A blood test designed to flag signals shared by more than 50 types of cancer detected disease in 173 of nearly 36,000 adults over 50 who had no cancer symptoms, according to results from the PATHFINDER 2 trial published Sept. 22 in Nature Medicine. Researchers say adding the test, called Galleri, to standard screening increased cancer detection roughly 6.5-fold compared with standard screening alone.
Of the 32,007 participants included in the trial's 12-month performance analysis, 287 received a positive signal from the test; cancer was ultimately confirmed in 173 of them, a positive predictive value of 60.3 percent. Specificity — the rate at which the test correctly identified people without cancer — was 99.64 percent. When the test did find something, it correctly predicted where in the body the cancer originated 91.3 percent of the time. More than half the cancers caught, 53 percent, were stage I or II, and about 70 percent were cancer types with no recommended screening test in the United States, according to Oregon Health & Science University, which enrolled the trial's largest single cohort.
A Complement, Researchers Say, Not a Replacement
The test works by sequencing cell-free DNA fragments that tumors shed into the bloodstream and looking for chemical patterns, mainly methylation signatures, that differ from healthy tissue. Nima Nabavizadeh, the trial's principal investigator, said the appeal is catching disease before it causes symptoms.
It finds many of these cancers while they're still asymptomatic and localized, when cure is still possible.
Nima Nabavizadeh, M.D., PATHFINDER 2 principal investigator
But the trial also underscores the test's limits. Its 12-month sensitivity — the share of all cancers diagnosed in participants that the test actually caught — was 39.3 percent overall, meaning most cancers diagnosed among trial participants during the year were not flagged by Galleri and were instead found through other means. Sensitivity was higher, 69.8 percent, for a pre-specified group of a dozen cancer types the test targets most directly. Nabavizadeh himself cautioned that the test is "not a replacement for standard screening, but a powerful complement to it." Follow-up after a positive result was generally not invasive: only 0.6 percent of participants needed an invasive procedure to confirm or rule out cancer.
The study was designed and funded by GRAIL Inc., the test's manufacturer, which also took part in data analysis and approved the manuscript for publication — a financial relationship independent researchers say should be weighed alongside the results. PATHFINDER 2 was also not a randomized trial comparing outcomes between tested and untested groups, so it cannot yet show whether using Galleri reduces cancer deaths. The test is currently offered as a lab-developed test rather than through full FDA approval, though an FDA advisory panel recently voted in favor of aspects of its use. Researchers say a randomized controlled trial measuring cancer mortality is still needed before health systems adopt multicancer blood screening broadly.
