A single vial of blood can now carry information that once required weeks of clinical observation to gather. A cholesterol panel, a hemoglobin A1c result, a tumor marker, or a genetic variant report are all biomarkers, and together they form one of the quietest but most consequential shifts in how modern medicine operates.
Physicians no longer wait solely for symptoms to declare themselves. They increasingly measure biological signals that reveal disease risk, confirm a diagnosis, or show whether a treatment is actually reaching its target, sometimes years before a patient would otherwise notice anything wrong.
The scale of this shift is easy to underestimate. Biomarker data now supports a majority of new drug approvals in the United States and Europe, with one analysis finding that well over half of all FDA and EMA approvals between 2015 and 2019 relied on biomarker evidence at some stage of development. That number reflects a broader transformation: biomarkers have moved from a research curiosity into a load bearing structure of clinical decision making, drug development, and regulatory review.
What Exactly Counts as a Biomarker?
In 2016, the FDA and the National Institutes of Health jointly published the Biomarkers, EndpointS, and other Tools resource, known as BEST, specifically to resolve years of inconsistent terminology across research and clinical practice. BEST defines a biomarker as a characteristic that is objectively measured as an indicator of normal biological processes, disease processes, or a biological response to an exposure or intervention, including a therapeutic one.
That definition draws an important line. A biomarker is a measurable, objective signal, not a subjective report of how a patient feels, functions, or survives. Blood pressure, a genetic variant, a tumor size measured by imaging, and a glucose level all qualify. A patient’s self reported pain score or quality of life questionnaire does not, even though both are valuable pieces of clinical information collected alongside biomarkers.
Biomarkers span a wide range of biological data types: molecular markers found in blood or tissue, histologic findings from biopsy samples, radiographic measurements from imaging studies, and physiologic readings like heart rate or blood pressure. What unites them is objectivity and reproducibility, not the specific method used to capture them.
The Biomarker Map: From Routine Blood Tests to Molecular Signals
Most people encounter biomarkers constantly without using the term. The table below maps the major categories clinicians and researchers work with today.
| Biomarker type | What it measures | Common example |
|---|---|---|
| Blood biomarkers | Circulating molecules, cells, or metabolic products | Hemoglobin A1c, cholesterol panel, complete blood count |
| Genetic biomarkers | Inherited or acquired DNA variants | BRCA1/BRCA2 mutations, HER2 gene amplification |
| Protein biomarkers | Circulating or tissue based proteins | Prostate specific antigen, troponin |
| Metabolic biomarkers | Small molecule byproducts of metabolism | Lactate, creatinine, glucose |
| Imaging biomarkers | Structural or functional measurements from scans | Tumor volume on MRI, bone density on DEXA |
| Physiologic biomarkers | Measurable body functions | Blood pressure, oxygen saturation, heart rate variability |
| Tissue and cellular biomarkers | Findings from biopsy or histology | Cell grade in a tumor biopsy, immune cell markers |
| Digital biomarkers | Data captured through sensors and wearables | Step count trends, heart rate patterns, sleep architecture |
The Seven Official Biomarker Categories, and Why the Distinction Matters
The BEST glossary formally defines seven categories of biomarkers, and the differences between them are not academic. Mixing them up can lead to a biomarker being used for a purpose it was never validated to serve.
Susceptibility or risk biomarker indicates the potential to develop a disease in someone who does not currently show signs of it, such as certain BRCA gene variants that raise breast and ovarian cancer risk without confirming that cancer is present.
Diagnostic biomarker detects or confirms an existing disease, the way sweat chloride concentration confirms a cystic fibrosis diagnosis, or glomerular filtration rate identifies chronic kidney disease.
Monitoring biomarker is measured repeatedly to track disease status or treatment response over time, such as HIV RNA levels used to guide antiretroviral therapy.
Prognostic biomarker estimates the likely course of a disease regardless of treatment.
Predictive biomarker identifies patients more likely to respond favorably or unfavorably to a specific intervention, with HER2 expression in breast cancer tissue serving as a textbook example since it determines eligibility for trastuzumab therapy.
Pharmacodynamic or response biomarker shows that a drug is biologically engaging its intended target.
Safety biomarker flags the likelihood, presence, or extent of a toxic effect from an exposure or treatment.
Diagnostic, Prognostic and Predictive Biomarkers Are Not the Same
The distinction between diagnostic, prognostic, and predictive biomarkers deserves special attention because clinicians and patients routinely confuse them, and the consequences of that confusion can affect treatment decisions.
Consider a hypothetical breast cancer scenario. A biopsy confirming malignant cells is diagnostic; it answers whether cancer is present. A separate marker indicating how aggressively that tumor is likely to grow, independent of what treatment is chosen, is prognostic; it answers how the disease is likely to behave. HER2 status, by contrast, is predictive; it answers which specific drug is likely to work, since HER2 positive tumors respond to HER2 targeted therapies like trastuzumab in a way HER2 negative tumors do not.
A biomarker can theoretically serve more than one of these roles, but assuming a diagnostic marker automatically predicts treatment response, or that a prognostic marker automatically guides drug selection, is a common and consequential error. A biomarker associated with a disease is not automatically a clinically actionable test; that leap requires its own validation.
How Biomarkers Support Precision Medicine
Biomarkers form the operational backbone of precision medicine, the practice of matching treatments to the biological profile of individual patients rather than treating a diagnosis as a single uniform category. Companion diagnostics, tests developed alongside a specific drug to identify which patients are likely to benefit, are one of the clearest examples of this connection.
Oncology has become the leading application area, where genetic and protein biomarkers routinely determine which targeted therapy a patient receives, but the model has expanded into cardiology, rare disease diagnosis, and increasingly psychiatry and neurology.
