Abstract
Cluster headache is a well-defined disorder in which patients suffer extremely painful headaches with clock-like regularity one to three times, or more, a day for perhaps several months followed by a period of remission (Table 1) (1). Cluster headache has been recognized for at least 250 years (2), yet it has remained a mechanistically poorly understood condition.
Clinical features of cluster headache simplified from the International Headache Society criteria (1).
The three major aspects of the pathophysiology of cluster headache are the trigeminal distribution of the pain, the autonomic features, and, more important, the episodic pattern of the attacks, which is in some ways the defining clinical signature of the disorder compared to migraine. These features raise the classic issues of the location of the lesion and the generic terminology that should be applied to these and related headaches, such as migraine. Ekbom's classic observation during angiography of a patient suffering an acute cluster headache demonstrating changes in the internal carotid artery (3) suggested a pathological focus in the region of the cavernous sinus. The arguments for this locus for the disease have been set out elsewhere (4, 5). Although both migraine and cluster headache are often referred to as “vascular headaches”, I would suggest that they should be referred to generically as “neurovascular headaches”. This proposal is partly based on the observation of brainstem activation during spontaneous migraine (6), and partly on studies of the trigeminovascular system (7), and, more importantly for cluster headache, it is based on the site of the dysfunction in the central nervous system.
Neuroendocrine changes in cluster headache
The overall impression of cluster headache is of a defect in cycling mechanisms. Kudrow (8, 9) was the first to point out that testosterone levels were altered in cluster headache patients during the bout. Nelson, shortly thereafter, confirmed this, reporting lower testosterone levels in patients with cluster headache during their on-phase (10). Leone et al. (11) identified reduced responses to stimulation by thyrorropin-releasing hormone, and there are interesting observations of disordered circadian rhythm for Cortisol, growth hormone, luteinizing hormone and prolactin. One area involved in human clock systems is the suprachiasmatic nucleus in the hypothalamic gray matter region, which sits at the base of the third ventricle. Melatonin is produced by the pineal gland and has a strong circadian rhythm regulated by the suprachiasmatic nucleus. Connections between the retina and the hypothalamus are thought to provide light cues for the circadian rhythm (12). The characteristic nocturnal peak of melatonin secretion is blunted during the active phase of cluster headache (13). The hypothalamic gray matter region is therefore of potential interest in studies of cluster headache.
Positron emission tomography studies of cluster headache
In order to study central nervous system activation, we recruited 9 male patients with chronic cluster headache, none of whom were using prophylactics. Cluster headache attacks were triggered with nitroglycerin spray and positron emission tomography (PET) scans performed with H2 15O. All subjects had 12–13 scans at 10-min intervals. We did three scans in rest, to get clear baseline measurements, and 1–3 scans after giving nitroglycerin, depending on how soon the headache developed. Most patients experienced acute pain within 20 min. We only recruited patients who had a reliable, rapid response to sumatriptan injections, so that we were sure we could resolve their headaches quickly (14) with a treatment known not to alter cerebral blood flow (6). We made 4–6 scans during the attack and 2 or 4 scans after relief from headache.
The PET data were realigned and normalized, and a correction was made for global changes in cerebral blood flow. A statistical comparison was made of regional cerebral blood flow changes between different conditions for each unit volume of the brain, with the resulting output being a statistical parametric map (SPM) (15). The results obtained were the aggregates of the data from 9 patients. The statistical analysis was designed to identify areas that were activated significantly only during headache and were not active during the rest period or during the nitroglycerin spray. Since cluster headache is a predominantly unilateral condition, the scans of patients with right-sided headache were mirrored to the left side to provide 9 sets of data distributed symmetrically with the side of the pain.
Visualization of three distinct types of activation during cluster headache
The areas that we observed to be activated fell into three categories: areas generally associated with pain, an area that seems specific to cluster headache, and vascular structures (16).
Pain
The anterior cingulate was significantly activated, as would be expected, since in most human PET studies of pain activation of the anterior cingulate is observed, perhaps as part of the affective response. We have also seen activation in the frontal cortex and insulae and ventroposterior thalamus contralateral to the side of the pain. In addition, we saw activation in the ipsilateral basal ganglia. This is not the first observation of basal ganglia changes associated with pain. It may simply relate to movement, the desire to move that is common in cluster patients, or even some deliberate inhibition of movement.
