5.1 Active Research: The Global Trial Landscape

Registries searched directly: ClinicalTrials.gov (API v2), EU CTIS and the legacy EU Clinical Trials Register/EudraCT, ANZCTR (Australia/NZ), Japan’s jRCT, China’s ChiCTR, and Korea’s CRIS. Japan’s UMIN-CTR could not be retrieved despite three attempts and is recorded as a genuine gap, not a zero. Unless flagged otherwise, every trial below carries the evidence tag preprint / trial (a registered but not-yet-peer-reviewed record); where a result has been published in a journal it is additionally tagged peer-reviewed.

5.1.1 The CGRP story: one win, four losses

This is the best-established recent trial programme in CH, and its pattern is the single most important fact in this chapter for a chronic-CH patient.

Drug	Trial	Population	Result	Source
Galcanezumab	NCT02397473	Episodic CH (n=109)	Positive. Weekly attacks −8.69 vs −5.22 placebo; LS mean diff −3.47 (95% CI −6.72 to −0.23), p=0.036	ClinicalTrials.gov
Galcanezumab	NCT02438826	Chronic CH (n=240)	Failed. −5.38 vs −4.59; LS mean diff −0.80 (95% CI −2.77 to 1.17), p=0.334	ClinicalTrials.gov
Fremanezumab	NCT02945046	Episodic CH (n=169)	Terminated for futility. p=0.9093 and p=0.1345 across doses	ClinicalTrials.gov
Fremanezumab	NCT02964338	Chronic CH (n=259)	Terminated for futility. p=0.2741 and p=0.3047	ClinicalTrials.gov
Eptinezumab (ALLEVIATE)	NCT04688775	Episodic CH (n=231)	Failed. −4.0 vs −4.6; mean diff 0.7 (95% CI −1.3 to 2.6), p=0.5048	ClinicalTrials.gov
Eptinezumab (CHRONICLE)	NCT05064397	Chronic CH (n=131)	Primary endpoint was safety, not efficacy — not a failed efficacy trial, just not an efficacy trial at all	ClinicalTrials.gov
Erenumab (CHERUB01)	NCT04970355	Chronic CH (n=101)	Failed — no results posted on the registry; published in JAMA Network Open 2025 peer-reviewed	ClinicalTrials.gov

Only galcanezumab works, and only in episodic CH. Every single chronic-CH CGRP antibody trial — across three different drugs — has failed or been abandoned for futility. This is widely discussed in headache-specialist circles as the field’s central unsolved puzzle for chronic sufferers specifically (see §5.3). (Editor’s note: full trial detail, the 2026 meta-analyses and the community’s divergent real-world experience are in Part V, §3.6.)

5.1.2 What’s genuinely new since the last pass (2024–2026 registrations)

Trial	Intervention	Population	Location	Status	Notes
NCT07714369	Zavegepant intranasal (gepant)	Episodic CH, acute	US (UTHealth Houston)	Not yet recruiting, est. start Oct 2026	First gepant tested as an acute CH abortive
CTIS 2025-521470-34-00 / 2025-521529-34-00	Candesartan (CandClus2), two Phase 3 trials	Episodic + chronic CH	Denmark, Norway	Authorised Nov 2025, not started	Largest new preventive-drug programme in CH; candesartan (an angiotensin-II blocker) was trialled once before, 2004
NCT06950281 / CTIS 2025-523293-17-00	Low-sodium oxybate (SUNCET, Xywav)	Chronic CH, nocturnal attacks	Netherlands (LUMC)	Authorised, not yet recruiting	Targets the nocturnal attack pattern directly — sleep-architecture, not clock-gene, mechanism
NCT05477459 / CTIS 2024-520305-39-00	LSD 25µg minidosing (CHIT)	Chronic CH (n=65)	Netherlands (Radboud UMC)	Recruiting	The live psychedelic CH trial right now; every-3-days dosing over 3 weeks
NCT06540651	Light therapy (Luminettes device)	Chronic CH (n=48)	France	Recruiting	The only interventional circadian-targeted CH trial found anywhere — see §5.1.5
NCT05264714	Rimegepant (gepant)	CH	US	Completed 31 Dec 2025	Results not yet posted
NCT07677137, NCT07113470	First-in-human neurostimulation devices (ONS; combined trigeminal+ONS “PRIMUS”)	Chronic CH	Belgium, Netherlands	Recruiting	Safety-phase device studies
NCT06787677, NCT06882278, ChiCTR2500106938	Sphenopalatine ganglion pulsed-radiofrequency / carotid sheath block	Episodic + chronic CH	China	Recruiting	A substantial Chinese procedural-treatment programme that is essentially invisible in Western headache literature
NCT07292090, NCT07348783	Yoga-based movement (YOURHEAD)	CH and/or migraine	Sweden	Active/Recruiting	Non-pharmacological, behavioural
NCT06917144	Repetitive TMS	Refractory chronic CH (n=8)	Spain	Completed	Results not posted

5.1.3 The psychedelics pipeline, in detail

This is the most active single therapeutic track in CH research right now, and it has become more crowded and more interesting since the last pass.

