MCED evidence

Multi-cancer early detection

Blood tests that look for a cancer signal across more than one cancer type. The same three levels used on MRD evidence apply here. A performance study is not a trial of fewer late-stage cancers.

Level 1

A prospective trial in which the ctDNA or MRD result assigns, withholds, or changes treatment. Randomized and single-arm adaptive trials both qualify. Positive and negative trials are listed.

Level 2

A large prospective observational cohort, or a correlative analysis from a phase 3 trial or other large study. The test informs prognosis or describes a treatment association. It does not assign therapy.

Level 3

A smaller cohort, a single-center series, a feasibility study, or an analytical validation. Useful for methods and early signals, not for changing practice on its own.

NegativeLevel 12026 · The New England Journal of Medicine
Effect of Screening with Multicancer Early-Detection Test on Late-Stage Cancer Diagnosis.
Sasieni P, et al. PMID 42776074. doi 10.1056/NEJMoa2505723
Randomized trial in England. 71,122 people were assigned to a blood multi-cancer test and 71,128 to usual care. The incidence of stage III or IV cancer among 12 prespecified types did not differ (incidence rate ratio 1.03, P = 0.63). The stage IV incidence rate ratio was 0.86.
Neutral or mixedLevel 22021 · Annals of Oncology
Clinical validation of a targeted methylation-based multi-cancer early detection test using an independent validation set.
Klein EA, et al. PMID 34176681. doi 10.1016/j.annonc.2021.05.806
Case-control validation from the Circulating Cell-free Genome Atlas study (NCT02889978), not a screening-outcomes trial. In 4,077 people (2,823 with cancer and 1,254 without), specificity was 99.5% and overall sensitivity was 51.5%, rising from 16.8% in stage I to 90.1% in stage IV.
Neutral or mixedLevel 22023 · The Lancet
Blood-based tests for multicancer early detection (PATHFINDER): a prospective cohort study.
Schrag D, et al. PMID 37805216. doi 10.1016/S0140-6736(23)01700-2
Prospective feasibility cohort (NCT04241796), not a mortality trial. A cancer signal was found in 92 of 6,621 people with a result (1.4%). Cancer was confirmed in 35 and not found in 57. Median time to diagnostic resolution was 79 days.
Neutral or mixedLevel 22026 · Nature Medicine
Performance and safety of a multi-cancer early detection test: the PATHFINDER 2 study.
Nabavizadeh N, et al. PMID 42773207. doi 10.1038/s41591-026-04618-w
Prospective performance study in 35,878 people aged 50 or older. Among 32,007 with analyzable results, the cancer detection rate was 0.54%, the positive predictive value was 60.3%, specificity was 99.64%, and sensitivity for all cancers was 39.3%. Endpoints were descriptive.
Neutral or mixedLevel 22020 · Science
Feasibility of blood testing combined with PET-CT to screen for cancer and guide intervention.
Lennon AM, et al. PMID 32345712. doi 10.1126/science.abb9601
DETECT-A, a prospective study of 10,006 women. Blood testing detected 26 cancers; 9 of 15 imaged with PET-CT were surgically excised. Standard screening found 24 more cancers, and 46 were found by neither. Futile invasive procedures occurred in 0.22%.
Neutral or mixedLevel 32018 · Science
Detection and localization of surgically resectable cancers with a multi-analyte blood test.
Cohen JD, et al. PMID 29348365. doi 10.1126/science.aar3247
Earlier CancerSEEK methods paper on a multi-analyte blood test for surgically resectable cancers. It is a detection study, not a population screening trial.
Neutral or mixedLevel 22023 · The Lancet Oncology
Multi-cancer early detection test in symptomatic patients referred for cancer investigation in England and Wales (SYMPLIFY).
Nicholson BD, et al. PMID 37352875. doi 10.1016/S1470-2045(23)00277-2
Prospective study in symptomatic people referred for urgent cancer investigation (ISRCTN10226380). When cancer was present, the predicted site of origin was accurate in 85.2% of cases. Sensitivity was 80.4% in the group referred for possible upper gastrointestinal cancer. This is a diagnostic-accuracy study, not a screening-mortality trial.