Cluster headache
The only activated area that is particular to cluster headache is the ipsilateral hypothalamic gray matter region. This is in contrast to the results of Hsieh et al. (17), who did PET scans on four cluster headache patients, two with right-sided and two with left-sided headaches. They observed cingulate cortex and frontal activation, but not hypothalamic activation or thalamic activation. In that study the data were not mirrored.
We observed ipsilateral activation at the base of the third ventricle in the hypothalamic gray matter region. This area is one of obvious interest because of its role in the control of circadian rhythm of neurons. More anatomical specificity requires further studies, which are ongoing.
Vascular change
An important aspect of this study is that we were able to visualize vessels in the region of the cavernous sinus/carotid artery. PET scans of capsaicin-induced pain (18) showed no hypothalamic activation, despite the fact that severe first-division pain is turned on by capsaicin. The capsaicin experiment did demonstrate flow changes in an area consistent with the cavernous sinus/ carotid artery, just as there are flow changes in the vessels in cluster headache. This implies that the activation of the carotid does not relate specifically to cluster headache, but that it is a trigemino-vascular autonomic reflex to first-division pain. The flow changes are, therefore, epiphenomena of the trigeminal activation, not part of the disease generation process in cluster headache.
Basic biology of cluster headache
The pain of cluster headache is very much a first-division of trigeminal phenomenon. Many of the autonomic features are due to seventh nerve activation, while the remaining changes are due to a transient cervical sympathetic deficiency. There are neuropeptide changes in the cranial circulation in both calcitonin gene-related peptide (CGRP) and vasoactive intestinal polypeptide (VIP) release during headache (19), which reflect the trigeminal and autonomic dimensions of the syndrome. It has been proposed that one might refer to such syndromes as trigeminal-autonomic cephalalgias (20), to emphasize the fact that these syndromes, such as cluster headache and paroxysmal hemi-crania, involve both a trigeminal and an autonomic neural activation in their expression. For cluster headache, a crucial aspect of the dysfunction appears to lie within the ipsilateral hypothalamic gray matter region, and it is clear that the carotid flow changes are driven by the ophthalmic division of the trigeminal nerve and are not due to cluster headache as such.
Conclusions
Recent results from PET scans during nitrate-induced cluster headache lend weight to the idea that the lesion in cluster headache, at least in the people who are “turned on”, is in the ipsilateral hypothalamic gray matter region (Fig. 1). Such a lesion may also play a role in the switch between remission and headache periods. The findings support the adoption of the term neurovascular headaches, rather than vascular headaches, for migraine, cluster headache, and related disorders, and thus offer a new perspective on the disease.
Elements of a neurobiologically based explanation for cluster headache. Pain afferents from the trigeminovascular system traverse the ophthalmic division of the trigeminal nerve, taking signals from the cranial vessels and dura mater. These synapse in the trigeminal nucleus caudalis and then project to the thalamus (ventroposterior) and lead to activation in cortical areas, including frontal cortex, insulae, and cingulate cortex, resulting in pain. There is reflex activation of the parasympathetic outflow via the facial (VIIth cranial) nerve, predominantly through the pterygopalatine (sphenopalatine) ganglion, which acts as a positive feedback system to dilate the vessels further and irritate trigeminal endings. This autonomic activation leads to lacrimation, reddening of the eye, and nasal congestion, while a local third-order sympathetic nerve lesion due to carotid swelling results in a partial Horner's syndrome. The key central nervous system site for triggering the pain and controlling the cycling aspects is in the posterior hypothalamic gray matter region, shown now to be active on PET scans in patients (16).
Footnotes
Acknowledgements.—
The author acknowledges the invaluable contributions of Drs A May, A Bahra, C Buchel, and Professor RSJ Frackowiak to the PET studies described herein. PET studies would not be possible without the willing participation of our patients, to whom I am grateful. The research program is funded by the Wellcome Trust and the Migraine Trust. PJG is a Wellcome Senior Research Fellow.