SPECIAL REPORT

Treatment of Bacterial Infections on the Cornea

By Terrence P. O'brien, MD

As sight-threatening corneal infections such as microbial keratitis present in offices all over the world, the challenge of eradicating infection while preserving the cornea remains a delicate balancing act. Successful management requires intimate knowledge of the relationship among organism, host and antibiotic. We have many tools to assist with diagnosis of infection and an expanding array of agents to fight infection, but all approaches require a balance between killing power and gentle healing. New medications and technologies still haven't completely resolved this problem.

In the past 20 years, we've seen rather significant advances in corneal surgeries but advances in therapies for corneal infection and adjunctive measures to prevent destruction of the cornea haven't been as pronounced. In fact, our basic approach to management of microbial keratitis has changed very little. There are a few new antibiotic options and a reconnection with older agents not used systemically. We have not, however, seen fundamental breakthroughs in understanding pathogenesis to identify novel targets leading to development of broad-spectrum treatments that could destroy the majority of keratitiscausing organisms, including increasingly resistant strains. We also lack strategies to prevent destruction of the corneal tissue, thus common infections are significant and contribute to structural and functional comorbidity.

We have a long way to go to reach optimal management strategies. Thankfully, the research pipeline may hold some significant promise.

Promising New Tools

In corneal infection, improvements in our ability to diagnose infections and their causes have exceeded advances in treatment. While some areas are still disappointing (the accuracy of clinical diagnosis of microbial keratitis still has a predictability of only about 75% in a standard clinical exam),¹ significant advances in noninvasive imaging techniques and femtosecond lasers are providing a whole new picture of what exists on the cornea.

Today, clinicians obtain material from the cornea for culture and often receive a negative result, even in cases where infection is clearly present. Fastidious organisms, sampling error, inadequate media and prolonged transport before processing are all factors that may reduce the likelihood of recovery organisms with culture. But with respect to ocular microbiology, conventional testing may only be scratching the surface of our potential ability to detect and identify pathogens on the surface of the eye. Now we wonder, “Is the patient really negative, or are we simply incapable of recognizing all of the many potential pathogens at work?” Advanced molecular microbiologic techniques are identifying totally new bacterial species on the ocular surface microbiome that could represent potential corneal pathogens. Many different genera of bacteria exist in the eye; robust microbiology techniques like high-definition imaging may reveal more pathogens that cause disease.

Non-invasive imaging techniques with tandem scanning confocal microscopy and anterior segment OCT have reached a higher resolution than ever before allowing direct visualization of yeasts, fungi and amoeba. Ultra-high resolution OCT now has a resolution of 1 to 3 microns. These technologies may allow us a direct view of bacterial organisms on the cornea, including many that are undocumented or not recognized as corneal pathogens.

Another promising tool to help us diagnose corneal infection by more precisely and safely sampling the cornea is the femtosecond laser. This technology is helping us to perform corneal biopsy in a precise, less invasive, more efficient and potentially less scarring manner.

These imaging and sampling techniques hold a great deal of promise for the future along with the advances in molecular microbiology. We can foresee them aiding in rapid diagnosis with higher sensitivity and specificity. However, the real-world disadvantages of these techniques are clear. Confocal microscopy requires skilled interpretation of scanned images. Ultra-high resolution OCT is still in the early stages and not widely commercially available. Few practices have femtosecond lasers as they are expensive and require highly skilled personnel to operate. Advantages of molecular microbiological rapid techniques include rapid confirmation of etiological agent, increased sensitivity and specificity, and detection of new/emerging pathogens. Disadvantages include the lack of well-defined standards, few commercial kits, greater expense, lack of an isolate for additional testing, requirement for skilled personnel and need for multiplex technology.

Figure 1. Ulcerative keratitis in association with soft contact lens wear due to Pseudomonas aeruginosa.

Despite these advances, a standard culture remains the gold standard, but has its challenges. It can be difficult for doctors to perform the cultures due to lack of required materials, and the practice may be too far from a lab to keep the culture viable during transport and before processing. Clearly, there's a need for more microbiological research into affordable, reliable, rapid point-of-care testing.

Overview of Treatments

Although there have been some notable advances, the basic approach to treating common bacterial infections of the cornea hasn't changed much. We used to treat infection with a “saturation bombing” approach, instilling one or more antibiotic drops every half hour in a fortified concentration to gain rapid control of organism replication and further spread into the corneal tissues. However, the emergence and global spread of antibiotic resistance have made us reevaluate proper antibiotic selection and use.

Fluoroquinolones have revolutionized the treatment of bacterial keratitis by providing commercially available, stable formulations that offer broad-spectrum, bactericidal, bioavailable and biocompatible action. These single-agent drugs can provide the clinical equivalent of combined therapies, and sometimes can be used as an alternative to combination fortified antibiotics with far less cytotoxicity. However, like older antibiotics, ocular pathogens are demonstrating increasing resistance to fluoroquinolones. Newer expanded-spectrum fluoroquinolones are another option and are an especially good choice with non-severe, peripheral, nonsight-threatening infections when the corneal cultures test negative. A broad-spectrum agent with bactericidal activity and good penetration into corneal tissues is more likely to destroy the unknown bacterial pathogen.

