Cohen · Science (New York, N.Y.) 2018 · case-control diagnostic accuracy study · n=1817

Detection and localization of surgically resectable cancers with a multi-analyte blood test.

Cited 2856 times in the scientific literature.

Level 4 - case-series / case-control

Case-control diagnostic accuracy study comparing patients with known cancer diagnoses against healthy controls

PubMed 29348365 · doi:10.1126/science.aar3247 · record verified 2026-08-28

What was done

Researchers evaluated CancerSEEK, a multi-analyte blood test assessing circulating proteins and mutations in cell-free DNA. The test was evaluated in 1,005 patients with nonmetastatic, clinically detected cancers across eight types (ovary, liver, stomach, pancreas, esophagus, colorectum, lung, and breast) and 812 healthy control individuals.

What was found

CancerSEEK was positive in a median of 70% of cases across the eight cancer types. For five cancers lacking average-risk screening programs (ovary, liver, stomach, pancreas, and esophagus), sensitivity ranged from 69% to 98%. Specificity exceeded 99%, with 7 of 812 healthy controls testing positive. The assay successfully localized the primary tumor to a small number of anatomic sites in a median of 83% of patients.

Why it matters

This study demonstrates the feasibility of combining circulating protein markers with cell-free DNA mutations into a single blood test capable of detecting and localizing multiple surgically resectable cancers simultaneously.

Limits

The case-control design evaluated patients with clinically confirmed, established cancers rather than an asymptomatic screening cohort, which typically overestimates diagnostic sensitivity. Performance in broader, prospective real-world populations and the impact on clinical outcomes remain unproven.

Cited by