Evolution of Biopharmaceutical Statistics
Mark Rothmann* (FDA), Yun Wang* (FDA), James Travis* (FDA)
*Office of Biostatistics, Office of Translational Science, Center for the Drug Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, Maryland, USA
Disclaimer: This article reflects the views of the authors and should not be construed to represent FDA’s views or policies.
Highlights
· Discover how statisticians are now integral to every stage of clinical trials, ensuring precise and reliable results.
· Learn about the revolutionary Bayesian methods that are enhancing the accuracy and reliability of clinical trial outcomes.
· Explore the collaborative programs led by the FDA that are fostering statistical innovation and paving the way for novel trial designs.
Statistical thinking has evolved into a proactive approach emphasizing pre-specification of design and analysis with detailed statistical analysis plans to provide clear interpretation of the results. The involvement of statisticians in clinical trials has evolved from being consultants providing sample size calculations and being consulted after data collection for data analysis to a collaboration where statisticians are involved during the whole process of a clinical trial, ranging from defining the research questions and endpoints, formulating the statistical hypotheses, justifying the study design, to monitoring the cumulative data for potential early stop due to safety, futility or efficacy or other adaptations, analyzing the data, interpreting the results, assessing the credibility of the findings, and ensuring valid conclusions about treatment efficacy and safety. A statistician’s role is crucial when non-traditional methods are involved, for example in designing platform trials or studies involving adaptive features or borrowing from external data, in analyzing data with complex statistical models or synthesizing evidence from multiple studies.
Historically, clinical trials focused on enrolling a homogeneous patient population to minimize the variability in outcomes. Now, clinical trials are encouraged to have participants in the trial consistent with the entire population who may get the drug should the drug be approved rather than a narrow segment of the population, leading to a more heterogeneous study population [1]. Considering and evaluating heterogeneous treatment effects is necessary in such clinical trials. Bayesian hierarchical models have been used that consider all data, not just subgroup-specific data, leading to increased precision in estimating treatment effects across subgroups [2,3].
Bayesian methods have also been applied in dose selection, the determination of non-inferiority margins, in pediatric extrapolation, adaptive clinical trials and in rare diseases. Regulatory agencies have become more open to non-traditional methods in medical product development [4].
Statisticians are key to leading and implementing clinical trial innovations, as seen in programs such as the Complex Innovative Trial Design Meeting program [5]. This program promotes statistical innovation by allowing the Food and Drug Administration (FDA) to publicly share and discuss the novel trial designs accepted by the program, including trial designs for medical products that have not yet been approved by the FDA. Statisticians also lead in a new, redesigned Bayesian Statistical Analysis demonstration program [6], which aims to increase experience in Bayesian statistical methods across sponsors and FDA clinical and statistical reviewers, including deepening an understanding of their applicability, opportunities, and challenges.
In addition to our regular interaction with the sponsors via different types of regulatory meetings, FDA statisticians participated in user fee negotiations and are involved in coauthoring numerous guidances. Statisticians took important or leading roles in the writing of several ICH guidances including ICH E9(R1) Statistical Principles for Clinical Trials: Addendum: Estimands and Sensitivity Analysis in Clinical Trials, ICH E11A Pediatric Extrapolation, ICH E17 General Principles for Planning and Design of Multi-Regional Clinical Trials and ICH E20 Adaptive Clinical Trials (draft). Within the FDA, some examples of statisticians-led guidances are Adaptive Designs for Clinical Trials of Drugs and Biologics, Adjusting for Covariates in Randomized Clinical Trials for Drugs and Biological Products, Interacting with the FDA on Complex Innovative Trial Designs for Drugs and Biological Products, Multiple Endpoints in Clinical Trials, Non-Inferiority Clinical Trials to Establish Effectiveness, and Statistical Approaches to Establishing Bioequivalence.
