Source: Linde M, Mulleners WM, Chronicle EP, McCrory DC. Valproate (valproic acid or sodium valproate or a combination of the two) for the prophylaxis of episodic migraine in adults. Cochrane Database Syst Rev. 2013;(6):CD010611.
Study Population: Patients 16 years or older with episodes of migraine headache on fewer than 15 days of every month
Efficacy Endpoints: 50% or greater reduction in migraines after one month of treatment
Harm Endpoints: Adverse effects such as fatigue, vertigo, nausea, tremor, and weight gain
Narrative: Migraines are the sixth most common cause of disability worldwide and affect more than 10 million U.S. adults annually.1,2 Daily antiepileptic medications are commonly used to prevent migraine headaches. A 2013 Cochrane review assessed evidence on the efficacy and tolerability of valproate medications (valproic acid [Depakene] or divalproex sodium [Depakote], or a combination of both) in migraine prophylaxis. Ten randomized trials enrolled 2,296 patients 16 years or older with migraines occurring on fewer than 15 days of every month. The studies compared valproate to a placebo or other medications. The analysis demonstrated that significantly more patients treated with divalproex sodium met their definition of success compared with those receiving placebo, with a number needed to treat (NNT) of 5 (95% confidence interval [CI], 4 to 8). Vertigo, nausea, and tremor were increased compared with placebo, whereas the frequency of other adverse effects (fatigue, weight gain, cumulative adverse effects) were not significantly different.3
Caveats: The 10 small trials in this review were judged overall to be at some risk of bias, often missing critical methodological elements such as allocation concealment (eight out of 10), double blinding (three out of 10), or complete reporting of outcomes (six out of 10). In addition, studies using varying forms of valproate and comparators were included and, in some cases, combined. The meta-analysis demonstrated superiority of valproate over placebo for migraine prevention. Similar to divalproex sodium, sodium valproate was found to have an NNT of 3 (95% CI, 2 to 9) for at least 50% reduction in migraine frequency compared with placebo. In two studies (one comparing divalproex sodium vs. propranolol and another evaluating sodium valproate vs. flunarizine) no significant differences in the proportion of responders were identified. Two small trials (n = 88) studying sodium valproate vs. topiramate (Topamax) found the latter medication to be slightly more effective, corresponding to a reduction of approximately one headache per 28 days with topiramate.3
Adverse effects are difficult to assess from these small studies. However, when assessed individually, statistically significant differences were identified for vertigo, nausea, and tremor (number to need to harm = 14, 7, and 14, respectively).3 Because migraine headaches tend to be more debilitating than these adverse effects, the data suggest good clinical tolerance with valproate and a favorable benefits-to-harms profile for migraine prevention. Of note, valproate medications are well-classified teratogens and physicians should consider this factor when treating women of reproductive age with these types of medications.4
Although further study is needed to better define the utility of this treatment compared with other medication options, the evidence suggests valproate medications are likely effective for preventing migraine, which agrees with current guideline recommendations.5 Therefore, we have classified this treatment as green, suggesting overall benefits outweigh harms. However, physicians should evaluate experienced adverse effects compared with prevention benefits for individual patients, because the data pertaining to adverse events are limited.
The original manuscript was published in Medicine by the Numbers, American Family Physician as part of the partnership between TheNNT.com and AFP.
Author: Adam R. Aluisio, MD, MS, and Shahriar Zehtabchi, MD
Published/Updated: November 1, 2016
The title bar is color-coded with our overall recommendation.