It has been widely reported that nonsteroidal anti-inflammatory drugs (NSAIDs) are effective in the prevention and treatment of macular edema after cataract surgery.1 In addition, the ESCRS PREvention of Macular EDema study has recently shown that the combination of a topical corticosteroid and a NSAID is more effective than when either agent is used alone to reduce the risk for developing cystoid macular edema after cataract surgery in nondiabetic patients.A Complications related to topical NSAIDs are relatively uncommon; however, punctate keratitis, infiltrative keratitis, persistent epithelial defect, corneal ulceration, corneal melting, and corneal perforation have all been reported with NSAID use.2,3 Corneal melting has been associated with diclofenac, ketorolac, bromfenac, and nepafenac, mostly when local or systemic factors predisposed the patient to develop corneal melting.2–10
We report a series of patients who developed visually significant corneal complications after the combined use of topical ketorolac (Acular) and neomycin/polymyxin B sulfate/dexamethasone (Maxitrol) after uneventful cataract surgery and, in some cases, without any identifiable predisposing factor.
Three patients presented to our eye casualty over a period of several months, each complaining of eye pain and reduced vision in the early postoperative period after uneventful cataract surgery, after which they had routinely been prescribed topical ketorolac and neomycin/polymyxin B sulfate/dexamethasone. Two patients had corneal melting and 1 had an atypical keratopathy. A retrospective investigation of 32 hospital sites in London identified 970 patients who had used this medication combination during the preceding year and, of those, 13 patients (14 eyes [1.3%]) developed severe corneal complications. Five patients developed corneal melting and 8 patients developed prolonged corneal epitheliopathy, which comprised irregular epithelial hypertrophy and punctate epithelial staining. The UK Medicines and Healthcare products Regulatory Agency was immediately informed, as well as relevant pharmaceutical companies, the Royal College of Ophthalmologists, and the UK Ophthalmic Pharmacists Group.
Table 1 shows the patient demographics, corneal complication, eyedrop regimen, and medical background. All patients had surgery completed between January 2015 and March 2016 and ranged in age from 72 to 99 years. Three patients were men and 10 were women. Patients were managed by different doctors during their initial presentation to eye casualty, which led to some variation in the initial treatment they received. Those patients who developed corneal melting are discussed in further detail later.
Blood tests were performed, where possible, to screen for any underlying autoimmune process that might have impacted the development of the adverse effect; Table 2 shows the blood test results for 12 patients. One patient did not consent for further investigation. Ten patients were positive for both herpes simplex virus and varicella zoster virus immunoglobulin G on serological testing, and 1 patient was positive for varicella zoster virus alone. Owing to the high rate of positive serology in the general population, we considered these results as insignificant in relation to the pathogenesis of reported adverse effects. Patient 4, who had a 50% inferior corneal melt, had a mildly raised rheumatoid factor and positive antinuclear antibody despite no known medical diagnoses. Two other patients had a positive antinuclear antibody test; 1 developed a corneal melt and 1 had prolonged epitheliopathy. No negative controls were tested.
An 85-year-old man with primary open-angle glaucoma, ischemic heart disease, and atrial fibrillation underwent uneventful cataract surgery in the left eye. The preoperative visual acuity was 6/36. He had been prescribed a routine postoperative combination of Acular three times a day and neomycin/polymyxin B sulfate/dexamethasone 4 times a day for 4 weeks. Postoperatively, he withheld the Latanoprost eyedrops that he was using at night for his glaucoma. On day 20, he attended the eye clinic complaining of discomfort and blurred vision. Examination confirmed stromal thinning and an epithelial defect of the left cornea. The topical medications at this initial review were changed to preservative-free dexamethasone 0.1% and chloramphenicol 0.5%, both 4 times a day. At last follow-up at 3 months postoperatively, this patient remains on long-term ocular lubricants and has a visual acuity of 6/36.
A 94-year-old woman with rheumatoid arthritis, hypothyroidism, and Raynaud disease underwent uneventful left eye cataract surgery. Her preoperative visual acuity was counting fingers. She had been prescribed a routine postoperative combination of ketorolac 3 times a day and neomycin/polymyxin B sulfate/dexamethasone 4 times a day for 4 weeks. On day 11, she presented to the eye clinic with a sore red eye and was diagnosed with a large corneal epithelial defect and stromal thinning. On day 43, the cornea was debrided and she continues on preservative-free topical dexamethasone 0.1%, moxifloxacin, and long-term ocular lubricants. Her visual acuity at 3 months was 6/60. Blood tests showed a raised erythrocyte sedimentation rate of 52, which could be attributable to her age and rheumatoid arthritis.
