Comparison of Innovative and Conventional Methods in Biosimilar Bridging Studies with Multiple References

Area of Science:

• Pharmaceutical Sciences
• Biostatistics
• Drug Development

Background:

• Biosimilar drug product assessment often involves bridging studies when multiple reference products exist.
• Conventional methods use 3-way pairwise comparisons for biosimilar bridging studies.

Purpose of the Study:

• To compare two innovative statistical methods with the conventional pairwise comparison approach for biosimilar bridging studies.
• To evaluate the efficiency and statistical power of new methods in biosimilarity assessment.

Main Methods:

• For parallel designs, the simultaneous confidence interval (CI) method was compared to the conventional approach.
• For crossover designs, multiplicity-adjusted Schuirmann's two one-sided tests (MATOST) was evaluated.
• Statistical power and type I error rates were key metrics for comparison.

Main Results:

• The simultaneous CI method demonstrated comparable statistical power to the conventional method in biosimilarity assessment.
• The MATOST method, using Holm and Bonferroni approaches, was found unfavorable due to large sample size requirements.
• While controlling type I error, MATOST's sample size inefficiency makes it less desirable.

Conclusions:

• The simultaneous CI method is a viable alternative for biosimilar bridging studies, offering similar statistical power.
• Conservative approaches within MATOST are not recommended for biosimilar bridging studies due to sample size demands.
• Optimized statistical methods are crucial for efficient and reliable biosimilar drug product assessment.