Abstract
Keywords
Introduction
The history of deep brain stimulation (DBS) in the headache fields started with the observation that the stimulation of specific brain areas provoked neurovascular headaches. Raskin et al. (1) reported that 15 (9%) of 175 patients who underwent DBS, mainly into the periaqueductal grey, for intractable low back pain developed severe headaches with migrainous features. Similarly, Veloso et al. (2) found that 15 (23·4%) of 64 patients developed chronic headache after stimulation of the periaqueductal area, sensory thalamus, or internal capsule; none of these patients was a headache sufferer before. These headaches started after a lag of about two months (mean) between implantation and headache development as the result of activation of neural pathways or threshold change (2). Sano et al. showed that posteromedial hypothalamotomy improved otherwise intractable facial pain secondary to cancer; at the same time they also observed that facial stimulation causes neuronal discharges in the same brain area (3).
In the late 1990s, neuroimaging studies changed the view of so-called vascular headaches, and in particular, cluster headache (CH). Activation of the ipsilateral posterior hypothalamic area during the attacks (4) and increased neuronal density in the same area (5) were observed in CH (6) patients and the posterior hypothalamic area was hypothesised to be the cluster generator (4).
These observations together with clinical features such as the striking clockwise regularity of attacks and seasonal recurrence of cluster periods (7), as well as the neuroendocrine abnormalities found in these patients (8), pointed to the brain as the place of origin of this neurovascular headache (for review, see May (9)).
In the late 1990s, a 38-year-old male suffering from chronic intractable CH came to our attention (10,11). He had at least four headache episodes per day mainly during the night, each accompanied by life-threatening hypertensive crises (up to 260/160 mmHg), massive oculo-facial phenomena and very aggressive behavior (10). Extensive investigations were all unremarkable. All drugs were tried with no benefit, including sumatriptan injection. The patient received four destructive operations on the right trigeminal nerve, after which right-sided attacks disappeared but left-sided CHs markedly worsened. In addition he developed intractable right-side anaesthesia dolorosa. During a hypertensive crisis the patient became blind in the right eye because of a vitreous humour haemorrhage and left trigeminal surgery was contraindicated to avoid corneal sequelae that could have rendered the patient completely blind. In July 2000 hypothalamic stimulation to control left-sided CHs was performed. Implantation alone and bipolar stimulation did not affect the crises; unipolar stimulation was started and after two weeks the left CHs gradually disappeared (11). Unknown to the patient, the stimulator was switched off and attacks recurred; when the stimulator was turned on the attacks disappeared (11). In May 2001 right CHs recurred (pain free on the left hypothalamic-stimulated side), again with the life-threating accompanying phenomena and completely resistant to all drugs. Hypothalamic stimulation was successfully started also on the right side with only minor transitory side effects (12).
Hypothalamic stimulation in short-lasting unilateral headache syndromes.
Patients underwent hypothalamic stimulation to treat intractable short-lasting headaches.
Abstract.
Efficacy on pain in first trigeminal branch only.
Mean.
One of the two patients suffered both from CH and SUNCT and both improved after hypothalamic stimulation. ePersonal communication, with authors' permission.
DBS: deep brain stimulation; SUNCT: short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing; CH: cluster headache; NR: not reported.
Efficacy and tolerability of hypothalamic stimulation
In CH, after a mean follow-up of 2.2 years, 65.7% (n = 46) had a 50% improvement or more and among these, 23 (32.9%) were pain free (Table 1). There is only one double-blind, randomised, controlled study investigating DBS efficacy in chronic CH (24); the main limitation of this interesting study is the short blind period, one month, too short to give clinically significant information on DBS efficacy. In the largest series (14), after an exceptionally long follow-up of 8.7 years (median), improvement occurred in 70% (12 of 17) of drug-resistant chronic CH patients: Six were almost pain free and six were transformed into episodic CH. Improvement could be maintained with the stimulators off but only after years of continuous stimulation; the latter observation suggested that hypothalamic stimulation can change disease course (14).
Continuous hypothalamic stimulation takes weeks to improve the condition while acute stimulation (136 attacks in 16 patients (34)) is not effective in resolving ongoing CH attacks, suggesting that hypothalamic stimulation acts by complex mechanisms requiring time to occur.
