Oral Acyclovir Therapy Accelerates Pain Resolution in Patients with Herpes Zoster: A Meta-analysis of Placebo-Controlled Trials

Meta-analysis of four double-blind, randomized, placebo-controlled trials of oral acyclovir (800 mg five times daily) for the treatment of herpes zoster was conducted to provide definitive assessments of the effect of acyclovir on the resolution of zoster-associated pain. The studies involved a total of 691 patients, and the analysis was performed on an intent-to-treat basis. A range of milestones of pain cessation were evaluated by means of Cox regression models with adjustment for relevant prognostic factors. The proportion of patients with postherpetic neuralgia at 3 and 6 months was also determined. Advancing age and more severe pain at presentation were associated with more prolonged pain. Acyclovir was clearly shown to accelerate pain resolution by all of the measures employed. Benefit was especially evident in patients 50 years of age or older. Fewer acyclovir recipients had postherpetic neuralgia at 3 or 6 months. Overall, the reductions of pain duration and prevalence were approximately twofold. The pain associated with herpes zoster is the principal reason why patients seek medical care [1]. The pain of herpes zoster has been variably described [2, 3], although by any definition the prevalence decreases in the 6 months after acute zoster. Traditionally, pain present while the rash persists or in the first 30 days from the onset has been termed acutepain. Pain present after these time points is generally termed postherpetic neural­ gia. However, for many patients who experience continuous pain, it may be impossible to regard it other than as a contin­ uum. The term zoster-associated pain [1] has therefore been adopted for the analysis of clinical trials evaluating antiviral agents for the treatment ofherpes zoster because it circumvents the need to select one ofseveral arbitrary definitions ofposther­ petic neuralgia [4-6] and it facilitates application of statistical principles, including intent-to-treat analysis (i.e., the inclusion of all randomized patients). For instance, a definition of post­ herpetic neuralgia that depends on a start time related to rash healing introduces inherent bias because studies have consis