LABORATORY SCIENCES Treatment of Experimental Bacterial Keratitis With Topical Trovafloxacin Irina S. Barequet, MD; Paul Denton, BS; Gerard J. Osterhout, MS; Suhas Tuli, MD; Terrence P. O’Brien, MD Objective: To investigate the therapeutic role of trova- Results: In vitro susceptibility study findings indicated floxacin mesylate, a newer-generation fluoroquinolone that the MIC of trovafloxacin was significantly lower than with an expanded spectrum of activity, in the treatment the MIC of ciprofloxacin and ofloxacin for S aureus, S of experimental bacterial keratitis. pneumoniae, and Haemophilus influenzae, lower than the MIC of ciprofloxacin and ofloxacin for Staphylococcus epi- Methods: Susceptibility studies were performed on vari- dermidis, and intermediate between ciprofloxacin and ous strains of ocular isolates to determine the minimum ofloxacin for P aeruginosa. Pharmacokinetic studies inhibitory concentration (MIC) of trovafloxacin com- showed a significant concentration of trovafloxacin in the pared with ciprofloxacin and ofloxacin, using the E-test treated corneas, especially in eyes with a denuded epi- method. Pharmacokinetic studies were performed by a thelium. All serum samples had undetectable trovafloxa- single topical administration of trovafloxacin to rabbit cin concentrations. Experimental keratitis studies showed eyes with either an intact or denuded corneal epithe- a statistically significant decrease of colony-forming units lium. Aqueous humor, vitreous, and corneal concentra- in trovafloxacin-treated eyes in the S aureus model and tions of trovafloxacin were determined at different time a similar decrease in the S pneumoniae and P aeruginosa points. Experimental bacterial keratitis studies were per- models. formed in rabbit eyes. Three identical studies were con- ducted using Staphylococcus aureus, Streptococcus pneu- Conclusions: Topical 0.5% trovafloxacin proved to be moniae,orPseudomonas aeruginosa. Therapy groups an effective ocular medication for the therapy of gram- included 0.5% trovafloxacin, 0.3% ciprofloxacin, 0.3% positive and gram-negative keratitis. ofloxacin, and isotonic sodium chloride solution. After 12 hours of drops administration, corneas were excised, Clinical Relevance: Trovafloxacin may provide an ex- homogenized, and serially plated. The main outcome mea- cellent therapeutic alternative in bacterial keratitis. sure was quantitative bacteriologic analysis for residual colony-forming units. Arch Ophthalmol. 2004;122:65-69 ICROBIAL KERATITIS travenous administration. Trovafloxacin continues to be a com- has broad spectrum antibacterial activity mon, potentially sight- against gram-negative, gram-positive, and threatening ocular anaerobic bacteria. In vitro and in vivo infection.1,2 Fluoroqui- studies8-13 in systemic diseases have dem- Mnolones are increasingly selected as ini- onstrated a greater activity against clini- tial, broad-spectrum agents for the topical cally important gram-positive organisms 3 From the Ocular Microbiology therapy of bacterial keratitis. The use of (most notably streptococci), while main- Laboratory, The Wilmer ofloxacin and ciprofloxacin, as single taining activity against gram-negative or- Ophthalmological Institute, agents, was shown in clinical trials to be ganisms, when compared with ciprofloxa- The Johns Hopkins University comparable to a combination of fortified an- cin and ofloxacin. School of Medicine, Baltimore, tibiotics in the treatment of acute bacte- The purpose of this study is to deter- Md (Drs Barequet, Tuli, and rial keratitis.4,5 However, these 2 quino- mine the microbiological efficacy of topi- O’Brien and Messrs Denton and lones, which currently are often used, have cal trovafloxacin for potential use in Osterhout); and Goldschleger limited in vitro activity against some gram- therapy of experimental bacterial kerati- Eye Institute, Sheba Medical positive organisms and various strains of tis. After performing susceptibility and Center, Tel Aviv University 6,7 Sackler School of Medicine, Pseudomonas aeruginosa. pharmacokinetic studies, we studied the Tel Hashomer, Israel Trovafloxacin mesylate, 7-(3- microbiological effect of topical trova- (Dr Barequet). The authors azabicyclo[3,1,0]hexyl)-naphthyridone, is floxacin on bacterial keratitis caused by have no relevant financial a newer-generation synthetic fluoroqui- either gram-positive organisms, such as interest in this article. nolone currently available for oral and in- Staphylococcus aureus and Streptococcus (REPRINTED) ARCH OPHTHALMOL / VOL 122, JAN 2004 WWW.ARCHOPHTHALMOL.COM 65 ©2004 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 pneumoniae, or a gram-negative organism (P aerugi- late of a methicillin-sensitive S aureus with a