cervix and thyroid, and melanoma of the skin had similar mortality expectations after surviving 10 years since diagnosis. This was consistent with another Australian report [13]. Our results of 5-year CS after surviving one year since diagnosis were also consistent with a recent international comparison of cancer survival in which NSW data for colorectum, lung and female breast cancer were also included [18]. Their published CS rates for the three cancers in NSW [18] were slightly higher than those reported here as their estimates were age-adjusted, which would have the effect of increasing overall survival by reducing the weight given to the poorer survival among older people. The overall consistency of these results [13,18] with those we reported for overall CS during an almost identical study period provides indirect confirmation of our findings. The strength of our study is that we presented age- and stage-specific CS in addition to the overall CS. By presenting age- and stage-specific CS estimates for 11 major cancers in one geographically defined population, clinicians, cancer patients and their support networks can compare temporal and age-specific patterns in CS across multiple cancer types, thus gaining a greater understanding of the ongoing survival expectations faced by cancer patients. Our results are consistent with many international studies including those in Europe [15,16] and North America [14,17] for a variety of individual cancer types. Our 5-year relative survival estimates conditional on surviving 5 years after diagnosis were very close to those for major cancers reported in Canada including cancer of the colorectum, lung, breast and melanoma [14]. The overall patterns of stage-specific CS for major cancer types from SEER data [17] and our data were also very similar particularly when accounting for any differences in age distribution: the increase in 5-year relative survival when conditional on more years already survived was greatest for later stage cancer and also the more fatal cancers but the stage differential in survival tended to reduce over time. Our stage-specific 5-year relative survival estimates conditional on having survived 5 years were very similar to those from SEER data [17] for localised and regional stage cancers of the colon and rectum, lung, breast and melanoma. However, considerable differences were observed for distant or unknown stages which may reflect different case mix in these two groups of patients due partly to more stringent criteria and data quality control being used in the SEER system than in Australia. Regarding the effect of age at diagnosis on CS, patients diagnosed at older ages tended to have lower relative conditional survival, but this effect reduced substantially after surviving 5–10 years for most cancers. The overall pattern of age-specific CS estimates were similar to those from other studies in Europe [15,16] and US [17], although different age categories make it difficult to compare specific estimates. When we reanalysed our data with age categories matching those in a European study [15], we found that age-specific CS was consistently higher in NSW, apart from identical results for thyroid cancer for the three younger age groups (data not shown). The overall survival differences for three major cancers of the colorectum, lung and female breast between Australia and Europe had been confirmed by a recent study comparing survival from four major cancers between Australia, Canada and several European countries [18]. As suggested in that study [18], the possible explanations for these survival differences may be due to later diagnosis or differences in treatment in the European population. As was noted in most previous studies, the greatest differences in conditional survival were for those cancers that had initially poor survival, such as lung or pancreatic cancer. This study confirmed a similar pattern for those cancers diagnosed at an advanced stage. While there was a substantial impact of disease stage on survival expectations at diagnosis, for most types of cancer this stage differential decreased as time since diagnosis increased, a pattern that has been reported in many international studies [2,8,15,17]. Unfortunately, because of the high initial mortality, the number of people who survive to enjoy this greater survival is low. However the fact remains that of the people who do survive more than ten years after diagnosis, many continue to have poorer survival expectations than the general population with the same age structure. This could relate to the impact of the co-morbidities associated with the initial cancer diagnosis (for example while smoking causes lung cancer it is also associated with increased risk of cardiovascular diseases); the late recurrences of the primary cancer or secondary tumours; or the late side effects of treatment [15,16]. These ongoing reduced survival expectations for people diagnosed with late stage cancer in particular has substantial implications for health care providers in providing regular surveillance and monitoring even when the patients have survived at least ten or more years after the initial cancer diagnosis. There are three widely used methods for estimating expected survival for relative survival analysis, commonly known as the Ederer I [30], Ederer II [27] and Hakulinen [31] methods. The Hakulinen method [31] was widely used in many international studies of cancer survival using population-based data including the most recent EUROCARE-4 study [32] and CONCORD study [33]. However, there is a growing consensus among researchers in relative survival analysis using population-based cancer registry data that the Ederer II method is more preferable [34-36], although relative survival estimates, using any of these methods, are generally very similar. Following this recommendation, we used the Ederer II [27] method in our estimation of relative survival. More recently, a modified Ederer II estimator has been proposed which is obtained by weighting the individual observations with their population survival [36]. The authors recommend the use of this new estimator when comparing cancer survival between countries because it is believed that this is the only unbiased estimator [36]. Another recent simulation study provided evidence that this estimator [36] is only unbiased for net survival when compared with other widely used estimators (including Ederer II [27] and Hakulinen [31] estimators) [37]. As the new estimator has not been used within the context of period analysis, future research in this area may be warranted. Strengths of this study are the statewide population-based cancer registry data including information about the stage at diagnosis and multiple cancer types. This makes our study more representative and comprehensive than many other studies. The similarity between our estimates for all stages of cancer at diagnosis and those for another Australian state [13] provide optimism that the stage-specific CS estimates can be generalised nationally. We were unable to adjust for treatment, which may have impacted survival estimates through initial remission of cancer but may also cause longer-term adverse complications. Although mortality information was obtained by matching against the National Death Index, it remains possible that some deaths were missed, thus artificially inflating survival estimates. However since the matching process was not conditional on cancer type, it is unlikely that this would influence comparisons of conditional survival across cancer types. In conclusion, these data provide important information for cancer patients, based on age and the stage at diagnosis, as they continue on their cancer journey. This information should be used by clinicians as a tool to make evidence-based decisions regarding follow-up, surveillance, or ongoing management according to their patient’s changing survival expectations over time.