Conference data

Slide and poster links, when added, may require an ASCO or ESMO member login.

Level 2positiveASCO Annual Meeting 2026
Safety and performance results from PATHFINDER 2, a registrational study of a multi-cancer early detection (MCED) test in an intended-use population.
Giridhar, Karthik · Abstract LBA10509 · LBA · PATHFINDER 2 · 35,878 enrolled; 32,007 with a 12-month cancer assessment analyzable for performance; 35,335 analyzable for safety
Galleri multi-cancer early detection test (targeted methylation cfDNA), as evaluated in PATHFINDER 2
Of 287 (0.9%) participants with a positive test, 173 were diagnosed with cancer within 12 months (detection rate 0.5%; PPV 60.3%, 95% CI 54.5-65.8%; specificity 99.6%). Twelve-month episode sensitivity was 69.8% in a prespecified 12-cancer subgroup, 66.2% in 6 aggressive cancers, and 39.3% across all cancers. Cancer-signal-origin accuracy was 91.3%. Of 151 MCED-detected new primaries, 80 (53.0%) were stage I-II. Five study-related adverse events occurred during diagnostic workup; none were serious. Median time from result to diagnostic resolution was 48 days.
Level 1mixedASCO Annual Meeting 2026
NHS-Galleri: Primary results from a randomised controlled trial to assess the clinical utility of a multi-cancer early detection (MCED) test in population screening.
Swanton, Robert Charles · Abstract LBA100 · LBA · NHS-Galleri · 142,924 enrolled participants in England, aged 50-79 years, randomized 1:1 after the year-0 blood draw
Galleri MCED test (cfDNA cancer-signal test)
The primary endpoint, a reduction in stage III/IV incidence among 12 prespecified cancers, was not met (706 vs 688 events; IRR 1.03, 95% CI 0.92-1.14; P=0.6324). A 14% reduction in stage IV cancers was observed after 3 years (342 vs 397; overall IRR 0.86, 95% CI 0.744-0.998). Stage I/II cancers increased by 16% (RR 1.16) and stage I-III by 19% (RR 1.19). Screen-detected cancers were 1,173 vs 290. Aggregate specificity was 99.55% and PPV 52.0%. No related serious adverse events were reported.
Level 2ESMO Congress 2025
Safety and performance of a multi-cancer early detection (MCED) test in an intended-use population: Initial results from the registrational PATHFINDER II study
Nima Nabavizadeh · Abstract LBA64 · Proffered Paper / late-breaking oral · PATHFINDER II
Slides WebcastSlides + webcast. An ASCO or ESMO member login may be required.
Level 3positiveASCO Annual Meeting 2023
Clinical evaluation of cancer signal origin prediction and diagnostic resolution following multi-cancer early detection testing.
Klein, Eric A. · Abstract 10520 · PATHFINDER · 39 participants with a cancer signal detected by both test versions
targeted methylation MCED test with cancer-signal-origin prediction (PATHFINDER / Galleri lineage)
Diagnostic evaluations were consistent with the predicted cancer signal origin in 30/39 (77%) cases. Diagnostic resolution was reached after initial evaluation in 32/39 (82%); the remaining 7 reached resolution after additional workup (3 with cancer, 4 without). Whole-body imaging was used in 27 cases and was equivocal or had no impact on resolution in 8/27 (30%).
Level 2positiveASCO Annual Meeting 2021
A prespecified interim analysis of the PATHFINDER study: Performance of a multicancer early detection test in support of clinical implementation.
Beer, Tomasz M. · Abstract 3070 · PATHFINDER · 4,011 of 4,047 participants analyzable in this interim cohort enrolled as of 6 October 2020
targeted methylation-based cfDNA MCED test (Galleri lineage; a more recent classifier than the version returned to investigators), predicting cancer signal origin
A cancer signal was detected in 38/4011 (0.95%). Using a conservative minimal PPV that treated all discordant positives as false positives, PPV was 43.3% (13/30, 95% CI 27.4-60.8%). Among 13 detected cancers, top cancer-signal-origin accuracy was 92.3% (12/13). Nine cancer types were represented among those 13 cases.