Spaces:
Running
Running
| <html> | |
| <head> | |
| <link rel="stylesheet" href="css/style_pha.css"></link> | |
| <title>Pharmacometricss</title> | |
| <meta charset="UTF-8"> | |
| </head> | |
| <body> | |
| <div class="container"> | |
| <div class="header"> | |
| <h1>DRC Pharmacometrics</h1> | |
| </div> | |
| <div class="nav"> | |
| <a href="https://drclab.github.io" target="_blank">DRC Lab</a> | |
| <a href="hed.html">DRC Higher Education</a> | |
| <a href="cpg.html">DRC CPG</a> | |
| <a href="genomics.html">DRC Biomics</a> | |
| <a href="pharmaco.html">DRC Pharmacometrics</a> | |
| <a href="sports.html">DRC Sports</a> | |
| <a href="about.html">About</a> | |
| </div> | |
| <div class="main"> | |
| <div class="content"> | |
| <h2>Pharmacometric Modeling</h2> | |
| <p><img src="https://th.bing.com/th/id/OIG.7pdij30TSuUnPie_xxgM?w=270&h=270&c=6&r=0&o=5&pid=ImgGn" height="220px" width="220px">Pharmacometrics is a scientific discipline that applies mathematical and statistical methods to study the effects of drugs and diseases in human populations. Pharmacometrics can help design optimal clinical trials, evaluate drug efficacy and safety, and support regulatory decision-making. Pharmacometrics can also provide insights into the mechanisms of drug action, the sources of variability in drug response, and the optimal dosing strategies for different patient groups. | |
| Pharmacometrics uses various types of models to describe and quantify the interactions between drugs, diseases, and patients. Some of the common models are:</p> | |
| <p>Pharmacokinetic (PK) models: These models describe how the body absorbs, distributes, metabolizes, and eliminates drugs over time. | |
| PK models can help estimate drug exposure and bioavailability in different tissues and organs.</p> | |
| <p>Pharmacodynamic (PD) models: These models describe how drugs affect biological processes and outcomes, such as biomarkers, symptoms, or quality of life. | |
| PD models can help estimate drug potency, efficacy, and tolerability in different patient populations.</p> | |
| <p>Systems pharmacology models: These models integrate PK and PD models with mechanistic information on the biological pathways and networks involved in | |
| drug action and disease progression. Systems pharmacology models can help understand the complex interactions between drugs and biological systems at | |
| multiple levels of organization.</p> | |
| <p> progression models: These models describe how the natural course of a disease changes over time, with or without treatment. | |
| Disease progression models can help predict long-term outcomes, identify risk factors, and evaluate treatment effects on disease modification or prevention.</p> | |
| <p>Trial simulation models: These models combine PK, PD, and disease progression models with information on trial design and execution. Trial simulation models | |
| can help optimize trial design, evaluate trial feasibility and success, and perform sensitivity and scenario analyses. | |
| </p> | |
| <p>Pharmacometrics is a rapidly evolving field that requires interdisciplinary collaboration among pharmacologists, clinicians, statisticians, engineers, | |
| and data scientists. Pharmacometrics can provide valuable information to support drug development and regulatory decisions that ultimately benefit patients | |
| and public health.</p> | |
| <h2>Drug-Disease Model</h2> | |
| <p>A drug-disease model is a mathematical representation of the interactions between a drug and a disease. It can be used to | |
| simulate the effects of different doses, regimens, and durations of treatment on the disease progression and outcomes. | |
| A drug-disease model can also incorporate various sources of variability and uncertainty, such as patient characteristics, | |
| pharmacokinetics, pharmacodynamics, adherence, and measurement errors. | |
| Drug-disease models can help in the design and analysis of clinical trials, as well as in the optimization and individualization | |
| of therapy. They can provide insights into the mechanisms of action and resistance of drugs, and identify potential biomarkers | |
| and surrogate endpoints. Drug-disease models can also support decision making and health economic evaluations by estimating the | |
| long-term benefits and risks of different treatment options. | |
| Drug-disease models are typically developed using a combination of mechanistic and empirical approaches. Mechanistic models are | |
| based on biological knowledge and assumptions about the underlying processes that govern the drug-disease interactions. | |
| Empirical models are based on statistical methods and data-driven techniques that fit the model parameters to the observed data. | |
| Both types of models have advantages and limitations, and often a hybrid approach is used to balance the trade-offs between | |
| complexity and realism. </p> | |
| <h2>Mono Exponential Elimination model</h2> | |
| <p>A mono exponential elimination model is a type of pharmacokinetic model that assumes that the drug is eliminated from the body at a | |
| rate proportional to its concentration. This means that the drug concentration decreases exponentially over time, following a single | |
| exponential decay function. The equation for the mono exponential elimination model is: | |
| <br> | |
| Cp(t) = Cp(0) * exp(-ke * t) | |
| <br> | |
| where Cp(t) is the plasma concentration of the drug at time t, Cp(0) is the initial plasma concentration of the drug, ke is | |
| the elimination rate constant, and exp is the exponential function. | |
| The mono exponential elimination model is the simplest compartmental pharmacokinetic model, and it can be applied to some | |
| drugs that distribute homogeneously and instantaneously in the body, and have a constant volume of distribution and a first-order | |
| elimination process. However, many drugs do not follow this simple model, and may require more complex models that account for | |
| multiple compartments, nonlinear | |
| elimination, absorption, distribution, metabolism, and other factors that affect the drug concentration-time profile.</p> | |
| <h2>Bayesian aggregation of average data</h2> | |
| <p>In drug development, it is often necessary to combine information from multiple sources, such as different clinical trials, | |
| to estimate the efficacy and safety of a new treatment. However, the available data may not be in a suitable format for direct | |
| pooling, as they may only report summary statistics (such as means and standard deviations) rather than individual patient data. | |
| In this <a href="https://arxiv.org/abs/1602.02055" target="_blank">paper</a>, Web et al.(2020) propose a Bayesian method for aggregating average data from multiple sources, using a hierarchical model that | |
| accounts for heterogeneity and uncertainty across studies. We illustrate the method with an example of combining data from four | |
| phase III trials of a new drug for treating rheumatoid arthritis. They show how the Bayesian approach can provide more precise and | |
| robust estimates of the treatment effect and its variability than conventional meta-analysis methods, and how it can | |
| facilitate inference on subgroups and individual patients. They also discuss some practical issues and challenges in | |
| implementing the method, such as prior specification, model checking, and computational efficiency.</p> | |
| </div> | |
| </div> | |
| <div class="nav_tech"> | |
| <a href="https://medium.com/paper-club/understanding-pyros-model-and-guide-a-love-story-7f3aa0603886" target="_blank">Pyro Model and Guide</a> | |
| <a href="https://mc-stan.org/" target="_blank">Stan</a> | |
| <a href="https://pytorch.org/" target="_blank">PyTorch</a> | |
| <a href="https://huggingface.co/" target="_blank">Hugging Face AI</a> | |
| <a href="https://openai.com/blog/chatgpt" target="_blank">ChatGPT</a> | |
| <a href="https://arxiv.org/" target="_blank">ArXiv</a> | |
| </div> | |
| <div class="footer"> | |
| <br> | |
| Copyright © Dulun Research & Consulting<br> | |
| Last updated on April 11th 2023<br> | |
| <a href="mailto:compute@dulun.com" target="_blank">Contact us</a><br> | |
| </div> | |
| </div> | |
| </body> | |
| </html> | |