Grail's Galleri blood test shows improved cancer detection, but survival impact remains unclear

Grail, a company betting its future on blood tests that can catch cancer before symptoms appear, released results on Friday from its largest study yet. The findings show the company has made real progress in detecting tumors early—but they also expose a gap that no amount of improved lab numbers can close: nobody yet knows if finding cancer sooner actually saves lives.

The Pathfinder 2 study followed nearly 36,000 adults over 50 for more than a year, testing Grail's Galleri blood test on a scale the company had never attempted before. The test hunts for cancer signals across dozens of tumor types at once, a fundamentally different approach from traditional screening, which targets one cancer at a time. When the researchers looked at what the test actually caught, they found it identified 40.4% of cancers in the study population. More than half of those were caught at stage 1 or 2, when treatment is typically less invasive and outcomes tend to be better. Notably, about three-quarters of the cancers detected were types—pancreatic, liver, head and neck—that don't have established screening programs today. For those cancers especially, early detection is rare.

The test's accuracy numbers improved noticeably from earlier versions. When someone got a positive result, there was a 62% chance they actually had cancer. That's a significant jump from the 43% to 50% range Galleri showed in previous studies. It means fewer people chasing false alarms, fewer unnecessary biopsies and procedures based on a mistake. The flip side was equally reassuring: when the test came back negative, it was right 99.1% of the time. People could feel confident they didn't have cancer—at least not one Galleri was designed to find.

Yet these numbers, however encouraging, sidestep the question that matters most to patients and doctors deciding whether to adopt the test. Does catching cancer earlier actually change the outcome? Does it save lives? The study measured whether Galleri could find tumors; it did not measure whether people who got those early diagnoses lived longer than people who didn't. That's the test that matters. A blood test that catches cancer at stage 1 instead of stage 4 sounds like an obvious win, but only if stage 1 detection leads to better survival. If the cancers Galleri finds are ones that would have been caught anyway through symptoms or other means, or if they're slow-growing tumors that wouldn't have killed the patient in the first place, then early detection becomes a different story—one about anxiety, cost, and unnecessary treatment rather than lives saved.

Experts in the field acknowledged the progress Grail has made in reducing false positives and improving sensitivity. But they also made clear that the real work—the studies that would actually prove the test saves lives—hasn't happened yet. That kind of evidence takes years to accumulate. It requires following large groups of screened and unscreened people over time, watching who gets sick, who gets treated, and who survives. Grail is not the only company racing to build that evidence. The liquid biopsy market is crowded and growing, with competitors also trying to prove their tests can reshape cancer screening. The company's announcement on Friday was a step forward in the technical race. Whether it's a step forward for patients depends on evidence that hasn't yet been gathered.