SPECIAL CORNEA ISSUE

Diagnosing & Managing Corneal Infections

Taking the proper steps pays off for patients.

BY JAMES McCULLEY, MD

Ophthalmology has the good fortune of being a medical specialty in which acute care is the exception rather than the norm. Although we've all seen our share of retinal detachments, traumatic cataracts and corneal foreign bodies, by and large the patient base of a comprehensive ophthalmologist is dominated by conditions of slow onset and gradual progression, thus affording us ample time to conduct appropriate diagnostic tests and to tailor the clinical management over time.

However, in acute corneal infection, our triage instincts must come into play during the initial encounter: time is of the essence, and our early decisions can weigh heavily on the final outcome for the patient.

Diagnosis and management of a corneal infection must be done on a case-by-case basis. While there are certain principles that are always followed, there are no hard-set rules to guide treatment. Below are the steps I take when a patient presents with a corneal infection.

Signs in a Patient's History

When the patient presents with a corneal infiltrate and other signs of ocular inflammation, the first task is to determine whether it is a sterile or infectious process. This can be accomplished, in part, by history and appearance. Items in a patient's history that relate to exposure (the environment in which he lives and general health status) influence my initial considerations. First, I find out if there has been exposure or trauma, and, if so, the nature of the precipitating event.

If it's exposure to vegetable matter, I'll be concerned about fungi, either alone or in a mixed infection. Exposure trauma raises questions and concerns, especially when it involves vegetation. If the patient is a contact lens wearer, I'm going to be most concerned about possible Pseudomonas or Acanthamoeba infection, but there are also quite a few minor corneal infiltrates associated with contact lens wear that may in fact not be truly active infectious processes (e.g., a single non-light blocking paracentral infiltrate is of less risk).

Herpes simplex virus dendritic ulcer.

Another concern arises if the patient is a healthcare worker; if so, hospital-acquired MRSA is a consideration. If the patient has been in a household or in an environment where community-acquired MRSA is a greater concern — for instance jail, military barracks, or if he participates on a sports team — situations where there's a good deal of human-human, human-surface or surface-human exposure, these would raise a red flag for possible community-acquired MRSA.

The ophthalmologist must also find out whether the patient is immune-challenged or if the immune system is very much intact. In immunocompromised patients, I'd worry more about an opportunistic type of infection. For example, AIDS patients have increased risk for some of the more unusual types of infections like microsporidium, which requires a very different kind of therapeutic approach.

Corneal Abrasion as Route of Infection

When a corneal epithelial defect provides a route for an invading infection, your treatment approach must target the indigenous ocular flora, which can be highly variable among patients. If the patient is living commensally with some significant pathogens in his flora, he can have a very significant infection from that flora. An abrasion that gets infected can be just as bad as serious trauma that creates an infection. Most organisms require a break in the epithelium or need to compromise the epithelium to cause a corneal ulcer. Most bacteria will not penetrate an intact healthy epithelium.

Considerations and Treatment

Typically, you will be looking at a non-light-blocking infiltrate. Infiltrate characteristics include epithelial status overlying, sharp vs. feathered edges, stromal melt, and intensity of infiltrate (light or non-light blocking). If it's light blocking, then you have to be more concerned that it could be an active infectious process.

The intensity, the extent and the location of the infection all play a major role. If it is peripheral and apparently a visually non-threatening infectious process, I may not culture it, as I do with all other ulcers. I also take into account whether the infiltrate is below the surface. If it's an intrastromal process, I biopsy it. If it's an ulcer, I can scrape it.

If I were in a private practice, a setting in which lab results are not routinely obtained in such circumstances, I might not culture so readily and would potentially select either a broad spectrum antibiotic or fortified therapy if it looked a little bit more threatening. Again, history influences this decision very heavily.

If the ulcer is either threatening the visual axis — in other words, if it is not rapidly and effectively controlled and might significantly decrease vision — or it looks like it would potentially be threatening perforation, I very definitely would culture. Again, it would depend on the history obtained, but I would treat with a fortified antibiotic like tobramycin. They are very effective against gram-negative bacteria. For additional coverage, I would also treat with a very good, strong gram-positive antibiotic, such as one of the cephalosporins or vancomycin.

If the infection looks like a fairly typical, relatively benign infectious process of the corneal periphery, I'll start a patient on an antibiotic administered every hour and will see them back in one to two days, depending on my level of concern. If I'm certain that the infection is limited to the cornea and has not penetrated the eye, topical medication alone will be sufficient. I use topical therapy because of its broader spectrum of efficacy and because of how well these agents penetrate. Topical administration delivers far more antibiotic directly to the cornea than systemic medications or or subconjunctival injections.

