3.3 Topiramate

3.3.1 Status

Third-line/tertiary option. Evidence level + = “low evidence”; “a possible alternative when verapamil or lithium is ineffective or not tolerated”; data “only from open observational studies” (PMC8748342). peer-reviewed There has never been a randomised placebo-controlled trial of topiramate in cluster headache.

3.3.2 Efficacy — and a big internal inconsistency in the literature

Leone M, Dodick D, Rigamonti A, et al. open trial, Cephalalgia 2003, DOI 10.1046/j.1468-2982.2003.00665.x. peer-reviewed n=33 analysed: only 21% (7/33) achieved >50% reduction — 26% of episodic, 10% of chronic. Seven patients took ≥200 mg/day for 15 days with zero improvement. p=0.3.

Other open series (PMC8748342): peer-reviewed

Study	n	Dose	Result
Initial open-label	10	not stated	Positive effect within 3 weeks
Open, consecutive	36 (26 ep, 10 chr)	100–150 mg/day	7/33 >50% reduction
Open	12	not stated	9/12 good or moderate effect
Prospective, Spain	26 (12 ep, 14 chr)	max 200 mg	15/26 remission of cluster period; 6/26 >50% reduction; mean time to remission 2 weeks

These results do not agree with each other. 7/33 (21%) vs 9/12 (75%) vs 15/26 remission (58%) in studies of the same drug at similar doses. No adequate explanation exists; the differences are most likely small-sample noise, differing outcome definitions, and open-label expectation effects. Contested.

2026 update

Shafiyev J, Gahramanov I. The Evaluation of Topiramate as a First-Line Medication for Cluster Headache Prophylaxis: A Prospective Observational Cohort Study. Clin Neuropharmacol 2026 Feb 23, online ahead of print, DOI 10.1097/WNF.0000000000000678, PMID 41805259. peer-reviewed Prospective observational, Central Military Hospital, Baku, Azerbaijan, Oct 2023–Dec 2024. n=51 adults (64.7% male, mean age 38.6), episodic and chronic, topiramate monotherapy 25 mg/day titrated weekly to max 100 mg. Significant reductions in VAS and HIT-6 over time in all patients. Chronic patients started with higher HIT-6 and improved more slowly than episodic. Somnolence 19.6%; no discontinuations.

Be sceptical of this one. The title frames topiramate as first-line, which no guideline supports. The study is observational with no control group, no randomisation, no blinding, and — critically — reports no attack-frequency outcome at all, only VAS and HIT-6, with no exact values, no p-values and no confidence intervals published in the abstract. For a condition where placebo response in prophylaxis trials is well documented (Nilsson Remahl AIM, Laudon Meyer E, Cordonnier C, Goadsby PJ. Placebo Response in Cluster Headache Trials: A Review, Cephalalgia 2003), an uncontrolled cohort showing “significant improvement over time” is close to uninformative. It is included for completeness, not because it changes anything.

3.3.3 Dosing and side effects

Start 25 mg/day, increase by 25 mg per week (slow titration explicitly reduces side-effect rate), target 100–150 mg/day, up to 200 mg (PMC8748342; Leone 2003). peer-reviewed

Adverse events: cognitive dysfunction, fatigue, dizziness, paraesthesia, mood swings, anxiety, weight loss, hair loss. Contraindications: kidney stones, glaucoma, hypercalcaemia, dose adjustment in renal impairment, pregnancy (PMC8748342).

One warning deserves to be pulled out of the table. That review states topiramate can cause mood swings and depression, and explicitly flags this as relevant “because of the increased risk of suicide in patients with cluster headache.” peer-reviewed Given CH’s grim nickname and documented suicidality, prescribing a mood-destabilising drug to this population is a real trade-off, not a footnote. If you try topiramate, have someone who will notice mood change before you do.

The cognitive effects (“dopamax” in community parlance) are the most common reason for discontinuation in practice.

3.3.4 Community view

Topiramate is not among the drugs the Rusanen synthesis singles out for either high or low self-reported efficacy — it is neither in the high-efficacy clade (LSD, psilocybin, ergoline alkaloids, methysergide, verapamil, corticosteroids) nor explicitly in the low clade (melatonin, valproic acid, gabapentin, amitriptyline, propranolol) (PMC9436841). citizen science The survey paper describes topiramate and melatonin as “usually recommended as tertiary options.” Community discussion is dominated by the cognitive side effects rather than by efficacy claims in either direction. community report