Predictors of a Negative Response to Topiramate Therapy in Patients With Chronic Migraine


  • John F. Rothrock MD,

  • Victoria A. Parada MD,

  • Renay Drinkard PA-C,

  • Richard M. Zweifler MD,

  • Kristin F. Key MPH

  • From the University of South Alabama, Mobile, Alabama.

Address all correspondence to Dr. John F. Rothrock, University of South Alabama Headache Center, 3401 Medical Park Drive, Building 3, Suite 205, Mobile, AL 36693.


Objective.—To identify variables predictive of a negative response to prophylactic therapy with topiramate in patients with chronic migraine.

Background.—While certain of the newer antiepileptic drugs (AEDs) have emerged as promising or definitely effective therapies for migraine prevention, we continue to lack biologic or clinical variables predictive of treatment response to these or other widely used prophylactic therapies.

Methods.—A consecutive series of 170 patients with IHS-defined migraine who were experiencing 15 or more days of headache per month were treated with topiramate according to a uniform dosing protocol. Variables examined for their potential value in predicting treatment response included age, gender, prior experience with prophylactic therapy, prior experience with divalproex sodium specifically, headache frequency and, if present, duration of chronic daily headache (CDH). A positive treatment response was defined as a 50% or greater reduction in headache days during the second treatment month relative to the patient's pretopiramate baseline. Only patients who completed the treatment phase and achieved the 50 mg BID target dose were analyzed (efficacy analysis). Each variable prospectively selected was evaluated in regards to treatment outcome via a paired t-test, and a multiple regression analysis of all variables subsequently was performed.

Results.—A total of 116 patients completed at least 60 days of treatment and consequently were available for analysis. In the efficacy analysis, 45 (38.8%) of the 116 responded positively to topiramate. Neither age nor gender influenced treatment response. Those patients with CDH of more than 6 months duration, patients who previously had tried and failed more than three prophylactic agents and patients who previously had failed to respond to divalproex sodium were more likely to be nonresponders, but after multiple regression analysis the only statistically significant predictor of a negative treatment response was CDH of more than 6 months duration (P < .001).

Conclusions.—Patients with chronic migraine who are treated with topiramate may respond positively at a rate approaching that reported from placebo-controlled trials involving topiramate or other AEDs administered to less severely afflicted migraineurs. Our analysis suggests that patients with chronic migraine least likely to respond to topiramate would be those with extensive and negative previous experience with prophylactic therapy, previous failure to respond to divalproex sodium, CDH, and, most notably, CDH of more than 6 months duration.