feature
Gaining Perspective in Incidence, Prevention & Treatment of Infection
Wound architecture, fluoroquinolones and pathogens impact trends in infection control.
BY FRANCIS S. MAH, M.D.

Trends in infection and infection prevention are clearly emerging now that enough time has elapsed to put into perspective the implications of fourth-generation fluoroquinolones, clear corneal incisions and post-LASIK keratitis. Regardless of a surgeon's skill or acumen, cataract and refractive surgery outcomes are only as good as the effort put forth to protect patients from persistent pathogens and resistant bacteria. While this has long been suspected, it has taken a decade of solid research to build the body of evidence that now supports that which was once simply an assumption.

Post-LASIK Keratitis

According to the literature, the incidence of post-LASIK keratitis has been reported to be anywhere from 1 in 1,000 cases to 1 in 3,000 cases. Most of us accept that the incidence probably falls squarely in between at about 1 in 2,000 cases per year. What is interesting is the change in pathogens that are causing these infections over time. For instance, according to a 2001 American Society of Cataract and Refractive Surgery (ASCRS) survey, roughly 50% of the pathogens that caused postop LASIK infections were mycobacterial species — chelonae, fortuitum and abscessus to name a few — and the survey mirrored the reports in the peer-reviewed literature.

Meanwhile, a follow-up ASCRS survey performed in 2004 indicated that mycobacteria were responsible for less than 5% of postop LASIK infections. While these findings are based on a survey that relies on respondents' memory, as opposed to strict medical records, the reduction from 50% to less than 5% is striking enough to suggest legitimacy.

Two Important Occurrences in the Ophthalmology Arena

Two important things happened during the 3-year interval between those two surveys. First, fourth-generation fluoroquinolones, gatifloxacin 0.3% (Zymar, Allergan) and moxifloxacin 0.3% (Vigamox, Alcon), became commercially available and there was rapid adoption. By 2005, approximately 90% of U.S. ophthalmologists were using these newer, stronger antibiotics. What's more, of those 5% of infections that were caused by mycobacteria, none of the patients were on moxifloxacin or gatifloxacin. They were reportedly on an older generation fluoroquinolone, such as ciprofloxacin (Ciloxan, Alcon), levofloxacin (Quixin, Santen) and/or ofloxacin (Ocuflox, Allergan). This would suggest that there is definitely a protective effect afforded by the fourth-generation fluoroquinolones, and this concept has been shown through in vitro and in vivo studies in the literature.

A second change during the 3-year period when mycobacteria's impact on postop LASIK plummeted is that the connection between mycobacteria and post-LASIK keratitis was widely reported in the literature and discussed at meetings, creating awareness among the ophthalmic community. Mycobacterium is ubiquitous, being found in all water supplies, but the knowledge that it was often responsible for postop keratitis has influenced trends in laser-suite sterility and operating-area protocol. For instance, using a bucket of ice to cool balanced salt solution was once acceptable, but this practice has been all but eliminated due to the link with post-LASIK mycobacterial infections.

Post-Cataract Endophthalmitis

The rate of postoperative cataract endophthalmitis had long been considered to be about 1 in 1,000 cases, with plenty of corroboration from the literature. However, a recent retrospective analysis of prospectively accrued data performed by Randall Olson, M.D., and colleagues at the University of Utah, indicated a much higher rate of post-cataract endophthalmitis; 1 in 400, when 10,000 cases were reviewed.1 This finding spurred a pair of retrospective analyses that looked not only at the rate of endophthalmitis, but also the factors that could be behind the rise in its incidence.

Peter J. McDonnell, M.D., and colleagues first looked at cataract surgery and endophthalmitis reports in the peer-reviewed literature and noted that the rate of endophthalmitis was decreasing before 1992, and then began to increase after 1992.2 They then compared the incidence before and after 1996, and the correlation was even stronger. There was a definite negative trend, less endophthalmitis risk, before 1996 and a definite positive trend, increasing endophthalmitis risk, after 1996. Their next study reviewed endophthalmitis and cataract surgery reviewed in Medicare reports, and they used the same time points to evaluate incidence: pre- and post-1992, and then pre- and post-1996. They once again saw an increase in endophthalmitis after 1992 and a stronger increase after 1996.3

They hypothesized that clear corneal incisions, which were introduced in 1991, could be responsible for the trend in increasing endophthalmitis rates, and that the steady increase from 1992 to 1996 was a result of the time that it took for widespread adoption of this type of wound architecture. Many other studies support this theory. Most notable among them is the randomized, prospective Japanese study led by Yasunori Nagaki, M.D., that showed a 4.6-fold increased risk of endophthalmitis with temporal corneal incisions vs. superior sclero-corneal incisions.4 This was particularly meaningful because it was done by a single skilled surgeon as opposed to a group where perhaps the surgical skill level and methods would be different among surgeons.

