Biostatistical Considerations in Development of Biomarker-Based Tests to Guide Treatment Decisions

Biomarkers require several levels of assessment, including assay method validation and assay qualification, before they can be used to guide treatment decisions for patients. They must also be evaluated in clinical trials to link them with biological processes and clinical endpoints to show that they are fit for purpose. This article briefly summarizes techniques that can be used for assay method validation and qualification. It also describes several clinical trial designs that have been described in the literature for evaluating biomarkers for guiding therapy. The concept of a prognostic factor is introduced as a consideration that must be taken into account when selecting appropriate clinical trial designs to distinguish biomarkers that predict for favorable response to a new therapy from biomarkers that predict for favorable response independently of therapy. The main focus is on Phase III trials, including both adaptive and nonadaptive designs.