What makes a good biomarker?

Abstract

Abstract:The last decade has seen an extraordinary amount of effort devoted in biomedical research to the field of biomarkers. There have been some notable successes with novel markers being adopted into clinical practice bringing clear clinical benefit to some patients -particularly with the increasing numbers of medicines being approved with companion diagnostics. However, it is fair to say that there has not yet been the numbers of clinically valuable biomarkers brought to medical practice that the research effort would seem to warrant. This paper evaluates examples of successful biomarkers, markers which might be considered partial successes and a few problematic examples and argues that more effort spent in the validation phase of marker development, and less in the discovery phase might be a more efficient way to allocate research resources.

Targeted and Untargeted Proteomics Approaches in Biomarker Development

“…Given the extensive research effort and resources that have been devoted to biomarker development in the past decades, it may be surprising how few biomarkers have ultimately achieved clinical adoption. Recent estimates suggest that only about 100 biomarkers are currently in clinical use, which amounts to only a small fraction of the potential biomarkers that have been reported in the literature . As of September 2018, the FDA biomarker qualification program listed only eight qualified biomarkers that have achieved regulatory acceptance specifically for facilitating drug development .…”

“…Further specific analytical and clinical measurement criteria must be met to obtain FDA approval of a specific diagnostic test in the form of a device or an in vitro diagnostic . Many reviews have therefore considered what makes a “good” biomarker in the hope of guiding translational research …”

“…At the stage of clinical utility, good biomarkers must serve the purpose of “improving patient outcomes.” Thus, a good biomarker must invariably be matched with an appropriate clinical problem. It must offer a “consequential” distinction and add to what is known from clinical history in terms of diagnosis, prognosis, or treatment.…”

“…A biomarker's potential for clinical utility is typically first assessed at the clinical‐qualification stage through clinical studies that are intended to link the biomarker to biological or clinical endpoints . Clinical sensitivity and specificity are assessed in samples with known outcomes, ideally using a rigorous study design and including different patient populations and laboratories …”

“…From a technical point of view, assays are expected to be implemented with low “no‐call” (ambiguous, borderline, non‐diagnostic results) and failure rates . The cost and invasiveness of testing are also important factors for clinical application.…”