• Ceruvia Lifesciences — BOL-148/NYPRG-101. preprint / trial Announced an $8 million founder-backed investment on 3 August 2026, described by CEO Carey Turnbull as fully funding “an integrated Phase 1/2 study” — a single-ascending-dose Phase 1 in healthy adults feeding directly into a randomised, placebo-controlled Phase 2 proof-of-concept in CH patients, within one protocol (BioSpace). A German Clinical Trial Application is expected September 2026, first patient enrolment Q4 2026, Phase 1 top-line data Q2 2027, feeding a planned US+EU Phase 3. As of this writing there is no registry entry anywhere for the trial — consistent with a CTA not yet filed. The programme’s human precedent is a 2010 Hannover open-label case series in which three doses of BOL-148 over ten days broke cluster cycles or converted chronic to episodic disease in several patients, with remissions lasting months (Karst et al., Cephalalgia 2010 peer-reviewed).
• Correction to the record: BETR-001 is a different company’s different molecule. preprint / trial BETR-001 belongs to BetterLife Pharma Inc. (Vancouver), not Ceruvia — it is the patented (6R,9R) active stereoisomer of 2-bromo-LSD, explicitly migraine-led, with cluster headache framed only as an orphan-designated “de-risking beachhead.” Its FDA pre-IND meeting is complete and a US IND is anticipated Q1 2027 (BetterLife release, 13 Jul 2026). As of August 2026, two independent 2-bromo-LSD programmes are running in parallel — Ceruvia (CH-first, EU-first) and BetterLife (migraine-first, US-first) — which is a materially richer picture than a single programme.
• Yale/VA Connecticut (Dr Emmanuelle Schindler). preprint / trial No new CH-specific efficacy trial has been registered since the original psilocybin study (NCT02981173, completed). The one new 2024–2026 Schindler-affiliated registration, NCT06464367 (“Mechanistic Studies of Psilocybin in Headache Disorders”), is in migraine, not CH — but it is worth knowing about because it is the first psilocybin study to prospectively instrument circadian rhythm (actigraphy) and sleep EEG as candidate mediators of the durable post-pulse benefit, i.e. it is explicitly testing a chronobiological mechanism for psychedelic efficacy in headache. The published follow-up to the original pulse-regimen extension paper (May 2024) has so far been in migraine, not CH (PubMed sweep).
• Clusterbusters/Yale DMT citizen-science survey — interim results now exist. citizen science Launched 2 August 2025 on the Clusterbusters forum, explicitly “in the tradition of citizen science,” in collaboration with Yale (Clusterbusters forum). An interim analysis — Schindler E, Lenaburg K, Wold R, “DMT use in Cluster Headache: Interim Analysis of an International Survey” — was published as a conference abstract in Neurology 2026;106, DOI 10.1212/wnl.0000000000215894 preprint / trial (online 10 June 2026, consistent with an AAN 2026 Annual Meeting presentation). No sample size or numerical results are publicly rendered on the abstract page. Notably, co-author Mr Lenaburg discloses being Clusterbusters’ Policy Director and Mr Wold discloses being Clusterbusters’ Executive Director — this is advocacy embedded directly in the authorship, not merely acknowledged.

5.1.4 PACAP: the strongest un-trialled target in headache medicine

peer-reviewed The single most striking finding of this research pass is a hard negative with a strong positive biology sitting right next to it. As of 13 August 2026, no PACAP-targeting drug has a cluster-headache trial registered, planned, or publicly discussed by any sponsor, anywhere — this was checked directly against ClinicalTrials.gov, EU CTIS, Lundbeck’s global pipeline page, Lundbeck’s own AHS June 2026 investor deck (which calls bocunebart “phase 3 ready” and mentions only migraine), Eli Lilly’s LY3451838 programme (which appears dormant, two completed non-CH trials only), and the newly-launched Slate Medicines ($130M Series A, February 2026, for SLTE-1009 — also migraine-framed).

