Objectives: To describe the clinical characteristics and laboratory findings of cluster headache patients whose headaches responded to testosterone replacement therapy.
Background: Current evidence points to hypothalamic dysfunction, with increased metabolic hyperactivity in the region of the suprachiasmatic nucleus, as being important in the genesis of cluster headaches. This is clinically borne out in the circadian and diurnal behavior of these headaches. For years it has been recognized that made cluster headache patients appear overmasculinized. Recent neuroendocrine and sleep studies now point to an association between gonadotropin and and corticotropin levels and hypothalamically entrained pineal secretion of melatonin.
Results: Seven male and 2 female patients, seen between July 2005 and February 2005, and between ages of 32 and 56, are reported with histories of treatment resistant cluster headaches accompanied by borderline low or low serum testosterone levels. The patients failed to respond to individually tailored medical regimens, including melatonin doses of 12 mg. a day or higher, high flow oxygen, maximally tolerated verapamil, antiepileptic agents, and parenteral serotonin agonists. All 9 patients demonstrated either abnormally low or low, normal testosterone levels. After supplementation with either pure testosterone in 5 of 7 male patients or combination testosterone/estrogen therapy in both female patients, the patients achieved cluster headache freedom for the first 24 hours. Four mail chronic cluster patients, all with abnormally low testosterone levels, achieved remission.
Conclusions: Abnormal testosterone levels in patients with episodic or chronic cluster headaches refractory to maximal medical management may predict a therapeutic response to testosterone replacement therapy. In the described cases, diurnal variation of attacks, a seasonal cluster pattern, and previous, transient responsiveness to melatonin therapy pointed to the hypothalamus as the site of neurological dysfunction. Prospective studies pairing hormone levels and polysomnographic data are needed.