The world is awash in data. This volume of data will continue to increase. In the pharmaceutical industry, much of this data explosion has happened around biomarker data. Great statisticians are needed to derive understanding from these data. This book will guide you as you begin the journey into communicating, understanding and synthesizing biomarker data. -From the Foreword, Jared Christensen, Vice President, Biostatistics Early Clinical Development, Pfizer, Inc.
Biomarker Analysis in Clinical Trials with R offers practical guidance to statisticians in the pharmaceutical industry on how to incorporate biomarker data analysis in clinical trial studies. The book discusses the appropriate statistical methods for evaluating pharmacodynamic, predictive and surrogate biomarkers for delivering increased value in the drug development process. The topic of combining multiple biomarkers to predict drug response using machine learning is covered. Featuring copious reproducible code and examples in R, the book helps students, researchers and biostatisticians get started in tackling the hard problems of designing and analyzing trials with biomarkers.
- Analysis of pharmacodynamic biomarkers for lending evidence target modulation.
- Design and analysis of trials with a predictive biomarker.
- Framework for analyzing surrogate biomarkers.
- Methods for combining multiple biomarkers to predict treatment response.
- Offers a biomarker statistical analysis plan.
- R code, data and models are given for each part: including regression models for survival and longitudinal data, as well as statistical learning models, such as graphical models and penalized regression models.
Table of Contents
Section I Pharmacodynamic Biomarkers
2. Toxicology Studies
3. Bioequivalence Studies
4. Cross-Sectional Profile of Pharmacodynamics Biomarkers
5. Timecourse Profile of Pharmacodynamics Biomarkers
6. Evaluating Multiple Biomarkers
Section II Predictive Biomarkers
8. Operational Characteristics of Proof-of-Concept Trials
with Biomarker-Positive and -Negative Subgroups
9. A Framework for Testing Biomarker Subgroups in
10. Cutoff Determination of Continuous Predictive
Biomarker for a Biomarker–Treatment Interaction
11. Cutoff Determination of Continuous Predictive Biomarker
Using Group Sequential Methodology
12. Adaptive Threshold Design
13. Adaptive Seamless Design (ASD)
Section III Surrogate Endpoints
15. Requirement # 1: Trial Level – Correlation Between
Hazard Ratios in Progression-Free Survival and Overall
Survival Across Trials
16. Requirement # 2: Individual Level – Assess the Correlation
Between the Surrogate and True Endpoints After Adjusting
for Treatment (R2
17. Examining the Proportion of Treatment Effect in AIDS Clinical
18. Concluding Remarks
Section IV Combining Multiple Biomarkers
20. Regression-Based Models
21. Tree-Based Models
22. Cluster Analysis
23. Graphical Models
Section V Biomarker Statistical Analysis Plan
Nusrat Rabbee is a biostatistician and data scientist at Rabbee & Associates, where she creates innovative solutions to help companies accelerate drug and diagnostic development for patients. Her research interest lies in the intersection of data science and personalized medicine. She has extensive experience in bioinformatics, clinical statistics and high-dimensional data analyses. She has co-discovered the RLMM algorithm for genotyping Affymetrix SNP chips and co-invented a high-dimensional molecular signature for cancer. She has spent over 17 years in the pharmaceutical and diagnostics industry focusing on biomarker development. She has taught statistics at UC Berkeley for 4 years.