- Design
- Systematic review with modified GRADE certainty assessment
- Population
- 217 randomised trials in adults with episodic or chronic migraine
- Primary outcome
- Change in monthly headache days, ≥50% responder rate, quality of life
- Effect
- High confidence: galcanezumab, erenumab (episodic); fremanezumab, galcanezumab, onabotulinumtoxinA (chronic)
A systematic review of 217 randomised trials by the American Academy of Neurology and American Headache Society graded the certainty of evidence for every pharmacological migraine preventive in adults. It informs a companion guideline.
For episodic migraine, evidence was high confidence for galcanezumab and erenumab, and moderate for atogepant, eptinezumab, fremanezumab, propranolol, topiramate and valproate. Amitriptyline, bisoprolol, flunarizine, fluoxetine, levetiracetam, metoprolol, nifedipine, pizotifen and telmisartan had low-confidence evidence of possible benefit. For chronic migraine, evidence was high confidence for fremanezumab, galcanezumab and onabotulinumtoxinA, and moderate for atogepant, eptinezumab, erenumab, topiramate and valproate. Head-to-head evidence was sparse and low confidence.
The practical point is that propranolol and topiramate stand on moderate-confidence evidence, not far behind the newer agents, and cost much less. Amitriptyline and flunarizine, widely used in India, rest on weaker evidence than their popularity suggests. Valproate and topiramate need particular care in women who could become pregnant.
- Offer prevention when migraine is frequent or disabling, and track headache days with a diary
- Propranolol or topiramate are reasonable first choices on moderate-confidence evidence
- Avoid valproate in women of childbearing potential; counsel on topiramate
- Reserve CGRP-targeted drugs for failure or intolerance of oral agents where cost is a barrier
- Give each preventive 2 to 3 months at an adequate dose before judging it
Why it matters
Some of the most commonly prescribed Indian first-line preventives rest on only low-confidence evidence.
Don't overread it
Certainty grades describe confidence in benefit against placebo; they do not rank drugs against each other.
The statistics, in plain English
GRADE confidence is a judgement about how likely further research is to change an estimate. 'High' means very unlikely; 'low' means the true effect may be substantially different. A drug with low-confidence evidence may still work well; the trials are simply too small, too few or too biased to be sure.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for neurology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free