I reserve systemic antibiotics for use only when fearful of the infectious process invading the eye; if so, I would treat with IV antibiotics and possibly intracameral antibiotics.

Even if it is a peripheral, relatively non-bothersome corneal ulcer, my university-based setting gives me the opportunity to scrape and culture it. If I did not have ready access to a microbiology lab, I might treat it empirically with a fourthgeneration fluoroquinolone as my first line of therapy. I would then move to dual therapy at a minimum, and add either formulated and/or fortified topical antibiotics.

If the infection appears very bothersome, or the culture sensitivities come back suggesting a need to reconsider my initial best-guess antibiotic, I potentially would see the patient every day or two. I will also look at antibiotic frequency, not just the antibiotic selection, but also the intensity of the antibiotics. Initially, I am going to be treating them with a minimum of my fortified dual therapy, every hour around the clock. If I'm even more concerned, then I might start in the first 12 to 24 hours to give them Q5 minute drops times four every hour so that I'm really loading the tissue. I don't give sub-conjunctival injections any longer, and I will not use systemic antibiotics or intravitreal/intracameral antibiotics unless I'm concerned that the infection has breached or could spread into the eye. In cases of suspected endophthalmitis, I'll culture the aqueous or the vitreous as well as scraping the ulcer.

The table below shows the AAO's recommendations for antibiotic therapy for patients with bacterial keratitis.¹

Follow-Up

The AAO's Preferred Practice Pattern guidelines for bacterial keratitis state that several clinical features suggest a positive response to antibiotic therapy:¹

► Reduced pain
► Reduced amount of discharge
► Lessened eyelid edema or conjunctival injection
► Consolidation and sharper demarcation of the perimeter of the stromal infiltrate
► Decreased density of the stromal infiltrate in the absence of progressive stromal loss
► Reduced stromal edema and endothelial inflammatory plaque
► Reduced anterior chamber cells, fibrin, or hypopyon
► Initial re-epithelialization
► Cessation of progressive corneal thinning

The guidelines go on to say that the follow-up exam frequency depends on the extent of the disease, "but severe cases (e.g., deep stromal involvement or larger than 2 mm with extensive suppuration) initially should be followed at least daily until stabilization or clinical improvement is documented."

The guidelines also mention that, "in general, the initial therapeutic regimen should be modified when the eye shows a lack of improvement or stabilization within 48 hours."

Keratoplasty

Surgery for acute infections has a very limited role; keratoplasty would be a last resort. In general, one is not going to perform surgery in an infected eye until the infection is brought under control and ideally eradicated, unless one has to do a keratoplasty to restore the integrity of the eye. Put another way — if there is a perforated ulcer, one has no choice but to operate if the eye has visual potential. If it is a blind eye, enucleation or evisceration would likely be the best course.

If the eye has visual potential and it is perforated or impending perforation, that would be when we would do a keratoplasty — and it would be termed more appropriately a tectonic keratoplasty. If you look at it like a tire, you have a blowout that must be patched to get the inner tube reinflated. If an eye is perforated, we have to put a patch on it to re-inflate the globe. That would be the only time that we would do keratoplasty as part of acute infection control.

The ideal is to control the infection and preferably eradicate it first. There are some situations where an infection may be smoldering and difficult to eradicate when it is very circumscribed, in which case we might consider in effect an excision of the affected tissue. If so, we need to be sure that we're getting into surrounding tissue that is free of infection.

Prior to considering keratoplasy, I will again reconsider the antibiotics being used. I'll look more carefully at the organisms identified on culture, review sensitivities and consider changing antibiotics or use a different group of antibiotics. Also, I will typically stop the patient's antibiotics for 24 hours if I feel like I need to reculture. They may not be effectively killing the organism, but they can stymie it enough so that it won't grow in culture.

If I'm losing the battle with an infection and I do keratoplasty, I immediately create an avenue for the infecting agent to get inside the eye. I might do a lamellar or a penetrating keratoplasty, but unless you have the infection controlled, you're compromising the eye further. Again, keratoplasty is not a viable front-line treatment in acute infections.

Effective diagnosis, appropriate culturing with determination of organism sensitivities and tailored antibiotic selection and administration are critical to successful outcomes in managing corneal ulcers. OM

Reference

1. American Academy of Ophthalmology. Preferred Practice Pattern Guidelines. http://one.aao.org/CE/PracticeGuidelines/PPP.aspx. Accessed May 12, 2010.

Dr. Jim McCulley is professor and chairman of the Department of Ophthalmology at UT Southwestern and is currently entering his 30th year of chairmanship. He is a recognized expert in cornea, external disease, keratorefractive and cataract surgery.