The initial choice of therapy is influenced by what you determine to be the most likely organisms at work and the resistance patterns encountered. For example, a few years ago, fortified cefazolin for coverage of gram-positive bacteria such as Staph and Strep combined with tobramycin for gram-negative bacteria was the initial empiric therapy, but over time, gram-positive corneal pathogens became increasingly resistant to cefazolin. Now about 40% of the Staphylococcal species recovered from the cornea in our micriobiology laboratory at Bascom Palmer Eye Institute are methicillin resistant, so vancomycin is our first choice, combined with either tobramycin or a fluoroquinolone as initial therapy. Resistance has changed the standard of initial treatment.

Severity and Choices

In all cases of suspected bacterial keratitis, we treat based on the severity of the patient's condition. Microbial keratitis is less severe if the ulceration is less than 2 millimeters in diameter, superficial rather than deep and peripheral rather than central in location. This calls for treatment with a single agent fluoroquinolone (levofloxacin 1.5%, besifloxacin 0.6%, gatifloxacin 0.5% or moxifloxacin 0.5%) and appropriately frequent follow-up visits.

More severe cases are greater than 2 millimeters in diameter, involve central cornea in mid to deep corneal substance, and threaten sight. We obtain material from the cornea for culture and vital staining, and then treat aggressively with combination therapy. Our goals are rapid cessation of replication, rapid elimination of invading pathogen(s), and simultaneous prevention of tissue destruction (especially stromal destruction, which can lead to loss of transparency and function).

We typically take a combination approach to treatment with a fortified antibiotic such as vancomycin 25 mg/ml and a fluoroquinolone (ciprofloxacin 0.3%, levofloxacin 1.5%, besifloxacin 0.6%, gatifloxacin 0.5% or moxifloxacin 0.5%). Drops are instilled every 5 minutes as a loading dose, and then every 30 minutes to 1 hour around the clock. We often use adjunctive agents, like a cycloplegic (cyclopentolate 2% or homatropine 5%) drop to reduce photophobia and light sensitivity. If the cornea is breaking down, we may add oral doxycycline, ethylene diamine tetraacetic acid (EDTA), vitamin C or corticosteriod eye drops in an attempt to block the action of collagenases, matrix metalloproteinases and other factors to prevent tissue destruction.

These methods battle the organism in severe cases. However, we're still not doing well with the host response and that last goal: preventing tissue destruction. We know more about the destructive factors released in the course of infectious keratitis, but we don't have novel agents that have been studied and approved to deal with them. Sometimes we can eliminate the microbe(s), but the cornea is destroyed in the process. In the future, we want to find a way to win the battle and the war.

Figure 2. History of prior radial/astigmatic keratotomy getting gasoline splashed into eye while filling boat. Internist prescribed topical prednisolone acetate 1% QID for “chemical conjunctivitis.” Patient developed ulcerative keratitis with deep, suppurative infiltrate adjacent astigmatic keratotomy incision. Cultures grew heavy Pseudomonas aeruginosa.

Figure 3. Ulceration of epithelium staining with fluorescein sodium and deep, suppuration involving astigmatic keratotomy incision.

Future Advances

Microbial keratitis is a complex interaction of the microbe, host and drug with the cornea. In the future, we need to learn more about those microbes, achieve a better understanding of the pathophysiology in the human cornea and improve the rapidity and reliability of diagnostic methods. Rapid, sensitive, exciting molecular biology techniques exist, but most people don't have access to them. Most of us still rely on the culture as the gold standard, as we try to close in on results that will guide therapy.

We need research and development of newer treatments as well. Fewer new antibiotics are being developed due to escalating costs and challenging regulatory pathways through the approval process. Better adjunctive strategies for control of organism and host-derived factors are necessary. Research is occurring in this area now. One study, the multicenter Steroids for Corneal Ulcers Trial (SCUT) investigated whether adding topical corticosteroids to existing treatments of bacterial corneal ulcers would improve vision.² Researchers concluded that although topical corticosteroids don't offer significant help, they are generally safe, and their adjunctive use may help prevent visual outcomes, especially in severe cases as with Pseudomonas species.

New delivery modalities may improve outcomes as well. It may be possible to deliver treatment to corneal tissue in new ways, whether it's activated by a laser, achieved through smart polymers, accomplished by iontophoresis, or inserted with nanoparticle technology for greater penetration.

In the future, we need a better way to prevent tissue destruction during the course of microbial keratitis. There is some promising use of collagen cross-linking with ultraviolet light and topical riboflavin for recalcitrant keratitis, but additional research on the mechanism of action and safety are required before widespread adoption can begin.^3,4 Bacterial infection of the cornea is common, sight-threatening and potentially devastating, making it important that we continue to explore all of these avenues to achieve better outcomes. ■

References

1. Dahlgren M, Lingappan A, Wilhelmus K. The Clinical Diagnosis of Microbial Keratitis. Am J Ophthalmol. 2007; 143(6): 940-944.
2. Srinivasan M et al. The Steroids for Corneal Ulcers Trial: Study Design and Baseline Characteristics. Arch Ophthalmol. 2012;130(2):151-157. (Accessed June 23 2012 at http://archopht.jamanetwork.com/article.aspx?articleid=1106553)
3. Makdoumi K, Mortensen J, Crafoord S. Infectious keratitis treated with corneal crosslinking. Cornea. 2010;29(12):1353-1358.
4. Iseli HP, Thiel M, Hafezi F, Kampmeier J, Seiler T. Ultraviolet A/Riboflavin Corneal Cross-linking for Infectious Keratitis Associated With Corneal Melts. Cornea: 2010;27(5):590-594.