Bayesian Bioequivalence Test Based on Skew Normal Model for Bioavailability Measure

根據偏斜常態分佈的貝氏生物相等性檢定

陳玉英(Yuh-Ing Chen) ； 黃啟軒(Chi-Shen Huang)

《中國統計學報》 53卷1期 (2015/03) Pp. 38-54

摘要

學名藥和專利藥的生物相等性評估為一般生物製藥廠主要進行項目之一，此試驗通常在2x2交叉設計下進行。為進行此試驗，須從藥物代謝動力學（PK）曲線計算而得的生物可用率參數，常用的生物可用率參數包括藥物濃度－時間曲線下的面積（AUC）和最大藥物濃度（C_(max)）。目前美國食品藥物管理局（2001）採用的評估方法是假設AUC或C_(max)為常態分布的條件下進行兩個單邊檢定（TOST）。然而，實際的試驗中C_(max)的對數通常是偏斜的，因此常態分布假設的TOST可能並不實用。因此，本文提出一個隨機效應和測量誤差變量皆為偏斜常態分布的混合效應模型，稱為SNMEM，並根據SNMEM提出兩種藥物生物相等性的貝氏檢定，並在文中以模擬研究和實例說明本文所提方法之應用。結果顯示，相較於常態分布假設的混合效應模型（NMEM），本文所提的SNMEM在進行貝氏檢定有較強的檢定力，因此更有利於檢測兩種藥物的生物相等性。

關鍵字

貝氏生物相等性檢定 ； 生物可用率參數 ； 交叉設計 ； 混合效應模型 ； 偏斜常態分布

並列摘要

Assessment of the bioequivalence between the test drug and patent drug is a major issue to the biopharmaceutical industry for marketing the test drug as a generic drug. To do so, a 2x2 crossover design is often conducted for a pharmacokinetic (PK) study where the bioavailability (BA) parameters such as the area under the concentration- time curve (AUC) and maximum concentration (C_(max)) are of primary interest. Both the BA parameters are conventionally estimated from each involved subject that are known as the BA measures. The bioequivalence test in current use is two one-sided tests (TOST) recommended by the U.S. FDA (2001) that assumes a normal distribution for the logarithm of the AUC measure. However, the TOST may not be of practical use since the logarithm of the C_(max) measure is usually skewed. Therefore, we consider, in this paper, a mixed-effects model for the skewed BA measure where the random effect and measurement error variable are both distributed according to skewed normal distributions, referred to as SNMEM. Bayesian tests based on the SNMEM are then proposed for the bioequivalence of the two drugs. A real data set involving two brands of theophylline tablets for respiratory diseases is further illustrated by applying the proposed SNMEM-based Bayesian bioequivalence tests. Through the data analysis, it is found that the proposed SNMEM-based Bayesian test is more powerful than the NMEM-based TOST or Bayesian test on detecting the bioequivalence of the two drugs under study.

並列關鍵字

Bayesian bioequivalence test ； bioavailability measure ； crossover design ； mixed-effects model ； skew-normal distribution

參考文獻

Azzalini, A.(1985).A class of distributions which includes the normal ones.Scandinavian Journal of Statistics.12,171-178.

Berger, R. L.,Hsu, J. C.(1996).Bioequivalence trials, intersection-union tests and equivalence con dence sets.Statistical Science.11,283-319.

Breslow, N.(1990).Biostatistics and Bayes.Statistical Science.5,269-284.

Burnham, K.P.,Anderson, D.R.(2002).Model Selection and Multivariate Inference: A Practical Information-Theoretic Approach.New York:Springer.

Chow, S. C.,Liu, J. P.(2000).Design and Analysis of Bioavailability and Bioequivalence Studies.New York:Marcel Dekker.