The pathogenesis of cluster headache is unknown (see, for example, Ref. 1). With regard to the main component of cluster headache, the pain itself, little information exists both as to the origin and as to the nature of the pain. Existing theories concerning cluster headache pathogenesis have been rather unproductive. In the following, evidence from seemingly totally unrelated fields that, nevertheless, possibly has a bearing on cluster headache pathogenesis will be reviewed.
On the tendency to bilaterality in cluster headacheCluster headache is in principle a unilateral headache. There is, however, a tendency for the headache to shift side, either during one single bout (circa 5%) or from one bout to another (circa 10%). All in all, approximately 15% of cluster headache cases thus shift side at some point along the time axis (2). Accordingly, there seems to be a "disposition" or "anlage" for the headache to develop on both sides in cluster headache, although the headache actually most frequently manifests itself on one side only.The prevalence of cluster headache in the form that is severe enough to lead to medical consultations may seem to be around 0.7% (3) in Western countries. The tendency to develop headache on the side opposite the pain side in a cluster headache patient may, therefore, seem to be around 200 times higher than for an individual to develop this type of headache de novo (4). Side shift is thus a conspicuous feature of this disease.On rare occasions, the headache may even be bilateral (5). Several variations of bilaterality have also been observed. Ekbom (6) has thus described a case with a most peculiar pattern, in which the headache during one bout, in addition to appearing on the original side, started appearing on the contralateral side; then "the headache shifted [side] in different attacks"; thereafter, a decrescendo occurred on the original side as opposed to a crescendo on the contralateral side, and, in the ultimate phase, attacks only appeared on the contralateral side.After operations on one side, pain attacks have a tendency to recur on the opposite, previously pain-free side. This was thus the case in 1 of 12 cases treated with the "combined procedure"-that is, sectioning of the greater superficial petrosal nerve and neurolysis of the sensory root of the trigeminal nerve (7).The attack of cluster headache as such consists of (at least) two components: the pain and the autonomic phenomena. The latter also show a tendency to bilaterality (8, 9) and, in fact, much more regularly so than does the pain. In patients with side shift there is sometimes a dichotomy between pain and autonomic phenomena, the latter persisting on the previous pain side (4). The schism may even be primary; that is, the pain has always been contralateral to the autonomic phenomena (10).This tendency to bilaterality in cluster headache is intriguing. A pathogenetic model of cluster headache must also include and explain the marked tendency to bilaterality of the pain and, above all, of the autonomic manifestati...