The Galleri test is a multi-cancer early detection (MCED) blood test that screens for a shared cancer signal across more than 50 types of cancer, particularly those lacking routine screenings.
The Galleri test (GRAIL, Inc.) is a blood-based multi-cancer early detection (MCED) test that analyzes methylation patterns of cell-free DNA (cfDNA) using machine learning to detect cancer signals across more than 50 cancer types and predict the likely tissue of origin in asymptomatic adults aged ≥50 years.
The test sequences cfDNA from a blood draw, targeting methylation patterns across >100,000 DNA regions.
The Galleri test requires a simple blood draw that looks for cell-free DNA (cfDNA) circulating in the bloodstream.
If a cancer signal is detected, the test analyzes the DNA pattern to predict where in the body the cancer started, guiding doctors on where to look for a diagnosis.
The test is prescription-only and recommended for adults aged 50 or older who have an elevated risk of cancer.
It is not for individuals who are pregnant, 21 or younger, or undergoing active cancer treatment.
Accuracy:
Galleri has a specificity greater than 99%, meaning the false-positive rate is less than 1%.
When a cancer signal is detected, the test accurately predicts the origin of the cancer in over 90% of cases.
The Galleri test does not detect all cancers, and both false-positive and false-negative results can occur.
It is designed to be used in addition to routine, standard-of-care screening tests (like mammograms, colonoscopies, or PSA tests), not as a substitute.
An AI-based classifier identifies aberrant methylation signatures associated with cancer and predicts the cancer signal origin (CSO) to guide diagnostic workup.
Test Performance
Specificity: 99.5%
Overall sensitivity: 51.5%, with marked stage dependence:
Stage I: 16.8%
Stage II: 40.4%
Stage III: 77.0%
Stage IV: 90.1%
Sensitivity for 12 prespecified high-mortality cancers (stages I–III): 67.6%
Cancer signal origin prediction accuracy: 88.7% of true positives
The PATHFINDER prospective study (n = 6,621 asymptomatic adults) reported a PPV (positive predictive value) of 38%, NPV ( negative predictive value) of 98.6%, and specificity of 99.1%.
A single-institution PATHFINDER analysis found sensitivity of 39% and specificity of 99.1%, with 6 of 15 false positives attributable to monoclonal B-cell lymphocytosis.
In the SYMPLIFY study of symptomatic patients referred for cancer investigation in the NHS, sensitivity was higher at 66.3% with specificity of 98.4%.
A large real-world series of 111,080 Galleri tests found a cancer signal detection rate of 0.91%.
Among 459 individuals with signal-detected results and reported outcomes, 258 had confirmed invasive cancer spanning 32 types.
Cancer signal origin prediction was correct in 87% of cases, with a median of 39.5 days from result receipt to cancer diagnosis.
Low early-stage sensitivity remains the primary limitation — the test detects only ~17% of stage I cancers, which is the population where early detection would provide the greatest benefit.
Most MCED-detected cancers are advanced stage — in the PATHFINDER study, only 48% of detected cancers were stage I–II (vs. 73% by usual care), and in the real-world Galleri series, 28% were stage I–II while 48% were stage IV.
No demonstrated mortality benefit.
The NHS-Galleri trial (n >140,000), the first and largest RCT of an MCED test, used late-stage cancer incidence reduction as its primary endpoint.
The trial did not meet its primary endpoint of reducing stage III–IV cancer incidence, though GRAIL reported a “substantial” reduction in stage IV diagnoses for 12 prespecified cancers.
It remains unclear whether MCED screening reduces mortality when accounting for lead-time and length-time bias.
Evidence is insufficient to evaluate the benefits, accuracy, and harms of MCD test screening, with no completed controlled studies report mortality outcomes.
The NCI Vanguard study, a 4-year evaluation of multiple MCED tests vs. standard of care, is underway and will inform the design of a definitive RCT to determine whether MCED screening reduces cancer mortality.
Mayo’s experience with >6,000 tests showed approximately 1% positive rate.