Comments on the Recent Results of the MCED Galleri Test
Topline results from the NHS-Galleri randomized trial of GRAIL’s Galleri multi-cancer early detection (MCED) blood test reported that the study did not meet its primary endpoint. Headlines and the market reaction were predictable. But a primary endpoint miss in a screening RCT should be treated as a call for rigor and transparency, not as a blanket verdict that the entire concept of MCED is finished.
Separate the headline from the scientific question
The NHS-Galleri trial was designed to evaluate a population-level outcome endpoint (a reduction in late-stage cancer incidence) rather than only cross-sectional clinical performance metrics. That design choice matters: stage-shift endpoints are influenced by time horizon, adherence, diagnostic pathways, and cancer biology and treatment regimens, not just by the test’s analytic and clinical detection accuracy.
FDA’s review and the approval pathway for MCED (e.g., sensitivity, specificity, tissue-of-origin prediction, and PPV).
Unlike many prior cancer screening tests, MCEDs are multi-cancer detection tests that is noninvasive only requiring a blood sample. However positive results require additional diagnostic testing to determine whether the result was a true signal indicative of the presence of cancer. These tests cost money and payers generally want to see value for these costs which is why demonstrating a reduction in cancers was an important endpoint.
Although, some stakeholders now demand answers about clinical utility (improvement in patient outcomes or a net patient benefit), FDA approval historically has not required evidence of improved patient outcomes. This creates a potential tension: insisting on utility evidence pre-approval may be viewed as moving the goalposts, but ignoring utility concerns may be viewed as insufficiently protective given the scale and complexity of multi-cancer follow-up workups.
A practical way to frame the regulatory reality is this: if the PMA package demonstrates robust clinical performance in the intended-use population and an acceptable benefit–risk profile for the claimed intended use, FDA has a clear pathway to approval. However, FDA can still narrow labeling (e.g., adjunct vs standalone), require post-market studies, and request additional analyses where uncertainties remain, particularly around downstream workup pathways and harms.
Bottom line:
Two things can be true at the same time:
(1) missing the primary endpoint is scientifically meaningful and demands transparency; and
(2) the goal of MCEDs, which is to detect cancers early enough to be curable or prolong survival, remains deeply important, especially for cancers with no screening today.