Set against that: interictal plasma PACAP-38 is elevated 34.3% overall in CH patients versus controls, and 49.8% in chronic CH specifically — the largest elevation of any CH subgroup studied — in a 205-patient Danish case-control study, partly funded by Lundbeck itself (Søborg et al., Eur J Neurol 2025 peer-reviewed). PACAP-38 infusion provokes cluster-like attacks in roughly 45% of CH patients in an active phase, without changes to CGRP, tryptase, or histamine — implying a genuinely distinct mechanism from the CGRP pathway that has now failed four times in chronic CH. The genetic case is weaker: the original ADCYAP1R1 association (from a small 99-patient Italian cohort) did not replicate in a larger 542-patient Swedish cohort, and ADCYAP1R1 is not among the loci found in the large 2023 international GWAS (§5.2.1).

Bottom line for a chronic-CH patient: PACAP is arguably the best-evidenced next drug target for CH — better evidenced pharmacologically than CGRP ever was for chronic disease — but nothing is in the clinic for CH specifically, and the earliest a CH-specific trial could plausibly start is after Lundbeck’s migraine Phase 3 programme reads out, with no announced timeline for that indication expansion.

5.1.5 Orexin antagonists and circadian-targeted drugs — the direct answer to a specific question

This chapter was originally asked to check specifically for orexin-pathway and circadian-system drugs. The direct answer:

• Orexin: a clean negative. peer-reviewed Registry sweeps of every dual/selective orexin antagonist (filanapant/vornorexant, seltorexant, suvorexant, lemborexant, daridorexant) against every headache condition returned zero cluster-headache trials and zero headache-indicated trials of any kind. The only orexin-antagonist trial ever conducted in any headache condition is Merck’s filorexant pilot in migraine (2015): 120 vs 115 patients, no significant benefit on monthly migraine days (difference −0.4, 95% CI −1.3 to 0.4), more somnolence on drug (13% vs 4%), and an explicitly negative conclusion (Chabi et al., Cephalalgia 2015). That trial is now 11 years old and nobody has re-tested the mechanism in CH, despite a genuinely plausible anatomical rationale (orexinergic neurons project to the periaqueductal grey, trigeminal dorsal horn, locus coeruleus, thalamus and spinal cord). The genetic signal for HCRTR2 (the orexin-2 receptor gene) has been reported in some small studies and not others, is explicitly called “unclear” in the definitive 2024 review (Stanyer, Hoffmann & Holland, Expert Rev Neurother 2024 peer-reviewed), and does not appear among the loci in the large 2023 GWAS at all. This is a genuine, unexploited gap in the field rather than a settled negative — the biology has never actually been tested in the right patients.
• Circadian drugs: thin but not empty. preprint / trial Two live trials genuinely target the circadian/nocturnal attack pattern: NCT06540651, a recruiting light-therapy trial in France explicitly framed around cluster headache being “a chronobiological disease” whose “aim is to re-adjust chronobiological rhythms,” and NCT06950281 (SUNCET), a Phase 2 low-sodium-oxybate trial targeting nocturnal chronic-CH attacks specifically. Neither is a clock-gene drug — light therapy targets circadian entrainment behaviourally, and oxybate targets sleep architecture pharmacologically. No melatonin trial and no clock-gene-modulating compound (REV-ERB agonist, CRY stabiliser, casein-kinase inhibitor, or similar) exists in any stage of headache drug development. The melatonin evidence base for CH remains two small, contradictory pilots from 1996 (positive) and 2002 (negative), described in a 2025 review as “not validated in high-quality clinical trials” (Burish et al., Cephalalgia 2025 peer-reviewed).

5.1.6 Regional registry coverage — a note relevant to an Australian patient

preprint / trial There are zero cluster-headache-specific trials natively registered in ANZCTR as of August 2026 — a broad health-condition search on “headache” returned 62 records and none was CH. The only CH trial physically running in Australia is the ClinicalTrials.gov-registered NCT05868044 PRIMUS device study (n=5, active, not recruiting). Australia’s one genuinely new CH research initiative — the MRFF-funded psilocybin pilot (“PEACE”) — is not yet in any registry either; recruitment had not started as of its September 2025 funding announcement (see §5.4.2 for the patient-advocacy chain that produced it).

Asian coverage: Japan’s jRCT lists exactly one substantive CH trial (the Japanese arm of eptinezumab’s ALLEVIATE trial); China’s ChiCTR lists three records including two national CH registries and one interventional carotid-sheath-block trial; Korea’s CRIS lists zero CH trials (confirmed as a real negative, not a broken search, because control searches on “headache” returned non-zero results).

Diagram: selected trial timeline