Randomised controlled trial of topical kanuka honey for the treatment of cold sores

ObjectivesTo investigate the acceptability of topical medical grade kanuka honey for treatment of recurrent Herpes Simplex Labialis (HSL) and the feasibility of an individual crossover design.DesignAn open-label, single group, crossover randomised controlled trial.MethodsFifteen participants aged 16 years or older with a doctor's diagnosis of recurrent HSL were randomised to receive, according to a crossover design, either 5% acyclovir cream or medical grade kanuka honey for their next two episodes of HSL. Randomised treatments were applied to the lips five times per day from the first symptoms of developing the cold sore, until the lesion resolved. The primary outcome was the lesion healing time, defined as the time from development of first symptoms of the cold sore to return of normal skin. A key secondary outcome was acceptability of the honey treatment on visual analogue scale (VAS) (0 = completely unacceptable to 100 = completely acceptable). Kaplan-Meier survival plots, estimates of median healing times, and Cox Proportional Hazards with a random effect for participants to take into account the crossover design compared time to healing between treatments.Results5/15 (33%) participants provided no data for either intervention, one provided complete data for only the honey intervention, and one provided data only on acceptability of the honey intervention. The Kaplan-Meier estimates of median healing time were similar for honey and acyclovir, both nine days. The hazard ratio for time to healing was 0.57 (95% CI 0.19 to 1.72), p = 0.32, favouring honey. The acceptability of honey was good, mean (SD) VAS 81.9 (24.2).ConclusionsTopical medical grade kanuka honey was highly acceptable to patients for the treatment of HSL. A large trial of kanuka honey in HSL may require an alternative recruitment site such as pharmacies, a parallel group rather than crossover design, and closer ongoing supervision of treatment and documentation of outcomes than used in our study.