Above all, do no harm

Optometric Management, Sep 2000 by Schmidt, Eric

When medications cause a corneal melt, do you know what to do?

We've devoted our careers to protecting and improving people's sight by any means necessary. Often this includes prescribing topical therapeutic agents. Occasionally though, as in this case, the very drugs used to alleviate a problem can do more harm than good. Read on to find out what you need to know about a potential medication danger.

THE CASE: Ms. Clark was always smiling and had a kind word for everyone. I enjoyed each visit with her immensely.

I'd treated Ms. Clark on four occasions for severe allergic keratoconjunctivitis. Twice it was so pronounced that I prescribed a steroid (fluorometholone acetate, eFLone) to quiet the corneal inflammation. The other two episodes resolved with olopatadine HCl (Patanol).

Ms. Clark had slowly growing cataracts. When I saw her last October, they'd progressed to limit her vision to 20/50 OD and 20/70 OS. Her eyes were fairly comfortable using just preservative-free Refresh 2 to 3 times a day, but her vision was blurry and now bothering her. I scheduled her for cataract surgery at our center in 3 weeks.

A bump in the road

One week before her surgery, Ms. Clark was back in my office. She was having problems with her allergies, and for the past few days her left eye had been irritated. It was red, sore and itchy, with some mucus discharge. She also said that it was sensitive to sunlight.

My examination showed that her vision was still 20/50 OD and 20/70 OS. The OD had trace bulbar injection, but no follicles or papillae. The cornea didn't stain with sodium fluorescein (NaFI); in fact, there were dry areas that exhibited negative staining scattered throughout it.

The OS revealed 2 bulbar injection with 1 papillae on the inferior and superior palpebral conjunctivae. The cornea OD showed a very heavy, diffuse pattern of hard, 3 punctate staining throughout the entire corneal epithelium. The anterior chambers were quiet and the cataracts were unchanged in both eyes.

These findings were similar to what Ms. Clark had shown earlier, except that she seemed more uncomfortable. I prescribed ketorolac tromethamine (Acular) OU q.i.d. to quiet the inflammation and make her feel better. I scheduled her for a follow-up in 6 days to make sure her cornea was clear before she underwent cataract surgery.

Still bumpy

Just 3 days later, Ms. Clark called, saying that she couldn't use the ketorolac tromethamine. She felt that it made her eyes redder and more sore and that her vision was blurrier. I asked her to come to the office that day. She was right - she had gotten worse. Her visual acuity (VA) was still 20/50 OD, but had decreased to 20/200 OS.

I re-examined her eyes and found that the conjuntivae were whiter than they'd been earlier, but that the cornea OS had developed a large melt centrally (see left), measuring 7 mm horizontally and 2 mm vertically and involving the visual axis. The defect stained almost completely with NaFl. I judged it to be about 50% thinned throughout the lesion. The rest of the cornea didn't stain.

Though it was only 50% thinned, this corneal melt was worrisome because of its size and central location. My immediate concerns were to stop its progress, prevent a corneal perforation and to minimize scarring, which would obscure vision.

The modus operandi

Corneal melts are defects of the cornea that arise quickly and progress rather slowly. They usually involve only the epithelium.

In the worst-case scenario, a stromal melt can progress posteriorly through Descemet's membrane and cause a perforation of the cornea. Corneal melts heal from the periphery toward the center. The initial task is to get the epithelium to return and cover the insult. After that, the threats of infection and perforation are greatly diminished. The melt will slowly fill in from the bottom up. A residual scar is possible but it's not inevitable.

Corneal melts were traditionally seen with rheumatologic conditions that affected the eye (rheumatoid arthritis, for example). In these "old" melts, the lesions were usually peripheral and associated with an immunemediated inflammatory response.

These peripheral melts were treated in ways ranging from lubricants to topical steroids to conjunctival resections or flaps. There was almost always an underlying systemic etiology.

Recently however, more than 200 cases of corneal melts with different characteristics have been reported. These "newer" melts aren't inflammatory in nature; no infiltrates were found, and they developed after use of the topical nonsteroidal anti-inflammatory drugs (NSAIDs) ketorolac tromethamine or diclofenac sodium. Most patients had dry eye before the melt.

Last year, the American Society of Cataract and Refractive Surgeons (ASCRS) issued a precautionary warning about a direct association between the rise of corneal melts and the use of NSAIDs for routine anterior segment surgery.

In looking more closely, it was found that most cases of corneal toxicity were associated with generic diclofenac sodium rather than with the name brand, Voltaren. As a result of this finding, this generic is no longer commercially available.