Source: Li Q, Chen N, Yang J, Zhou M, Zhou D, Zhang Q, He L. Antiviral treatment for preventing postherpetic neuralgia. Cochrane Database Syst Rev. 2009 Apr 15;(2):CD006866. Review. PubMed PMID: 19370655.
Efficacy Endpoints: Incidence of PHN 6 months after zoster diagnosis, pain severity score at 6 months
Harm Endpoints: Serious (life-threatening or requiring prolonged hospital admission) or non-serious medication adverse events
Narrative: Post-herpetic neuralgia (PHN) is a syndrome of continued and chronic pain after the development of shingles, which is a reactivation of the varicella zoster (chickenpox) virus. The likelihood of developing shingles increases with age, with estimates of 10-20% incidence throughout one’s lifetime. The risk of developing post-herpetic neuralgia also increases with age, and is rare in patients under 30. PHN is exquisitely painful and debilitating, and unfortunately few effective treatments exist. It is theorized that treatment with anti-viral drugs before the PHN develops may reduce the risk of PHN developing.
This systematic review includes 1211 patients in 6 trials. All trials included patients with zoster for 72 hours or less, and trials had many exclusion criteria, including immunocompromised state. Most trials used acyclovir, while one trial used famciclovir. The review fails to find a benefit to the acyclovir at 4 or 6 months, though interestingly at one month the analysis does suggest a benefit. The one trial that used famciclovir (at two different doses) also failed to show a benefit, with the Forrest plots in one case favoring placebo at a level that achives statistical significance.
Harms were uncommon in both arms of all trials and did not reach statistical significance.
The Cochrane review concludes that acyclovir is not helpful in the prevention of PHN, but that larger trials with famciclovir may be useful, especially when looking at subgroups like age at time of zoster diagnosis or immunocompetence status.
Caveats: Acyclovir has very well documented compliance issues, and this is no surprise when considering its recommended five-times-a-day dosing regimen. This may have led to compliance/adherence issues among patients given acyclovir, affecting efficacy In addition, the apparent suggestion of a benefit at 1 month is interesting, and may indicate that antivirals delay, rather than reducing, the incidence of PHN.
It is worth noting that the fairly robust number of subjects overall and narrow confidence intervals in this review suggest that there is no benefit to antiviral agents, however these data do not exclude the possibility of a small benefit that has been, as of yet, undetected. If this benefit exists it may also therefore be too small to be clinically meaningful.
Harms of antivirals have tended to be uncommon in trial data, although most trial data for these agents are like the trial data reviewed here —they have been generated in attempts to detect benefits rather than harms. It often takes large post-marketing datasets to detect dangerous or important harms of such compounds. Because few if any pharmacologic agents are truly devoid of biologic harms, it seems likely that harms do exist. This should be considered by clinicians and patients in deciding whether to utilize agents that appear to have either no benefit or else a theoretical benefit that trial data have yet to detect.
Author: Graham Walker, MD
Published/Updated: November 23, 2010
The title bar is color-coded with our overall recommendation.
If you have suggestions, requests, or questions about a particular NNT review, please send us a message and we’ll try to address it as soon as possible.