To promote collaborations among the FDA, industry, and academia, the Center for Drug Evaluation and Research (CDER) is engaging in a number of scientific public and private partnerships and consortia [7]. For example, FDA statisticians actively contributed to the American Statistical Association (ASA) Biopharmaceutical (BIOP) Safety Working Group, Heart Failure Collaboratory, Type I Diabetes Consortium. FDA has co-sponsored symposiums/workshops to address challenges and opportunities in assessing and communicating heterogenous treatment effects [8,9], advance the development of pediatric therapeutics [10], and explore novel endpoints for rare disease drug development [11]. FDA also launched programs to incorporate the patient’s voice in drug development and evaluation [12] and enable the integration of real-world data/evidence in regulatory decision-making [13]. Regulatory and industry statisticians have worked together to organize multiple annual statistical meetings. They also work together in many scientific working groups, including on Alzheimer's disease, oncology, metabolic disorders, Bayesian methods, cell & gene therapy, pediatric drug development, and many more.
To succeed today in biopharmaceutical statistics, one should have a strong foundation in statistical methodology, knowledge in clinical trials, capability to analyze and interpret data, along with excellent collaboration, communication and problem-solving skills. In the era of artificial intelligence (AI), a willingness to learn the latest scientific advancement and embrace different approaches is fundamental for adapting to a continually evolving biopharmaceutical industry and regulatory landscape. For example, at one time SAS was the primary statistical package used for statistical analysis in regulatory submissions and reviews. Now, it is increasingly common to see submissions using more diverse approaches and software, particularly when using complex methods.
References
1. FDA draft guidance: Diversity Action Plans to Improve Enrollment of Participants from Underrepresented Populations in Clinical Studies 2024.
2. FDA Impact Story (2019). Using innovative statistical approaches to provide the most reliable treatment outcomes information to patients and clinicians. Available at https://www.fda.gov/drugs/regulatory-science-action/impact-story-using-innovative-statistical-approaches-provide-most-reliable-treatment-outcomes.
3. Wang, Y., Tu, W., Koh, W., Travis, J., Abugov, R., Hamilton, K., Zheng, M., Crackel, R., Bonangelino, P., Rothmann, M. Bayesian Hierarchical Models for Subgroup Analysis. Pharmaceutical Statistics. 2024; 23:1065–1083. https://doi.org/10.1002/pst.2424.
4. Ionan, A.C., Clark, J., Travis, J., Amatya, A., Scott, J., Smith, J.P., Chattopadhyay, S., Salerno, M.J., Rothmann, M. Bayesian Methods in Human Drug and Biological Products Development in CDER and CBER. Ther Innov Regul Sci 2023; 57 436-444. https://doi.org/10.1007/s43441-022-00483-0.
5. Complex Innovative Trial Design Meeting Program, https://www.fda.gov/drugs/development-resources/complex-innovative-trial-design-meeting-program.
6. CDER Center for Clinical Trial Innovation (C3TI), https://www.fda.gov/about-fda/center-drug-evaluation-and-research-cder/cder-center-clinical-trial-innovation-c3ti.
7. CDER Scientific Public Private Partnerships and Consortia, https://www.fda.gov/drugs/science-and-research-drugs/scientific-public-private-partnerships-and-consortia.
8. Assessing and Communicating Heterogeneity of Treatment Effects for Patient Subpopulations: Challenges and Opportunities, https://www.jhsph.edu/research/centers-and-institutes/center-of-excellence-in-regulatory-science-and-innovation/news-and-events/Critical-Issues-in-Heterogeneity-of-Treatment-Effect.html.
9. Heterogeneity of Treatment Effects in Clinical Trials: Methods and Innovations, https://mrctcenter.org/news-events/heterogeneity-of-treatment-effects-in-clinical-trials-methods-and-innovations/#1602863324215-1289c9d5-a82a.
10. Pediatric Science and Research Activities, https://www.fda.gov/science-research/pediatrics/pediatric-science-and-research-activities.
11. Rare Disease Endpoint Advancement Pilot Program Workshop: Novel Endpoints for Rare Disease Drug Development, https://healthpolicy.duke.edu/events/rare-disease-endpoint-advancement-pilot-program-workshop-novel-endpoints-rare-disease-drug.
12. CDER Patient focused drug development, https://www.fda.gov/drugs/development-approval-process-drugs/cder-patient-focused-drug-development.
13. CDER and CBER Real World Evidence Program, https://www.fda.gov/science-research/real-world-evidence/center-biologics-evaluation-and-research-center-drug-evaluation-and-research-real-world-evidence.