Will FDA approve mitapivat for sickle-cell disease by Nov. 1?

TL;DR

I assign a 75% chance that FDA approves mitapivat for at least some sickle-cell patients by the end of November 1, 2026. FDA had seen the mixed Phase 2 and Phase 3 results before steering Agios toward accelerated approval, and the subsequently agreed confirmatory trial is now recruiting (Agios pre-sNDA update; ClinicalTrials.gov). The main threats are the failure to prove fewer pain crises or less fatigue and a late labeling or REMS issue that pushes an otherwise favorable decision beyond the hard deadline (RISE UP results; prior mitapivat delay).

Context

Agios submitted the sickle-cell sNDA on May 12, 2026. FDA accepted it for Priority Review on July 7, 2026, with accelerated approval requested and a November 1 goal date. The resolution is forgiving on label scope: approval for any sickle-cell subgroup counts.

Mitapivat is already approved for PK deficiency and thalassemia, so FDA knows the molecule, manufacturing system, dose and general safety profile (FDA thalassemia approval). That lowers basic product and CMC risk, but the 100-mg twice-daily dose carries a boxed warning and restricted REMS for hepatocellular injury in thalassemia (FDA label).

Evidence

The historical backbone is favorable. FDA's September 2026 longitudinal report found a first-cycle Complete Response rate of only 7% among 272 NDA/BLA efficacy-supplement filings in FY2024, giving a loose 93% no-CR starting anchor (FDA report, FY2008–FY2024). FDA also acted by the goal date on 86 of 90 priority efficacy supplements in the FY2024 cohort, or 96% (FY2025 PDUFA performance report). These are broad reference classes. They include many routine supplements with cleaner evidence than this application.

The product-specific regulatory sequence is the strongest positive evidence. After Agios presented both RISE UP phases, it reported on March 31, 2026 that FDA recommended a confirmatory-trial proposal to support accelerated approval. Agios then reported agreement with FDA on that trial before filing on May 12. These are sponsor accounts rather than public FDA meeting minutes, but they indicate that FDA did not treat the pain-crisis miss as an automatic bar after seeing the data.

RISE UP Phase 3 randomized 207 patients aged at least 16 years for 52 weeks, 138 to mitapivat and 69 to placebo. Hemoglobin response occurred in 40.6% versus 2.9%, with p<0.0001; mean hemoglobin improved by 0.769 versus 0.026 g/dL, and indirect bilirubin also improved strongly (November 19, 2025 results). The signal replicated in the 79-patient Phase 2 portion: response at the selected dose was 50.0% versus 3.7% on placebo (peer-reviewed Phase 2 report). The pivotal study was not designed to require both primary endpoints to win; its prespecified framework allowed success through either hemoglobin response or pain crises (RISE UP design).

The direct clinical evidence is weaker. Annualized pain crises were 2.62 with mitapivat versus 3.05 with placebo, a nonsignificant 14% reduction with p=0.1213. PROMIS Fatigue was effectively null, at −2.72 versus −2.25 with p=0.7112 (RISE UP results). Later analyses found fewer transfused patients, 23.9% versus 40.6%, and 0.70 versus 1.59 red-cell units per randomized patient (EHA 2026 analysis). That is clinically supportive, but transfusion was not one of the trial's multiplicity-controlled primary or key secondary endpoints.

FDA policy supports the filing theory. Its December 2024 draft guidance says an increase in hemoglobin greater than 1 g/dL after 24 weeks has been considered reasonably likely to predict improvement in how sickle-cell patients feel or function (FDA accelerated-approval guidance). FDA's current surrogate table also lists hemoglobin response for accelerated approval in sickle-cell disease, although the listed mechanism is an HbS polymerization inhibitor rather than a PK activator (FDA surrogate table). Surrogate acceptance remains product-specific.

Voxelotor is the central warning. FDA accelerated-approved it in 2019 after a 51.1% versus 6.5% hemoglobin-response result (FDA voxelotor approval). It was withdrawn in September 2024 after postmarketing studies showed more vaso-occlusive crises and deaths among treated patients (FDA withdrawal alert). Mitapivat has a different mechanism, and its randomized crisis rate moved in the favorable direction, but Oxbryta gives FDA a concrete reason not to equate higher hemoglobin automatically with net benefit.

A weak negative is the same-class comparator etavopivat. In April 2026, Novo Nordisk reported that it met both Phase 3 co-primary endpoints, including a 27% reduction in vaso-occlusive crises and a 48.7% versus 7.2% hemoglobin response (Novo Nordisk results). Etavopivat was not approved by the cutoff, so it does not remove the current unmet need. It does make mitapivat's pain-crisis miss harder to dismiss as an unavoidable feature of the drug class.

The SCD safety data were broadly balanced: serious adverse events occurred in 20.3% of mitapivat patients and 29.0% of placebo patients, and the trial showed no liver pattern suggestive of the hepatocellular injury seen in thalassemia (RISE UP safety results). Still, FDA previously extended mitapivat's thalassemia review by three months for a requested REMS, then approved it 16 days after the extended goal date while labeling and REMS work continued (extension disclosure; December 8 update; December 23 approval). The existing operational REMS makes a repeat delay less likely, not impossible.

My model assigns 82% to approval in the current review cycle and about 92% to approval occurring by the strict deadline conditional on an otherwise favorable review. The internal calculation is 0.820 × 0.915 = 0.7503. The substantive estimate sits below the broad efficacy-supplement base rate because of the failed clinical endpoints, Oxbryta and liver risk; the timing estimate sits below FDA's 96% reference rate because of mitapivat's own recent REMS-related delay.

What's non-obvious

The obvious read is that mitapivat failed one of two co-primary endpoints and therefore has a weak application. That misses the study design and FDA's subsequent conduct. The trial was built to succeed through either primary endpoint, FDA's draft guidance still recognizes the exact type of hemoglobin response observed, and FDA reportedly chose the accelerated-approval route after reviewing the mixed results (trial design; FDA guidance; pre-sNDA interaction).

The opposite mistake is to apply the 93% efficacy-supplement base rate almost mechanically. This application asks FDA to accept a surrogate after the most prominent prior use of that surrogate ended in withdrawal. The strict deadline also makes the same drug's prior labeling and REMS delay more useful than the general approval base rate (FDA Oxbryta alert; prior mitapivat review delay).

Uncertainties

• FDA's private view of whether hemoglobin response is adequate for a PK activator specifically is unknown. Public guidance is favorable but nonbinding and case-specific (FDA guidance; surrogate table).
• The status of labeling, REMS amendments, manufacturing review, inspections and any major amendment is confidential. FDA can remain on track, extend the date or miss it without those details becoming public well in advance (FDA PDUFA communication process).
• The full Phase 3 dataset remains sponsor-reported rather than contained in a peer-reviewed paper or public FDA review package (November 2025 results; June 2026 presentation).
• REIGNITE was recruiting with a target of 159 participants, a July 21, 2026 start and an August 4 registry update, but the public record does not show FDA's private assessment of enrollment pace or final postmarketing milestones (ClinicalTrials.gov).