Purpose
To determine the impact of topical corticosteroid use, before the diagnosis of Acanthamoeba keratitis (AK), on visual outcomes.
Methods
A medical record review of patients diagnosed with Acanthamoeba keratitis at Moorfields Eye Hospital, London between January 1991 and April 2012 was conducted. Patient demographics, clinical findings and management details were collected. Suboptimal visual outcome was defined as final visual acuity ≤ 20/80, corneal perforation, or need for keratoplasty. For bilateral cases, one eye was randomly excluded from analysis. A multivariate logistic regression model, optimized for the use of steroids prior to diagnosis, was used to generate odds ratios for a suboptimal visual outcome.
Results
Data for 209 eyes of 196 patients were available. AK was diagnosed with culture (94 eyes, 48.0%), histopathology (27 eyes, 13.8%), confocal microscopy (38 eyes, 19.4%) or a typical clinical course and treatment response (37 eyes, 18.9%). Information on visual outcomes and prior steroid use was available for 174 eyes (88.8%). In multivariate analysis, prior steroid use was associated with poorer vision (odds ratio, OR 3.90; 95% confidence interval, CI, 1.78-8.55). Additional independent risk factors for poorer visual outcomes were older age (60+, (OR 8.97; 95% CI 2.13-37.79) and late stage disease at presentation (OR 5.62; 95% CI 1.59-19.80).
Conclusions
Steroid use prior to the diagnosis of AK, most often due to the misdiagnosis of herpetic keratitis, is predictive for a suboptimal visual outcome. Incorporating AK in the differential diagnosis of suspected herpetic keratitis is essential, particularly in contact lens wearers, in older individuals and in those presenting with progressive/indolent keratitis. Topical steroid use should be avoided until Acanthamoeba has been excluded. Accurate diagnosis requires culture, PCR and confocal microscopy.
Keywords: 402 Acanthamoeba •
487 corticosteroids •
462 clinical (human) or epidemiologic studies: outcomes/complications