Despite the lack of large prospective studies, there is a wealth of experimental data and retrospective clinical evidence that favors the use of preoperative, intraoperative, and postoperative antibiotics to reduce the risk of endophthalmitis.

The choice of antibiotic is important. The ideal antibiotic has broad-spectrum activity, is fast at killing organisms, has good tissue penetration, and is nontoxic and well tolerated by the eye.

ANTIBIOTIC SPECTRUM

Fluoroquinolones became very popular in the early 1990s because of their broad spectrum. Unlike the aminoglycosides, ciprofloxacin (Ciloxan; Alcon Laboratories, Inc., Fort Worth, Texas) and ofloxacin (Ocuflox, Allergan Inc., Irvine, CA), when they were first introduced, had excellent coverage against both gram-positive and gram-negative organisms. However, during the years, this activity against gram-positive organisms has become compromised by the emergence of resistant strains.1-3 There is increased resistance in ocular isolates, a situation that reflects the increased resistance of bacteria in the population due to the widespread use of these drugs for the treatment of systemic disease, as well as the prevelent use of ciprofloxacin in the animal-feed industry.

Recently, new fluoroquinolone antibiotics have been introduced in ophthalmology that are considered the next generation beyond ofloxacin and ciprofloxacin. These drugs are gatifloxacin (Zymar; Allergan, Inc.) and moxifloxacin (Vigamox; Alcon Laboratories, Inc.). Zymar is a 0.3% solution preserved with benzalkonium chloride (BAK) 0.005%, and Vigamox is a 0.5% solution of moxifloxacin without a preservative. The spectrum of activity for these drugs is significantly improved for the treatment of gram-positive organisms.7 This improvement is due to the multiple enzymes targeted by these drugs, which inhibit topoisomerase II and topoisomerase IV in a number of gram-positive organisms.8,9 The differences in spectrum between the two drugs are not great and do not significantly favor one more than the other.

PHARMACODYNAMICS

Fluoroquinolones exhibit concentration-dependent killing. This killing is faster than that with aminoglycosides, penicillin, and vancomycin, which are time-dependent “killers.” With the fluoroquinolones, it is not uncommon to see 3 logs of killing at two to four times the MIC within 2 to 4 hours for gram-positive organisms. Fast killing limits (1) the accumulation of harmful bacterial toxic products, (2) the time that therapeutic levels need to be maintained in the target tissue to ensure sterilization, and (3) the emergence of resistant strains.

The formulation may also be important in preoperative use. In studies by Hyon et al,10 Zymar was able to reduce the numbers of bacteria in Staphylococcus species cultures by 99.9% in 15 minutes, a rate that is far superior to Vigamox.11 This rate is probably due to the rapid action of BAK, which is present at 0.005%. BAK might also provide a major advantage when antibiotic prophylaxis is started within hours of cataract surgery, or when a rapid reduction in bacterial numbers is beneficial.

TISSUE PENETRATION

Human and rabbit studies show that both gatifloxacin and moxifloxacin in the commercial preparations of Zymar and Vigamox penetrate well into ocular tissues. In studies that my colleagues and I and others have performed, both in rabbit and in human models, gatifloxacin 0.3% and moxifloxacin 0.5% had essentially the same intrinsic penetration into the cornea and aqueous with the epithelium intact.6,12 Because Vigamox is formulated at 0.5%, which is 60% higher than Zymar, the resultant concentrations are generally higher for Vigamox.6,12 In either case, the concentrations on the surface of the eye, as well as in the superficial cornea, are probably more than adequate to kill organisms in the timeframe required after topical dosing. With respect to aqueous levels, both drugs can provide concentrations that are above the MIC in the aqueous (Figure 1). However, a more reasonable approach to ensure adequate aqueous levels is to directly deposit the antibiotic into the aqueous at the time of cataract surgery. There have been a number of advocates for intracameral antibiotics, including a study that resulted in an excess of 25,000 cases without a single incidence of endophthalmitis while using intracameral vancomycin.13 Gimbel et al14 reported similar results.

The idea of intracameral antibiotics is a good one, and I believe that choosing the correct antibiotic is important. Vancomycin only attacks gram-positives and does not have broad-spectrum activity like the fourth-generation fluoroquinolones. In addition, vancomycin is a time-dependent killer, and it may take 8 to 12 hours to get 3 logs of killing with bacteria that are growing at a logarithmic phase.15

Gatifloxacin and moxifloxacin have both gram-positive and gram-negative coverage and are relatively fast at killing. Preliminary studies with intracameral gatifloxacin (Tequin; Bristol-Myers Squibb Company, Princeton, NJ) in both the rabbit model16 and human model17 demonstrated safety. In the future, controlled studies to evaluate the safety and efficacy of intracameral gatifloxacin will be performed and may identify a more appropriate method of intracameral antibiotic prophylaxis.

