AHS 2026: Cluster headache remains a CGRP challenge—but this migraine drug shows promise
Cluster headache has proven a tough nut to crack for CGRP-targeted therapies. Now, a small pilot study presented at the American Headache Society's 68th Annual Scientific Meeting suggests a familiar migraine drug may have a role to play, offering an early signal that warrants a closer look.
A first look at a gepant in cluster headache
Galcanezumab earned an FDA nod for episodic cluster prevention, but trials of other CGRP-targeting medications have shown mixed results, and no gepant has been tested prospectively in cluster until now. So any new mechanism getting a toehold here draws attention, even from a very small dataset.
Investigators presented a prospective open-label pilot of rimegepant 150 mg in patients meeting ICHD-3 criteria for episodic or chronic cluster headache.
The trial enrolled 20 patients; 11 (nine episodic, two chronic) provided complete treatment data. The design was hybrid: patients used a single dose at onset of a severe attack for acute assessment, monitoring headache severity every 15 minutes, then took 150 mg every other morning for eight days (four doses total) while headache frequency, severity, and rescue medication use were tracked against a week-long baseline.
The headline numbers: Mean headache frequency dropped by 4.0 attacks per week from baseline to the treatment period, with average attack duration shortened by 14.9 minutes. Out of the 11 patients, three achieved greater than 50% reduction in attack frequency; a fourth had a 40% reduction.
On the acute side, six of 13 patients who tried the first dose improved by at least one point on a 5-point Likert scale within 30 minutes. Patient satisfaction skewed positive—eight of 13 were "very" or "somewhat" satisfied, and nine of 13 said they would want to use rimegepant again. Safety was unremarkable aside from one patient with asymptomatic liver-function elevations above three times the upper limit of normal.
Signal detected, but questions remain
What can this tell us? Honestly, not much beyond feasibility and a signal worth chasing. There is no placebo arm, no blinding, and an N of 11 completers—a setup where expectation bias, natural cluster-cycle waning, and regression to the mean all have room to operate. The every-other-day dosing strategy is novel and pragmatic, but its independent contribution to the frequency reduction is impossible to parse here.
What the pilot does establish is that a randomized, placebo-controlled trial is logistically doable in this population and that the safety profile doesn't raise immediate red flags. For now, rimegepant remains a migraine drug—but the CGRP story in cluster just got a new, if very early, chapter.