Other studies that have shown a greater incidence of endophthalmitis with clear corneal incisions include a Canadian retrospective study that reported a 2.6- to 3.5-fold increased risk of endophthalmitis with clear corneal incisions vs. superior scleral tunnels;5 a 15-fold increase in the risk of endophthalmitis with clear corneal vs. scleral tunnel incisions reported by Kurt Buzard, M.D.;6 and a 3.36-fold increased risk of endophthalmitis with corneal incisions vs. scleral tunnels that was reported by Cooper and colleagues in St. Louis.7 These studies tell a story that indicate a 2 to 4 times increased risk of endophthalmitis with clear corneal incisions when compared to scleral tunnel incisions.

Meanwhile, a recent a report out of Bascom Palmer Eye Institute that reviewed more than 35,000 intraocular cases over a 7-year period, showed a decrease in the rate of endophthalmitis. This report by Harry W. Flynn, M.D., and colleagues found that even the rate following over 21,000 cataract surgeries seemed to be declining in their surgical center. Unfortunately, with only eight post-cataract cases of endophthalmitis, no comment can be made about the potential difference between clear corneal and scleral tunnel incisions.8

Non-Surgical Wounds

Before the advent of fluoroquinolones in 1991, the most feared pathogen in the non-surgical arena was Pseudomonas aeruginosa. Along with other gram-negative bacteria, P aeruginosa was prevalent in contact lens wearers. The fear was that a patient would contract a gram-negative infection that would perforate, and vision would be lost. The referrals for gram-negative infections far outweighed those of gram-positives. At that time, the agents typically used to fight keratitis were cefazolin and tobramycin. Since the advent of fluoroquinolones in 1991, far fewer patients are presenting with severe, vision threatening, suppurative bacterial keratitis. I suspect that corneal specialists would agree that although P aeruginosa is still feared, it is probably much less of an issue due to the availability of the fluoroquinolone class of antibiotic.

While the newer generation of antibiotics covers gram-positive resistant bacteria better than previous antibiotics, there is no decrease in their gram-negative coverage, according to our findings at the University of Pittsburgh, Charles T. Campbell Ophthalmic Microbiology Laboratory. So as far as looking for a commercially available drug for prophylaxis of either refractive surgery or cataract surgery, or even for empiric therapy for keratitis, the first drug to reach for is a fourth-generation fluoroquinolone.

The Strengths of Moxifloxacin and Gatifloxacin

While fluoroquinolones as a class are weakest against gram-positive resistant bacteria such as Staphylococcus aureus and Staphylococcus epidermis, the fourth-
generation fluoroquinolones moxifloxacin and gatifloxacin cover these better than previous generations, and it is generally acknowledged that moxifloxacin has an advantage over gatifloxacin against the gram-positive resistant bacteria. Gatifloxacin, on the other hand, seems to have a slight advantage in combating gram-negatives.

Undoubtedly, resistance will eventually emerge even against the newest antibiotics, but another advantage inherent in these drugs compared with older fluoroquinolones is their chemical makeup (a C7 side chain that is somewhat larger) endowing them with delayed resistance — gatifloxacin has a cyclic structure whereas moxifloxacin has a bi-cyclic ring. This means that,theoretically, the bacteria will become resistant to the fourth generation later, so the antibiotics will be useful for a longer time. Perhaps instead of 1 or 2 or 5 years, fourth-generation fluoroquinolones will remain effective for 10 to 20 years.

Differences in Commercial Formulations in Moxifloxacin and Gatifloxacin

While moxifloxacin and gatifloxacin are almost identical in their make-up — except for that C7 side-chain distinction — their commercial formulations, Vigamox and Zymar have a few more differences, specifically concentrations and preservatives. Vigamox also has a near-neutral pH which enhances solubility. An advantage of the Vigamox formulation is that it enables higher concentrations in the tissues so less is required to do the same job as the lower concentrated Zymar.

