Background: Eye disease due to herpes simplex virus (HSV) commonly presents as epithelial keratitis.
Objectives: To compare the relative effectiveness of antiviral agents, interferon, and corneal débridement in the treatment of acute HSV epithelial keratitis.
Search strategy: We searched CENTRAL (The Cochrane Library 2010, Issue 4), MEDLINE (January 1950 to October 2010), EMBASE (January 1980 to October 2010), Latin American and Caribbean Literature on Health Sciences (LILACS) (January 1982 to October 2010), Zetoc (British Library's Electronic Table of Contents), System for Information on Grey Literature in Europe (openSIGLE), Biosciences Information Service (BIOSIS), the metaRegister of Controlled Trials (mRCT) (www.controlled-trials.com), ClinicalTrials.gov (www.clinicaltrials.gov), Japan Information Center of Science and Technology (JICST-EPlus), and China Academic Journals database (CAJ) via China National Knowledge Infrastructure (CNKI) with citations confirmed using China/Asia On Demand (COAD). There were no language or date restrictions in the search for trials. All databases except CNKI and COAD were last searched on 27 October 2010, CNKI and COAD were searched on 1 April 2010. We also searched literature digests, conference proceedings and reference lists.
Selection criteria: Of 152 eligible studies,106 comparative treatment trials involving 5872 eyes with dendritic or geographic epithelial keratitis were analysed for corneal healing over two weeks.
Data collection and analysis: Interventions were compared at 14 days after trial enrolment by calculating a risk ratio (RR) that was adjusted with indirect RR, assessed by an inconsistency index (I(2) ) and supplemented by a seven-day RR and a hazard ratio (HR).
Main results: Idoxuridine, though uncertainly better in healing outcome than control because of few trials with 14-day follow up, allowed earlier corneal re-epithelialisation. Vidarabine resulted in a significantly better outcome than placebo in one trial (RR 1.96; 95% CI 1.10 to 3.49). Compared to idoxuridine, in combined direct and indirect analyses, vidarabine (RR 1.11; 95% CI 1.03 to 1.19), trifluridine (RR 1.31; 95% CI 1.20 to 1.42), acyclovir (RR 1.23; 95% CI 1.16 to 1.31), brivudine (RR 1.38; 95% CI 1.18 to 1.61), and ganciclovir (RR 1.40; 95% CI 1.25 to 1.57) were significantly more effective. Trifluridine (RR 1.12; 95% CI 1.04 to 1.21) and acyclovir (RR 1.11; 95% CI 1.05 to 1.19) appeared more effective than vidarabine. No significant differences were found in comparisons between acyclovir, trifluridine and brivudine. The comparison of ganciclovir to acyclovir was limited by heterogeneity and possible publication bias. The joint use of two topical antivirals (RR 1.00; 95% CI 0.89 to 1.12) and the use of oral acyclovir alone (RR 0.92; 95% CI 0.79 to 1.07) or combined with a topical antiviral (RR 1.08; 95% CI 0.99 to 1.17) appeared as effective as topical antiviral therapy. Compared to antiviral monotherapy, the combination of an antiviral with interferon (RR 1.03; 95% CI 0.99 to 1.07) or with débridement (RR 1.04; 95% CI 0.95 to 1.14) did not yield significantly better outcomes but may have accelerated healing. The corneal epithelial healing outcome was improved when antiviral therapy was added to débridement (RR 1.21; 95% CI 1.04 to 1.42).
Authors' conclusions: Trifluridine and acyclovir are more effective than idoxuridine or vidarabine, and similar in therapeutic effectiveness. Brivudine and ganciclovir are at least as effective as acyclovir. While not improving outcome, the combination of interferon and an antiviral agent may speed healing. The effectiveness of corneal epithelial débridement is improved by an antiviral agent.