Bioequivalence Statistical Analysis: A Bootcamp for Drug Development

This five-day "bootcamp" workshop offers a practical and comprehensive introduction to bioequivalence analysis in drug development. Participants will learn to design, analyze, and interpret bioequivalence studies using statistical methods while selecting appropriate study designs and approaches for various scenarios. Key skills covered include statistical inference, power and sample size planning, data analysis, and interpreting study results for informed decision-making. The workshop primarily utilizes the Pumas software, including PumasCP, while also addressing a few SAS examples since these appear in some regulatory documents. Participants will explore the influence of regulatory guidelines, tackle common challenges such as missing data and multicenter trials, and gain insight into statistical methods, assumptions, and algorithms. Additionally, the workshop introduces virtual bioequivalence and its practical applications. Through hands-on projects and case studies, attendees will acquire valuable, actionable experience for their professional work.

Instructors: Andreas Noack, PhD, and Yoni Nazarathy, PhD

Workshop Objectives

1. 1. Understand Bioequivalence Study Design
   Develop the ability to design bioequivalence studies for various scenarios, aligning with regulatory guidelines and industry best practices.
2. Apply Statistical Inference
   Learn to perform and interpret statistical inference using linear models, mixed models, nonparametric models, and other relevant approaches for bioequivalence analysis.
3. Apply Specific Techniques
   Learn how to use specific bioequivalence techniques for highly variable drugs and narrow therapeutic index drugs.
4. Plan Power and Sample Size
   Gain skills in power analysis and sample size determination for study planning.
5. Analyze and Interpret Data
   Use statistical methods and software tools to analyze bioequivalence data, interpret results, and address challenges like missing data.
6. Utilize Software for Analysis
   Build proficiency with Pumas and PumasCP for bioequivalence analysis.
7. Explore Advanced Scenarios
   Understand advanced bioequivalence scenarios, including virtual bioequivalence, multicenter trials, and other complex study designs, and their practical applications in drug development.
8. Gain a Deep Statistical Understanding
   Strengthen understanding of statistical principles and methodologies underpinning bioequivalence analysis. This includes a basic understanding of the development of bioequivalence approaches over the years.

Target Audience

The workshop is for professionals involved in pharmaceutical development, research, and regulatory affairs, including scientists and researchers from the pharmaceutical industry, regulatory affairs professionals, statisticians working in the pharmaceutical domain, as well as academics and students interested in bioequivalence.

Agenda

Day 1: Introduction to bioequivalence with an end to end example for generic drug approval

We consider a complete simple example of a clinical trial for generic drug approval where drug concentration data is recorded for both a reference and a test formulation. After visiting pharmacokinetic principles relevant to bioequivalence, with a birds eye view, we use PumasCP to carry out NCA (Non-compartmental analysis) to reach a dataset suitable for bioequivalence analysis. We then demonstrate the analysis and its implications, including understanding of basic concepts such as geometric means (and their ratios), coefficient of variation, and basic statistical assumptions. Here we see the TOST (Two One Sided t-Test) for the first time. In passing, we also review basic statistical concepts including descriptive statistics, probability distributions, hypothesis testing, and confidence intervals.

As with all days of the workshop, we undertake a hands-on approach, where we alternate between theoretical discussion and hands on activities. We will use PumasCP this day.

Day 2: The basic guidances and statistical models

This day focuses on two main topics:

(1) Understanding guidances by the US FDA, EMA, Indian regulators, and others.

(2) Unpacking statistical models. With respect to regulators we visit the key regulatory documents and learn to navigate key sections of these documents, also with the help of AskPumas.

With respect to statistical models we understand the basics of linear models, mixed models, nonparameteric analysis, and a few other variants, all used for bioequivalence analysis. The statistical bootcamp is "from the ground up", where we revisit the statistical meaning of the associated hypothesis tests and inference.

During this session, we will use Pumas command line with the `Bioequivlance` package for the first time and continue to use PumasCP.

Day 3: Study designs, study planning, and more on statistical models

We take a deep dive into study designs including parallel designs, cross over designs, replicate designs, and partial replicate designs. In each case we consider uses, pros, and cons. We also learn how to use Pumas' `BioequivalencePower` package for study planning. In particular we take a deeper look at concepts of hypothesis tests, understanding the tradeoff of sample size and statistical power under various assumptions in the bioequivalence context. We also deepen our understanding of the statistical models explored previously.

We make heavy use of Pumas command line with hands on tasks and exercises.

Day 4: Highly variable drugs, narrow therapeutic index drugs, and dealing with special cases in data.

We advance to reference scaling techniques used both for highly variable drugs and narrow therapeutic index drugs. We study the guidances for these techniques, carry out hands on approaches using Pumas and PumasCP, and investigate some of the theoretical considerations for such studies. We also visit guidances, including product specific guidances for such cases. There are multiple cases to consider for reference scaling analysis and this day provides a thorough coverage of these cases. An additional component of the day is the use of Pumas and Julia for data manipulation including handling of missing data and transformations. Finally, we carry out a simple Monte Carlo simulation study of bioequivalence analysis to understand some of the considerations that have driven bioequivalence research.

Pumas code is provided for this simulation study and we tweak this code as an exercise.

Day 5: Additional topics and bioequivalence trends

This final day is focused on completing our understanding of the existing popular bioequivalence analysis techniques and considering new trends and further challenges. In particular we study virtual bioequivalence and model-based approaches for waiving bioequivalence studies with a demonstration using Pumas. We also deal with additional cases that can occur in practice such as multicenter trials, multiple dosages, and multiple references. We close with a discussion of the historical progression of bioequivalence analysis, in view of the methods explored during the workshop.

Also check out other Workshops in Manipal, Nonprofit events in Manipal.