Value-based assessment of pharmacodiagnostic testing from early stage development to real-world use

Abstract

Disease etiology may be regarded as a consequence of both genotypic and biochemical phenomena, which impact individual patients in different ways. Disease prognosis, beneficial treatment response, and susceptibility to adverse drug effects are often intimately tied to individual biology. Clinical and genetic biomarkers applied individually or in concert are increasingly used to stratify patient populations in terms of prognosis, therapeutic benefit, or safety. As a result, clinical trialists are challenged to design studies that reflect these determinants of outcome, to optimize the patient's eventual clinical course both in the trial and in actual practice. These designs are informed both by preclinical studies and by real-world research that can establish proof of concept for a novel biomarker and provide a basic understanding of the relationship between biomarker and clinical outcome. As clinical and real-world studies unfold, a deeper understanding of the nature of the biomarker and its potential uses in drug development is gained. Specifically, one can eventually define the biomarker as prognostic (i.e., predicts disease progression), predictive (predicts treatment response or adverse outcome(s)), or exhibiting both prognostic and predictive properties. One must further validate the performance of these emerging biomarkers, again in both the trial and real-world environments. The eventual adoption of the biomarker as a useful pharmacodiagnostic test is premised upon this early translational research. In this article, the development and validation of predictive and prognostic biomarkers is discussed by using selected examples that highlight factors contributing to the valuation of biomarkers and their application to personalized medicine in the real world.

Keywords: biomarkers; drug industry; economics; personalized medicine; pharmacodiagnostics.