I have a young, female patient who wears a contact lens in one eye but presented with diffuse corneal edema (800µm) in both eyes. She is taking amantadine for her Parkinson’s-type tremors secondary to use of risperidone. Is the amantadine responsible for her acute corneal edema?
“Bilateral corneal edema in a young, monocular contact lens patient is a head-scratcher,” says Brett G. Bence, OD, Director of Optometry at Northwest Eye Surgeons in Seattle.
Dr. Bence recently saw a patient with a similar presentation. The patient was a 30-year-old white female referred from her optometrist for a second opinion concerning recent corneal edema. She was a high myope (-18D), with a large angle esotropia and hand motion vision in the amblyopic eye.
She wore a soft contact lens in her better eye, with a visual acuity of 20/100. Her best-corrected visual acuity (VA) prior to corneal edema was 20/30. Her medical history was significant for bipolar disorder, anxiety, hypothyroidism, gastroparesis and exercise-induced asthma. Her medications included Nexium, risperidone, acetaminophen with codeine and trazodone.
“Review of her health history reveals an unfortunate but anticipated side effect of risperidone use—muscle tremors,” says Dr. Bence. “For treatment of these Parkinson-like tremors, she was prescribed 150mg amantadine BID. About two to three months prior to this visit, she independently increased the dose to 150mg three times a day to lessen the tremors without her prescribing physician’s knowledge.”
A reasonable differential diagnosis for corneal edema can include endothelial dystrophy; infectious and/or inflammatory causes, such as HSV endotheliitis, interstitial keratitis or keratouveitis; endothelial cell loss from trauma or chronic corneal hypoxia secondary to contact lens overwear; or toxic endothelial dysfunction from systemic medications.
In this particular case, corneal endothelial dystrophy seemed unlikely, “due to the patient’s young age, absence of clinical guttata and the seemingly abrupt onset,” says Dr. Bence.
Endotheliitis or keratouveitis from HSV was also unlikely, due to the absence of keratic precipitates, endothelial white cells, concomitant anterior uveitis/AC cells or prior HSV keratitis.
Aside from corneal stromal edema, the patient did not present with any signs of interstitial keratitis (e.g., peripheral-to-central corneal neovascularization, mild conjunctival injection, endothelial and stromal precipitates, and scarring).
“She had no history of corneal trauma, corneal surgery or endothelial damage,” says Dr. Bence. Her referring optometrist shared concern that contact lens overwear may have caused the corneal edema in the contralateral amblyopic eye, noting that she may not have reported contact lens wear on that eye. “However, the patient adamantly denied this,” he says.
The working diagnosis became toxic endothelial dysfunction secondary to amantadine use. Her prescribing psychiatrist was contacted and amantadine was discontinued. She was also asked to stop contact lens wear, use Muro 128 ophthalmic ointment (Bausch + Lomb) HS and initiate one drop Durezol (difluprednate, Alcon) TID OU.
Following one week of treatment, she showed slight improvement; however, after two weeks, pachymetry showed her corneal edema regressed significantly to 605µm in the amblyopic eye and 521µm in her better eye. Her best VA improved to 20/60 with glasses. Durezol use was tapered, but Muro 128 was continued.
“Interestingly, her optometrist reported that at follow-up a few weeks later, her refractive error had reduced from -18D to -16D,” he says. “This raises the possibility that chronic, perhaps subclinical, corneal edema was slowly altering the curvature and refractive error.”
Toxic corneal endothelial dysfunction from systemic medications is uncommon. In this particular case, amantadine induced bilateral endothelial dysfunction with secondary corneal edema.
“We can be lulled to sleep on some of these mystifying cases,” says Dr. Bence. “Reviewing side effects of systemic medications remains a vital consideration when ocular history and findings don’t immediately answer the question.”