In head-to-head studies that evaluated fluoroquinolone concentrations in humans, Vigamox penetrates into the conjunctiva, cornea and aqueous at least two to three times higher levels than Zymar. The advantage here is that if you are treating a corneal ulcer, for example, or trying to prevent an infection in the cornea or anterior chamber after refractive surgery or cataract surgery, then higher concentrations of antibiotic afford better protection and/or better treatment.

Tissue concentrations are a critical component of evaluating antibiotic efficacy. Not only is spectrum of coverage and level of potency important, but concentrations achieved in the target tissues are equally important. For fluoroquinolones, clinical efficacy can be predicted by calculating the C_max (peak concentrations) to minimum inhibitory concentrations (MIC) ratio. If the ratio equals or exceeds 10:1, clinical efficacy will be achieved for fluoroquinolones. Each class of antibiotic has a
specific ratio which needs to be achieved for clinical efficacy, but one can see that efficacy can be achieved either by decreasing the MIC, or increasing the tissue concentrations. This is the advantage of the fluoroquinolone class and for the newest generation, gatifloxacin and moxifloxacin specifically.

Another difference in the commercial formulations is that Zymar is preserved with benzalkonium-chloride (BAK), whereas Vigamox is self-preserved. Vigamox's commercial formulation enabled it to pass all of the U.S. pharmacopoeia requirements without the need for supplemental preservatives. Interestingly, some studies, including our own, suggest that the addition of BAK provides Zymar with a slightly faster kill rate and potentially lower MICs in vitro, but the clinical significance remains to be seen. Here again, lowering the MIC may be a mechanism of improving clinical efficacy since tissue concentrations are less than Vigamox. This effect could be important for ocular surface clearance of bacteria, but would have no effect in the intraocular tissues.

As future studies are done and elucidation of ocular standards for efficacy are described, it is crucial to test not only the drug, but also the commercial formulations. Unlike our systemic counterparts, formulations can make a difference in toxicity (e.g., BAK), tissue concentrations (e.g., solution vs. ointment, drug concentration) and may have an impact on potency.

Clear Corneal Wound Infiltrates

Numerous studies have shown that clear corneal incisions may not be as forgiving as scleral tunnels. Dr. McDonnell performed a study of clear corneal incisions using India ink to detect possible flow of surface fluid along the incision.9 Histologic examination revealed India ink particles in all incisions for up to three-fourths of the length of the wound. In one cornea, ink particles were present at least 73% of the way to the internal endothelial end of the incision. They showed that lower pressures inside the eye will lead to the India ink almost crawling through the corneal incisions and the ink can actually be seen entering the eye. The India ink is comparable to bacteria in size, so this study shows that in corneal incisions that become infected, it is probably bacteria that is on the surface of the eye that tracks through the incision and into the eye, thus leading to endophthalmitis. The take-home message is that you need your postoperative-prophylactic antibiotic to clear the surface of the eye of bacteria as rapidly as possible and you want an antibiotic that can actually get into the tissues, the cornea and the anterior chamber so that it can actually kill the bacteria in the track.

Case Reports of Clear Corneal Infiltrate Infections

Case reports of clear corneal infiltrate infections at New York Eye and Ear Infirmary and Wills Eye Hospital10 indicated that the infections typically occurred at least a week after the surgery, and that the majority of cases were caused by a gram-positive bacteria such as streptococci species, which is typically a more devastating pathogen than S epidermidis in terms of prognosis. The timing of the infections suggests that they occur just as patients discontinue their week-long postop antibiotics. We had two similar corneal infiltrate infections leading to endophthalmitis at the University of Pittsburgh. Both cases, which occurred years apart, started the day after the patient was told to stop the prophylactic antibiotic. Both cases were caused by the streptococci species, and both cases resulted in devastating vision loss.

