FDA Advisers Weigh GRAIL's Galleri After NHS Trial Missed Primary Endpoint
核心洞察
An FDA advisory panel meets Sept. 23 to vote on GRAIL's premarket approval application for Galleri (搜索), the first multi-cancer (搜索) early detection test reviewed by such a committee.
The NHS-Galleri (搜索) trial in 142,942 people did not significantly reduce stage III or IV diagnoses across 12 prespecified cancer (搜索) types, though stage IV diagnoses fell 14%.
Positive predictive value was 52% in NHS-Galleri (搜索), meaning roughly half of people told a cancer (搜索) signal was detected were ultimately confirmed to have cancer.
An FDA advisory panel meets Wednesday, Sept. 23, to discuss and vote on GRAIL's application for premarket approval of its Galleri (搜索) blood test. It is the first time an FDA advisory committee has reviewed such an application for a multi-cancer (搜索) early detection test. The public docket drew 232 comments before it closed on Sept. 16, according to the Federal Register notice, which also said the agency intended to post background materials no later than two business days before the meeting.
The test is already sold by prescription, and the only randomized trial behind the application delivered a mixed result. Nothing about Wednesday's vote changes current screening guidelines.
The Randomized Trial Behind the Vote Delivered a Split Result
The sharpest point of debate is the NHS-Galleri (搜索) trial in England, the largest randomized study of the test. The trial enrolled 142,942 people ages 50 to 77 who had no cancer (搜索) symptoms. Half received annual Galleri testing for three years on top of routine screening.
According to the American Society of Clinical Oncology, the trial did not reach its primary goal of significantly reducing cancers diagnosed at stage III or IV among 12 prespecified cancer (搜索) types, a group that did not include breast or cervical cancer.
Other results pointed in a more favorable direction. The Galleri (搜索) group had 14% fewer stage IV diagnoses among those 12 cancers, four times as many cancers found through screening (1,173 versus 290), and 20% fewer cancers found through emergency presentations. GRAIL funded the trial.
"While the Galleri (搜索)-NHS study results show some encouraging trends toward tumor downstaging, it is important to recognize that the trial did not statistically reduce late-stage cancers by its predefined primary endpoint," said Dr. Julie R. Gralow, ASCO's chief medical officer. She added that longer-term follow-up is needed to fully weigh the population-level benefits against the risks.
No trial has yet shown that Galleri (搜索) screening lowers cancer (搜索) deaths. According to GRAIL's announcement of the meeting, the application also relies on the single-arm PATHFINDER 2 study of 25,490 U.S. participants, which measures test performance but has no comparison group, along with first-year data from more than 70,000 people in the NHS-Galleri intervention arm.
FDA staff raised no major concerns about the multi-cancer (搜索) test ahead of the advisory committee meeting, according to a Reuters report.
A Positive Signal Starts a Diagnostic Chase
In NHS-Galleri (搜索), specificity was 99.55%, meaning very few people without cancer (搜索) received a positive result. But the positive predictive value was 52%, according to the trial abstract in the Journal of Clinical Oncology. Roughly half of the people told a cancer signal was detected were ultimately confirmed to have cancer.
That leaves the other half facing scans, specialist visits and sometimes biopsies before learning they are likely cancer (搜索)-free. ASCO listed short-term anxiety about a positive result and the burden of follow-up testing among the trial's reported effects.
A negative result carries its own caution. The proposed indications for use published in the Federal Register state: "A test result of No Cancer (搜索) Signal Detected does not rule out the presence of cancer, and individuals should continue with guideline-recommended single-cancer screening tests."
Intended Population and Cost Barriers
The proposed intended use covers adults 50 and older, the age group in which cancer (搜索) risk climbs steeply. People with symptoms such as unexplained weight loss, bleeding, a new lump or persistent pain are not the audience for a screening test and need a diagnostic workup regardless of any blood result.
Cost is the other barrier. GRAIL sets the retail price at $950, and most buyers pay out of pocket. Congress voted in February to allow Medicare to reimburse multi-cancer (搜索) detection tests starting in 2028, but only after the tests clear full FDA review, NPR reported. Approval alone would not guarantee private insurance coverage.
Patients considering the test can ask what follow-up a positive result would trigger, what their insurer would pay for that follow-up, and whether they are current on colonoscopy, mammography, cervical screening and, for eligible current and former smokers, lung CT. Skipping or delaying a recommended screening in favor of a blood test would be premature based on current evidence.
Wednesday's Vote Is Advice, Not Approval
The FDA meeting page lists the session from 9 a.m. to 6 p.m. Eastern at the agency's White Oak campus in Silver Spring, Maryland, with an online option for the public. The panel will discuss the application, make recommendations, and vote. Its recommendation does not bind the FDA's final decision, though the agency often follows its advisers.
Several questions remain open. The FDA's background materials will show where agency reviewers agree or disagree with the company; any approval could carry a narrower label than GRAIL requested, and mortality results from the U.K. trial are years away.