Aims To test the hypothesis that Acanthamoeba keratitis (AK) outcomes differ for different topical antiamoebic therapies (AAT) and to provide the detailed patient outcome data.

Methods A retrospective cohort study of 227 patients developing AK between 25 July 1991 and 10 August 2012. Inclusion criteria required a complete record of AAT treatment for both the primary outcome of a medical cure rate at 12 months and the secondary outcome of Snellen visual acuity ≤6/24 and/or surgical intervention. Analysis used multivariable regression to control for differences in baseline disease characteristics for both primary and secondary outcomes with unadjusted analyses for other outcomes. Subjects were categorised for analysis both by the AAT used at baseline and also by mutually exclusive AAT (patients exposed to all the drugs in each group, and no others, for some period). AAT categories were PHMB monotherapy, PHMB+diamidine, PHMB+chlorhexidine+diamidine, diamidine monotherapy and other AAT.

Results Analysis by baseline AAT showed no notable difference between treatments for both a medical cure at 12 months in 60.79% (138/227) or for a poor outcome in 49.34% (112/227). When AATs were analysed by mutually exclusive groups, PHMB monotherapy provided the best outcomes. These findings are subject to bias requiring careful interpretation. Overall cure rates for the 214 subjects with resolved outcomes were 94.27% (214/227), median time to cure 5 months (IQR 3.25–9.00 months) and range 1–26.24 months.

Conclusion PHMB 0.02% monotherapy for the initial treatment of AK is as effective as biguanide+diamidine combination therapy. Chlorhexidine monotherapy was too infrequent for comparison. The outcome data are the most detailed available.