Bioequivalence Assessment for Locally Acting Drugs: A Framework for Feasible and Efficient Evaluation

TL;DR

This paper introduces a new statistical framework for more effective and feasible bioequivalence testing of locally acting drugs, addressing complex multivariate data challenges and improving test power and size control.

Contribution

It develops an adjusted TOST procedure that accounts for multivariate data, ensuring size control and increased power, with demonstrated superior performance through simulations and case studies.

Findings

Enhanced bioequivalence testing accuracy for locally acting drugs.

Adjusted TOST improves power while maintaining size control in multivariate settings.

Method outperforms traditional approaches in simulations and real case studies.

Abstract

Equivalence testing plays a key role in several domains, such as the development of generic medical products, which are therapeutically equivalent to brand-name drugs but with reduced cost and increased accessibility. Promoting access to generics is a critical public health issue with substantial societal implications, but establishing equivalence is particularly challenging in multivariate settings. A notable example refers to locally acting drugs designed to exert their therapeutic effects at a localized area where they are administered rather than being absorbed into the bloodstream, where complex experimental protocols lead to reduced sample sizes and substantial experimental noise. Traditional approaches, such as the Two One-Sided Tests (TOST), cannot adequately tackle the complex multivariate nature of such data. In this work, we develop an adjustment for the TOST procedure by simultaneously correcting its significance level and equivalence margins to ensure control of the test size and increase its power. In large samples, this approach leads to an optimal adjustment for the univariate TOST procedure. In multivariate settings, where we show that an optimal adjustment does not exist, our proposal maintains equal marginal test sizes and overall size control while maximizing power in important cases. Through extensive simulation studies and a case study on multivariate bioequivalence assessment for two antifungal topical products, we demonstrate the superior performance of our method across various scenarios encountered in practice.