Skip navigation
Acute Pain | Chronic Pain | General

Treatment of postherpetic neuralgia

Clinical bottom line:

Tricyclic antidepressants are effective in relieving postherpetic neuralgia (odds ratio 0.15 (0.08 to 0.27)). Topical capsaicin is also associated with effective pain relief at the recommended dose (odds ratio 0.29 (0.16 to 0.54)).

There is currently insufficient evidence to determine whether transcutaneous nerve stimulation (TENS), benzodiazapines, antiviral agents, anti-prostaglandins and acupuncture are effective. Current evidence suggests that they are not, and these treatments should not be used. Weak evidence suggests that vincristine iontophoresis may have some benefit, but side effects are unpleasant. More proven treatments are therefore recommended.

This has been defined as pain persisting in the dermatomes affected by herpes zoster (shingles) after the disappearance of the rash caused by the infection.

Up to 15% of untreated patients have persistent pain one month after healing of an acute herpetic rash. One quarter of these (4% of total) still have pain at one year. The risk of postherpetic neuralgia increases sharply with age, and can be as high as 50% in patients aged over 60 years and 75% in those aged over 75 years.

Treatment during the acute phase has been covered in a separate review (Lancaster et al, 1995). This review considers treatment during the chronic phase of the condition.

Systematic review

Volmink J, Lancaster T, Gray S, Silagy C. Treatments for postherpetic neuralgia: A systematic review of randomized controlled trials. Family Practice. 1996; 13: 84-91. ISSN: 0263-2136.

Inclusion criteria were randomised controlled trials of postherpetic neuralgia, where pain had persisted for at least one month after onset of herpes zoster; at least two treatment groups.

Pain data were extracted. Expected event rates were subtracted from observed event rates for each trial. Estimate of effect size was calculated using odds ratios with 95% confidence intervals.


Tricyclic antidepressants

Five trials were included. Three trials compared a tricyclic (amitriptyline or desipramine) with placebo. Three of three trials demonstrated better pain relief with a tricyclic. The odds ratio was 0.15 (0.08 to 0.27). However, desimpramine was associated with some potentially serious side effects. One trial also examined maprotiline, and found no benefit over amitriptyline.

Transcutaneous electrical nerve stimulation

One low quality trial compared combination clomipramine and carbamazepine with transcutaneous electrical nerve stimulation (TENS). Combination therapy was better than TENS, with an odds ratio of 0.15 (0.03 to 0.7).


One trial compared lorazepam with placebo and amitriptyline. Lorazepam was not associated with any benefit: odds ratio 1.00 (0.24 to 4.18). It was associated with sedation and low mood, including onset of severe depression in some patients.


One small trial demonstrated no benefit of acylovir treatment compared with placebo; odds ratio 1.16 (0.21 to 6.47). However, group size may be too small to demonstrate an effect.

Topical capsaicin

Three placebo-controlled trials were included. Two of three trials demonstrated significant benefit of capsaicin (0.075% for six weeks) over placebo. The remaining trial was large, but used a weaker preparation (0.025% for four weeks). Capsaicin was associated with skin reactions which lessened during treatment.


One trial compared benzydamine with placebo. There was no significant benefit, with an odds ratio of 1.2 (0.37 to 3.92). Rashes were more frequent in benzydamine patients.


One small trial of transdermal vincristine iontophoresis showed significant benefit over saline-only placebo: odds ratio 0.05 (0.01 to 0.26). However skin irritation and painless burns were common in both groups.


One trial showed no benefit of acupuncture over placebo (mock TENS): odds ratio 0.92 (0.28 to 3.00). Failure to complete treatment was high because of the pain associated with acupuncture.

Related topics