OCULAR SAFETY AND TOXICITY

Zymar and Vigamox both appear to be safe for treatment of conjunctivitis, and the FDA has approved them for such use. In off-label antibiotic usage for surgical prophylaxis, subtle differences in toxicity may be more important. Donnenfeld et al17 showed that moxifloxacin induced conjunctival hyperemia and edema, caused the pupil to constrict with topical dosing, and was generally less comfortable for patients. He suggested that topical Vigamox drops caused an increased release of prostaglandin, thereby producing these results. Studies from the University of Utah have shown delayed epithelial healing in a prospective study of penetrating keratoplasty in humans when comparing Vigamox to Zymar.18 This finding was counterintuitive to many, because Zymar is preserved with 0.005% BAK and Vigamox is not preserved. Nonetheless, the findings showed a statistically significant delay in healing in the Vigamox group when compared to the Zymar group. In a prospective human study of penetrating keratoplasty,17 a statistically significant result was that Vigamox delayed epithelial healing when compared to Zymar (Table 1). Experimental studies with rabbits have also shown delayed epithelial healing in a rabbit PRK model with topical Vigamox as opposed to Zymar.19

With cataract surgery, changes in wound healing may also be important. Beuerman20 from Louisiana State University reported that Vigamox delayed maturation of rabbit keratectomy wounds and implied the agent could delay the healing of cataract wounds in humans. To date, this delay has not been demonstrated or examined in human cataract surgery, but it is a concern.

OTHER CONSIDERATIONS

Zymar with BAK can kill fungi and yeast that could contaminate the bottle at a much faster rate than antibiotics without a preservative.21 This killing rate could also be important for elderly cataract patients, who often have difficulty handling small bottles. The Vigamox bottle is packaged with a warning on the box that cautions the user to avoid touching the tip.

RECOMMENDATIONS

I believe the fourth-generation fluoroquinolones have utility in the perioperative period in cataract surgery to reduce the risk of endophthalmitis. This would be considered an off-label use and, as such, should be driven by the judgment of the physician and by the evidence-based literature. Based upon the kill curves, I believe starting the fourth-generation fluoroquinolones 1 to 3 days preoperatively is valuable. I prefer Zymar because it has BAK, which facilitates much faster killing on the ocular surface. Zymar is better tolerated with less superficial ocular inflammation, and it allows faster healing in the PRK model reported by Solomon et al,22 as well as in animal studies.19 Early observations in human cornea transplant also suggest better healing with Zymar as opposed to Vigamox.16

Intraoperative vancomycin has been well accepted and is used by more than 30% of ophthalmologists. I believe in the future gatifloxacin may be an excellent alternative because of its broader spectrum of coverage and faster rate of killing. Preliminary studies in the rabbit, as well as human studies, show relative safety.17 Widespread use and adaptation by the general ophthalmologist should await larger, peer-reviewed prospective studies.

CONCLUSION

It is important to continue antibiotics in the postoperative period until the wound is secure. As shown in a Utah study, the antibiotic choice was important to reduce endophthalmitis.23 McDonnell et al24 showed that clear-corneal, no-stitch wounds were not stable in an experimental model. When pressure was put on the eye and then released, some eyes actually aspirated tear film fluid. I recommend using Zymar q.i.d. for 4 to 7 days until the epithelial barrier has been firmly reestablished and the wound is stable.

Robert W Snyder, MD, PhD, is Professor of Biomedical Engineering at the University of Arizona in Tucson. He receives research support from Allergan Inc. and participates on their Speaker's Bureau. Dr. Snyder may be reached at (520) 722-4700; snydereyedoc@aol.com.