Biomarker discovery in this context typically starts with an omics scale screen, genomic, proteomic, metabolomic, or transcriptomic data drawn from patient cohorts, searching for signals associated with disease state or treatment outcome. What emerges from that screen is only a candidate.
Validating a biomarker across different populations, disease stages, and clinical settings is a separate, often multi year process, and skipping that step is one of the most common reasons promising biomarkers fail to translate into reliable clinical tools.
When a Biomarker Can Mislead
No biomarker is perfectly accurate, and readers who encounter biomarker based headlines benefit from understanding a few statistical realities before drawing conclusions.
Every diagnostic test has a sensitivity, the proportion of people with the disease who test positive, and a specificity, the proportion of people without the disease who test negative. A highly sensitive but poorly specific biomarker generates many false positives, unnecessarily alarming patients and triggering further testing.
A highly specific but poorly sensitive one misses real cases. Predictive value, how likely a positive or negative result is to be correct, depends heavily on how common the condition is in the population being tested, which is why a test that performs well in a high risk population can perform poorly when applied broadly to a low risk one.
Surrogate endpoints raise a related concern. Researchers sometimes use a biomarker, such as tumor shrinkage or a reduction in a lab value, as a stand in for a harder to measure clinical outcome like survival or quality of life. A treatment can improve a surrogate biomarker without meaningfully improving how long or how well a patient lives, which is why regulators increasingly require confirmatory evidence connecting a surrogate marker to a real clinical benefit before relying on it alone.
Confounding variables, population differences, biomarker variability between labs, and simply using an unvalidated biomarker outside the population it was studied in round out the list of ways biomarker data can mislead even careful readers.
From Discovery to Clinical Use
A biomarker’s path from laboratory finding to routine clinical tool generally follows seven stages: a biological observation suggesting a signal worth investigating, identification of a candidate biomarker through omics or targeted research, analytical validation confirming the test measures what it claims to measure accurately and reproducibly, clinical validation confirming the biomarker correlates meaningfully with the relevant disease state or outcome, regulatory review where applicable, clinical implementation into routine practice, and ongoing post market monitoring to confirm real world performance matches what trials suggested. Many candidate biomarkers identified in early research never complete this pipeline, and that attrition is a feature of a rigorous validation process, not a failure of it.
Where Biomarker Research Is Heading
Several converging trends are reshaping how biomarkers are discovered and used. Multiomics approaches combine genomic, proteomic, and metabolomic data rather than analyzing a single layer in isolation, capturing interactions a single omics layer would miss. Liquid biopsy technology detects circulating tumor DNA and other markers from a simple blood draw, offering a less invasive alternative to tissue biopsy for certain cancers.
Digital biomarkers captured through wearable devices, from heart rate variability to gait patterns, are being studied as early indicators for conditions ranging from atrial fibrillation to neurodegenerative disease. Artificial intelligence is accelerating the identification of candidate biomarkers from large, complex datasets, though AI generated candidates still require the same validation rigor as any other biomarker before clinical use.
How Patients Should Interpret Biomarker Results
A single biomarker result rarely tells a complete clinical story on its own. Reference ranges printed on a lab report describe statistical norms across a population, not a personalized risk threshold calibrated to an individual’s age, sex, medical history, or other risk factors. A value slightly outside the reference range is not automatically alarming, and a value inside it does not guarantee the absence of disease.
Clinicians interpret biomarker results alongside symptoms, medical history, and often repeat testing over time, and patients benefit from asking what a specific result means in that broader context rather than searching for a single number to define their health status.
The growing sophistication of biomarker science has genuinely expanded what medicine can detect, predict, and treat, from confirming rare genetic conditions in infancy to selecting cancer therapies based on a tumor’s molecular fingerprint. That expansion, however, has not eliminated the fundamental discipline biomarkers require: a signal is only as useful as the validation behind it, and the difference between an interesting biological association and a clinically trustworthy test remains one of the most important distinctions in modern medicine.
As multiomics, liquid biopsy, and AI assisted discovery continue to generate new candidate biomarkers at an accelerating pace, that discipline, not the volume of new signals identified, will determine how much of this progress actually reaches patients in a form they can rely on.
FAQ
Q: What is a biomarker in simple terms?
A: A biomarker is a measurable biological characteristic, such as a blood test result, genetic variant, or imaging finding, used as an objective indicator of a normal process, a disease, or a response to treatment.
Q: What are the main types of biomarkers?
A: The FDA-NIH BEST glossary defines seven categories: susceptibility or risk, diagnostic, monitoring, prognostic, predictive, pharmacodynamic or response, and safety biomarkers.
Q: What is the difference between a biomarker and a diagnostic test?
A: A biomarker is the underlying measurable characteristic, while a diagnostic test is the validated clinical tool built around measuring that biomarker for a specific medical purpose.
Q: What is a predictive biomarker?
A: A predictive biomarker identifies which patients are more likely to respond favorably or unfavorably to a specific treatment, such as HER2 status guiding eligibility for HER2 targeted breast cancer therapy.
Q: What is a prognostic biomarker?
A: A prognostic biomarker estimates how a disease is likely to progress regardless of which treatment a patient receives, distinguishing it from a predictive biomarker tied to treatment response.
Q: Are genetic markers biomarkers?
A: Yes, inherited or acquired DNA variants are a major biomarker category, used for risk assessment, diagnosis, and increasingly for selecting targeted therapies in precision medicine.
Q: Can biomarkers detect cancer?
A: Certain biomarkers can help detect, stage, or monitor cancer, but most require confirmation through additional testing such as imaging or biopsy rather than serving as a standalone diagnosis.
Q: Are biomarker tests always accurate?
A: No single biomarker test is perfectly accurate. Sensitivity, specificity, and the population being tested all affect how reliable a positive or negative result actually is.