Gepant Selection: One Agent or Two for Acute and Preventive Migraine Care
A man with established coronary disease has never had an acute migraine treatment that actually works, because the standard acute class is contraindicated for him. Four gepants exist and were not all studied for the same job — the question is whether one drug can honestly do both, or whether asking it to is a real compromise.
R.D. keeps the stent card from two years ago in his wallet, tucked behind his insurance card, a habit he picked up without really deciding to after the cardiologist told him to carry it in case he ever needed emergency care somewhere without his records. He is 52, a warehouse operations manager, and has had migraine with aura since his twenties — manageable for most of that time with over-the-counter ibuprofen and a dark room, until a drug-eluting stent went into his left anterior descending artery after an anterior STEMI. Nobody has been comfortable prescribing him a triptan since. His migraines have also, independently, gotten worse: nine days a month now, missing at least one shift monthly, occasionally driving into work anyway on a headache he shouldn’t have driven through. He has also let his amlodipine refill lapse twice in the past year, on a drug he has taken for years without complaint — ordinary attrition rather than defiance, and a fact that will matter once the question becomes how many prescriptions he is asked to keep alive at once.
Triptan avoidance in established coronary disease isn’t a formality — 5-HT1B agonism produces real coronary vasoconstriction, and the class carries a genuine contraindication in documented ischemic heart disease, his exactly. That leaves a real gap: no acute therapy that works for him, and no preventive at all. Gepants close both ends of it by antagonizing the CGRP receptor instead, without the vascular effects that rule out triptans for him.
Four gepants exist and they do not do the same job: ubrogepant and zavegepant are acute-only, atogepant is preventive-only, and rimegepant alone carries both indications, with trial data behind an as-needed dose and a scheduled every-other-day one from the same tablet. His nine migraine days a month sit awkwardly against the division between them. Nine is episodic migraine by definition, which puts him inside the enrolled population of both trials in question — ADVANCE tested atogepant in patients with four to fourteen migraine days a month, and rimegepant’s scheduled-dosing trial covered a comparable band — so neither drug is being extrapolated onto him. But nine is also the frequency at which headache clinics conventionally stop treating attacks one at a time and start treating the pattern. The trials describe him. Which side of his own threshold he sits on they do not address, and neither does either label — that judgment has never been made for him by anyone, including him.
First real acute and preventive plan since the stent
I’d build this as two purpose-built agents, not one dual-purpose one. Atogepant has its own dedicated preventive trial, ADVANCE, with a real, clinically meaningful drop in monthly migraine days against placebo — that’s a different evidence base than a drug repurposed for prevention on top of an acute indication. And for the acute side, zavegepant’s intranasal route has the fastest onset of any gepant, which matters directly for a man who needs to keep driving a shift, not just eventually feel better.
I don’t disagree that ADVANCE is the stronger trial on paper. But look at his own chart before defaulting to the two-agent plan — two documented lapses on amlodipine refills in the past year, on a drug he’s taken for years. That’s not a hypothetical adherence concern, it’s a demonstrated pattern, and it’s exactly the kind of patient where one prescription, one refill cycle, matters more than the marginal trial-data advantage.
Rimegepant’s BHV3000-305 data support both the PRN and the scheduled every-other-day regimen from the same tablet. It won’t match atogepant’s numbers in a head-to-head that’s never actually been run, but it’s real support for one drug covering both roles.
You’re both right about different things, and I don’t think we need to decide the whole regimen today. He has an acute problem right now that neither of you disputes — start rimegepant PRN for that alone.
What makes this safe to defer rather than decide up front is something neither trial argument turns on: gepants haven’t shown the medication-overuse risk triptans and opioids carry. That’s not a hopeful extrapolation from mechanism — rimegepant’s own 52-week open-label safety study tracked frequent as-needed use without the escalation in headache frequency that defines the problem, and Lipton’s 2026 review treats absence of medication-overuse association as an established class property. So how often he actually reaches for it over the next two months is real, usable data, not a danger sign the way it would be with a triptan. If he’s using it more than eight or ten days a month, that itself tells us he needs scheduled prevention, and we decide between atogepant and continuing rimegepant then, with his own pattern in hand instead of guessing at it now.
Agreed: rimegepant started PRN today for his acute migraines, with explicit tracking of days-used-per-month; a two-month follow-up set to review that pattern directly rather than estimate it.
That frequency itself becomes the basis for adding scheduled prevention — the Headache Medicine Specialist would move to atogepant at that point; the Clinical Pharmacologist would first try scheduling rimegepant's own every-other-day regimen before adding a second agent, not yet resolved between them.
The Headache Medicine Specialist wants it available from day one for a mid-shift attack rather than waiting for rimegepant to prove too slow; the Primary Care Physician preferred not adding a second acute agent before there's real data he needs it. Left open.