In Vitro Efficacy and Pharmacodynamic Indices for Antibiotics against Coagulase-Negative Staphylococcus Endophthalmitis Isolates

Purpose: To compare pharmacodynamic indices and minimal inhibitory concentrations for vancomycin, gatifloxacin, moxifloxacin, linezolid, and combined quinupristin and dalfopristin for historic and current human coagulase-negative staphylococcus (CoNS) endophthalmitis isolates. Design: Experimental study. Participants: Fifty-nine CoNS endophthalmitis isolates retrieved from patients at the Bascom Palmer Eye Institute from 1993 through 2006. Methods: Coagulase-negative staphylococcal endophthalmitis isolates were recovered from the microbiology specimen bank, rehydrated, and processed for susceptibility testing using standard microbiological protocols. E tests were used to determine and compare mean inhibitory concentration for 50% of isolates (MIC₅₀) and mean inhibitory concentration for 90% of isolates (MIC₉₀) values. Peak concentration (C_max) was defined as the maximum attainable aqueous concentration using topical or oral therapy, or both. The MIC₅₀ and MIC₉₀ values for each antibiotic are the minimum concentrations that inhibit 50% and 90% of CoNS endophthalmitis isolates, respectively. Significance was determined by the McNemar test. Pharmacodynamic indices (C_max/MIC) were calculated using determined MIC values and published intraocular drug concentrations for topical and oral dosing. The pharmacodynamic index was defined as the achievable aqueous humor concentration of an antibiotic divided by the concentration of the antibiotic required to inhibit a specified percentage of microbiologic isolates. Main Outcome Measures: Pharmacodynamic indices for new and conventional antibiotics. Results: General in vitro susceptibility patterns in descending order were vancomycin (100%), linezolid (100%), quinupristin and dalfopristin (98%), moxifloxacin (48%), and gatifloxacin (47%). The corresponding MIC₅₀ and MIC₉₀ values were vancomycin, 2 μg/ml and 3 μg/ml, respectively; linezolid, 1 μg/ml and 4 μg/ml; quinupristin and dalfopristin, 0.25 μg/ml and 0.5 μg/ml; moxifloxacin, 0.75 μg/ml and ≥32 μg/ml; and gatifloxacin, 2 μg/ml and ≥32 μg/ml. Pharmacokinetic indices (C_max/MIC₉₀) for topical dosing were all <1. There was a significant difference in the percent of isolates susceptible to combined quinupristin and dalfopristin, vancomycin, and linezolid compared with moxifloxacin and gatifloxacin. There was no statistical significance in CoNS susceptibility between the tested fluoroquinolones. Conclusions: Vancomycin, linezolid, and combined quinupristin and dalfopristin were more effective in vitro than fluoroquinolones against CoNS in the current study. Reported aqueous concentrations of the antibiotics used in this study failed to provide adequate coverage for 90% of the CoNS endophthalmitis isolates.