1. Jensen HG, Felix C. In vitro antibiotic susceptibilities of ocular isolates in North and South America. In Vitro Antibiotic Testing Group. Cornea. 1998;17:79-87.
2. Goldstein MH, Kowalski RP, Gordon YJ. Emerging fluoroquinolone resistance in bacterial keratitis: a 5-year review. Ophthalmology. 1999;106:1313-1318.
3. Alexandrakis G, Alfonso EC, Miller D. Shifting trends in bacterial keratitis in south Florida and emerging resistance to fluoroquinolones. Ophthalmology. 2000;107:1497-1502.
4. Chu Y. Penetration of gatifloxacin ophthalmic solution 0.3% into aqueous humor patients undergoing cataract surgery. Poster presented at: The ARVO Annual Meeting; April 28, 2004; Fort Lauderdale, FL.
5. Price FW and Price MO. Penetration of gatifloxacin ophthalmic solution 0.3% into aqueous humor patients undergoing cataract surgery. Poster presented at: The ARVO Annual Meeting; April 29, 2004; Fort Lauderdale, FL.
6. McCulley JP. Comparative penetration of 2 fourth-generation fluoroquinolones into the aqueous humor of humans. Paper presented at: The AAO/SOE Joint Meeting; October 25, 2004; New Orleans, LA.
7. Kowalski RP, Dhaliwal DK, Karenchak LM, et al. Gatifloxacin and moxifloxacin: an in vitro susceptibility comparison to levofloxacin, ciprofloxacin, and ofloxacin using bacterial keratitis isolates. Am J Ophthalmol. 2003;136:500-505.
8. Fukuda H, Kishii R, Takei M, Hosaka M. Contributions of the 8-methoxy group of gatifloxacin to resistance selectivity, target preference, and antibacterial activity against Streptococcus pneumoniae. Antimicorb Agents Chemother. 2001;45:1649-1653.
9. Takei M, Fukuda H, Kiishii R, Hosaka M. Target preference of 15 quinolones against Staphylococcus aureus, based on antibacterial activities and target inhibition. Antimicrob Agents Chemother. 2001;45:3544-3547.
10. Hyon JY, Joo MJ, Hose S, et al. Comparative efficacy of topical gatifloxacin with ciprofloxacin, amikacin, and clarithromycin in the treatment of experimental Mycobacterium chelonae keratits. Arch Ophthalmol. 2004;122:1166-1169.
11. Eser I, Hyon JY, Hose S, et al. Comparative antimicrobial efficacy of preserved and preservative-free topical fourth generation fluoroquinolones against various strains of staphylococcus. Poster presented at: The ARVO Annual Meeting; April 29, 2004; Fort Lauderdale, FL.
12. Levine J, Noecker RJ, Lane LC, et al. Comparative penetration of moxifloxacin and gatifloxacin in rabbit aqueous. J Cataract Refract Surg. 2004;30:2177-2182.
13. Gills JP. Antibiotics in irrigating solutions. J Cataract Refract Surg. 1987;13:344.
14. Gimbel HV, Sun R. Prophylactic intracameral vancomycin and CME. Ophthalmology. 2000;107:1614-1615.
15. Snyder R, Krueger T, Nix DE. Vancomycin kill curves for ocular isolates. Paper presented at: The Ocular Microbiology Immunology Group Meeting; November 10, 2001; New Orleans, LA.
16. Snyder RW, Chang M, Hare W, et al. Intraocular safety of gatifloxacin in a rabbit model. Paper presented at: The Ocular Microbiology Immunology Group Meeting; November 15, 2003; Anaheim, CA.
17. Donnenfeld ED, Snyder RW, Kanellopolous AJ, et al. Safety of prophylactic intracameral gatifloxacin in cataract surgery. Paper presented at: The Ocular Microbiology Immunology Group meeting; November 15, 2003; Anaheim, CA.
18. Moshifar M. The effect of the fourth-generation fluoroquinolones on corneal re-epithelialization after penetrating keratoplasty. Cornea. In press.
19. Gao, J, Siemasko KF, Vu C, et al. The effect of fourth-generation fluoroquinolones on rabbit corneal wound healing. Poster presented at: The ARVO Annual Meeting; April 29, 2004; Fort Lauderdale, FL.
20. Schmidt LP, Beuerman RW. Comparison of gatifloxacin and moxifloxacin in healing of a linear incision in the rabbit cornea. Poster presented at: The ARVO Annual Meeting; April 26, 2004; Fort Lauderdale, FL.
21. Rupp D, Reeves T, Kapadia S, Anger C. The antimicrobial preservative efficacy of Zymar and Vigamox against yeast isolates. Paper presenteted at: The Ocular Microbiology Immunology Group Meeting; November 15, 2003; Anaheim, CA.
22. Solomon R, Donnenfeld E, Perry H, et al. The effect of gatifloxacin 0.3% and moxifloxacin 0.5% on corneal epithelial wound healing following photorefractive keratectomy. Paper presented at: The AAO/SOE Joint Meeting; October 25, 2004; New Orleans, LA.
23. Jensen MK, RG Fiscella MK. Comparison of endophthalmitis rates over four years associated with topical ofloxacin vs. ciprofloxacin. Poster presented at: The ARVO Annual Meeting; May 9, 2002; Fort Lauderdale, FL.
24. McDonnell PJ, Taban M, Sarayba M, et al. Dynamic morphology of clear corneal cataract incisions. Ophthalmology. 2003;110:2342-2348.