Biomarkers

#Define the characteristic and the measurement

The FDA-NIH BEST definition describes a biomarker as a measured characteristic that indicates a biological process, a disease process or a response to an exposure or intervention. FDA includes molecular, histologic, radiographic and physiologic characteristics. Biomarkers are therefore not limited to blood tests or genes.

A useful description names the characteristic, the source or material and the measurement method. A biomarker can also be a panel of characteristics. BEST distinguishes this kind of indicator from an assessment of how someone feels, functions or survives. A measurement can supply information without being a complete diagnosis or a direct measure of benefit.

Evidence: FDA: biomarkers and qualification

#State the purpose before interpreting the label

BEST separates susceptibility or risk, diagnostic, monitoring, prognostic, predictive, pharmacodynamic or response, and safety biomarkers. These labels describe the question being asked of a measurement. Being informative for one purpose does not establish that the same measurement is useful for all of the others.

In FDA drug development, a context of use combines the biomarker category with its specified intended use. The population and setting can be part of that description. FDA qualification supports a stated interpretation within that context; it is not a blanket endorsement of every use or the measurement method itself. This page explains the distinction, not a regulatory filing process.

Evidence: FDA: biomarkers and qualification / FDA: biomarker context of use

#Separate prognosis from prediction of an intervention effect

BEST defines a prognostic biomarker in relation to the likelihood of a future event, recurrence or progression among people who have the disease or condition of interest. A predictive biomarker identifies people more likely than similar people without the marker to experience a favourable or unfavourable effect from a particular exposure or intervention.

These questions are related but not interchangeable. Better outcomes in one biomarker group after treatment can reflect prognosis rather than a different treatment effect. BEST explains why comparison with a control and consideration of groups with and without the marker matter to that distinction. This is a research-design concept, not a prediction about an individual patient.

Evidence: FDA-NIH BEST: prognostic biomarker / FDA-NIH BEST: predictive biomarker

#Check measurement validity and meaning separately

BEST distinguishes analytical validation, whether a test measures what it is intended to measure, from clinical validation, whether the biomarker can measure or predict the relevant clinical concept. Validation is tied to a purpose. Discovery findings alone are typically not enough to establish a biomarker as fit for that use.

Cancer markers illustrate why context matters. NCI explains that noncancerous conditions can raise some tumor markers and that not everyone with a particular cancer has a raised marker. NCI also describes limitations of cancer-treatment biomarker tests, including differences between cells and changes over time. These cancer examples do not define every biomarker. A result needs the evidence and context for its actual purpose.

Evidence: FDA-NIH BEST: validation and intended purpose / NCI: tumor markers / NCI: biomarker testing for cancer treatment

#Do not equate marker change with clinical benefit

A biomarker may show a biological response without directly measuring how someone feels, functions or survives. FDA describes a surrogate endpoint as a substitute for a clinical outcome in a trial. Evidence is needed to support the relationship between that substitute and clinical benefit in the intended setting.

Even a validated surrogate does not capture every benefit or harm. FDA notes that an improvement in one setting can mislead in another because effects outside the surrogate may matter. A changed number is therefore not automatically proof that a treatment helps overall. This is general research education, not test selection, personal result interpretation or a Mynd testing service.

Evidence: FDA: biomarkers and surrogate endpoints / FDA: biomarkers and qualification

Source note

The sections above were checked against the linked sources. No clinical review has been performed. This is general research education, not a clinical guideline.