Clinical Question: Which is more effective for the treatment of postherpetic neuralgia: opioids or tricyclic antidepressants (TCAs)?

Setting: Outpatient (any)

Study Design: Crossover trial (randomized)

Synopsis: Adults with neuropathic pain for at least three months after herpes lesions had resolved were recruited through physician referral and advertisements. Patients had to have pain of at least 4 on a scale from 1 to 10 to be included. Of 103 patients who were screened, 76 were randomized and 71 started the study, but only 44 completed all three study periods.

In this crossover trial, patients received an opioid, a TCA, or a placebo for each of three eight-week study periods, with a one-week washout period between phases. The order in which they received the drugs was randomly assigned using concealed allocation. Outcomes were assessed twice a week via telephone, and at a clinic visit at the end of each study period.

The opioid was sustained-release morphine (titrated to a maximal dosage of 240 mg per day) and was compared with nortriptyline (titrated up to 160 mg per day) or placebo in two or three divided doses per day. Patients who did not tolerate morphine could be switched to methadone, and those who could not tolerate nortriptyline were switched to desipramine. Patients who took at least one dose of a medication and gave a rating of pain were included for that study period. The primary outcomes were pain intensity (zero to 10), pain relief (zero to 100 percent), and cognitive function (using the Wechsler Adult Intelligence Scale and other validated measures).

The mean daily dosages were 91 mg for sustained-release morphine and 89 mg for nor-triptyline. Opioids and TCAs both were more effective at reducing pain intensity than placebo (1.9 versus 1.4 versus 0.2 points on the 10-point scale; P <.01). This is a clinically meaningful difference, but barely. The percentage of pain relief also was similar between opioids and TCAs, and both were better than placebo (38 percent versus 32 percent versus 11 percent; P <.001). TCAs slightly reduced some cognitive measures; opioids and placebo did not. Significantly more patients preferred opioids to TCAs (54 percent versus 30 percent). Constipation, nausea, and drowsiness were common in the patients taking opioids, and dizziness was common in the patients receiving TCAs. Most important, patients in the opioid group were more likely to drop out during or after the opioid phase (n = 20) than during the TCA (n = 6) or placebo (n = 1) phase.

Bottom Line: Opioids and TCAs provide similar pain relief for patients with postherpetic neuralgia, but TCAs are generally better tolerated. (Level of Evidence: 1b)