生物类似药研究中相似性的主成分分析评价(英文)

Principal component analysis for assessing biosimilarity in biosimilar studies

为了评价仿制生物制品(follow-on biologics)的生物类似性,美国食品药品监督管理局(FDA)在2012年2月颁布了3个指导原则草案。在有关"科学考量"(scientific consideration)的指导原则草案中,FDA要求产品的相似性必须在安全性、纯度和有效性等各特征区中具有"证据链完备性"(totality-of-the-evidence)。然而,即使在同一特征区内,也会有各种因素对生物类似性的确证产生不同程度影响。为在生物类似药物的特征区内提供类似性的证据,本研究采用了包括主成分分析(PCA)在内的不同加权分析法进行了类似性的试验分析和比较。并在2×2交叉设计基础上进行了广泛的模拟研究,以对所提出的方法在有限样本条件下的评价性能进行研究。由于在特征区域内也存在不同变量间的相关性,权重主成分分析(PCAweight)评价的把握度最高。该方法可进一步拓展应用于生物类似药物各种特征区域间的生物类似性的"证据链完备性"评价。

For assessment of biosimilarity of follow-on biologics, the United States Food and Drug Administration （US FDA）published three draft guidances in February 2012. As indicated in the draft guidance for scientific consideration , FDA requires totality-of-the-evidence in biosimilarity across various functional areas such as safe- ty, purity and efficacy. However, even within a functional area, there are various factors that will differently influence the demonstration of biosimilarity. To provide evidence of biosimilarity within functional areas, different weighting methods including principal component analysis （ PCA ） are proposed and compared. Extensive simulation studies based on 2 × 2 crossover designs were conducted to study the finite sample performance of the proposed methods. Considering there is correlation among different variables within a given domain, PCA-weight method performs best with highest power. The proposed method can be extended to provide totality-of-the-evidence of biosimilarity across functional areas.