Interest in deep brain stimulation for refractory cluster headache was ignited after imaging studies showed increased blood flow (indicating activation) in the posterior hypothalamus during cluster headache attacks and increased neuronal density in the same brain area. Hypothalamic implantation of stimulating electrodes for intractable chronic cluster headache was first performed in 2000, with good results. So far, more than 60 drug-resistant chronic cluster headache patients received hypothalamic implantation.[7–20] The approach has also been extended to patients with other types of trigeminal autonomic cephalalgia.[21–24] The overall success rate (patients pain-free or with ≥50% improvement) for published cases is approximately 61%, and the mean follow-up is 2.4 years (Table 1). In the largest series to date (n = 19), with a mean follow-up of 8.7 years (range: 6–12 years), long-lasting improvement was present in 71% (12 of 17) with six patients persistently almost pain free and another six no longer experiencing daily attacks, however, episodic attacks interspersed with long-lasting remission. In five of the six almost pain-free patients, the stimulators have been switched off for a median of 3 years (range: 3–4 years) – after several years during which stimulation was necessary for relief. Five patients did not improve, four had bilateral cluster headache and three developed tolerance after experiencing relief for 1–2 years. Adverse events were electrode displacement (n = 2), infection (three electrode infections and one infection of impulse generator), electrode malpositioning (n = 1), transient nonsymptomatic third ventricle hemorrhage (n = 1), persistent slight muscle weakness on one side (n = 1) and a seizure (n = 1). In the first few years after implantation, most patients experienced headache recurrence a short time after the stimulator was switched off or if the battery ran out.
Other smaller studies have reported broadly comparable findings.[8,10–19] In the only sham-controlled trial, 11 drug-resistant chronic cluster headache patients were randomized to either real or sham posterior hypothalamic stimulation. The randomized phase lasted just 1 month during which active and sham stimulation were compared. A 1-year open phase followed. The primary outcome was weekly-attack frequency. Pain intensity, sumatriptan injections, emotional impact (Hamilton Depression), quality of life (Short Form 12), behavior changes, and homeostatic and hormonal functions were also monitored. There were no significant differences between the active and sham stimulation patients for attack frequency, pain intensity, sumatriptan injections or other variables, during the randomized phase. By contrast, 1 year after continuous hypothalamic stimulation (open phase), three patients were pain free and another three had >50% reduction in attack frequency. One subcutaneous infection, one transient loss of consciousness and one case of micturition syncope occurred. There were no changes in hormonal function or electrolytic balance. As expected, the short randomized phase did not show deep brain stimulation efficacy or differences between the groups. However, the open phase demonstrated long-term efficacy in over 50% of patients. Ethical constraints probably prevented the sham stimulation from continuing beyond 1 month.
Posterior hypothalamic stimulation was also effective in three patients suffering from short-lasting neuralgiform headache attacks with conjunctival injection and tearing (SUNCT), which, like cluster headache, is a form of trigeminal autonomic cephalalgia.[21–23] The first patient became pain free, but needed additional prophylaxis with lamotrigine. The second patient experienced a substantial reduction in attack frequency (from 120 to 25 per day after 1 year). The third patient also had a reduction in attack frequency, from 30 per day to a few sporadic attacks, after 15 months of continuous stimulation. Walcott and coworkers reported that a patient with chronic paroxysmal hemicrania had obtained relief after posterior hypothalamic stimulation.
Hypothalamic stimulation has been tested for the acute treatment of cluster headache. When an attack occurred, treatment consisted of switching the stimulator on (if it was switched off) or increasing stimulation intensity (if it was switched on). A total of 136 attacks were treated, of which 108 were assessable (intolerable side effects arose during the remaining 28 attacks). A clinically significant reduction in pain intensity (≥50%) was noted in only 25 attacks (23%), suggesting posterior hypothalamic stimulation is not useful for the acute treatment of cluster headache.
The long experience of posterior hypothalamic stimulation for intractable cluster headache makes it possible to conclude that it produces long-lasting improvement in a considerable proportion of patients and that stimulation is well-tolerated for many years after implantation. In some patients, after several years during which stimulation was necessary for relief, a persistent almost pain-free condition can be maintained without stimulation. It is also now evident that tolerance can occur after long-lasting improvement in some cases. Bilateral chronic cluster headache seems to predict poor response to hypothalamic stimulation.
Nevertheless, posterior hypothalamic stimulation is not without risk. One patient died from intracerebral hemorrhage along the electrode track a few hours after implantation and another had a subclinical hemorrhage. Three serious adverse events occurred in the 11 patients enrolled in the randomized controlled trial. The risk of hemorrhage is approximately 3% and is in the upper part of the range reported for deep brain stimulation in movement disorders. To minimize this risk Seijo et al. slightly modified the hypothalamic target to avoid the lateral ventricle wall, without affecting efficacy. In another neuroimaging study, the anatomical location of the stimulating electrodes did not differ between responders and nonresponders. Other reported adverse events are panic attack with dysautonomia, oculomotor disturbances,[7–20] intraoperative transient ischemic attack, subcutaneous infection, transient loss of consciousness with hemiparesis and micturition syncope, erectile dysfunction and paroxysmal sneezing. In view, therefore, of the invasive nature of hypothalamic stimulation and risk of serious adverse events, the authors suggest that this treatment should be tried only after at least 1 year of occipital nerve stimulation has proved ineffective.
Posterior hypothalamic stimulation tends to improve sleep by suppressing nocturnal headache attacks. When the stimulator is switched on and amplitude increased slowly, heart rate, blood pressure and respiratory rate under resting conditions do not change. However, abrupt changes in stimulation voltage may produce autonomic manifestations and oculomotor disturbances.
With regard to the mechanism of action of hypothalamic stimulation as a treatment for chronic cluster headache, when the technique was first introduced, the hypothesis was that the inhibitory action exerted by high-frequency electrode stimulation might reduce supposed hypothalamic activation during a headache attack. However, as noted, acute hypothalamic stimulation does not abort acute cluster headache attacks,[7–20] suggesting – together with the latencies for a prophylactic effect – that stimulation works by mechanisms more complex than mere inhibition of activation.
The transformation from chronic to sporadic/episodic cluster headache in some of patients still receiving long-term stimulation, suggests that hypothalamic stimulation may act on circuits involved in the chronicization of the condition. Hypothalamic stimulation has been shown to increase blood flow through pain matrix areas, such as the anterior cingulate, the insula and the frontal lobe – just as cluster headache attacks do. It has been suggested that these structures play a major role in the chronicization of pain, probably through long-term potentiation. Hypothalamic stimulation may reset these areas, perhaps revoking long-term potentiation, with the result that the chronic course of cluster headache is interrupted and reverts to a sporadic/episodic form.
It is also probable that environmental, genetic and behavioral factors influence the chronicization of cluster headache attacks.
Involvement of pain matrix brain areas is also suggested by the observation of increased cold pain threshold in the territory of the ipsilateral ophthalmic nerve in 11 patients with drug-resistant chronic cluster headache undergoing posterior hypothalamic stimulation.
Future Neurology. 2013;8(4):457-467. © 2013 Future Medicine Ltd.