Diagnostic Uncertainty in Drug Development

Abstract

The traditional drug development framework relies on target selection and disease classification criteria that are not designed to capture the entire universe of patients that could benefit from a novel therapy but the majority of patients with shared features of a disease. This framework fits the current model of clinical development and regulatory approval, but it must evolve when studying complex diseases which have multifactorial etiology and are associated with high levels of heterogeneity and diagnostic and scientific uncertainty. Our ability to treat complex diseases can only improve if we understand the multiple aspects of the endotype (i.e., underlying molecular mechanisms and clinical characteristics) and evaluate their influence on responses to treatment in discrete subgroups of patients. We must look toward a new era of clinical trial types in which modern data-driven approaches fueled by comprehensive sets of real-world data and evolving trends in patient selection and stratification allow us to increase the representativeness of real-world populations. We must meet the needs of more patients than those who fulfill strict diagnostic, classification, and response criteria that do not always reflect everyday clinical practice.