廣義樞紐量在生物對等性評估、遺傳率及生物相似性產品檢定方法應用之研究

Abstract

本論文的主要目的是將廣義樞紐量應用於線性化的二次動差評估標準信賴上界的建立及應用，包括: (1) 體外生物等效性(In vitro Bioequivalence)評估 (2) 遺傳率(Heritability)檢定 (3) 生物相似性(Biosimilar)產品評估 對於一些局部作用的藥品，並不會被吸收到血液中。美國FDA建議利用體外的生物等效性試驗來衡量這些產品兩種藥物產品是否提供相同的治療效果 (2003 U.S. FDA draft guidance)。然而，在美國FDA指南中所建立的生物等效性檢定方法中，生物等效性檢定標準(BE criterion)並未考慮到試驗中不同的變異來源。因此我們考慮在巢式設計之下，利用修正大樣本漸進（modified large-sample）與廣義樞紐量(generalized pivotal quantities)方法，來建立(1-α)100%的信賴上界進行檢定。 遺傳率由各種不同的育種數據組成總變異，因此包含不同的變異來源。因此，也可以利用相同的模式和方法，來進行遺傳率檢定方法的建立。 另外，由於小分子仿製藥和大分子生物製劑產品的一些本質上的區別，小分子藥物的生物等效性的評估方法並不能直接應用到評估生物相似性產品上。在本論文中，考慮在2×3附加對照組的交叉設計之下，同樣利用廣義樞紐量來建立個體生物等效性指標(Individual bioequivalence; IBE)的(1-α)100%信賴上界。 本論文亦進行了模擬數據的研究，以評估上述方法在各種變異參數組合情況下的表現，並計算實例進行比較方法間的優劣，最後進行數據結果與方法表現上的討論。

The objectives of this dissertation is to apply the GPQ approach to constructing the upper confidence limit of the linearized criteria based on second moments for evaluation of the in vitro bioequivalence, heritability testing and biosimilar drug product. Some locally acting products are not intended to be absorbed into the bloodstream. The in vitro bioequivalence studies based on machinery experiments are suggested to measure these products whether the two drug products provide the same therapeutic effect (FDA, 2003). However, the linearized criterion recommended in the draft FDA guidance does not take in consideration different sources of variation. Therefore, we apply the modified large sample (MLS) and generalized pivotal quantities (GPQ) methods to derive the (1-α)100% upper confidence limit for the linearized criterion for evaluation of in vitro bioequivalence with consideration of different sources of variation under the two-stage nested random-effect model. The heritability is defined as the ratio of additive genetic variance to phenotype variance. The hypothesis of interest is to evaluate whether the ratio is greater than some pre-specified value. As results, the same methodology for evaluation of in vitro bioequivalence can be used for assessment of heritability. Due to some fundamental differences between small molecule generics and large molecule biological products, the standard methods for bioequivalence assessment of small molecule drug products cannot be directly applied to assessing biosimilarity of biosimilar products. We proposed to assess biosimilarity by constructing a (1-α)100% upper confidence bound for the Individual bioequivalence (IBE) criterion based on the method of GPQ under a 2×3 extra-reference crossover design. Simulation studies were conducted to evaluate the performance of these proposed methods in terms of size and power under various scenarios. Numerical examples illustrate the application of in vitro bioequivalence, nested random-effect model, modified large sample method proposed methods.