Combination therapy for chronic migraine

Neurology 2012;78:940–941 Chronic migraine (CM) is a frequent and challenging condition for practicing neurologists. It affects around 2% of adults and its prevalence in middleaged women reaches 5%.1 Management of CM is generally unsatisfactory and largely not evidencebased, in part because until recently CM has not been considered a true entity, but mostly a consequence of analgesic overuse. Evidence that distinguishes the conditions includes findings that only one-third of CM patients in the general population overuse symptomatic medications and that preventives for CM work better than analgesic withdrawal2 (and even in the absence of analgesic withdrawal3). These results support the idea that CM is a biological entity possibly due to a still unknown genetic background. As a direct consequence of this rapidly evolving concept, specific guidelines for clinical trials in CM were developed in 2008,3 which have allowed demonstration that topiramate4 and onabotulinumtoxinA5 are efficacious over placebo in CM. In fact, analgesic withdrawal, which should not be forgotten in the presence of medication overuse, improves CM together with these preventives. While very encouraging, these interventions may not be enough for numerous CM patients. From a mean of 20 headache days per month, reduction in controlled trials was between 2.5 and 6.4 days for topiramate and 7.7 days for onabotulinumtoxinA.4,5 What could we offer to patients with an insufficient response? Combinations of medications often provide improved effectiveness in other conditions. Not surprisingly, then, headache specialists have been trying to improve patients’ outcomes by combining 2 medications that are known to be efficacious in (episodic) migraine prevention, hypothesizing a synergistic effect of the mechanisms of action of different pharmacologic groups. Taking into consideration the data coming from open studies, many of us believe that some of those combinations, especially the ones joining a neuromodulator and a -blocker, help CM patients.6,7 In this issue of Neurology, however, Silberstein et al.8 are the first to pour cold water on this concept, showing that in conditions such as migraine, often using subjective evaluation measures, we have to be very cautious when extrapolating data from clinical practice. In a multicenter, randomized, placebocontrolled trial, Silberstein et al. found no additional benefit from combining therapeutic doses of propranolol LA with topiramate for treatment of inadequately controlled CM (at least 10 headaches/ month). In fact, the combination of topiramate and LA propranolol did not improve (vs placebo) any of an extensive number of efficacy endpoints at 3 or 6 months or quality of life, as measured by Migraine Specific Quality of Life or by the sensitive Migraine Disability Assessment test. Other variables, such as previous use of topiramate, analgesic overuse, or prior treatment for depression, did not predict an advantage of the combination therapy. Although the authors conclude that no safety concerns were identified with the combination, there was a tendency for patients on propranolol vs placebo to report more adverse events. This could explain why subjects assigned to placebo had, on average, a greater—but not statistically significant—improvement in quality of life at 3 and 6 months compared to subjects in the propranolol arm. Is this the end of combination therapy for our CM patients? Of course not. These negative results come from just one controlled trial. Even though it is fully understandable that the independent review board recommended the end of enrollment after the preplanned interim analysis of 171 patients (86% of the planned sample size), it is also fair to say that the final number of patients in the 2 arms (52 in the placebo and 64 in the propranolol group) was rather low to draw definitive conclusions, especially taking into account the high placebo effect. Just as a recent example, onabotulinumtoxinA required 2 large trials, with more than 2,000 patients, to demonstrate significant separation from placebo in the treatment of comparable CM patients. There are other limitations