Verapamil is the preventive medication of choice for both prolonged episodic and chronic cluster headache (6,22,38). Doses in the range of 240 to 360 mg twice daily are common, with an occasionally higher dose used in refractory cases. These larger doses are considerably higher than those for cardiac disease or hypertension. Verapamil may cause heart block by slowing atrioventricular node conduction. PR interval prolongation on the electrocardiogram (ECG) may serve as a coarse indicator of impending heart block. Although formal guidelines for verapamil use are not available, it is reasonable to start with 240 to 480 mg per day (depending on body surface area) after a normal ECG is demonstrated. The daily dose may be increased by 80 to 120 mg each week or two, but it is suggested that an ECG be performed prior to each increase and remains normal. Dose escalation may continue until the attacks disappear, side effects occur, or the maximum daily dose of 960 mg is reached. Standard shorteracting verapamil preparations seem to be more effective than extended-release formulations (32). The most common side effect is constipation, but dizziness, distal edema, nausea, fatigue, hypotension, and bradycardia also occur. β-Blockers must not be given concurrently.

Lithium is generally an effective cluster headache preventive medication but less reliably so for episodic cluster headache than chronic cluster headache (6,14,33). Most patients benefit from 600 to 1,200 mg per day. Tolerability is assessed after 3 to 4 days at 300 mg twice a day, and additional 300 mg per day administrations added until the headaches disappear. Lithium serum levels have to be checked to prevent side effects and should be measured 12 hours after the last dose. In the chronic forms of the disease, lithium should be slowly tapered off at 6- to 12-month intervals to detect those patients who have become episodic. Among the first side effects to appear are agitation, postural tremor of the hands, insomnia, weakness, nausea, thirst, slurred speech, and blurred vision. Toxicity is signaled by nausea, vomiting, anorexia, diarrhea, confusion, nystagmus, ataxia, extrapyramidal signs, and
seizures. Hypothyroidism and polyuria (nephrogenic diabetes insipidus) can occur with long-term use. Polymorphonuclear leukocytosis may occur and be mistaken for occult infection. Renal and thyroid function tests are performed prior to and during treatment. Because of the narrow therapeutic window, side effects have to be carefully looked for. In warmer weather, patients should be warned not to become dehydrated, which increases the chance of toxicity.

Methysergide is not available in the United States. Clinical experience demonstrates that methysergide is an effective preventive for cluster headache, although efficacy findings derive only from open trials (9,34). Starting dose is 1 mg per day to minimize side effects and increments should be by 1 mg (in a thrice daily regimen) every 3 to 5 days. If tolerated, methysergide can be raised up to 12 mg/d. Common short-term side effects are nausea, vomiting, dizziness, muscle cramps, abdominal pain, and peripheral edema. Because of its vasoconstrictive properties, coronary or peripheral arterial insufficiency are both contraindications and troublesome side effects (but uncommon) and require drug discontinuation. Prolonged treatment may rarely cause retroperitoneal, pulmonary, pleural, or cardiac fibrosis (25). As a consequence, methysergide is most appropriate for patients with short cluster periods (less than 4 months). When used for prolonged periods, this risk can be minimized by 1-month drug holidays every 6 months, gradually tapering off the drug.

In cases of continuous administration, careful monitoring including yearly echocardiogram, chest x-ray, and abdominal magnetic resonance imaging (MRI) are recommended (62). To reduce the risks of vasoconstrictive effects, it is recommended to administer injectable sumatriptan a few hours apart from methysergide intake and sparingly. In the United States, methysergide is no longer available and methylergonovine, an active metabolite, is a good alternative. It is started at 0.2 mg/d and raised to three times a day rapidly. The maximum dose is usually 0.4 mg three times a day. The same cautions apply. In the United States, it is contraindicated to use any triptan when a patient is taking an ergot.

A dose of 2 to 4 mg per day may be tried for 2 or 3 weeks (66) when verapamil, lithium, and methysergide are ineffective. If the attacks are predictable, the drug should be taken 30 to 60 minutes beforehand; for attacks awakening the patient during the night, 1 to 2 mg (tablets or suppositories), may be prescribed before retiring. Dr. Lee Kudrow frequently prescribed it twice daily for extended periods with no problem.

In an open-label study of 54 patients with intractable cluster headache (23 episodic, 31 chronic) repetitive intravenous dihydroergotamine rendered all patients headache free (50). Twelve months later, 83% of episodic cluster headache and 39% of chronic cluster headache patients remained headache free. This strategy is often effective in chronic cluster headache patients who do not respond to other preventives.

Pizotifen is not available in the United States. An open-label study (65) and one controlled trial (13) reported pizotifen to be moderately effective.

Corticosteroids (prednisolone, prednisone, and dexamethasone) are the most rapidly effective preventive agents for cluster headache (2,8). Both oral prednisone and prednisolone are given to a maximum of 60 mg once daily for 5 to 10 days and thereafter the dose is decreased by 5 to 10 mg every 3 days. In long-lasting cluster periods and chronic forms, relapse is almost invariable as the dose is tapered.

Open-label studies with sodium valproate or divalproex sodium report effectiveness in 54 to 73% of cluster headache patients (19,23,28). The use of this drug in cluster was first reported by Arieh Kuritsky in the 1980s. A double-blind, placebo-controlled, parallel group study of sodium valproate (1,000 to 2,000 mg/d) in cluster headache prevention found no significant difference between the 50 patients in the treatment group (37 episodic, 11 chronic, 2 unspecified) and the 46 in the placebo group (36 episodic, 6 chronic, 3 unspecified) (15). Other studies are needed to further explore usefulness of valproate in cluster headache prevention.

In the first four open-label studies on cluster headache prevention with topiramate, it was administered in the range 25 to 200 mg per day in a total of 30 episodic and 22 chronic cluster headache patients (17,35,52,72). The drug was moderately or markedly effective in 37 patients (70%). Two of the studies reported that topiramate was associated with rapid improvement (within 1 to 4 weeks) (35,72). In these studies, some patients were given topiramate with corticosteroids and others entered the study at the end of their cluster period (72). These observations could explain, at least in part, the observed rapid improvement. In an open-label study of 33 cluster headache patients (23 episodic cluster headache, 10 chronic cluster headache), 21% had a reduction of more than 50% in headache frequency during the 20-day study period. The response was not dose related (40).