Thirty-four per cent of CH patients do not improve after hypothalamic stimulation (Table 1). Similarly, in the largest study, less than one-third of the patients did not improve, 80% (four of five) of whom had bilateral CH, and 60% (three of five) developed tolerance after experiencing relief for one to two years, showing that tolerance can occur after persisting improvement (14). Bilateral CH seems to predict poor response to hypothalamic stimulation (14).
Other otherwise intractable TACs benefited from hypothalamic stimulation: After at least one year of follow-up, three SUNCT patients (29–31) and one chronic paroxysmal hemicrania patient (32) markedly improved. One patient suffering both from CH and SUNCT also improved after hypothalamic stimulation (27).
A certain improvement was obtained in one patient suffering from unbearable CH-like attacks who was affected by ipsilateral trigeminal infiltration produced by an aggressive angiomiolypoma (28).
Hypothalamic stimulation has been successfully tried also in symptomatic (multiple sclerosis) trigeminal neuralgia unresponsive both to drugs and destructive neurosurgery on the trigeminal nerve (33). The authors reported that hypothalamic stimulation improved pain only in the first trigeminal branch, suggesting that differing trigeminal branches may undergo differing control mechanisms and pathways.
A blind-to-the-patient interruption of DBS has been achieved in many cases, and relapses have been observed on each interruption during the first years of stimulation (11,14). When the stimulator has been turned back on, attacks improved again (11,14).
Reported adverse events were electrode displacements, infections, electrode malpositioning, small non-symptomatic third ventricle haemorrhage, persistent slight muscle weakness, seizure, diplopia and vertigo with higher amplitudes, panic attack, transient ischaemic attack, high-frequency tremor, cervical dystonia, and changes in appetite and thirst (see Table 1 for details) (11–26). Syncopes have also been reported (12,13) even if only a slight increase in sympathico-excitatory tone in the head-up-tilt test was found (35). In one small series, one patient died of brain haemorrhage (15). Sleep pattern disruption has been reported in two patients with pre-existing sleep disturbances (27) but in another series sleep pattern was not affected by hypothalamic stimulation (36).
Because of its invasiveness, hypothalamic DBS should be proposed in chronic intractable TACs only after other, less-invasive, neurostimulation procedures have been tried (37).
Insights into CH and TAC pathophysiology after hypothalamic stimulation
Acute hypothalamic stimulation did not improve ongoing acute CH attacks (34), while under continuous hypothalamic stimulation efficacy is obtained weeks after its start (11–32). These observations suggest that hypothalamic stimulation acts by complex mechanisms in CH prevention (14). In a positron-emission tomography (PET) study, hypothalamic stimulation modified brain activity in the ipsilateral trigeminal system as well as in brain areas belonging to the ‘pain matrix’ (38). Fontaine et al. (39) examined the correlation between hypothalamic electrode position and efficacy in hypothalamic-implanted chronic CH patients. The authors reported that the contact coordinates and the structures were not significantly different between responders and non-responders, suggesting that failure of hypothalamic stimulation treatment in cluster headache may be due to factors unrelated to electrode positioning (39). They also suggested that the therapeutic effect is probably not related to direct hypothalamic stimulation; the latter might exert its effect by modulating either a hypothalamic CH generator or the anti-nociceptive system (39). An increased cold pain threshold in the ipsilateral to the pain first branch of the trigeminal nerve (40) has been observed in hypothalamic-stimulated chronic CH patients. This could also explain why hypothalamic stimulation in trigeminal neuralgia produced improvement only on the first branch trigeminal pain (33).
Article highlights
Deep brain stimulation of certain brain areas, mainly the periaqueductal grey, was observed to produce headache with migrainous and/or cluster-like features. Long-term continuous hypothalamic stimulation improves many drug-resistant chronic cluster headache patients as well as other trigeminal autonomic cephalalgias; the procedure also improved pain in the first trigeminal branch in trigeminal neuralgia. These observations need confirmation in randomised, controlled studies. These observations point to a key role of the posterior hypothalamic area in the pathophysiology of unilateral short-lasting headaches, possibly by regulating the duration rather than triggering the attacks. Because of its invasiveness, hypothalamic stimulation can be proposed only after other, less-invasive, neurostimulation procedures have been tried.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.