Biomarkers in Drug Development

Biomarkers in Drug Development

Scope note: These notes cover biomarkers as drug-development and clinical-evaluation tools — the FDA-NIH BEST framework categories, the biomarker–surrogate–clinical-endpoint hierarchy, analytical and clinical validation, regulatory qualification, biomarker modalities (genomic/proteomic/metabolomic/imaging/digital), and applications (target engagement, dose-finding, enrichment/stratification, go/no-go, companion diagnostics, safety monitoring). Surrogate-endpoint trial design per se is treated only insofar as a biomarker becomes a surrogate; the mechanics of surrogate-endpoint trials are a separate topic.

1. Definition and conceptual foundations

• A biomarker (biological marker) is a defined characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or responses to an exposure or intervention, including therapeutic interventions — molecular, histologic, radiographic, or physiologic characteristics are all types of biomarkers (Atkinson 4e Ch.17, p.323, p.325).
• Critically, a biomarker is not an assessment of how an individual feels, functions, or survives — that distinction is the fault line separating a biomarker from a clinical endpoint (Atkinson 4e Ch.17, p.325).
• In clinical medicine biomarkers are used to aid diagnosis, monitor disease status, assess efficacy and safety of an intervention, and provide information on prognosis or risk; in drug development they acquire additional uses — target engagement, efficacy/proof-of-concept, and drug safety signalling (Atkinson 4e Ch.17, p.323).
• Biomarker use and the practice of medicine "go hand-in-hand" historically — from the thermometer (fever) to next-generation DNA sequencing (NGS) driving precision definition of cancers (Atkinson 4e Ch.17, p.323).
• Rational biomarker use increases the probability of success of therapeutic programs, and clinically validated surrogate endpoints enable faster availability of new therapies (Atkinson 4e Ch.17, p.323).

Historical drivers and cautionary tales

• Cardiac Arrhythmia Suppression Trial (CAST): antiarrhythmic therapy (encainide, flecainide) suppressed ventricular ectopy (the biomarker) yet increased mortality — the definitive illustration that a biomarker response can be dangerously discordant with clinical outcome, and that suppression of a biomarker is not proof of benefit (Atkinson 4e Ch.17, p.323).
• The AIDS epidemic was likely the single biggest factor accelerating biomarker-as-surrogate use: of many candidate markers, only suppression of HIV RNA proved reliable enough to underpin accelerated antiretroviral approval; combination antiretroviral therapy made traditional clinical endpoints impractical, prompting FDA guidance on plasma HIV-RNA for antiretroviral approval (Atkinson 4e Ch.17, p.323).
• 2004 FDA white paper "Innovation or Stagnation: Challenge and Opportunity on the Critical Path to New Medical Products" highlighted the need for a reliable toolkit (including biomarkers) to predict safety/effectiveness, and launched the Critical Path Initiative (CPI) — a national strategy to modernise product development and make it more predictable and less costly (Atkinson 4e Ch.17, p.324; Atkinson 4e Ch.35, pp.683–684).
• The white paper also birthed a formal Biomarker Qualification Program and a broad push toward precompetitive public–private partnerships to generate biomarker evidence (Atkinson 4e Ch.17, p.324).

Standardising the language — the BEST resource

• Consensus biomarker definitions are recent: an NIH expert working group (>20 years ago) produced preferred definitions and a conceptual framework for surrogate endpoints (Biomarkers Definitions Working Group, Clin Pharmacol Ther 2001) (Atkinson 4e Ch.17, p.324).
• The FDA-NIH Joint Leadership Council later prioritised harmonising translational terminology and developed BEST (Biomarkers, EndpointS, and other Tools) Resource — a glossary clarifying definitions and the hierarchical relationships/dependencies among terms; it is an intentionally "living" document updated periodically with stakeholder input (Atkinson 4e Ch.17, p.324).