高变异药生物等效性研究中的样本量估计

Sample sizes of reference-scaled average bioequivalence for highly variable drugs: implementation of FDA approach

参比制剂校正的平均生物等效性(reference-scaled average bioequivalence,RSABE)是高变异药物生物等效性评价的重要方法。RSABE的等效性界值与参比制剂的个体内变异相关,其等效性界值是一个随机变量,因此无法通过双单侧检验的效能函数进行样本量估计。高变异药物RSABE的样本量估计目前尚无公认方法,本文结合实例介绍了平均生物等效性(ABE)和高变异药RSABE研究设计中对样本量的考虑和样本量的估计过程,为申办方和研究者进行生物等效性研究中提供参考。

Reference-scaled average bioequivalence(RSABE） is the preferred alternative approach of average bioequivalence(ABE） for highly variable drugs. As the bioequivalence limit of RSABE is scaled by intra-subject coefficient of variance(CV） which is a random variable,the limit is not a constant,so that the sample size calculations based on the power function of Schuirmann＇s two one-sided tests procedure is not applicable for RSABE approach recommended by FDA. Sample size was obtained by simulations in most current trials for highly variable drugs. Here the procedure of sample size estimation for ABE and RSABE was presented through examples to provide a reference to bioequivalence trials.