FDA advisers weigh Grail's multi-cancer blood test amid efficacy doubts

A blood test that aims to detect multiple cancers from a single draw faces its first major regulatory test on Wednesday, as a panel of U.S. Food and Drug Administration advisers votes on whether its benefits outweigh its risks for people aged 50 and older.

The test, developed by California-based Grail Inc., hunts for fragments of DNA that tumors shed into the bloodstream — an approach that could mark a shift from the current generation of screenings, which typically target one cancer at a time, such as mammograms and colonoscopies.

In briefing documents released ahead of the meeting, FDA staff did not flag major safety concerns but asked the panel to consider whether the test identifies enough early-stage cancers to justify marketing it for "early detection." The agency is not bound by the advisers' recommendation.

Galleri is already available in the U.S. through a regulatory pathway for certain laboratory-developed tests, with patients typically paying $700 or more out of pocket. Formal approval would broaden access by paving the way for coverage by Medicare and private insurers.

The review has drawn scrutiny over several unusual features of Grail's application. The FDA regards the company's current test as a different product from the version used in two large studies, prompting Grail to test frozen blood samples from those trials with the newer version. The submission also includes only the first year of data from a 140,000-person U.K. study, which did not meet its primary goal of showing a statistically significant reduction in late-stage cancers — a shortfall not cited in the FDA review but likely to surface in Wednesday's discussion.

According to study data released by the FDA this week, the test correctly identified cancer in more than 66% of patients who received a "cancer signal detected." Over the one-year period, it detected about 30% of cancers later diagnosed by physicians. The U.K. trial, run with the National Health Service, was not designed to show that screening reduced cancer deaths, which would have required far longer follow-up.

A separate U.S. analysis found the test flagged a cancer signal in 287 people, with cancer eventually diagnosed in 173 of them. It performed better at picking up less common cancers, such as liver and ovarian cancers, than common ones like prostate cancer — a pattern that raises questions about its population-level impact.

Oncologists remain divided on whether the technology is ready for broad use. Dr. Badrinath Konety of Allina Health Cancer Institute in Minneapolis said the approach holds "a lot of promise" but has urged caution with patients who ask about it. "The benefit is it may give you a signal," he said. "The flip side now is you've got to chase that signal," which can mean additional, costly tests that may not yield clear answers.

Another concern is that people who receive an all-clear from DNA testing might skip guideline-recommended screenings such as mammograms or colonoscopies. Johns Hopkins scientist Nickolas Papadopoulos called the test "pretty promising" but said that if the FDA clears it, doctors must understand its pros and cons — and ensure their patients do too. Among the open questions, he said, is whether testing once a year or once every three years is enough to catch cancers early. "We're still learning," he said.