A 78-year-old man with no systemic or ocular history underwent cataract surgery in the left eye, during which there was an anterior capsule tear. There were no further sequelae of this surgical complication, and an intracapsular posterior chamber intraocular lens was inserted as planned. He had been prescribed a routine postoperative combination of ketorolac 3 times a day and neomycin/polymyxin B sulfate/dexamethasone 4 times a day for 4 weeks. On day 6 postoperatively, he presented to the eye clinic with a large corneal epithelial defect and an 80% stromal melt (Figures 1, A, B, and C). The loose corneal epithelium was debrided at that time. Topical medication was changed to preservative-free dexamethasone 0.1%, moxifloxacin, and ocular lubricants. Three days later, a 4-week course of oral doxycycline and a reducing regimen of oral prednisolone were prescribed. This man continued on ocular lubricants and has a visual acuity of counting fingers. His preoperative corrected visual acuity was 6/12.
An 81-year-old woman with a history of blepharitis and hypertension had uneventful right eye cataract surgery. Her preoperative visual acuity was 6/12. She had taken 3 months of oral lymecycline preoperatively to control her blepharitis. She had been prescribed a routine postoperative combination of ketorolac 4 times a day and neomycin/polymyxin B sulfate/dexamethasone 4 times a day for 4 weeks. Four days postoperatively, she attended the eye clinic with pain and blurred vision. At this time, she was diagnosed with a corneal epithelial defect. Medication was changed to hourly topical moxifloxacin, ganciclovir ointment 0.03%, 400 mg oral acyclovir 5 times a day, and ocular lubricants. Over the course of 9 months, this patient had recurrent filamentary keratitis requiring debridement and corneal stromal melt (Figure 2). Fourteen months after the initial surgery, her visual acuity remains counting fingers and she remains on ocular lubricants, preservative-free dexamethasone 0.1% once daily, and acetylcysteine 5% 4 times a day.
A 72-year-old woman with osteoarthritis, thrombocytopenia, and asthma had uneventful right eye cataract surgery. Her preoperative visual acuity was 6/9. She had been prescribed a routine postoperative combination of ketorolac 3 times a day and neomycin/polymyxin B sulfate/dexamethasone 4 times a day for 4 weeks. Eight days postoperatively, she presented to the eye clinic with a 4-day history of blurred vision. She was immediately diagnosed with a corneal melt and central corneal perforation, which was sealed using standardized corneal gluing techniques. Four days later, the right eye became more painful, and she was diagnosed with endophthalmitis. A tectonic corneal graft was performed alongside routine endophthalmitis management. Her final visual acuity is perception of light.
This case series highlights the risk of drug-related complications following the use of routine medications after uneventful cataract surgery.
NSAIDs reduce inflammation through the inhibition of cyclooxygenases, reducing the synthesis of prostaglandins. In 1999, the ASCRS surveyed its members to investigate any complications that were associated with topical NSAID use postoperatively.B The results of this survey demonstrated that generic diclofenac was mostly associated with either corneal melting or punctate epitheliopathy.B Varying degrees of keratitis were reported with the use of Acular in that report, but no cases of corneal melting.B
Although the exact mechanisms leading to corneal melting with NSAIDs are not clearly identified, a proposed mechanism is the upregulation of matrix metalloproteinases that have proteolytic and collagenolytic activity and destroy the corneal extracellular matrix.2,11 It has also been reported that topical NSAIDs reduce corneal sensation through their effect on the corneal nociceptor response, which, though helping postoperative pain, can also result in delayed wound healing and re-epithelialization.2
It is well documented that some topical antibiotics and preservatives can be toxic to the corneal epithelium. Lazarus et al.12 studied ophthalmic preparations and the associated epithelial toxicity to the cornea. On reviewing antibiotic preparations, gentamicin and neomycin were significantly more toxic than tobramycin. With regard to preservatives, benzalkonium chloride (BAK) was the most toxic. Relevant to our case series, Acular contains both ketorolac and BAK, whereas Maxitrol contains neomycin and BAK.
