Will FDA grant PMA approval to GRAIL's Galleri multi-cancer early detection test by December 31, 2026?

TL;DR

I assign a 21% chance that FDA grants original PMA approval to Galleri by December 31, 2026. Galleri will probably get a panel outcome that keeps approval alive, but the September 23 meeting leaves only 99 calendar days, and GRAIL told investors it still expected potential approval in the first part of 2027 (FDA meeting notice, earnings-call transcript). The randomized NHS trial missed its primary endpoint, while the unpublished assay bridge and post-submission data updates create enough review risk that early 2027 is the base case (ASCO abstract, GRAIL 10-Q).

Context

GRAIL submitted Galleri’s final modular-PMA module on January 29, 2026, and later reported that FDA had accepted the application for review. FDA has scheduled its Molecular and Clinical Genetics Panel for September 23 to consider a broad prescription-only indication covering adults aged 50 or older, with cancer-signal-origin prediction and diagnostic follow-up after positive results (GRAIL, SEC filing, FDA).

At the September 20 cutoff, FDA’s meeting page contained only two conflict-waiver documents. The FDA and sponsor briefing books, voting questions, roster, and agenda had not appeared; FDA said background material was intended to be available by September 21, so the absence is neutral rather than adverse (FDA event page, Federal Register notice).

Evidence

The historical backbone points to eventual approval being much easier than approval inside this deadline. A study of post-2010 PMA panels found that 48 of 52 applications were eventually approved, but the mean panel-to-decision interval under the newer voting system was 243 days—far longer than Galleri’s 99-day window (FDLI study).

The three closest identified original cancer-screening PMAs reviewed by Galleri’s panel show the range:

Shield proves that approval inside 99 days is operationally possible. Cologuard shows that even a clean cancer-screening approval can miss the deadline, while Epi proColon shows how a divided evidentiary package can remain unresolved for years.

The complete MDUFA V history available for completed panel-reviewed PMA cohorts is also slow. These data cover original PMAs and panel-track supplements and measure a MDUFA decision, which need not be an approval:

Galleri would need approval roughly 336 calendar days after the final module. Modular review is favorable because earlier modules may already have been examined, but the recent cohort data show that sponsor-response time commonly pushes panel cases beyond that mark.

MDUFA V sets a goal of a decision within 320 FDA days for 90% of panel-reviewed PMAs and says FDA will seek a decision within 60 days of the panel recommendation, as resources permit. With no clock stops, a January 29 filing would put the 320-day point in mid-December; however, FDA days exclude sponsor holds, and an approvable or not-approvable action would satisfy the process goal without satisfying this question (MDUFA V commitment).

The strongest case-specific timing signal is GRAIL’s own guidance. When an analyst asked on August 5 whether a fall panel changed the previously discussed early-2027 timeline, CEO Josh Ofman said the expectation was unchanged and that potential approval was expected in the first part of 2027 (transcript). Management can guide conservatively, but it knows more than outsiders about FDA questions, labeling, inspections, amendments, and review-clock interruptions.

The scientific evidence supports a workable panel outcome but not a clean consensus. FDA’s 2023 general MCED panel favored randomized studies, tissue-of-origin guidance, continued conventional screening, cancer-specific benefit-risk assessment, and specificity above roughly 99%; it also said stage shift could support an early-detection claim even though members disagreed over whether mortality must be the primary endpoint (FDA summary). Galleri meets several of those conditions.

PATHFINDER 2 enrolled 35,878 adults, with 32,007 in its 12-month performance analysis. It reported 60.3% positive predictive value, 99.6% specificity, 39.3% all-cancer episode sensitivity, 69.8% sensitivity for 12 high-mortality cancers, and 91.3% cancer-signal-origin accuracy; 0.6% of 35,335 safety participants underwent an invasive procedure after a positive result (ASCO abstract). Those results support safety and clinical validity, but 39.3% overall sensitivity makes false reassurance and labeling central issues.

The 142,924-person NHS-Galleri randomized trial is more problematic. Its primary combined Stage III/IV endpoint failed: 706 cases in the intervention arm versus 688 in control, an incidence-rate ratio of 1.03. Stage IV disease fell from 397 to 342 cases, an incidence-rate ratio of 0.86, but that was a secondary endpoint with a confidence interval barely below 1.0 (ASCO abstract). This gives FDA a plausible benefit argument, but not the result expected from an uncomplicated pivotal success.

The submitted device is also an updated assay rather than the exact version used in the principal trials. GRAIL says its PMA contains a bridging analysis, but no public concordance or performance results are available (GRAIL). That bridge could be routine, or it could be the hidden issue that drives labeling changes or another information request.

My scenario model is:

The 35.0% timing estimate after a workable favorable outcome is below the 50% observed across Shield and Cologuard because Galleri is broader, its randomized primary endpoint failed, management still guides to 2027, and unresolved amendment or bridge work may remain. A narrower label and post-approval requirements can still produce approval, but they do not guarantee that negotiations finish by year-end.

What's non-obvious

The sponsor’s timing signal is more useful than the statutory-looking deadlines. Ofman reaffirmed the first-part-of-2027 expectation on August 5, and the September 23 panel date was publicly announced two days later; I infer that GRAIL probably already knew the approximate panel schedule when it answered the analyst’s question (transcript, announcement). That does not rule out a positive surprise in December, but it argues strongly against treating the 60-day aspiration as the base case.

There is also a quiet clock risk. GRAIL’s August filing says certain full PATHFINDER 2 and NHS-Galleri results presented after the January submission were included in the PMA, while MDUFA V permits substantial extensions for unsolicited major amendments submitted after substantive interaction; FDA-requested updates do not receive the same treatment (GRAIL 10-Q, MDUFA V commitment). The public record does not disclose how FDA classified those updates. This may explain why management expects 2027 despite the nominal calendar allowing a December decision.

Uncertainties

• FDA’s briefing package and voting questions were not public at the cutoff. They could reveal that the remaining questions are mostly labeling, which would raise the forecast, or that FDA doubts effectiveness or the assay bridge, which would lower it sharply (FDA event page).
• The assay-bridging results are not public, so outside observers cannot verify that the PMA version preserves the trial version’s specificity, sensitivity, and cancer-origin performance (GRAIL).
• The exact FDA filing date, accumulated sponsor-response days, amendment classification, manufacturing-inspection status, proposed final label, and post-approval-study negotiations are not public. Any one of these could move an otherwise favorable application past December 31 (FDA PMA process, MDUFA V commitment).