1,000 people screened
Galleri study assay (MCED-V2)
PATHFINDER 2 · 35,350 people
Asymptomatic adults aged 50+ at 32 U.S. and Canadian sites; no clinical suspicion of cancer or recent cancer diagnosis/treatment.
Target finding: Cancer diagnosed within 12 months
Each view describes its own study population and definition. These cohorts are not a ranking of screening methods.
One dot ≈ one person. Dots are rounded; exact proportions remain in the labels, including findings smaller than one dot.
A positive result is a signal. A negative result is not a clean bill of health.
From screening to the recorded finding
12-month cancer-status assessment; some participants did not enter the completed-follow-up performance analysis. · Values per 1,000
Published counts & scaling
| Reported category | Study count | Per 1,000 |
|---|---|---|
| Positive; cancer diagnosedCancer was diagnosed after a cancer-signal-detected result. | 173 | 4.89 |
| Positive; no cancer diagnosedNo cancer diagnosis in the completed-follow-up analysis; this is not proof of being cancer-free indefinitely. | 114 | 3.22 |
| Positive; follow-up incomplete305 total positive results minus 287 positives in the completed-follow-up performance set. Derived by subtraction; not classified as false positive. | 18 | 0.51 |
| Negative; cancer diagnosedCancer was diagnosed within the assessment window despite no signal detected. | 267 | 7.55 |
| Negative; no cancer diagnosedNo cancer diagnosed in the completed-follow-up analysis; this is not a general healthy status. | 31,453 | 889.76 |
| Negative; follow-up incomplete35,045 total negative results minus 31,720 negatives in the completed-follow-up performance set. Derived by subtraction; not classified as no cancer. | 3,325 | 94.06 |
| Total | 35,350 | 1,000.00 |
Per-1,000 people values are the source count ÷ cohort size × 1,000, displayed to two decimal places. Dot allocations use largest-remainder rounding, so the drawing has exactly 1,000 dots. Do not infer that a small category was absent from its rounded dot count.
Diagnostic workup
Separate safety population: 35,335 participants, including 290 positive-result participants. These workup counts must not be merged into the grid branches.
- Invasive procedures
- 213 of 290Positive-result participants with at least one invasive procedure; 295 procedures in total.
- Diagnostic resolution
- 48 daysKaplan–Meier median; 95% CI 43–56 days. Fifteen participants did not reach resolution.
- When cancer was diagnosed
- 37 daysMedian; IQR 24–61 days.
- When no cancer was diagnosed
- 70 daysMedian; IQR 38–121 days.
Study context & sources
Asymptomatic adults aged 50+ at 32 U.S. and Canadian sites; no clinical suspicion of cancer or recent cancer diagnosis/treatment.
12-month cancer-status assessment; some participants did not enter the completed-follow-up performance analysis.
Counts show study-cohort experience scaled to 1,000, not a prediction for an individual or the general population.
A cancer diagnosis after a negative blood test can be an interval cancer; these data do not establish that it was already detectable at the blood draw.
Detection and diagnostic performance do not establish a reduction in cancer deaths.
This grid retains all 35,350 analyzable tests; published performance uses 32,007 participants. The 3,343 outside that performance set remain explicitly unclassified.
MCED-V2 was the prospective study assay. Revised-assay retrospective FDA bridging estimates are a different analysis.
Single-arm, sponsor-funded study; diagnostic costs were covered.
The dot grid shows screening findings; it does not by itself establish that screening extends life. A negative result does not establish that a person is healthy. Different cohorts, target definitions and reference assessments cannot be compared as though they were randomized against each other.
Source date: 2026-09-22 · Cohort reference reviewed 2026-10-03