predetermined nosa) and compared its efficacy with that of ciprofloxa- MIC50 (inhibits 50% of strains) of 0.5 µg/mL for ciprofloxa- cin and ofloxacin. cin. The other 2 experiments were performed in an identical fashion by inoculating either S pneumoniae or P aeruginosa. Twelve hours after the injection, the rabbits were random- METHODS ized to 4 treatment groups, with 6 rabbits in each group: (1) ciprofloxacin, 3 mg/mL (Ciloxan 0.3%; Alcon Laboratories, Fort BACTERIAL STRAINS Worth, Tex); (2) ofloxacin, 3 mg/mL (Ocuflox 0.3%; Aller- gan, Irvine, Calif); (3) trovafloxacin mesylate, 5 mg/mL (Trovan- Bacterial strains used in this study were clinical ocular isolates IV; Pfizer, New York, NY); and (4) preservative-free 0.9% non- collected from corneal specimens at the Wilmer Ophthalmo- bacteriostatic isotonic sodium chloride solution for intravenous logical Institute, Baltimore, Md. use (control). Trovafloxacin eyedrops were prepared using the intravenous solution that was transferred into a dropper bottle ANTIMICROBIAL SUSCEPTIBILITY STUDIES under a sterile hood. The eyedrops were administered every hour for 12 hours. Minimum inhibitory concentrations (MICs) of trovafloxacin, One hour after the last dose of eyedrops, the rabbits were sys- ciprofloxacin, and ofloxacin were determined for various ocu- temically anesthetized and then humanely killed by intracar- lar isolates of S aureus, S pneumoniae, Staphylococcus epidermi- diac injection of pentobarbital sodium (Beuthanasia-D Spe- dis, P aeruginosa, and Haemophilus influenzae. The E-test method cial; Schering-Plough Animal Health Corp, Kenilworth, NJ). was used for determining the MIC for each strain. The bacte- The rabbits’ corneas were excised using a sterile, disposable, rial suspension was prepared by collecting the clinical isolates 7.5-mm corneal trephine and irrigated with 3 mL of phosphate- from a blood agar plate. The isolate sample was adjusted with buffered saline to eliminate any residual antibiotic or debris on 0.9% nonbacteriostatic isotonic sodium chloride solution to the surface of the corneas. The corneal buttons were immedi- achieve the same density as a 0.5 McFarland standard (1 ϫ 108 ately homogenized and then serially diluted before plating in colony-forming units/mL). The appropriate E-test strip was duplicate on blood agar for S aureus and S pneumoniae and on placed on a Mueller-Hinton II agar plate (BBL, Cockeysville, Mueller-Hinton II agar for P aeruginosa. The specimens were Md) inoculated with the suspension of bacteria. The plates were incubated at 35°C for 24 hours before quantitative bacterio- incubated for 16 to 24 hours at 37°C, and the MIC was read logic analysis. from the scale on the side of the strip at the point where the ellipse of growth inhibition intercepted the strip. All the tests RESULTS were performed in duplicate. PHARMACOKINETIC STUDIES ANTIMICROBIAL SUSCEPTIBILITY STUDIES The pharmacokinetic studies were performed on rabbit eyes. The MICs for the various ocular strains are given in Twelve rabbits were used for this study. Epithelial removal (to Table 1. The MIC for P aeruginosa strains was signifi- promote antibiotic entry and to simulate human ulcerative kera- cantly lower for ciprofloxacin compared with trova- titis) was evaluated. In each rabbit, one eye underwent re- floxacin (P=.008) and ofloxacin (P=.007) and signifi- moval of the central (7.5 mm) corneal epithelium, and in the cantly lower for trovafloxacin compared with ofloxacin fellow eye the epithelium remained intact. A single drop of trova- (P=.007). floxacin was applied to both eyes. Rabbits were humanely killed The MIC for S aureus strains was significantly lower at 15, 30, 60, 120, and 240 minutes (at each time point, 3 rab- Ͻ bits were euthanized), and samples were obtained immedi- for trovafloxacin compared with ciprofloxacin (P .001) ately thereafter from aqueous humor, vitreous, and the central and ofloxacin (PϽ.001), and no significant difference was cornea (7.5 mm). The tissue concentration of trovafloxacin was found between the MIC for ciprofloxacin and ofloxacin determined using a high-performance liquid chromatography (P=.10). The MIC for H influenzae strains was signifi- assay.14 cantly lower for trovafloxacin (P=.005) and ciprofloxa- cin (P=.005) compared with ofloxacin. No significant dif- BACTERIAL KERATITIS STUDIES ference was found between trovafloxacin and ciprofloxacin (P=.44). The MIC for S epidermidis was lower for trova- In these studies, 72 (24 in each experiment) New Zealand white floxacin than ciprofloxacin; however, this was not sta- rabbits, weighing 2.0 to 2.25 kg, were used in accordance with the guidelines for animal experimentation established by the tistically conclusive (P=.07). The MIC for S pneumonia