Most corneal ulcers respond to single fluoroquinolone therapy

Issue: November 1996

In conjunction with this issue's special focus on cornea and external disease, Primary Care Optometry News features a round-table discussion on ocular infection. Participants included moderator Kenneth R. Kenyon, MD; Eduardo C. Alfonso, MD; David G. Hwang, MD; and Terrence P. O'Brien, MD.

Kenneth R. Kenyon, MD: Is the problem of overuse, abuse and resistance something that is allowing your university-based referral practice to encounter more disasters, even though the community is knocking off the easy ones, or do you feel that the community-based standard of approach is the appropriate one?

David G. Hwang, MD: In a university-based practice, one certainly is seeing a select group, but we encountered eight cases during a 2-year period in which empiric monotherapy with a fluoroquinolone, with treatment varying from every hour to every 3 hours, was inadequate to treat streptococcal keratitis, even if the infiltrate was relatively small (less than 1 mm in diameter). When we recultured these patients, we were able to isolate viable strep species in three of them. More interestingly, when we switched from monotherapy to combination therapy (i.e., cefazolin and tobramycin or an equivalent, such as vancomycin and tobramycin in penicillin-allergic patients), all cases improved. All eight of these eyes had other risk factors predisposing to infection, such as presence of a corneal graft or steroid use.

Kenyon: Would you say that it is not a risk to take an approach that might involve culture and monotherapy?

Hwang: I think mono therapy is appropriate for most cases of corneal ulcers, particularly when the ulcers are small, when they do not involve the visual axis, when the etiologic organism is felt to be gram-negative or Staphylococcus species and when the patient does not have other risk factors that predispose to local immunocompromise.

Kenyon: Dr. Alfonso, do you concur with those remarks?

Eduardo C. Alfonso, MD: I agree with Dr. Hwang that monotherapy certainly treats most of the community-acquired and community-seen cases of keratitis. I also agree with your statement that this oversimplification is beginning to pose certain problems regarding the more difficult cases coming to us at a later stage.

Before, when fortified antibiotics were the main treatment, it was more likely that cases would be referred on an early basis, cultures would be done and those that were difficult cases, such as a fungal Acanthamoeba mycobacteria or an anaerobic case, would receive the appropriate treatment much sooner before structural damage to the cornea was significant. We are seeing cases of, for example, fungal keratitis that come in quite late after having received a fluoroquinolone for seven to 10 days with poor response and then referred for appropriate cultures and modification of therapy.

Kenyon: What is the appropriate management strategy in a community-based case of presumed bacterial keratitis (of mild to moderate severity) as defined by an epithelial defect, a stromal infiltrate and a relatively acute onset, that has not been treated or only minimally treated?

Terrence P. O'Brien, MD: The most recent Preferred Practice Pattern manuscript provided by the American Academy of Ophthalmology on bacterial keratitis provides some useful guidelines for the practicing community ophthalmologist in choosing the initial management strategy. The Preferred Practice Pattern calls for the clinician to make a clinical severity judgment assessment and, if the keratitis is mild or moderate, then it is appropriate perhaps for empiric monotherapy to occur as an evolving standard of care.

If, on the other hand, it is a central suppurative keratitis of larger than 2 mm, which falls into the severe group, a culture-guided approach would seem prudent as the continuing standard of care under those circumstances. If the clinician is unable to adequately obtain material, culture accessories or both to adequately perform a microbiological work-up of such a severe keratitis, then the patient should probably be quickly referred to a center that has those capabilities.

Alfonso: One key issue that we need to consider is with what risk factor the patient is presenting. In our specific area, for example, if a patient presents with a mild-to-moderate infiltrate outside the visual axis, and gives a history of having been out in the backyard cutting weeds and felt something get into the eye, this immediately lights up a risk factor, indicating that the patient may have a nonbacterial fungal keratitis. We need to consider those existing risk factors when we decide to use monotherapy empiric treatment vs. culturing the patient, and then guide our treatment based on the findings of the microbiology laboratory.

