Abstract
The two-arm parallel design that is commonly used to assess biosimilarity has the drawback that it does not take into account the inherent variability within the reference products. The three-arm parallel design was proposed to solve this problem. The purpose of this article was to extend the previous results to the time-to-event endpoints in the exponential model, the Cox proportional hazard model, and the restricted mean survival time. The relative distance is defined, and the corresponding test procedures are developed based on asymptotic theory. The Type I error rates and powers are investigated theoretically and empirically. Supplementary materials for this article are available online.
Supplementary Materials
The methodology proposed in this article is implemented in the R package Biosimilar3ArmSurv. If the models and clinical trial design described in this article are used to evaluate biosimilarity, this Biosimilar3ArmSurv package can be used to calculate the sample size required for each arm. Also, the empirical Type I error rates and the empirical powers can be investigated through simulation under various assumptions. This Biosimilar3ArmSurv package is available for free download on the journal website.