Two biotech companies — Karyopharm and Rezolute — two very different stumbles toward the same regulator, and an FDA that may be rewriting the rulebook on what counts as enough evidence to approve a drug.
Karyopharm Therapeutics on Monday reported topline results from the Phase 3 SENTRY trial of selinexor in combination with ruxolitinib for frontline myelofibrosis, according to a Karyopharm investor press release. The verdict was split: 50% of patients on the combination achieved a spleen volume reduction of at least 35% at week 24, compared to 28% on ruxolitinib alone — a statistically significant win on the first co-primary endpoint. Spleen shrinkage is clinically meaningful in myelofibrosis; an enlarged spleen causes pain, early satiety, and systemic symptoms that devastate quality of life.
But the second co-primary endpoint — symptom improvement as measured by absolute total symptom score — looked nearly identical in both arms (9.89 points versus 10.86 points), and the difference was not statistically significant. The drug hit the spleen target and missed the symptom one. Mixed is the accurate word.
There was an encouraging signal on overall survival: the hazard ratio was 0.43, with a nominal p-value of 0.0222 — just barely under the threshold, and with a wide confidence interval that spans 1.0 on one side. Karyopharm will keep following patients to see if the signal holds. On bone marrow fibrosis improvement, progression-free survival, and hemoglobin stabilization, there was no meaningful difference between arms as of the February 20 data cutoff. Safety was manageable but not clean: grade 3+ adverse events occurred in 70% of the combination arm versus 50% on ruxolitinib alone, and discontinuation rates were 15% versus 9%.
Karyopharm said it plans to meet with the FDA to discuss an sNDA filing. The company expects to present additional data at a medical conference and submit a manuscript to a peer-reviewed journal.
Meanwhile, Rezolute — which has no approved products and whose stock crashed in December when its sole pipeline candidate failed a Phase 3 trial — is also heading back to the FDA. The companys drug ersodetug missed the primary endpoint of its sunRIZE trial in congenital hyperinsulinism: reductions in hypoglycemia events were not statistically significant compared to placebo, which Rezolute attributes to a pronounced study effect — behavioral changes during the trial that confounded the measurement. CGM data showed consistent, clinically significant improvements, but the pre-specified Week 24 endpoint came up short.
And yet: the FDA did not dismiss the filing. At a Type B meeting on March 17, the agency acknowledged the challenges of measuring hypoglycemia in a heterogeneous pediatric population where self-monitored glucose is necessary for standard of care, according to Rezolute's FDA meeting outcomes. FDA encouraged Rezolute to submit the full dataset — including CGM endpoints and preliminary observations from the ongoing open-label extension — for independent agency review. A decision on whether a marketing application can proceed is expected in the second half of 2026.
These two cases arrive at FDA door at a moment when the agency itself is reconsidering what evidence it needs. In February, Commissioner Marty Makary and CBER Director Vinay Prasad published a paper in the New England Journal of Medicine arguing that the traditional two-pivotal-trial standard should become a default rather than a requirement — one adequate and well-controlled study, with confirmatory evidence, may be sufficient. The same month, FDA issued a draft guidance on a plausible mechanism framework for individualized therapies targeting rare genetic diseases, facilitating a path for single trials where randomized controls are not feasible, according to an Arnold & Porter analysis.
The policy shift is real. Whether it helps either of these specific drugs is a different question. Selinexor path to approval depends on whether FDA reads the OS signal as sufficiently compelling to offset a missed symptom endpoint and meaningful toxicity. Ersodetug path depends on whether FDA reviewers find the CGM data and mechanistic story convincing enough to overcome a missed primary endpoint in a rare disease with no approved therapies.
Myelofibrosis is a rare, progressive bone marrow cancer. Current standard of care includes ruxolitinib, fedratinib, and momelotinib — the latter approved in 2023 specifically for patients with anemia. A new drug entering this space needs a clear reason to exist. Selinexor case rests on the OS signal and the spleen volume result; the toxicity profile will matter to prescribers.
Congenital hyperinsulinism is rarer still — a genetic disorder causing dangerously low blood sugar in infants and children, with limited treatment options and significant risk of brain damage from recurrent hypoglycemia. Ersodetug mechanistic argument is clean: it is a monoclonal antibody that acts downstream from the pancreas, potentially effective across all forms of hyperinsulinism. If it works, it works broadly.
Both stories are ultimately about the same thing: a regulator signaling that the evidence bar is negotiable, and companies racing to test exactly how low it can go.