Kenyon: Assuming that no extraordinary risks pointed you in the direction of a fungus, for example, would the approaches suggested by Dr. O'Brien of single fluoroquinolone therapy in the absence of culture be an appropriate approach in your view?

Alfonso: Yes. This is the way I would guess 95% of corneal ulcers are being treated successfully. The number of ulcers, for example, referred to a tertiary care center such as ours has markedly decreased since the availability of non-fortified monotherapy.

Kenyon: I think it is clearly a trend at our hospital center as well as in our clinical practice that most ulcers are getting stopped in their tracks by nonfortified monotherapy.

Hwang: I would generally agree with most of the statements, although I do lament that we may be sacrificing the 5% for the 95%, which is to say that we are not culturing some of these cases that are going to end up being atypical cases or resistant cases that pose problems in the subsequent decision-making algorithm. In an ideal world, one would like to be able to use fluoroquinolone mono therapy, which represents cost savings, added convenience and increased accessibility for the patient, yet does not sacrifice the diagnostic information that one can obtain from culturing these patients.

I would also like to bring up the issue of emerging resistance. Fluoroquinolones have two areas that pose potential concern. One is the issue that we have referred to, which is the hole in their coverage of streptococcal species, and the other is the issue of the epidemiology of fluoroquinolone resistance and antimicrobial resistance.

We in ophthalmology have felt ourselves to be relatively immune to this phenomenon, thinking that the high levels of drug achieved through topical dosing would present barriers to the emergence of resistance, but we have found, like the emergence of resistance that was seen with the introduction of ciprofloxacin and other fluoroquinolones for systemic use, a similar emerging pattern of increasing prevalence of fluoroquinolone resistance among ocular isolates.

One study that we have done showed an increase from a baseline of about 2.5% prior to the introduction of ciprofloxacin to an overall prevalence of resistance of 10%. We attribute this rise during a 2-year period since the introduction of fluoroquinolones to topical fluoroquinolone use rather than systemic fluoroquinolone use.

Although this is conjectural, we do have some supportive data based on the fact that we see a very high conversion to fluoroquinolone resistance among ocular isolates in patients who are on routine antibiotic prophylaxis following cataract surgery.

Alfonso: We are seeing a similar trend. We do continue to recognize, though, that through our ability to extrapolate what we see in the microbiology laboratory in vitro to the clinical situation, significant disparity can be found. In spite of the fact that we see the increasing resistance in the in vitro testing, that does not seem to be related to what we are seeing in the clinical setting with the treatment of cases.

Hwang: The concern really is not over the use of fluoroquinolones for the case of presumptive bacterial keratitis, but the use of fluoroquinolones for routine cases of relatively mild or minor infections, particularly when the dosing is done at a low level once or twice a day for weeks to months.

This is the area where we have a tremendous potential to see the emergence of resistance through selection. If fluoroquinolones are used judiciously for more severe infections for limited periods of time at appropriate dosing regimens recommended by the manufacturer, this is the appropriate place for them. Indiscriminate use, when other conventional antibiotics continue to be available, will accelerate their status as being just another drug that unfortunately has lost much of its clinical utility because of resistance.

Kenyon: Are we agreed then to consider fluoroquinolones as the big guns that they are and hence use them sparingly, not chronically?

O'Brien: Yes. We should take this message into the clinic and certainly not use the fluoroquinolones for prophylaxis on a long-term basis, either following keratoplasty or even after cataract or keratorefractive surgery.

Kenyon: Let's shift the discussion slightly to the appropriate and inappropriate conjoint application of steroids in the treatment of bacterial keratitis: when to add steroids, when not to add them and when never to use them.

