Indoleamine Hallucinogens in Cluster Headache: Results of the Clusterbusters Medication Use Survey

The investigators conducted an anonymous online survey modelled after previously published instruments. Recruitment targeted cluster headache communities and clinics: invitation emails were sent to clusterbusters.org registrants in May 2012 and the survey was advertised on clusterheadache.com, relevant Facebook pages and headache clinic listings. Participation was voluntary and no personal identifying information was collected. The questionnaire contained 41 items covering demographics, headache characteristics, substance use, and medications; most responses used fixed-choice lists while several items allowed free-text elaboration. Responders rated medication effectiveness on a four-tier scale (not effective, partially effective, moderately effective, completely effective), though these levels were not rigidly defined. Of 651 respondents, 558 completed the survey; further analysis was restricted to 496 participants who indicated their diagnosis had been verified by a neurologist or headache specialist. Demographic details reported include sex (366 men, 73.8%; 130 women, 26.2%) and other characteristics referenced but not fully presented in the extracted text. Analysis was primarily descriptive. For group comparisons the researchers combined the lower two efficacy categories (not effective + partially effective) and compared them with the higher two (moderately effective + completely effective) using Fisher’s exact test; GraphPad software was used for statistical testing. Free-text answers on dosing, regimens and effects were categorised and reported narratively. The extracted text does not report any ethical review statements or response-rate calculations beyond raw counts.

The final analysed sample comprised 496 respondents with clinician-verified diagnoses. Abortive treatments: subcutaneous triptan injection was rated most effective in comparing categories and was statistically superior to high-flow oxygen, oral and intranasal triptans, and psilocybin in the pooled efficacy comparison (reported p-values all < 0.01). High-flow oxygen was more effective than oral and intranasal triptans (p < 0.0001) but did not differ from psilocybin (p > 0.4). Psilocybin was rated significantly more effective than oral and intranasal triptans (p < 0.0001). Ergot derivatives (cafergot/ergotamine and intravenous DHE) were significantly less effective than high-flow oxygen, triptan injection and psilocybin in these comparisons. Preventive treatments: the most commonly used preventatives included verapamil, prednisone, melatonin, topiramate and psilocybin. Prednisone showed greater pooled efficacy than verapamil (p < 0.008). Psilocybin, LSD and lysergic acid amide (LSA) were reported to be significantly more effective than verapamil, prednisone and the non-hallucinogenic ergot methysergide (p-values ranging from < 0.0001 to < 0.03 depending on the comparison). Direct comparisons among indoleamines indicated that psilocybin was statistically similar to LSD and to the non-hallucinogenic congener BOL, but more effective than LSA and DMT; LSD was similar to BOL and superior to LSA and DMT. BOL (n = 10 responders) produced at least moderate protection in 60% of those who tried it. Effects on cluster periods and remission: in free-text responses about agents that shortened or aborted a cluster period (199 responders, 264 responses) psilocybin was the most frequently cited (67 responses, 33.7%) and was identified significantly more often than any other response (p < 0.001). For medications reported by respondents to have induced remission from chronic cluster headache (60 responders, 80 responses), psilocybin was again the most commonly listed (18 responses, 30.0%) and more frequently cited than verapamil, steroids, LSD and LSA (p-values < 0.05 for these comparisons). Dosing and regimens: verapamil doses reported ranged from 120 to 1020 mg daily (n = 84), with 480 mg daily most commonly reported (n = 18) and about one-third of verapamil reports exceeding 480 mg. Psilocybin doses reported as dried mushrooms for abortive use ranged from 0.1 to 5 g (n = 14) and for prevention 0.1 to 6 g (n = 57); some doses were below typical recreational ranges. Reported LSD doses for aborting were 150–200 µg (n = 2) and for prevention 100–300 µg (n = 8). LSA dosing was reported as numbers of seeds (wide ranges depending on plant source). Use frequency for hallucinogens varied: some used them daily to weekly for abortive purposes (n = 23) whereas prevention regimens were typically infrequent, from every few weeks to twice yearly (n = 80); eight responders explicitly wrote that a single dose was used. Adverse effects: conventional medications yielded a range of adverse reports in free-text entries (e.g. rebound headaches with sumatriptan/ergot/steroids, nausea with opioids, cardiac issues with verapamil, steroid-induced mania and avascular necrosis in one case). Indoleamine hallucinogens attracted relatively few negative reports: one person attributed headache to LSD, and a few reported transient sickness, wooziness or abdominal discomfort with LSA. The extracted text notes that side-effects were not systematically solicited and fewer than 30 respondents mentioned adverse effects in free-text sections.

Schindler and colleagues interpret their findings as consistent with earlier reports that indoleamine hallucinogens—particularly psilocybin and LSD—are perceived by many patients to be effective for aborting attacks, preventing attacks and in some cases shortening cluster periods or inducing remission. The survey reproduced expected findings for conventional therapies: high-flow oxygen and subcutaneous triptans were commonly used and found effective for acute attacks, and verapamil and steroids showed partial preventative benefit similar to prior surveys. Psilocybin was tried as an abortive by roughly one-third of respondents who attempted it, with about two-thirds reporting at least moderate benefit and one-third reporting complete benefit; at pooled levels of efficacy psilocybin compared favourably to high-flow oxygen and exceeded the efficacy of oral and intranasal triptans, though injectable triptan remained superior for very rapid abortive effect. The authors highlight that preventative use of psilocybin and LSD was reported to provide at least moderate protection in over 70% of those who tried them, with complete preventative response in roughly 40%, percentages larger than those reported for conventional preventatives in this survey. The non-hallucinogenic LSD congener BOL also performed well in the limited number of respondents who used it, prompting the suggestion that hallucinogenesis per se may not be necessary for therapeutic benefit. In addition, many respondents described infrequent dosing schedules or single-dose effects, distinguishing these regimens from daily dosing required by conventional preventatives. The paper acknowledges numerous limitations: online recruitment introduces selection bias towards Internet users and those aware of or interested in alternative therapies; recall bias is likely given variable illness durations; diagnostic validity is imperfect despite restricting analysis to those who reported specialist verification; some reported attack frequencies exceeded diagnostic criteria limits, raising the possibility of misclassification; the four efficacy levels were not standardised, allowing subjective interpretation; comparisons are confounded because many respondents tried multiple treatments and samples were not independent; and dosing, formulation and purity of illicit substances could not be verified. The authors conclude that controlled experimental studies are required to establish efficacy and safety under standardised conditions and note that non-hallucinogenic congeners such as BOL could permit exploration of pharmacological effects independent of subjective hallucinogenesis.

The Clusterbusters medication use survey adds observational support that indoleamine hallucinogens—psilocybin, LSD and LSA—are rated by many people with cluster headache as effective for abortive and preventative purposes, and that these agents were reported to shorten cluster periods or induce remission in some cases. Respondents often described benefit from modest and infrequent dosing, and a non-hallucinogenic congener (BOL) showed similar preventative effects in a small number of users. The authors recommend controlled clinical studies to establish the effects and safety of these compounds in cluster headache, with BOL proposed as a candidate to dissociate potential therapeutic mechanisms from hallucinogenesis.