4.2 Sphenopalatine ganglion (SPG) stimulation — Pulsante / Pathway CH-1 and CH-2

4.2.1 What it is

A miniature implanted neurostimulator placed via a trans-oral, gingival-buccal approach into the pterygopalatine fossa, with a lead abutting the sphenopalatine ganglion. It has no battery: the patient holds a hand-held remote controller against the cheek, which powers the implant by radiofrequency and delivers on-demand stimulation during an attack. peer-reviewed The SPG was targeted because of its role in the trigemino-autonomic reflex producing lacrimation, nasal congestion and rhinorrhoea; high-frequency stimulation is thought to block parasympathetic outflow (PMC8665918, 2021).

historical SPG blockade (not stimulation) is much older — “SPG blockades had been used to terminate cluster headache since the early 1900s”, though they were “clinically difficult to perform” and never widely adopted. Devoghel reported 120 blocks with an 85% response rate; Pipolo et al. reported 15 endoscopic SPG blocks with 54% complete remission lasting 1–28 months (Trigeminal Autonomic Cephalalgias: Beyond the Conventional Treatments, Curr Pain Headache Rep 2014;18(8):438 / PMC4119587).

4.2.2 Pathway CH-1 — the pivotal European trial

peer-reviewed Schoenen J, Jensen RH, Lantéri-Minet M, Láinez MJA, Gaul C, Goodman AM, Caparso A, May A. “Stimulation of the sphenopalatine ganglion (SPG) for cluster headache treatment. Pathway CH-1: a randomized, sham-controlled study.” Cephalalgia 2013;33(10):816-830, DOI 10.1177/0333102412473667, PMID 23314784 / PMC3724276, NCT01255813. Funded by Autonomic Technologies Inc (ATI).

• 32 patients with refractory chronic CH implanted; 28 completed the experimental period. Each attack randomised to full, sub-perception, or sham stimulation.

Outcome at 15 min	Full stimulation	Sham	p
Pain relief	67.1%	7.4%	<0.0001
Pain freedom	34.1%	1.5%	<0.0001
Sub-perception stimulation (pain relief)	7.3%	—	0.96 vs sham

• Acute responders: 9/28 (32%) — 9 of 14 eligible patients, 64%.
• Frequency responders: 12/28 (43%), with a mean 88% reduction in attack frequency.
• Therapeutic responders (acute and/or frequency): 19/28 (68%).
• Rescue medication use 31.0% (stimulation) vs 77.4% (sham).

The sub-perception arm is an underappreciated strength of this trial: it provides an active-but-imperceptible control, which is very hard to achieve in neuromodulation and largely rules out a pure paraesthesia-expectation effect.

peer-reviewed CH-1 safety — and this is where the honesty is required:

• Sensory disturbance in 26/32 patients (81%); localised loss of sensation in 19/32.
• Resolution averaged 82–97 days (range 12–259 days); 6 cases had not resolved.
• Pain 12/32 (38%); trismus 5 (16%); swelling 7 (22%); infection 2; paresis 2.
• 5 device- or procedure-related serious adverse events; 5 late-onset explants or lead revisions (Schoenen 2013).

An 81% rate of facial sensory disturbance, mostly transient but permanent in some, is a real cost. The 2014 review characterises it the same way: “81% reported transient facial sensory disturbance. Five cases required surgical revision for lead migration or explantation” (PMC4119587).

4.2.3 Pathway CH-2 — the US confirmatory trial

peer-reviewed Goadsby PJ, Sahai-Srivastava S, Kezirian EJ, Calhoun AH, Matthews DC, McAllister PJ, Costantino PD, Friedman DI, Zuniga JR, Mechtler LL, Popat SR, Rezai AR, Dodick DW. “Safety and efficacy of sphenopalatine ganglion stimulation for chronic cluster headache: a double-blind, randomised controlled trial.” Lancet Neurol 2019;18(12):1081-1090, DOI 10.1016/S1474-4422(19)30322-9, PMID 31701891, NCT02168764.