In a rheumatology population, Maxitrol has been reported to be an independent causative factor in the development of corneal melts after cataract surgery.13 This highlights that when used even as a single agent, these topical medications can lead to adverse effects on the cornea. In our reported series, 3 patients had rheumatoid arthritis, 1 of whom developed corneal melting. A deficient tear film also has been reported as a risk factor for developing corneal melting if a precipitating factor such as a topical NSAID is used.14 Two patients in our case series had a diagnosis of dry-eye disease, and another patient used long-term preserved glaucoma eyedrops, which can independently adversely affect the quality of the ocular surface, specifically the corneal epithelium. Marcon et al. report a case of sterile corneal melting in a 78-year-old man with dry eyes who presented 2 months after cataract surgery.10 The patient had been using a combination of ketorolac and tobramycin/dexamethasone (Tobradex) 4 times daily. Other ocular and systemic risk factors recognized to predispose the cornea to NSAID-related adverse effects have been reported including keratoconjunctivitis sicca, limbal stem cell deficiency, neurotrophic keratitis, persistent epithelial defects, rosacea, Sjögren syndrome, and erythema multiforme, none of which were identified in our cohort of patients.5–7
Given the variety of patients in our cohort and across the literature, we conclude that the risk of developing adverse effects, including corneal melting, with the use of topical NSAIDs is multifactorial. This case series suggests that there may be a combined toxic effect on the cornea when using topical neomycin/polymyxin B sulfate/dexamethasone with ketorolac after cataract surgery. We recognize that these topical medications are widely and successfully used for a variety of clinical indications; hence, further experimental studies would need to be performed to investigate any potential mechanisms or causal relationships. In the meantime, we propose that the combined use of topical NSAIDs and other agents such as neomycin and benzalkonium should be considered with added caution, especially in populations at risk, to prevent the rare event of corneal melting and the associated permanent visual morbidity.
1. Hoffman RS, Brage-Mele R, Donaldson K, Emerick G, Henderson B, Kahook M, Mamalis N, Miller K, Realini T, Shorstein NH, Stiverson RK, Wirostko B. Cataract surgery and nonsteroidal anti-inflammatory drugs. J Cataract Refract Surg 2016;42:1368–1379
2. Prasher P. Acute corneal melt associated with topical bromfenac use. Eye Contact Lens 2012;38:260–262
3. Asai T, Nakagami T, Mochizuki M, Hata N, Tsuchiya T, Hotta Y. Three cases of corneal melting after instillation of a new nonsteroidal anti-inflammatory drug. Cornea 2006;25:224–227
4. Lin J, Rapuano C, Laibson P, Eagle R, Cohen E. Corneal melting associated with use of topical nonsteroidal anti-inflammatory drugs after ocular surgery. Arch Ophthalmol 2000;118:1129–1134
5. Flach A. Topically applied nonsteroidal anti-inflammatory drugs and corneal problems: an interim review and comment. Ophthalmology 2000;107:1224–1226
6. Guidera A, Luchs J, Udell I. Keratitis, ulceration, and perforation associated with topical nonsteroidal anti-inflammatory drugs. Ophthalmology 2001;108:936–944
7. Montes-Mollon M, Perez-Rico C, Beckford-Tongren C, Castro M, Pareja-Esteban J, Romero-Garcia A. Corneal melting and topical nonsteroidal anti-inflammatory drug (NSAID). A case report. Arch Soc Esp Oftalmol 2009;84:311–314
8. Khalifa Y, Mifflin M. Keratitis and corneal melt with ketorolac tromethamine after conductive keratoplasty. Cornea 2011;30:477–478.
9. Mian S, Gupta A, Pineda R. Corneal ulceration and perforation with ketorolac tromethamine (Acular®) use after PRK. Cornea 2006;25:232–234
10. Marcon A, Rapuano C, Tabas J. Tissue adhesive to treat 2-site corneal melting associated with topical ketorolac use. J Cataract Refract Surg 2003;29:393–394
11. Brejchova K, Liskova P, Cejkova J, Jirsova K. Role of matrix metalloproteinases in recurrent corneal melting. Exp Eye Res. 2010;90:583–590
12. Lazarus HM, Imeria PS, Botti RE, Mack RJ, Lass JH. An in vitro method which assesses corneal epithelial toxicity due to antineoplastic, preservative and antimicrobial agents. Lens Eye Toxic Res 1989;6:59–85
13. Lockington D, Sharma R, Mantry S, Ramaesh K. Maxitrol-induced corneal melts after routine cataract surgery in rheumatology patients. Clin Exp Ophthalmol 2015;432:188–189
14. Flach A. Corneal melts associated with topically applied nonsteroidal anti-inflammatory drugs. Trans Am Ophthalmol Soc 2001;99:205–212
OTHER CITED MATERIAL
“PRevention of Macular EDema After Cataract Surgery (PREMED) study,” presented at ESCRS Congress, Lisbon, Portugal, Journal of Cataract and Refractive Surgery, 2017
NSAID Adverse Reaction Report (Press release): American Society of Cataract and Refractive Surgery, USA, 1999