Statistics from Altmetric.com
Abuse of topical anaesthetic agents is a rare but potentially dangerous cause of toxic keratitis. Persistent use of topical anaesthetics may result in severe pain, hyperaemia, mucopurulent discharge, chronic epithelial defects, stromal infiltrates with oedema and neovascularisation, and even hypopyon or hyphaema.1 Permanent corneal scarring and visual loss may lead to penetrating keratoplasty and enucleation.2 Patients have obtained anaesthetic drops from eye care practitioners and pharmacists, from theft of the practitioners' offices, and over the counter in some developing countries.2–4 We present a case of toxic keratitis secondary to topical anaesthetic abuse that is unique in two aspects. Firstly, the patient received the drops directly from an emergency department physician and, secondly, the patient developed an ulcerative keratitis months after the cornea had re-epithelialised.
A 24 year old woman presented with complaints of decreased vision and severe eye pain bilaterally. Two weeks before presentation, she had a history of intensely pruritic allergic conjunctivitis. She was examined by a community hospital emergency department physician, who reportedly treated her for a corneal abrasion in the right eye. Two ophthalmologists subsequently treated her for presumed bilateral bacterial keratitis that worsened despite aggressive topical antibiotic therapy, and the patient was referred to the UMDNJ-New Jersey Medical School ophthalmology department. The patient was not a contact lens wearer, and she denied tap or pool water exposure. Past medical history was unremarkable.
On examination, visual acuity was hand movements in both eyes. The eyelids were oedematous and the palpebral and bulbar conjunctiva were severely hyperaemic. The corneas had relatively symmetrical 5 by 7 mm epithelial defects with underlying dense stromal ring infiltrates (Fig 1). The stromas were mildly thinned, and endothelial folds and fine keratitic precipitates were noted. Anterior chamber examination revealed a mild cell and flare reaction, but no hypopyon. Iris, pupils, and lens examinations were unremarkable, and retinal examination revealed normal macula, disc, and blood vessels in both eyes.
Initial differential diagnosis included bacterial, fungal, viral, and Acanthamoeba infection. However, after repeated review of the history, the patient admitted that the initial emergency department physician had given her a bottle of topical anaesthetic and told her “not to tell anyone.” The patient confessed to using the oxybuprocaine (proparacaine) drops every hour in both eyes over the previous 2 weeks. The anaesthetic drops were discarded, and treatment with erythromycin ointment, diclofenac drops, and homatropine 5% was instituted. A bandage soft contact lens reduced the patient's pain. Corneal cultures, performed to rule out concurrent or subsequent infection, were negative. Over the next week, visual acuity improved to 20/80 in the right eye and 20/40 in the left, though residual bilateral 4 mm epithelial defects with stromal rings remained. The patient was then lost to follow up for 2 months.
Follow up examination revealed a residual epithelial defect in the right eye, and healed epithelium in the left eye. Over the next month, the epithelial defect resolved in the right eye, but bilateral ring-like stromal haze persisted. Four months later the patient presented with a new 2 by 3 mm epithelial defect with underlying stromal infiltrate and neovascularisation in the left eye (Fig 2). The patient vehemently denied any further use of topical anaesthetic drops. Cultures were performed, and treatment with ofloxacin and erythromycin was instituted. The cultures were negative, and the epithelial defect slowly resolved. The underlying corneal stroma remained opacified and neovascularisation progressed. At last examination, visual acuities were 20/40 in the right eye, and counting fingers at 1 foot in the left eye.
Topical anaesthetic agents are routinely used in ophthalmic examinations and procedures. The most commonly used preparations include oxybuprocaine and amethocaine (tetracaine). They are topical ester anaesthetics with fast onset of action (10–20 seconds), and relatively short durations of action (10–15 minutes).1 The agents act by interfering with sodium transmission through neuronal membranes, thereby stabilising the transmembrane potential, inhibiting the generation of action potentials, and blocking the sensation of pain.5
The toxic effects of topical anaesthetics are both direct and indirect. Epithelial cell organelles, desmosomes, and cytoskeletal elements can be directly damaged by these agents, altering cellular metabolism and function.4,6,7Indirectly, loss of epithelial microvilli can cause tear film instability, promoting corneal desiccation and inhibiting re-epithelialisation.4,6,7 Secondary neurotrophic changes may then occur.
Electron microscopy of corneal endothelial cells exposed to topical anaesthetics shows loss of the normal mosaic of hexagonal cells and abnormalities of the intercellular junctions and apical attachments.3 Immune complexes released from the cornea may be recognised by the body as antigen, causing the Wessely immune ring infiltrate seen clinically.4,5 A combination of these mechanisms may result in the clinical manifestations presented in our case.
Patients who have abused topical anaesthetic agents often present in extreme pain. There may be eyelid oedema with mucopurulent discharge and conjunctival hyperaemia. Like our patient, large chronic corneal epithelial defects with underlying stromal infiltrates are often initially misdiagnosed as Acanthamoeba or Gram negative bacterial infection.3,6 The stroma may be oedematous or thinned. Keratitic precipitates and an anterior chamber reaction are common. Cultures are most often negative, but superinfection may occur. Infectious crystalline keratopathy, which is usually associated with concomitant use of topical corticosteroids, has been reported in association with topical anaesthetic abuse.2 In that report, Kintner et al diagnosed Streptococcus viridans by corneal biopsy.2 Concurrent or subsequent infection with Pseudomonas and Bacillus has also been reported.5 Stromal neovascularisation and scarring complicate the healing and may be responsible for permanently decreased visual acuity. Patients may undergo conjunctival flaps, penetrating keratoplasty, and even enucleation for intractable pain after anaesthetic abuse.4,5
We are unaware of previous reports of a late occurring corneal ulcer after resolved toxic keratitis, and could find no reference to it in a computerised search utilising Medline. While the initial visual acuity improved to 20/40 in the left eye, late ulceration resulted in stromal scarring and neovascularisation that ultimately led to the poor visual acuity. The patient now requires penetrating keratoplasty to restore vision. The initial epithelial toxicity may have caused unstable epithelial attachment to the underlying stroma, leaving it susceptible to erosion and ulceration. Close long term follow up is necessary for these patients.
Patients who abuse topical anaesthetic agents obtain the drugs in various ways. In a report from Saudi Arabia, several patients sustained toxic keratopathy after the anaesthetic was dispensed over the counter.3 In countries where dispensing is more restricted, patients have been inappropriately prescribed topical anaesthetics by optometrists, ophthalmologists, and even otolaryngologists.2 Theft of the drugs from practitioners' offices may also contribute to the abuse.2 An association between topical anaesthetic abuse and psychiatric disturbances has been reported.2 In our case, the patient vehemently insisted that an emergency department physician gave her the oxybuprocaine. If true, then the dangers of topical anaesthetic abuse need to be more widely understood throughout the entire medical community. This information may ultimately lead to a decreased incidence of this rare but preventable entity.
Presented as poster at the American Society of Cataract and Refractive Surgery Meeting, Boston, MA, USA May 2000.
The authors have no financial interest in any product mentioned in this report.