The U.S. Headache Consortium guidelines for the treatment of migraine summarize data from clinical studies of various drugs used in preventive therapy of migraine. The analysis included alpha₂ agonists, anticonvulsants, antidepressants, beta blockers, calcium channel antagonists, non-steroidal anti-inflammatory drugs (NSAIDs), serotoninergic agents, hormonal agents, feverfew, magnesium and vitamin B₂ (riboflavin). Members of the consortium graded the quality of the evidence for the use of a particular drug based on the findings from clinical trials. The bulk of the analysis was based on the technical review of the treatment of migraine by the Agency for Healthcare Research and Quality (AHRQ, formerly the Agency for Health Care Policy and Research). See the accompanying table on page 2536 for a comprehensive list of specific drugs and the strength of the evidence for their use in the prophylaxis of migraine.

Summary of Evidence for Specific Drugs for Prophylaxis

The following highlights the findings from clinical studies of specific agents:

• Alpha₂ agonists—Sixteen trials of clonidine and one trial of guanfacine were reviewed and suggested that alpha₂ agonists are minimally and not conclusively efficacious. Three of 11 placebo-controlled trials of clonidine showed a significant difference between placebo and treatment arms, but the magnitude of the effect was small.

• Anticonvulsants—Strong support of the efficacy of divalproex and valproate was found in five studies of these agents. The headache consortium found that the evidence was weaker in support of the use of other anticonvulsants, such as carbamazepine, clonazepam and gabapentin.

• Antidepressants—Clinical studies indicate that amitriptyline is the only antidepressant that has shown fairly consistent efficacy in the prevention of migraine. This agent has also been evaluated more frequently than other antidepressants. One study revealed that fluoxetine was significantly better than placebo in the prevention of migraine, but another study failed to duplicate this finding. A randomized placebo-controlled study, reported in 1998, showed that fluoxetine may be beneficial in the prevention of migraine.

• Beta blockers—The AHRQ report on drug therapies for the prevention of migraine included analysis of 74 controlled trials of beta blockers. Propranolol was investigated in 46 studies and metoprolol in 14 studies. The consortium found consistent evidence for the efficacy of propranolol, 120 to 240 mg daily, in the prevention of migraine attacks. Studies indicate that beta blockers with intrinsic sympathomimetic activity are ineffective for preventing migraine.

• Calcium channel antagonists—The review included 45 controlled trials of calcium channel antagonists, including 11 trials of nimodipine, five trials of nifedipine, three trials of verapamil and one trial of nicardipine. Three placebo-controlled trials of nimodipine showed no significant difference between the active agent and placebo, but two trials showed large and statistically significant differences in favor of nimodipine. In two of three placebo-controlled trials of verapamil, significant differences were found with the calcium channel blocker, but the relevance of the findings is uncertain because of high dropout rates in the two studies. In a placebo-controlled trial that included a propranolol arm, no significant differences in the effects on headache frequency were noted between verapamil and propranolol.

• NSAIDs—The AHRQ review included 23 controlled trials of 11 different NSAIDs. A meta-analysis of five placebo-controlled trials of naproxen or naproxen sodium suggests a modest but statistically significant effect on headache index or frequency. In several studies that compared NSAIDs with propranolol and metoprolol, no differences between these agents were found.

• Serotoninergic agents—The review included 13 controlled trials of ergot derivatives. According to the U.S. Headache Consortium, the evidence is insufficient for the efficacy of ergotamine or the combination of ergotamine, caffeine, butalbital and belladonna alkaloids (Cafergot compound) in the prevention of migraine. The guidelines also state that the usefulness of methysergide is now limited because of its association with retroperitoneal and retropleural fibrosis.

• Hormone therapy—The review included six controlled trials on the use of estrogens and/or progestogens for the prevention of migraine. Two placebo-controlled trials of estradiol, administered premenstrually as a gel or patch, suggest that a relatively high dosage (1.5 mg per day of the gel form) may be effective in women whose migraines are associated with their menstrual cycle. Evidence does not point to a benefit in patients whose migraines are not related to the menstrual cycle.

• Feverfew—Three trials of feverfew suggest that this herbal remedy may have an effect on migraine. One trial of a group of patients who used feverfew showed that withdrawal of the herb was followed by a statistically significant increase in headache frequency. Another study of patients who had never used feverfew revealed a statistically significant difference between the feverfew group and the placebo group. In a double-blind, randomized, crossover trial, feverfew was found to be associated with a significant reduction in pain intensity and other symptoms such as nausea, vomiting, photophobia and phonophobia.

• Vitamins and minerals—Two studies showed benefits of magnesium over placebo, whereas a third study failed to show any benefit. In one study of high-dose (400 mg) vitamin B₂ (riboflavin), a significant benefit was noted three and four months after initiation of vitamin B₂ supplementation.