PS34: Recent Advances in Statistical Methodology for Biosimilar and Bioequivalence Studies

In the evolving landscape of healthcare, generics and biosimilars play a pivotal role in enhancing accessibility and reducing costs. Representing 90% of U.S. prescriptions but less than 18% of drug expenses, generic and biosimilar drugs saved the U.S. over $400 billion in 2023 alone, according to the Association for Accessible Medicines. However, the rapid expansion of biosimilar and generic drug markets has not been matched by equivalent advancements in statistical methodologies for bioequivalence and biosimilar studies. This session focuses on recent innovations in statistical approaches, showcasing cutting-edge techniques developed through collaboration among regulatory agencies, industry, and academia. The first speaker is from industry, discussing the utilization of real-world evidence (RWE) to improve the design and execution of biosimilar trials. A novel matching method is presented to identify comparable patient groups from real-world data, offering a cost-effective way to enhance the estimation of equivalence margins when historical data is limited. This approach integrates robust statistical frameworks with advanced data extraction techniques to repurpose existing clinical and observational data effectively. Through a detailed case study, the presentation demonstrates how data-driven methodologies not only improve trial efficiency but also address patient heterogeneity and increase confidence in biosimilarity assessments, paving the way for broader adoption of biosimilars. The second speaker is from academia, introducing an innovative two-stage adaptive trial design for biosimilar development. This hybrid design integrates pharmacokinetic (PK) and clinical studies under the fundamental assumption that PK similarity predicts clinical similarity. The first stage focuses on PK endpoints, and interim decisions guide whether to proceed with clinical assessments in the second stage. This flexibility allows trials to adapt dynamically to interim results, avoiding duplicative efforts, reducing resource utilization, and increasing the likelihood of successful biosimilar approval. Additionally, a novel biosimilarity index derived from this framework is introduced, providing a unified statistical measure that aligns with regulatory guidelines while maintaining scientific rigor. The adaptability of this approach addresses challenges unique to biologics, including complex manufacturing processes and variations in patient response. The final presentation, delivered by an FDA statistician, delves into the innovative application of Bayesian dynamic borrowing methods in bioequivalence studies. These advanced methods leverage prior data, such as results from fasting studies, to inform subsequent fed pharmacokinetic studies for high-risk generic products. By employing dynamic priors like power priors and mixture priors, this methodology can potentially reduce sample size requirements while preserving the validity of study conclusions. Additionally, simulation will be conducted to compare operating characteristics of Bayesian dynamic borrowing approaches, traditional frequentist methods, as well as adaptive designs. Recommendations will be provided to select appropriate statistical approaches under varying scenarios. This research will help to alleviate the burden on generic drug applicants and accelerate the availability of cost-effective therapies for patients. Lastly, a brief forward-looking discussion on the integration of artificial intelligence (AI) and machine learning (ML) into biosimilar and bioequivalence research will be provided by an FDA statistical leader. Collectively, these presentations exemplify how statistical innovation and collaboration between agencies, industry, and academia can drive progress in the development of generic and biosimilar drugs. By addressing key methodological challenges, fostering interdisciplinary partnerships, and embracing technological advancements, this session underscores the critical role of statistics in promoting safe, effective, and affordable therapies for the future.