Background: Estimated conditional survival for cancer patients diagnosed at different ages and disease stage provides important information for cancer patients and clinicians in planning follow-up, surveillance and ongoing management. Methods: Using population-based cancer registry data for New South Wales Australia, we estimated conditional 5-year relative survival for 11 major cancers diagnosed 1972-2006 by time since diagnosis and age and stage at diagnosis. Results: 193,182 cases were included, with the most common cancers being prostate (39,851), female breast (36,585) and colorectal (35,455). Five-year relative survival tended to increase with increasing years already survived and improvement was greatest for cancers with poor prognosis at diagnosis (lung or pancreas) and for those with advanced stage or older age at diagnosis. After surviving 10 years, conditional 5-year survival was over 95% for 6 localised, 6 regional, 3 distant and 3 unknown stage cancers. For the remaining patient groups, conditional 5-year survival ranged from 74% (for distant stage bladder cancer) to 94% (for 4 cancers at different stages), indicating that they continue to have excess mortality 10-15 years after diagnosis. Conclusions: These data provide important information for cancer patients, based on age and stage at diagnosis, as they continue on their cancer journey. This information may also be used by clinicians as a tool to make more evidence-based decisions regarding follow-up, surveillance, or ongoing management according to patients' changing survival expectations over time.

Background: Survival estimates for cancer patients are traditionally reported from the time of diagnosis such as five-year survival. It is useful for answering questions that many people ask about their prognosis when first diagnosed with cancer. For cancer patients who have already survived a number of years, survival expectations at diagnosis are too pessimistic because they include people who have already died. An ongoing question among these surviving patients is “now that I have survived for x years, what is the probability that I will survive another y years”. Over the past decade, the concept of conditional survival (CS) has emerged to directly address this question, because it provides cancer patients with survival expectations based on people who have reached a similar point in their cancer journey. However, numerous previously published CS estimates have focused on one or a few cancer types, including cancer of the head and neck [1], stomach [2], colon [3-5], rectum [6], lung [7,8], breast [9] and melanoma of the skin [10-12]. Only a few published studies provided estimates for many cancer sites [13-18], and an even smaller number have included stratification by age group and stage at diagnosis [15-17]. Ellison et al. [14] acknowledged that a stratification of conditional survival estimates by age group at diagnosis provides more relevant clinical information for clinicians and cancer patients. Similarly other studies have acknowledged the limitation of excluding information about stage at diagnosis [13,15]. This has been shown to be an important prognostic factor for survival outcomes [19]. While it has been suggested that the impact of stage reduces and can disappear for long term conditional survival [16], there are currently no published Australian data describing conditional survival outcomes according to the stage at diagnosis. This paper provides conditional survival estimates from New South Wales (NSW), Australia stratified by age group and stage at diagnosis for 11 major cancers. Discussion and conclusion: The pre-publication history for this paper can be accessed here: http://www.biomedcentral.com/1471-2407/12/460/prepub

