Biomarkers and Mutations
Educational orientation
Expanded educational edition: 2026-09-10. This page explains concepts; it does not interpret genomic results.
By the End of This Chapter
Distinguish prognosis, prediction and unknown variant significance.
Learning matrix and gaps in the short edition
Core Idea
A biomarker is a measurable feature that can help describe disease, risk, prognosis, or treatment relevance. A mutation is a DNA change; some mutations help drive cancer, some are passengers, and some have no clear meaning in a given context.
Modern oncology often uses biomarkers from tissue, blood, imaging, or molecular assays. The meaning depends on the cancer type, test method, evidence level, and clinical setting.
Step-by-Step Explanation
1. Measure, interpret, use
A biomarker is a measurable characteristic; a DNA variant is only one possible type. Analytical validity asks whether the test measures correctly. Clinical validity asks whether the measurement relates to a condition or outcome in the studied context. Clinical utility asks whether using it improves relevant decisions or outcomes. Good measurement alone does not answer the other two questions.
2. Meaning depends on context
A prognostic marker is associated with disease course; a predictive marker informs differences in an intervention's effect according to marker status. “Predictive” here does not merely mean that an algorithm makes predictions. A variant of uncertain significance must not become a conclusion about pathogenicity or response. Tumor type, exact alteration, method, and evidence belong to the interpretation.
3. A present alteration is not necessarily a driving alteration
A driver alteration contributes to properties favoring tumor development or maintenance; a passenger may accompany evolution without causing those properties in that context. Frequency, functional effect, and biological context help investigate the distinction. But “driver” does not mean treatable target, and “rare” does not necessarily mean important. A variant may be detected with analytical confidence while its clinical significance remains uncertain. Separating detection, biological role, and possible use prevents a variant list from becoming an unjustified decision list.
Supporting reading: NCI: Biomarker Testing for Cancer Treatment.
The Concept Illustrated
Worked Example
A fictional study finds more recurrence in the group with marker M but does not compare interventions by M status. This may support a prognostic association, subject to design and confounding; it does not show benefit from choosing a therapy by that marker. The invalid leap is turning an outcome correlation into a treatment rule.
Essential Vocabulary
- Driver: alteration that contributes to cancer behavior.
- Passenger: alteration not known to drive disease.
- Predictive biomarker: informs differences in an intervention's effect according to marker status.
- Prognostic biomarker: associated with disease course or outcome in a defined context; it does not alone establish benefit from choosing an intervention.
What This Does Not Mean
A mutation listed in a report is not automatically actionable. Evidence, indication, and assay validity matter.
Builder/Data Lens
Variant annotation needs genome build, transcript, allele fraction, test type, and knowledgebase version.
Check Your Understanding
Does finding a molecular target prove an applicable therapy exists?
Answer and reasoning
No. Biological plausibility, preclinical evidence, human studies, guideline use, and approval for an indication are different states. Approval requires checking indication, authority, country, and date.
Optional Deeper Exploration
A variant database is an evidence source, not an automatic decision engine. Preserve version, assertion context, evidence strength, and conflicting interpretations.
Go Deeper
References
- National Cancer Institute. Biomarker Testing for Cancer Treatment
- National Cancer Institute. The Genetics of Cancer