• 21 US centres; 93 randomised (45 SPG, 48 control); enrolment July 2014 – February 2017.
• Pain relief at 15 minutes: 62.46% (95% CI 49.15–74.12) with SPG stimulation vs 38.87% (95% CI 28.60–50.25) control; OR 2.62 (95% CI 1.28–5.34), p=0.008.
• 9 serious adverse events by the end of the open-label phase: 3 procedure-related (aspiration during intubation, nausea/vomiting, venous injury), 1 both device- and procedure-related (infection), 5 unrelated.

Note the sham response: 38.87% pain relief in the control arm of CH-2, versus 7.4% in CH-1. That is an enormous difference in control-arm performance between two trials of the same device, and it substantially narrows the treatment effect in the US trial. The absolute gain in CH-2 is about 24 percentage points; in CH-1 it was about 60. Contested — but the most likely explanation is a different control condition and a different (US, higher-expectation) population, not a different device.

4.2.4 Pathway R-1 — the European registry (non-Anglophone-led)

peer-reviewed Barloese M, Petersen A, Stude P, Jürgens T, Jensen RH, May A. “Sphenopalatine ganglion stimulation for cluster headache, results from a large, open-label European registry.” J Headache Pain 2018;19(1):6, DOI 10.1186/s10194-017-0828-9, PMID 29349561 / PMC5773459.

Explicitly non-Anglophone: 12 centres — 10 in Germany, 1 in Denmark (Danish Headache Center/Rigshospitalet lineage, Barloese and Jensen), 1 in Austria. This is the German/Danish/Austrian real-world dataset.

• 97 enrolled (88 chronic, 9 episodic), implanted September 2012 – March 2015; 85 evaluated at 12 months (78 chronic, 7 episodic).
• Chronic CH frequency responders: 55% (43/78).
• 23/78 (29%) were attack-free at 12 months.
• Acute responders: 32% (27/85).
• 13,600 attacks treated: effective therapy in 39%, pain freedom in 26%.
• Therapeutic responders (chronic): 65% (51/78); ≥75% response 47% (37/78); ≥30% response 74% (58/78).
• 74% were able to stop, reduce, or remain off all preventive medication (summarised in PMC8665918).
• 8 lead repositions; 4 initial implant failures.

peer-reviewed Long-term data: Jürgens TP et al., Cephalalgia 2017 — 18-month outcomes (SAGE); and Barloese M et al., J Headache Pain 2016;17(1):67, DOI 10.1186/s10194-016-0658-1 — 24-month outcomes: of 33 chronic CH patients followed 24 months, 30% experienced at least one period of attack remission, beginning after approximately 3.5 months of treatment, with HIT-6 improving by a mean 12.5 points (PMC4961666; as summarised in PMC8665918). At 24 months: acute responders 45%, frequency responders 33%, total response 61%.

This is the most interesting biological finding in the SPG literature: a device designed purely as an acute abortive turned out to have a preventive effect that emerged over months. That was not the hypothesis.

4.2.5 Guideline position

peer-reviewed The EAN 2023 guideline gives SPG stimulation a STRONG recommendation with MODERATE quality evidence — the strongest evidence grade of any device in this chapter — citing pain relief in approximately 65% of attacks and stating it is “efficacious in 60%–70% of patients”. And then, in the same document: “The method is not available at the time of writing.” Table 18 records: “At the time of publishing, not available in Europe” (EAN guideline 2023, PDF).

peer-reviewed The American Headache Society’s 2016 CH treatment guideline (Robbins MS et al., Headache 2016) assigned SPG stimulation a Level B recommendation. (Wiley’s site blocks automated retrieval; this citation is included for completeness but I was not able to fetch and verify the AHS text directly — treat the Level B grading as unverified in this document.)

preprint / trial NHS England declined routine commissioning of SPG stimulation for refractory chronic CH in adults in 2018 (NHS England clinical commissioning policy document).

4.2.6 The commercial collapse — and why this section matters more than the efficacy numbers

preprint / trial Autonomic Technologies Inc (ATI), of Redwood City / Mountain View, California, received European approval to market the Pulsante SPG Microstimulator in February 2012 and collapsed by the end of 2019. Its UK subsidiary, Autonomic Technology Limited, was formally dissolved on 11 May 2021 (UK Companies House, company 07239097).

