Abstract
Purpose::
toxic effects of topical medications on the ocular surface are documented by several in vivo and in vitro studies. However, few data are available about the incidence and the clinical characteristics of drug-induced toxic keratopathy (TK).This study aims to describe frequency, clinical presentation and management of drug-induced keratopathy.
Methods::
retrospective study on patients with corneal ulcer referred to our tertiary referral center from 2003 to 2005. Diagnosis was based on history, clinical findings, microbiological exams and, occasionally conjunctival biopsy. Toxic keratopathy was diagnosed in presence of negative microbiological exams, absence of identified causes and evidence of corneal healing after discontinuation of all topical treatments except of preservative-free artificial tears. History, clinical findings, visual acuity, symptoms and therapeutic regimen were recorded. The number and the class of topical drugs, the number of instillations per day and the number of days of treatment have been related to the healing time of corneal ulcer by Spearman rho test.
Results::
226 consecutive corneal ulcers were included: 70 (31%) were infective, 135 (60%) non-infective and 21 (9%) classified as toxic keratopathy (14 male,7 female; mean age 53± 20 years).Ulcerative TK was associated with corneal oedema (43%), superficial neovascolarization (33%), stromal reaction (19%) and deep neovascularization (5%). All patients complained symptoms of ocular surface discomfort. Patients were receiving 1 to 7 different topical drugs, (administration ranging from 5 to 25 times daily) for a mean time of 21.7±12.5 days. All ulcers completely healed after discontinuation of topical treatments (mean healing time: 9.2±5.9 days). Healing time was significantly related to the number of topical drugs used (p=0.029, rho=0.48), to the frequency of drug instillation (p=0.042, rho=0.047) and to the duration of treatment (p=0.049, rho=0.43), but not to any specific drug. At follow-up visits (10.8±11 months), 5 patients showed clear cornea (mean VA=1/1) and 16 patients showed cornea scarring (mean VA=0.4/1±0.3/1).
Conclusions::
Despite the increasing diagnostic and therapeutic tools, TK continues to represent a frequent and misdiagnosed disease. It is mandatory to perform a prompt diagnosis because cornea healing and outcomes are related to treatment duration and frequency of topical drug instillation.
Keywords: drug toxicity/drug effects • cornea: clinical science • keratitis