Takeda has terminated ENDYMION 2 (NCT05163314), the open-label Phase 3 extension that was to follow patients on its epilepsy candidate soticlestat (TAK-935) over the long term. The registry now lists the study as terminated, with the sponsor stating the closure was “due to negative results from phase 3 SKYLINE and SKYWAY studies and unrelated to patient safety.”
The extension had enrolled 352 participants with Dravet syndrome or Lennox-Gastaut syndrome since March 2022, gathering long-term safety and tolerability data. Its fate was decided upstream, by the two pivotal trials feeding it.
The pivotal misses
SKYLINE, the Dravet trial (n=144, ages 2–21), measured percent change from baseline in convulsive seizure frequency per 28 days as its primary endpoint. Per Takeda’s topline release, soticlestat missed that endpoint, with a reported p-value of 0.06 — a narrow miss of statistical significance. The company noted nominally significant results on secondary measures, including responder rate and global impression of improvement (p ≤ 0.008), but these do not rescue a failed primary.
The extension closure was “unrelated to patient safety” — it followed the efficacy failures, not a tolerability signal.
SKYWAY, the Lennox-Gastaut trial (n=270, ages 2–55), measured percent change in Major Motor Drop seizure frequency per 28 days as its primary endpoint. That endpoint was also missed.
Takeda said it “will engage with regulatory authorities to discuss the totality of the data generated by these studies to determine next steps.” With both Phase 3 primaries failed and the safety extension now closed, soticlestat’s path in these treatment-resistant childhood epilepsies is, at minimum, sharply narrowed. The secondary and exploratory signals are hypothesis-generating, not confirmatory.