Skip navigation

Anticonvulsants for migraine prophylaxis

 

Clinical bottom line

A number of anticonvulsant drugs have been tested in over 1,300 patients for the ability to reduce the frequency of migraine attacks. The proportion of patients with attack frequency reduced by 50% doubles with anticonvulsants, and half the patients treated have at least this level of relief. The number needed to treat for one to have migraine frequency cut by half is 3.8 (3.2 to 4.6).

Reference

E Chronicle, W Mulleners. Anticonvulsant drugs for migraine prophylaxis. Cochrane Database of Systematic Reviews 2004, issue 3.

 

Background

Anticonvulsant drugs were considered to be useful in reducing the number of migraine attacks in the mid 1990s. Subsequently a number of trials with different drugs have been conducted.

Methods of the review

Results

Ten good quality trials reported at least 50% reduction in migraine frequency as an outcome in trials comparing anticonvulsant with placebo. Results of individual trials are shown in Figure 1. There was a high degree of consistency despite four different anticonvulsants being used.

Figure 1: Individual trials comparing anticonvulsant with placebo for at least 50% reduction in migraine

Table 1 has the results for all drugs combined, and for the individual drugs, which were used at common anticonvulsant doses. Most information was available for divalproex sodium/sodium valproate, and topiramate, with numbers needed to treat of about 4. Overall 46% of patients had effective reduction in migraine frequency with anticonvulsants, compared with 20% with placebo, with a relative benefit of 2.3 (2.0 to 2.8) and an NNT of 3.8 (3.2 to 4.6).

Table 1: Combined results, and for individual drugs

 
Number of
At least 50% reduction in frequency with
  
Anticonvulsant
Trials
Patients
Anticonvulsant
Placebo
NNT
(95% CI)
Divalproex/valproate
5
647
43
21
4.5
(3.5 to 6.6)
Topiramate
3
514
50
22
3.5
(2.8 to 4.9)
Carbamazepine
1
93
58
10
2.1
(1.6 to 3.3)
Gabapentin
1
87
46
16
3.3
(2.1 to 8.4)

 

Eight trials with 841 patients indicated that 1.4 migraine attacks were prevented every 28 days with use of anticonvulsant.

Adverse events

Adverse events recorded commonly included nausea (NNH 7), asthenia/fatigue (NNH 12), tremor (NNH 12), weight gain (NNH 16), and dizziness/vertigo (NNH 16). There was considerable variability between trials. Overall 14% of patients withdrew because of adverse events with anticonvulsants.

Comment

Anticonvulsants are effective in reducing the frequency of migraine attacks, though adverse events are a problem. These may be reduced by titration of doses over a period of time, or by changing anticonvulsants if they are a problem. Little additional information has become available since 2004.