Bromfenac-induced neurotrophic keratitis in a corneal graft

Discussion

Topical NSAIDs have proved their efficacy in many conditions like controlling postoperative pain and inflammation, preventing development of cystoid macular oedema following cataract surgery, maintaining intraoperative mydriasis and also off-label use for ocular inflammatory conditions like allergic conjunctivitis, pingueculitis and episcleritis.1–3 Being an effective alternative to topical steroids, absence of side effects like cataract and glaucoma makes its use quite prevalent in ophthalmic practice. Topical NSAIDs and steroids have shown similar efficacy in controlling inflammation after cataract surgery, hence making the former a safer, but not inferior, alternative.4

As with almost all available medications, they are not devoid of adverse reactions, which range from mild symptoms like redness, transient burning and stinging sensation to more serious ones like superficial punctate keratitis, epithelial defects, corneal melts, ulcerations and perforations.5 Mechanisms for such inadvertent side effects associated with topical NSAIDs are not clearly elucidated; speculated hypothesis includes activation of matrix metalloproteinases, impairment of wound healing and neurotrophic effect, resulting from analgesic action of these drugs, the latter being unlikely in the absence of ocular surface disease.6 7

Bromfenac is a phenylacetic acid derivative that has been approved by the Food and Drug Administration as 0.09% ophthalmic solution for two times per day instillation in the treatment of pain and inflammation after cataract surgery.6 Bromfenac is considered comparatively safer than other NSAID drops with regards to ocular surface complications. Reports of corneal melt associated with bromfenac are rare, with less than four cases reported in the literature.8 However, this complication is frequently associated with use of alternative NSAIDs like diclofenac and ketorolac.5 9 10 Although the use of NSAIDs is implicated in corneal melts in many case reports, development of neurotrophic keratitis following use of these agents has not reported as yet. Risk factors like concurrent use of topical steroids, undiagnosed dry eye disease,11 previous history of ocular surgery like pterygium surgery using bare sclera technique, preservatives present in topical medications, immunological disorders like rheumatoid arthritis12 or Sjogren’s syndrome13 and systemic diseases like diabetes mellitus can all contribute to the multifactorial pathogenesis of PED in these cases.

Neurotrophic keratitis is characterised by epithelial breakdown that progresses from punctate keratitis and focal epithelial loss in early stages (stage 1) to persistent non-healing epithelial defects (stage 2), stromal ulceration and melting in late stages (stage 3).14 The very fact that the anterior stroma was affected indicated an advanced stage of neurotropic keratitis in our patient, which warranted immediate intervention with cessation of the inciting bromfenac eyedrops.15 The possible factors that led to the development of NK in our patient can be attributed to trigeminal damage secondary to iatrogenic injury, chronic use of preserved eyedrops for glaucoma with the concurrent use of topical steroids and bromfenac solution adding to the mishap. Reduced corneal sensation, secondary to healed viral keratitis and multiple corneal transplantation, added a neurotrophic component to the epithelial defect. The cascade of inflammatory events accompanying the PED and the healing response which followed eventually led to a central macular corneal opacity.

The very first step in management of neurotrophic keratitis involves cessation of all topical medications, and starting the patient on preservative-free lubricants.15 The rationale behind the former is to avoid potential toxicity to the epithelium by the active compound and/or preservatives used in various eye drops. If drug-induced, as in our case, cessation of the inciting agent brings the destructive pathogenic processes to a halt, allowing repair of the damaged regions by normal healing response of the eye. Topical medications, like antibiotics and antiglaucoma medications in our patient, should be used in preservative-free formulations, if deemed absolutely necessary. These medications were started in our patient considering the high risk of infection, owing to advanced stage of disease and stromal involvement. Topical steroids should not be used in an attempt to control the inflammatory damage, as it impairs stromal healing and increases the risk of corneal melt and perforation. Surgical modalities like tarsorrhaphy, use of conjunctival flap, tissue adhesive or amniotic membrane transplantation have been described to improve tissue healing and prevent perforation.16 These methods were not considered in our patient, owing to the satisfactory response seen with conservative management, including topical lubrication, antibiotics and night time eye patching.

This unfortunate complication could easily have been avoided by proper communication between the three parties involved: the patient, the practitioner and the pharmacist. The very fact that doctors are stereotyped and take the brunt of jokes for their illegible handwriting is nothing to be laughed at or felt proud of; it may sometimes lead to improper medication and dosage,17 and occasionally fatality.18 The pharmacists are not less to blame, with lack of appropriate qualification and unethical practices not uncommon in India.19 Patients need to be aware of these fallacies prevalent in medical practice, as it is them who use these medications, reap their benefits and suffer their deleterious effects. Doctors need to educate the patients regarding the medications and their use; this will go a long way in preventing such avoidable mishaps, alleviating some load off the healthcare system.