Type I Error Inflation of Blinded Sample Size Re-Estimation in Equivalence Testing

Biosimilar products are a relative newcomer in the pharmaceutical industry. It was only in 2006 that the first biosimilar product, Omnitrope[textregistered] (somatropin), a human growth hormone produced by Sandoz was approved by the European Medicine Agency, followed later the same year by the US Food & Drug Administration, and in 2009 by Pharmaceuticals and Medical Devices Agency in Japan. The clinical component of a biosimilar development borrows extensively from traditional drug development. However, when it comes to clinical study designs, this may not always be applicable. This article investigates Type I error violations that occur when blinded sample size reviews are applied in equivalence testing as used in biosimilar drug development. We give a derivation which explains why such violations are more pronounced in equivalence testing than in the case of superiority testing. In addition, the amount of Type I error inflation is quantified by simulation as well as by some theoretical considerations. Nonnegligible Type I error violations arise when blinded interim reassessments of sample sizes are performed particularly if sample sizes are small, but within the range of what is practically relevant.生物类似药（biosimilar）是制药行业中的相对新兴品类。直至2006年，首款生物类似药产品——山德士（Sandoz）生产的人生长激素类药物奥曲肽®（Omnitrope，somatropin）——获得欧洲药品管理局批准，同年晚些时候获美国食品药品监督管理局批准，2009年获日本药品医疗器械局批准。生物类似药研发的临床环节大量借鉴了传统药物研发模式，但在临床研究设计层面，该思路并非总能适用。本文针对生物类似药研发中采用的等效性试验，探讨了实施盲法样本量复核时出现的Ⅰ类错误（Type I error）偏离问题。我们通过推导过程阐释了为何此类错误偏离在等效性试验中比在优效性试验中更为显著。此外，我们通过模拟实验与部分理论分析，对Ⅰ类错误的膨胀幅度进行了量化。当开展盲法样本量中期重新评估时，尤其是在样本量较小但处于实际应用相关范围内的情况下，会出现不可忽视的Ⅰ类错误偏离问题。