Background: Estimated conditional survival for cancer patients diagnosed at different ages and disease stage provides important information for cancer patients and clinicians in planning follow-up, surveillance and ongoing management. Methods: Using population-based cancer registry data for New South Wales Australia, we estimated conditional 5-year relative survival for 11 major cancers diagnosed 1972-2006 by time since diagnosis and age and stage at diagnosis. Results: 193,182 cases were included, with the most common cancers being prostate (39,851), female breast (36,585) and colorectal (35,455). Five-year relative survival tended to increase with increasing years already survived and improvement was greatest for cancers with poor prognosis at diagnosis (lung or pancreas) and for those with advanced stage or older age at diagnosis. After surviving 10 years, conditional 5-year survival was over 95% for 6 localised, 6 regional, 3 distant and 3 unknown stage cancers. For the remaining patient groups, conditional 5-year survival ranged from 74% (for distant stage bladder cancer) to 94% (for 4 cancers at different stages), indicating that they continue to have excess mortality 10-15 years after diagnosis. Conclusions: These data provide important information for cancer patients, based on age and stage at diagnosis, as they continue on their cancer journey. This information may also be used by clinicians as a tool to make more evidence-based decisions regarding follow-up, surveillance, or ongoing management according to patients' changing survival expectations over time.

[CONTENT] Conditional survival | Relative survival | Cancer registry | Australia [SUMMARY]

[CONTENT] Conditional survival | Relative survival | Cancer registry | Australia [SUMMARY]

[CONTENT] Conditional survival | Relative survival | Cancer registry | Australia [SUMMARY]

[CONTENT] Conditional survival | Relative survival | Cancer registry | Australia [SUMMARY]

[CONTENT] Conditional survival | Relative survival | Cancer registry | Australia [SUMMARY]

[CONTENT] Conditional survival | Relative survival | Cancer registry | Australia [SUMMARY]

[CONTENT] Adolescent | Adult | Aged | Aged, 80 and over | Breast Neoplasms | Colorectal Neoplasms | Female | Humans | Male | Middle Aged | Mortality | Neoplasm Staging | Neoplasms | New South Wales | Population Surveillance | Prostatic Neoplasms | Registries | Survival Analysis | Survival Rate | Time Factors | Young Adult [SUMMARY]

[CONTENT] Adolescent | Adult | Aged | Aged, 80 and over | Breast Neoplasms | Colorectal Neoplasms | Female | Humans | Male | Middle Aged | Mortality | Neoplasm Staging | Neoplasms | New South Wales | Population Surveillance | Prostatic Neoplasms | Registries | Survival Analysis | Survival Rate | Time Factors | Young Adult [SUMMARY]

[CONTENT] Adolescent | Adult | Aged | Aged, 80 and over | Breast Neoplasms | Colorectal Neoplasms | Female | Humans | Male | Middle Aged | Mortality | Neoplasm Staging | Neoplasms | New South Wales | Population Surveillance | Prostatic Neoplasms | Registries | Survival Analysis | Survival Rate | Time Factors | Young Adult [SUMMARY]

[CONTENT] Adolescent | Adult | Aged | Aged, 80 and over | Breast Neoplasms | Colorectal Neoplasms | Female | Humans | Male | Middle Aged | Mortality | Neoplasm Staging | Neoplasms | New South Wales | Population Surveillance | Prostatic Neoplasms | Registries | Survival Analysis | Survival Rate | Time Factors | Young Adult [SUMMARY]

[CONTENT] Adolescent | Adult | Aged | Aged, 80 and over | Breast Neoplasms | Colorectal Neoplasms | Female | Humans | Male | Middle Aged | Mortality | Neoplasm Staging | Neoplasms | New South Wales | Population Surveillance | Prostatic Neoplasms | Registries | Survival Analysis | Survival Rate | Time Factors | Young Adult [SUMMARY]

[CONTENT] Adolescent | Adult | Aged | Aged, 80 and over | Breast Neoplasms | Colorectal Neoplasms | Female | Humans | Male | Middle Aged | Mortality | Neoplasm Staging | Neoplasms | New South Wales | Population Surveillance | Prostatic Neoplasms | Registries | Survival Analysis | Survival Rate | Time Factors | Young Adult [SUMMARY]

[CONTENT] survival cancers initially | patients survival expectations | comparison cancer survival | survival estimates cancer | conditional survival cancer [SUMMARY]

[CONTENT] survival cancers initially | patients survival expectations | comparison cancer survival | survival estimates cancer | conditional survival cancer [SUMMARY]