Statistical Innovation in Biosimilar and Bioequivalence Studies
Chairs: Fairouz Makhlouf (FDA) and Shein-Chung Chow (Duke University)
Speaker: Wanjie Sun, PhD (FDA)
Title: Multiplicity Control in Bioequivalence/Biosimiilar Studies
Abstract: For generic drugs, a three-way crossover bioequivalence (BE) study is often used to compare two generic (T) formulations against the common brand-name (R) formulation. Adjustment for multiplicity in equivalence testing, however, is little researched. The new ICH M13A guidance mentioned multiplicity control for equivalence but did not recommend any specific methods. In this paper, we evaluate the applicability of traditional multiplicity adjustment methods in equivalence testing. We propose three revised methods (Bonferroni, Holm and Hochberg) which are applied on not only p-values but also the more commonly used confidence intervals in equivalence testing. We also apply the ‘two-at-a-time’ rule as recommended by regulatory agencies and incorporate the correlation among test statistics in simulation. All of these are advances compared to current multiplicity control methods for equivalence. Simulation shows that our proposed methods in a three-way crossover study greatly improve power and reduce needed sample size compared to conducting two two-way crossover studies, control the family-wise error rate at a desired level, and only slightly increase the required sample size compared to no alpha adjustment. Therefore, we recommend our revised Bonferroni, Holm or Hochberg method in a three-way crossover design when assessing the BE of 2 Ts to 1 R.
Speaker: Grace Zhou (Duke University)
Title: A Novel Two-Stage Adaptive Parallel Study Design for Biosimilars or PK Bioequivalence with Unblinded Sample Size Re-estimation and Optional Effect Size Re-Estimation
Abstract: In comparative clinical endpoint trials, biosimilar products need to demonstrate that there are no clinically meaningful differences between the test biosimilar and the reference biologic, typically using a parallel, two-arm study. Separately, for generic drugs with a long half-life, a parallel, two-arm study can be used in PK bioequivalence studies to demonstrate bioequivalence (BE) between a generic drug and the reference drug. While a fixed design has traditionally been used in both studies, it relies on initial assumptions made for sometimes unknown study design parameters, potentially resulting in over- or under-powered trials when those assumptions do not hold. Previously, Fuglsang (2014) proposed a two-stage adaptive design for parallel two-arm PK BE studies where the alpha adjustment was based on simulation results. Later, Maurer et al (2018) used combination methods which analytically controls the Type 1 error rate in a two-way crossover PK BE study. In this work, we apply Maurer’s methods to a two-stage adaptive parallel study design for PK BE trials or comparative clinical endpoint trials for biosimilars with unblinded sample size re-estimation and optional effect size re-estimation. Simulation results show that our proposed method controls the overall type I error rate under all simulated scenarios, at the same time, save on costs by reducing the average sample size when the study is over-powered and potentially save a study from certain failure by increasing the average sample size when the study is under-powered. By helping sponsors cut cost and improve the success rate, the proposed adaptive design can make comparative clinical endpoint studies for biosimilars and PK bioequivalence more affordable, increasing the accessibility of biosimilars and generic drugs to the public.
Speaker: Ying Yuan, PhD (University of Texas MD Anderson Cancer Center)
Title: A Bayesian Group Sequential Design for Randomized Biosimilar Clinical Trials with Adaptive Information Borrowing from Historical Data
Abstract: At the time of developing a biosimilar, the reference product has been on market for years and thus ample data are available on its efficacy and characteristics. We develop a Bayesian adaptive design for randomized biosimilar clinical trials to leverage the rich historical data on the reference product. This design takes a group sequential approach. At each interim, we employ the elastic meta-analysis predictive (EMAP) prior methodology to adaptively borrow information from the historical data of the reference product to make go/no-go decision based on Bayesian posterior probabilities. In addition, the randomization ratio between the test and reference arms are adaptively adjusted at the interim with the goal to balance the sample size of the two arms at the end of trials. Simulation study shows that the proposed Bayesian adaptive design can substantially reduce the sample size of the reference arm, while achieving comparable power as the traditional randomized clinical trials that ignore the historical data.
Discussant: Shein-Chung Chow, PhD (Duke University)
