Therapeutic and optical keratoplasty in the management of Acanthamoeba keratitis: risk factors, outcomes, and summary of the literature.

Objective: To report the risk factors for and outcomes of therapeutic and optical keratoplasty in the management of Acanthamoeba keratitis (AK).

Design: Retrospective case series.

Participants: A total of 50 eyes of 196 patients with retrievable medical records, diagnosed with AK at Moorfields Eye Hospital, London, underwent keratoplasty between January 1991 and April 2012.

Methods: Patient demographics, initial clinical examination findings, and management details were collected. The ophthalmic characteristics of patients who underwent keratoplasty for AK were compared with those who did not. Patients undergoing therapeutic keratoplasty were compared with those undergoing optical keratoplasty for baseline characteristics, management details, and visual outcomes. A multivariate logistic model was used to derive the odds ratios of a poor visual outcome in all keratoplasty patients.

Main Outcome Measures: Poor visual outcome was defined as final visual acuity of 20/200 or worse. Secondary outcomes of interest included number of clinic visits and the need for additional intraocular surgery.

Results: Of the 196 AK patients, a total of 50 patients (25.5%) underwent penetrating or anterior lamellar keratoplasty, 10 of whom (20%) underwent repeat procedures. Of these 50 patients, 26 (52%) had therapeutic keratoplasty, predominantly for corneal perforation. The remaining 24 patients (48%) underwent optical keratoplasty for visual rehabilitation. Thirty-seven (80.4%) patients in the keratoplasty group initially were misdiagnosed as having herpes simplex keratitis versus 59 (41.8%) patients who did not require a keratoplasty (P < 0.001). Final visual outcomes were significantly better in the optical group compared with the therapeutic group, with 13 (54.2%) achieving visual acuity of 20/30 or better versus 7 (26.9%), respectively. On multivariate analysis, beginning therapy at a hospital other than Moorfields and undergoing a therapeutic, rather than an optical, keratoplasty were associated significantly with a poor visual outcome from keratoplasty.

Conclusions: The prognosis of keratoplasty differs markedly when performed for therapeutic purposes compared with visual rehabilitation. Where possible, keratoplasty should be delayed until such time as the eye is uninflamed and medically cured of Acanthamoeba.