Biomarker

A biomarker, short for biological marker, is a measurable biological indicator used to assess a normal physiological process, a pathological process, or a pharmacological response to a therapeutic intervention. Biomarkers can be molecular (such as proteins, metabolites, or genetic variants), cellular (such as circulating tumor cells), physiological (such as blood pressure or heart rate), or imaging based (such as tumor size on MRI).

Biomarkers play a central role in modern clinical and translational research. They serve multiple functions: as diagnostic tools (identifying the presence of a disease), prognostic indicators (predicting the likely course of a disease), predictive markers (identifying which patients are likely to respond to a specific treatment), and surrogate endpoints (serving as stand ins for clinical outcomes in trials where the true endpoint takes too long to observe). The use of biomarkers can accelerate drug development by enabling earlier assessment of efficacy and safety, improve patient stratification in clinical trials, and support personalized medicine approaches that tailor treatments to individual biological profiles. Regulators can also accept biomarkers as trial endpoints: in the US, for example, a validated surrogate endpoint can support traditional approval, while accelerated approval can rest on a surrogate endpoint that is reasonably likely to predict clinical benefit, with confirmatory trials required (since the Food and Drug Omnibus Reform Act of 2022, the FDA can require that they be underway when it grants accelerated approval). This makes biomarker research a high priority area.

Biomarker development follows a defined pipeline that begins with discovery and moves through analytical validation, clinical validation, and clinical utility assessment. In the discovery phase, researchers identify candidate biomarkers through genomic, proteomic, metabolomic, or imaging studies. High throughput technologies such as genome wide association studies, mass spectrometry, and next generation sequencing have greatly expanded the number of candidate biomarkers.

Analytical validation confirms that the biomarker can be measured accurately and reproducibly using a specified assay or measurement technique. This involves establishing the assay's sensitivity, specificity, precision, linearity, and stability. Clinical validation demonstrates that the biomarker reliably reflects the biological process or clinical condition it is intended to measure in relevant patient populations. This typically requires testing in independent cohorts and across diverse populations.

Clinical utility assessment, the final and most demanding step, demonstrates that use of the biomarker leads to improved health outcomes compared to standard care without the biomarker. This may involve randomized trials comparing biomarker guided treatment strategies with standard approaches.

Biomarkers are classified by their intended use. The FDA-NIH BEST glossary lists seven categories: susceptibility or risk, diagnostic, monitoring, prognostic, predictive, pharmacodynamic or response, and safety biomarkers; a surrogate endpoint is usually a biomarker used as a trial endpoint. Four common uses are described here. Diagnostic biomarkers detect the presence of a disease (such as troponin for myocardial infarction, or HbA1c for diabetes). Prognostic biomarkers predict the course of a disease regardless of treatment (such as tumor grade in cancer, or the Gleason score in prostate cancer). Predictive biomarkers identify patients likely to respond to a specific treatment (such as HER2 status predicting response to trastuzumab, or KRAS mutations predicting lack of response to anti EGFR antibodies in colorectal cancer). Surrogate endpoints are usually biomarkers that substitute for clinical outcomes in clinical trials (such as tumor shrinkage as a candidate surrogate for survival, or LDL cholesterol as a surrogate for cardiovascular events).

Biomarkers are relevant whenever research involves the measurement of biological processes, disease diagnosis or prognosis, treatment response prediction, or clinical trial design. They are particularly important in drug development, where validated biomarkers can serve as endpoints for early phase trials, enabling faster and smaller studies. They are also central to precision medicine initiatives that aim to match patients with the most appropriate treatments based on their biological profiles. Biomarkers should not be treated as substitutes for clinical outcomes unless the evidence shows that changes in the biomarker reliably predict changes in those outcomes.

A common error is using a biomarker as a surrogate endpoint in a clinical trial without adequate validation that changes in the biomarker reliably predict changes in the clinical outcome of interest. Many promising biomarkers have failed at this step, leading to incorrect conclusions about treatment efficacy. Another mistake is assuming that a statistically significant association between a biomarker and a clinical outcome in a discovery cohort will replicate in independent populations; many biomarker associations fail to replicate due to overfitting, population differences, or analytical variability.

Researchers also sometimes conflate diagnostic and predictive biomarkers. A biomarker that is diagnostic for a disease (identifying who has it) may not be predictive of treatment response (identifying who will benefit from a specific therapy), and vice versa. Finally, neglecting to account for pre analytical variables (how the sample is collected, processed, and stored) can introduce measurement error that undermines the biomarker's performance.

Researchers developing a blood test for early detection of pancreatic cancer identify a panel of four protein biomarkers that, when combined, achieve a sensitivity of 89% and a specificity of 92% in a discovery cohort of 300 patients. They then validate the panel in an independent cohort of 500 patients from a different institution, where it achieves a sensitivity of 82% and a specificity of 88%. A clinical utility study is planned to determine whether screening with the biomarker panel in high risk individuals (those with hereditary pancreatitis or familial pancreatic cancer) leads to earlier detection and improved survival compared to standard surveillance with imaging alone.