Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Botulinum toxin for migraine prevention

Botulinum toxin RCT
Trial design

A randomised trial tells us that botulinum toxin injected into the forehead reduces migraine frequency or intensity [1]. How are we to interpret this? Is it the best thing since sliced bread, or does the evidence leave us with that 'ho..hum' feeling?

The first thing to do is ask what criteria are important for clinical trials of migraine prophylaxis. A review that draws heavily on the work of others helps thinking [2] because it examines important issues about trial design. Trials should meet several critical criteria, some obvious, some less so:

There may be other considerations for different treatments, but, simply put, trials lacking these characteristics should be treated with caution.

Botulinum toxin RCT

Trial design

The trial was randomised, double blind and compared placebo with two strengths of botulinum toxin over three months. We are told that non-completers were dismissed from the study, but not how many or what treatment. The trial would score 2 out of a possible 5 points on a common quality scoring scale in which trials scoring 2 or less may be subject to bias.

Adult patients had to have migraine diagnosed to International Headache Society criteria, with 2-8 episodes a month of at least moderate severity over a four-week baseline period. Patients could continue using concurrent prophylactic medicines. Exclusions were sensible, and included more than 15 headache days a month, and symptomatic medication overuse.

Treatments were injections of placebo vehicle, or 25U or 75U symmetrically into the forehead muscles. Follow up was over three months, during which patients kept diaries, including migraine frequency and intensity. Assessments were made at one, two and three months


There were 123 participants with 122 completers. Migraine frequency averaged 4.0 to 4.8 per month in the three groups. Baseline severity was the same.

For the favoured outcome of reduction in migraine frequency, statistical differences occurred only with 25U at three months, but not with 25U at one or two months, or with the higher dose of 75U at any time. Results were given for 50% reduction in migraine frequency or reduction in frequency of two episodes a month. Formal statistical significance was barely attained, with a p value of 0.46, though the relative risk included 1 (Table 1). Different results in one or two patients would have rendered the result nonsignificant (Table 1).

Table 1: Migraine frequency - 25U botulinum toxin and placebo. Effect of a single patient change in success (treatment or placebo groups) on results

At least 50% reduction in migraine frequency
Botulinum toxin Placebo Relative risk
(95% CI)
Actual results 19/42 10/41 1.9 (0.98 to 3.49)
One fewer with treatment 18/42 10/41 1.8 (0.93 to 3.35)
One more with placebo 19/42 11/41 1.7 (0.92 to 3.10)

For the secondary outcome of migraine severity, statistically significant benefit was attained for 25U at one and two months, but not at three months, and not at any time for 75U.

Adverse events occurred more frequently with 75U (50% of patients) than with placebo (25%), and we are told that with 25U the rate was not significantly different from placebo.


There was a dose response for adverse events, with higher rates with 75U. This dose at no point or with any endpoint showed a benefit compared with placebo. The 25U dose was only effective on migraine frequency, and then barely, at three months. On migraine severity statistical significance, again bare, was at one and two months.

The simple fact is that with one or two patients giving different responses, this would have been declared a negative trial. It does not inspire confidence, especially as this is the only randomised controlled trial for this intervention in this indication and the quality of reporting allows for the possibility of bias, as well as it being financed by the manufacturer. Given the not inconsiderable expense, justifying this to any purchasing authority would be a hard task.

The trial fulfilled most or all of the criteria for prophylactic interventions in migraine. The intervention just didn't make the grade.


  1. S Silberstein et al. Botulinum toxin type A as a migraine preventive treatment. Headache 2000 40: 445-450.
  2. V Limmroth, MC Michel. The prevention of migraine: a critical review with special emphasis on beta-adrenoceptor blockers. British Journal of Clinical Pharmacology 2001 52: 237-243.
previous or next story in this issue