Why a Promising Cancer Biomarker Is Not Yet a Screening Test

A biomarker study may report excellent sensitivity, specificity or area under the curve without proving that the assay is ready for population screening. The distinction depends on how participants were selected, when samples were collected and what decisions the result is expected to change.

Case-control performance starts with known groups

Early studies commonly compare stored samples from people known to have a disease with samples from selected healthy or benign controls. This design is useful for discovering whether a biological signal exists. It can also make the groups more different from each other than the people encountered in routine care.

Real clinics include early and atypical disease, unrelated illnesses, medications, pregnancy and technical variation between collection sites. A threshold that separates curated groups may lose accuracy when applied consecutively to everyone who qualifies for testing.

Prevalence changes what a positive result means

Sensitivity measures how often a test detects the condition among people who have it. Specificity measures how often it returns a negative result among people who do not. Positive predictive value asks a different question: among everyone who tested positive, how many truly have the condition?

When a disease is rare, even a highly specific test can produce more false positives than true positives. This is why predictive values calculated for a high-risk clinic cannot be assumed for the general population. Screening proposals must be evaluated in the population where they will actually be used.

Prospective validation changes the order

In a prospective study, participants are enrolled and tested before researchers know who will develop or already has the target condition. The threshold should be fixed in advance. Investigators then follow a defined clinical pathway to confirm diagnoses and count false positives, false negatives and indeterminate results.

Multicenter testing also shows whether sample handling, equipment and laboratory practice affect results. Reproducibility is essential if a test is expected to guide decisions outside the laboratory that developed it.

Clinical utility is more than diagnostic accuracy

A screening test should help people live longer or avoid serious disease, or improve decisions enough to justify its harms and costs. Researchers must ask whether it finds disease early enough to change treatment, whether follow-up procedures cause harm and whether false reassurance delays care.

A biomarker can therefore be scientifically important before it becomes clinically useful. Strong retrospective performance supports the next study. It does not by itself support changing medical guidance.