Alfonso: Steroids are a wonderful drug, but we can also appreciate the disasters that can be created when you suppress the immune response with steroids and do not have adequate control of the replicating microorganism. I certainly use corticosteroids in the treatment of bacterial keratitis. I reserve their use for when I feel adequate control of the infectious process, and that usually means treating with topical antibiotics for at least 48 to 72 hours in the mild to moderate cases, seeing that an adequate response has been obtained and then adding the corticosteroids at the same time that I continue to treat vigorously with the antimicrobials.

I never use the corticosteroid when I have stopped the antimicrobial early in the process for fear of worsening the process. Some of the difficulties that have arisen with the use of corticosteroids are when no control of the replicating microorganism is obtained with an antimicrobial and the corticosteroids are added. I have a good clinical sense that they do decrease the destruction of the cornea by the microorganisms and the immune response to them and that they also make the patient more comfortable and decrease the inflammatory response; therefore, making the whole ocular surface look better and feel better.

Kenyon: Dr. Alfonso, have you found any application of steroids appropriate in the management of a presumed or proven gram-negative microbial keratitis?

Alfonso: Yes, they can be used, but you want to make sure that you have the gram-negative infection under control, and you never want to use the corticosteroid in the absence of good control of the microbial process. Again, you may want to delay in a severe case the use of corticosteroids until you feel very comfortable that you have been able to control the infectious process. This, in some severe cases, may mean that you do not use them until 3 or 4 weeks after the onset of the keratitis. At some point, you may want to use them to decrease some of the scarring that takes place as a secondary effect of the infection.

Kenyon: Dr. Hwang, any opinions on the use and abuse of steroids in this setting?

Hwang: The comments of Dr. Alfonso are very judicious and are based on his experience. I have much less experience, and my viewpoint generally has been to use antimicrobial therapy alone.

It is the rare case in which we will initiate corticosteroid therapy. We generally have not been using steroids at all in patients with gram-negative infection (especially Pseudomonas because of the documented risk of relapsing infection.

In exceptional circumstances of gram-positive bacterial keratitis after the infection is clearly under control, we still use steroids judiciously to control presumed post-infectious sterile inflammation.

Kenyon: Any comments with respect to nonsteroidal anti-inflammatories?

Alfonso: It is common that after you have treated a patient with microbial keratitis with appropriate antimicrobial therapy the pain tends to ease up after 2 to 3 days of therapy, but in some of the more chronic cases that require more prolonged treatment (for example, mycobacterial keratitis), I have had some limited experience of using a nonsteroidal preparation such as diclofenac with a subjective impression that the patients do not complain of pain as much.

O'Brien: I was very excited about that potential application for the analgesic effect of topical nonsteroidals for patients with Acanthamoeba keratitis and severe radial keratoneuritis. Unfortunately, my experience in using these agents was that the patients continued to have rather severe pain despite the use of both topical diclofenac and ketorolac tromethamine.

I have not had wide experience with using nonsteroidals for bacterial keratitis. I am aware of several experimental models with Pseudomonas keratitis that have not shown any deleterious effects of recurrence, either with nonsteroidals or steroidal topical treatment regimens.

Hwang: I do not have any experience with nonsteroidal control of pain in bacterial keratitis.

Kenyon: Nor do I, yet it is clear that these preparations have evolved as a panacea in the acute management of excimer photorefractive surgery, and patients who would previously experience severe pain for several days now seem to be at the same low pain level as patients who are undergoing simple incisional surgery. This observation has prompted modification of my strategy for the recurrent erosion patient requiring epithelial debridement. In the past, these patients would have to be pressure patched for at least 48 hours before they would tolerate a bandage soft contact lens. Now the application of diclofenac for a period of perhaps 48 hours can get them through the bandage lens adaptation period with ease and without the anticipated 2 a.m. onset of the tight lens syndrome.

Alfonso: I do the same. Patients with recurrent erosion who I place on a contact lens treatment are also put on diclofenac, and that seems to help them tolerate the lens.