Skip navigation
Acute Pain | Chronic Pain | General

Anticonvulsants for chronic pain

Clinical bottom line:

Anticonvulsants are useful in the management of trigeminal neuralgia, with a NNT of 2.6 (2.2 to 3.3) for 50% pain relief at 5-14 days. There is evidence that anticonvulsants are useful in managing diabetic neuropathy at two weeks, but there is currently little evidence available to assess longer term benefit. There is evidence that anticonvulsants are useful in migraine prophylaxis. There is preliminary evidence that anticonvulsants are useful in a number of other pain conditions, but this is based on single trials.

In all cases, the relatively high risk for minor adverse effects of anticonvulsants should be considered when prescribing.

Anticonvulsant drugs have been used in pain management since the 1960s, mainly for the relief of trigeminal neuralgia. Anticonvulsants are also prescribed as adjuvant drugs in other pain syndromes, and in combination with antidepressants, for example, in the treatment of post-herpetic neuralgia. Serious side effects are associated with these drugs, including death from haematological reactions. Impaired mental and motor function are common adverse effects.

Systematic review

McQuay, H.; Carroll, D.; Jadad, A. R.; Wiffen, P.; Moore, A. Anticonvulsant drugs for management of pain: a systematic review. BMJ. 1995; 311(7012): 1047-52. ISSN: 0959-8138.

Inclusion criteria were randomised, double-blind, controlled trials of anticonvulsants in analgesia; pain not induced experimentally or by other drugs.

Information was extracted from original reports, and odds ratios were calculated. These were used to calculate NNTs for at least 50% pain relief where possible.


Trigeminal neuralgia

Three placebo controlled trials of trigeminal neuralgia were included. These were all with carbamazepine. Three of three trials showed significant pain relief with carbamazepine with doses ranging from 400 mg/day to 2.4 g/day (for 5 to 14 days).

When these data were pooled and compared with placebo, the NNT was 2.6 (2.2 to 3.3), and the NNH for minor adverse effects was 3.4 (2.5 to 5.2).

Trials with active arms found carbamazepine to be significantly better than tizanidine, no different to tocainide and significantly less effective than pimozide.

Diabetic neuropathy

Three placebo controlled trials of diabetic neuropathy were included (one with carbamazepine and two with phenytoin). Two of three found significant improvement at two weeks compared with placebo. The third trial found no difference in mean scores in a 23 week treatment period. It was only possible to pool the data from the two significant trials. This gave an NNT of 2.5 (1.8 to 4.0) and a NNH for minor adverse effects of 3.1 (2.3 to 4.8).

Migraine prophylaxis

Three placebo controlled trials were included (carbamazepine 3 tablets/day for six weeks, sodium valproate 800 mg/day for eight weeks, and clonazepam one to two tablets/day for 60 days). Two of three trials found anticonvulsants to be significantly better than placebo. The clonazepam trial found no significant difference. It was only possible to pool data from the two significant trials. This gave an NNT of 1.6 (1.3 to 2.0) and an NNH of 2.4 (1.9 to 3.3).

Table: NNT and relative benefit for anticonvulsants compared with placebo for neuropathic pain conditions

Condition Number of trials Anticonvulsant improved/total Placebo improved/total Relative benefit (95%CI) NNT (95%CI)
Diabetic neuropathy 3 56/68 26/68 1.9 (1.4 to 2.7) 2.5 (1.8 to 4.0)
Trigeninal neuralgia 3 178/315 41/224 3.1 (2.3 to 4.1) 2.6 (2.2 to 3.3)
Migraine prophylaxis 2 63/74 17/77 3.7 (2.4 to 5.9) 1.6 (1.3 to 2.0)
Other pain syndromes 1 5/14 1/15 5.4 (0.7 to 40) not calculated
Other pain syndromes

One trial of phenytoin 300 mg/day for six weeks showed no benefit over placebo for irritable bowel syndrome.

One trial of carbamazepine up to 800 mg/day improved central pain after stroke in 5/14 patients compared with 10/15 on amitriptyline and 1/15 on placebo.

One trial of clonazepam (mean 0.375 mg per day) was significantly better than placebo at 30 days.

One trial showed intramuscular gold to be more effective than phenytoin for rheumatoid arthritis in a 24 week trial.

One trial showed phenytoin 200 mg/day to be as effective as buprenorphine and buprenorphine plus phenytoin 100 mg/day for cancer pain.

One trial showed carbamazepine plus clomipramine improved postherpetic neuralgia in 9/16 patients compared with 3/13 on transcutaneous electrical nerve stimulation.

Adverse effects

There were 16 drug-related trial withdrawals vs. 2 in the placebo groups (placebo controlled trials only) Where adverse effects were reported, the incidence was 25-50% in each trial. Drowsiness, dizziness and disturbance in gait were the most common problems.

Related topics