The Proper Course of Action

These cases can resolve with better results as long as they are diagnosed prior to concluding the antibiotic prophylaxis. Obviously, the most important factor in diagnosis is a meticulous slit lamp exam. When you encounter a postoperative infection similar to these, you should take a culture from the incision and it should be done in a sterile operating room. You should follow the culture with intravitreal injections of antibiotics, as well as a thorough irrigation of the incision with vancomycin and amikacin. Following the culture and irrigation, hourly antibiotics should be prescribed and these patients should be examined frequently. We were able to treat several cases prior to progression to full-blown endophthalmitis. Most patients, in fact, retained excellent vision throughout the resolution of the corneal incision abscess.

These infections raise all sorts of questions: Despite adequate penetration and concentrations of antibiotics in the cornea, are we just delaying the inevitable if the wound is compromised? Do these incisions need to be tested better? Do patients need to be kept on antibiotics longer? Reporting the timing of clear corneal infections and the pathogens responsible for them provide fodder for discussion at this point and will eventually play into the evolution of prophylaxis and treatment.

Pathogen Trends

As far as pathogen trends go, the most feared pathogen today is probably the gram-positive methicilin-resistant
S aureus (MRSA). If there is a weak point to the fluoroquinolone class of antibiotics, it is the development of resistance in gram-positive bacteria. We may be seeing a slight chink in the antibiotic armor with the expansion of fluoroquinolone-resistant MRSA. The growing incidence of MRSA has been a known factor in the infectious disease arena for several years, but the threat is only recently becoming relevant to ophthalmologists.

Several years ago, a study reported by Eric Donnenfeld, M.D., and colleagues showed that MRSA was a common infection after refractive surgery in healthcare workers. When MRSA is a concern or it is cultured, vancomycin is the gold standard. Unfortunately, MRSA is becoming increasingly common in the general population. Three years ago, there was an approximately 20% carriage rate of MRSA in the community; today that number has grown to approximately 50%. MRSA is spreading and we have to figure out how to prevent MRSA infections and how to treat them. As research and development of newer antimicrobials try to answer this next problem, moxifloxacin and gatifloxacin remain the broadest spectrum, most potent antibiotics with the best tissue concentrations that are commercially available. 

Francis S. Mah, MD is assistant professor of Ophthalmology, University of Pittsburgh, and co-medical director of the Charles T. Campbell Ophthalmic Microbiology Laboratory.

References

1. Jensen MK, Fiscella RG, Crandall AS, Moshirfar M, Mooney B, Wallin T, Olson RJ.A retrospective study of endophtalmitis rates comparing quinolone antibiotics. Am J Ophthalmol. 2005;139(1):141-148.

2. Taban M, Sarayba MA, Rao B, et al. Dynamic morphology of clear corneal cataract incision. Invest Ophthalmol Vis Sci. 2003;2578.

3. M. Taban, M.A. Sarayba, B.Rao, J.Zhang, R.Schiffman, Z.Chen, P.J. McDonnell, Dynamic Morphology of Clear Corneal Cataract Incisions. Ophthalmology 2003;110:2342–2348

4. Nagaki Y, Hayasaka S, Kadoi C, et al. Bacterial endophthalmitis after small incision cataract surgery: effect of incision placement and intraocular lens type. J Cataract Refract Surg. 2003;29:20-26.

5. Colleaux KM, Hamilton WK. Effect of prophylactic antibiotics and incision type on the incidence of endophthalmitis after cataract surgery. Can J Ophthalmol. 2000;35:373-378.

6. Buzard K, et al. Endophthalmitis: Scleral tunnel v. clear corneal incision. In: Blue Line Incision and Refractive Phacoemulsification. Thorofare, NJ: Slack Inc; 2001.

7. Cooper BA, et al. Case-control study of endophthalmitis after cataract surgery comparing scleral tunnel and clear corneal wounds. Am J Ophthalmol. 2003;136(2):300-305.

8. Eifrig CW, Flynn HW Jr, Scott IU, Newton J.Acute-onset postoperative endophthalmitis: review of incidence and visual outcomes (1995-2001). Ophthalmic Surg Lasers. 2002;33(5):373-378.

9. M. Taban, M.A. Sarayba, B.Rao, J.Zhang, R.Schiffman, Z.Chen, P.J. McDonnell, Dynamic morphology of clear corneal cataract incisions. Ophthalmology. 2003;110:2342–2348.

10. Cosar CB, Cohen EJ, Rapuano CJ, Laibson PR. Clear corneal wound infection after phacoemulsification. Arch Ophthalmol. 2001;119:1755-1759.