Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Anticonvulsant drugs reduce pain in trigeminal neuralgia and diabetic neuropathy and are effective for migraine prophylaxis

ACP J Club. 1996 Mar-April;124:35. doi:10.7326/ACPJC-1996-124-2-035

Source Citation

McQuay H, Carroll D, Jadad AR, Wiffen P, Moore A. Anticonvulsant drugs for management of pain: a systematic review. BMJ. 1995 Oct 21;311:1047-52. [PubMed ID: 7580659]



To determine, using meta-analysis, the effectiveness and adverse effects of anticonvulsant drugs in the management of pain.

Data Sources

Studies were identified by searching MEDLINE (1966 to February 1994), by hand searching 40 medical journals and 9 specialty journals published between 1950 and 1990, by scanning bibliographies of relevant papers, and by contacting experts.

Study Selection

Studies were selected if they were randomized controlled trials of the analgesic effects of anticonvulsant drugs. Studies were excluded if they were about experimental pain or about anticonvulsant drugs used to treat pain produced by other drugs.

Data Extraction

Data were extracted about the pain condition and the number of patients studied; the anticonvulsant drug used and the treatment regimen; comparison treatment; duration of the study and follow-up; outcome measures and results; and minor and major adverse effects. Studies were scored independently for quality.

Main Results

37 studies were identified, and 20 met the selection criteria. The only placebo-controlled study in acute pain found sodium valproate to have no analgesic effect. 3 placebo-controlled trials investigated pain relief with carbamazepine for trigeminal neuralgia. At 5 to 14 days follow-up, more patients had improved with carbamazepine than with placebo (56% vs 18%, P < 0.001). {This absolute risk improvement (ARI) of 38% means that 3 patients would need to be treated (NNT) for 5 to 14 days to provide pain relief for 1 patient, 95% CI 2 to 3; the relative risk improvement (RRI) was 209%, CI 132% to 316%.}* More patients who received carbamazepine had adverse effects than did those who received placebo (47% vs 18%, P < 0.001). 2 placebo-controlled trials investigated anticonvulsant drug use in treating diabetic neuropathy. Anticonvulsant drug use led to more pain relief than did placebo (82% vs 43%, P < 0.001) {ARI 39%; NNT 3, CI 2 to 4; RRI 93%, CI 46% to 165%}*. More patients who received anticonvulsant drugs had adverse effects than did those who received placebo (38% vs 5%, P < 0.001). 2 placebo-controlled trials investigated anticonvulsant drug use for migraine prophylaxis. Anticonvulsant drug use led to more improvement than did placebo (85% vs 22%, P < 0.001) {ARI 63%; NNT 2, CI 1 to 2; RRI 286%, CI 158% to 502%.}* More patients who received anticonvulsant drugs had adverse effects than did those who received placebo (53% vs 11%, P < 0.001). Phenytoin had no effect on the irritable bowel syndrome, and carbamazepine had little effect on pain after stroke. Clonazepam was effective in 1 study for temporomandibular joint dysfunction.


Anticonvulsant drugs are effective for management of pain in trigeminal neuralgia and diabetic neuropathy and for migraine prophylaxis but are associated with adverse affects.

Source of funding: Oxford Regional Health Authority and Pain Research Funds.

For article reprint: Dr. H. McQuay, Oxford Pain Relief Unit, Churchhill Hospital, Oxford OX3 7LJ, England, UK. FAX 44-1865-226-060.

*Numbers calculated from data in article.


The management of pain is a major clinical challenge in the 1990s. The most recent reminder of this challenge is a study published in JAMA that shows inadequate pain control in many seriously ill patients hospitalized in several major medical centers (1). A rarely discussed barrier to the effective management of pain is the difficulty of treating neuropathic pain.

McQuay and colleagues have done an extensive review of the medical literature in an effort to show the efficacy of anticonvulsant drugs in the treatment of neuropathic pain. Their search produced only 20 randomized studies that were eligible for analysis, including 17 that dealt with chronic nonmalignant pain.

Although the results of this meta-analysis did show that anticonvulsant drugs are effective in the treatment of neuropathic pain, this finding is tempered by the high rate of nonresponse, 43%, in patients treated with carbemazepine.

Challenges to the analyses were raised by the authors on the lack of standardized data reporting across articles. Investigators in this field must improve the quality of their reporting, for example, by defining a clinically important level of pain relief (i.e., 50%) that would constitute a clear response to the medication under study.

The intent of this meta-analysis was to better define the role of anticonvulsant drugs in the treatment of neuropathic pain, and McQuay and colleagues have accomplished this. More important, they have succeeded in showing that much still needs to be done before clinicians can confidently provide effective management to patients with neuropathic pain.

Barry M. Kinzbrunner, MD
Vitas Health Care CorporationMiami, Florida, USA


1. SUPPORT Principal Investigators. A controlled trial to improve care for seriously ill hospitalized patients. JAMA. 1995;274: 1591-8.

Updated Commentary

Recently, a new anticonvulsant, gabapentin, has been reported to effectively treat neuropathic pain in a variety of clinical settings. Doses as low as 300 mg/day have been effective for some patients, with the common therapeutic range of the medication being 600 to 1800 mg/day (1). The medication should be started at low doses and increased over several days to reduce the risk for side effects, which include sedation and dysequilibrium. The advantage of gabapentin over carbamazepine and other more commonly used anticonvulsants in this setting is its lower toxicity profile. Gabapentin has been found to be particularly useful in the treatment of AIDS-related neuropathic pain because of its lack of interaction with antiretroviral medications and, unlike carbemazepine, its lack of bone marrow toxicity.

1. Weinreb NJ, Kinzbrunner BM, Clark M. Pain management. Chapter 6 in Kinzbrunner BM, Weinreb MJ, Policzer JS. Twenty Common Problems in End of Life Care. New York, McGraw Hill, 2001;91.