[PEER-REVIEWED — journalism in Nature] The consequences were documented in detail by Liam Drew in Nature’s immersive feature “Abandoned: the human cost of neurotechnology failure”:

• “More than 700 other people” were left with an implanted ATI device and no manufacturer support (Nature — Abandoned).
• After ATI’s closure, “users and physicians could no longer access the proprietary software needed to recalibrate the device and maintain its effectiveness.”
• The featured patient, Markus Möllmann-Bohle (German; first cluster headache 1987 at age 22, chronic from 2006, up to 8 hour-long attacks per day, implanted 2013), used stimulation about an hour at a time, 5 or 6 times daily, which was “enough to prevent attacks from becoming debilitating”. He said: “There is still no medication reliable enough to help me live a pain-free life without the device.”
• He replaced the hand-held unit’s battery several times, repaired a faulty charging port himself, sourced replacement batteries from a US firm which then stopped making them, and had his most recent battery custom-made by a Chinese company. The battery “was never intended to be accessible to the user”. He stated the stimulator had to be replaced every few months, and after one year at the latest, because of battery failure.
• A second patient, Timothy White, credited the device with allowing him to complete his medical training; after ATI’s bankruptcy he switched to a migraine drug at triple the recommended dose. He noted that after the company closed, “no one was collecting data on security, safety, complications, or side effects.”

[COMMUNITY-REPORT / HISTORICAL] This is the clearest documented case in the CH field of patients being stranded by a company failure, and it is the central fact a patient should weigh before consenting to any implanted CH device. It is not unique: Nature documents Nuvectra (bankruptcy 2019, ≥3,000 spinal-cord stimulator implants, replacement surgery around US$40,000), Second Sight (2020, ~350 retinal implant users), and Stimwave (Nature — Abandoned). Neurotech Reports has separately catalogued the pattern of headache-device market failures (Neurotech Reports).

4.2.7 Current status, 2025–2026

preprint / trial ATI’s intellectual property was acquired in late 2020 by Unity HA of Effingham, Illinois, which renamed itself Realeve in August 2021 with Jon J. Snyder as CEO (BusinessWire, 18 August 2021).

preprint / trial In April 2021 the company obtained FDA Breakthrough Device designation for the Pulsante SPG Microstimulator System for “treatment of acute headache pain associated with chronic cluster headaches”. Its own materials describe Pulsante as “a development stage therapy” and state that “the Pulsante SPG Microstimulator has treated over 700 patients in the US and EU to date” (Realeve / Unity HA Breakthrough Device announcement).

Breakthrough Device designation is not approval. It is an expedited-review pathway. As of the material I could retrieve, Pulsante is not FDA-approved and not CE-marked.

preprint / trial Snyder told Nature in July (2022): “Since we do not have FDA or CE mark approval yet, we are unable to market the therapy and provide support.” Snyder subsequently departed and a consulting firm took temporary control; interim CEO Peter Donato said Realeve had gained approval in Denmark to distribute replacement devices and software to existing users, hoping deliveries could begin in the second half of 2023, with talks underway with three other European countries. Nature does not state that deliveries began (Nature — Abandoned).

preprint / trial Company activity is ongoing but has broadened away from cluster headache. In September 2023 Realeve announced a multimillion-dollar grant supporting a Cleveland Clinic post-stroke recovery study (BusinessWire, 20 September 2023). In May 2025, manufacturing partner Promex announced that it and Realeve had been recognised by MedTech Breakthrough for the Pulsante Micro-Neurostimulator System (Promex Industries, 8 May 2025) — evidence that the device is still in active development and manufacturing engineering as of 2025. Realeve’s founder and CEO is now listed as Dr Peter Bonutti (Realeve, January 2025).

Bottom line as of this writing (2026): SPG stimulation has the best trial evidence of any invasive CH neuromodulation — two positive sham-controlled RCTs plus a positive registry — and is commercially unavailable in Europe, the US and Australia. There is no route for an Australian patient to obtain it. Its resumption is unknown; Realeve has been “about to relaunch” since 2021.

4.2.8 Community sentiment

community report Community sentiment on SPG stimulation is bifurcated in a specific way: strongly positive about the therapy, deeply bitter about the company. Möllmann-Bohle continued to jury-rig batteries for years rather than give up the device (Nature). That combination — a therapy patients will hand-build batteries to keep using, made by a company that dissolved and left them without software — is the defining story of SPG stimulation, and it should carry more weight in a personal decision than the CH-1 response rates.