论文部分内容阅读
为了检测仿制生物制品与其原研生物制品的相似性,目前主要采用基于相对差异指标的统计方法进行评价。常用的统计推断主要基于双序列(two-sequence)、双周期(two-period)的交叉设计,或者在忽略原研药批次间变异情况下的双臂平行设计。本文采用三臂平行设计(three-arm parallel design),分别测定仿制生物制品与原研生物制品的绝对差异,原研生物制品不同批次间的绝对差异,并通过该两绝对差异之比来评价仿制生物制品与原研生物制品的生物相似性概率(biosimilar probability)。这种基于生物相似性概率的评价标准,在大样本下具有渐进正态的性质,给出了检验的功效和临床试验中样本量的计算公式。本文通过模拟研究和样本量估算,证明了生物相似性概率的统计检验法具有良好的检验效能。
In order to test the similarity between a generic biological product and its original biological product, the statistical method based on the relative difference index is mainly used to evaluate the similarity. Commonly used statistical inferences are mainly based on two-sequence, two-period crossover designs, or parallel arms design, ignoring variability between batches of the original drug. In this paper, the three-arm parallel design is used to determine the absolute difference between the original biological products and the original biological products, the absolute difference between the original batches of different biological products and the ratio of the two absolute differences Biosimilar probability of the product to the original biological product. The evaluation criterion based on the biological similarity probability has the property of asymptotic normality under the large sample, and provides the test efficacy and the formula for calculating the sample size in the clinical trial. In this paper, through the simulation research and sample size estimation, we prove that the statistical test of biological similarity probability has a good test efficiency.