LETTER

Propionibacterium acnes has become an important pathogen in nosocomial infection, including endophthalmitis. The treatment of endophthalmitis presents many diagnostic and therapeutic challenges. Although P. acnes has demonstrated in vitro susceptibilities to many beta-lactam agents, clindamycin, and macrolides, intravitreal treatment success in patients has been inconsistent.

While quinolones are not considered the drugs of choice for anaerobic infections, we previously demonstrated high-level in vitro susceptibilities of P. acnes to sparfloxacin, ofloxacin, and ciprofloxacin (7). Moxifloxacin is a newer fluoroquinolone with reported enhanced antibacterial activity against anaerobes (2, 3). We conducted this study to determine the susceptibility of P. acnes eye isolates to moxifloxacin using the E test. The E test has been shown to be a reliable method for MIC susceptibility testing of anaerobic bacteria (1, 7). If these data indicate in vitro susceptibility to this newer quinolone, further studies using an experimental rabbit model of P. acnes endophthalmitis would be warranted.

Twelve clinical isolates of P. acnes originally recovered from eye specimens submitted to the Division of Microbiology of the Long Island Jewish Medical Center, New Hyde Park, N.Y., were used in this study (7). Additionally, two strains of P. acnes (ATCC 11827 and 11828) and one strain of B. fragilis (ATCC 25285) served as controls. Other conventional control organisms were also tested in accordance with the National Committee for Clinical Laboratory Standards (NCCLS) guidelines (4, 5) to ensure that experimental conditions were met.

Anaerobic organisms were initially grown on commercially prepared sheep blood agar (Becton Dickinson Microbiology Systems, BBL, Cockeysville, Md.) for 48 h at 35°C under anaerobic conditions (BBL GasPak system). Inocula for the susceptibility studies were prepared by suspending a 48-h culture in reduced brucella broth (BBL) to achieve the density of a 1.0 McFarland nephalometer standard. Susceptibility testing was performed on sheep blood-enriched Mueller-Hinton agar. MICs were determined by means of the E test (AB Biodisk North America, Piscataway, N.J.) following a 48-h incubation at 35°C in an anaerobic chamber in accordance with NCCLS guidelines (6). MIC was defined as the point on the epsilometer scale at which the ellipse of growth inhibition intercepted the test strip. Any intercept value which fell between two points on the scale was rounded up to the next higher value. All isolates, including control organisms, were tested in triplicate for susceptibility to moxifloxacin (Bayer Pharmaceuticals, West Haven, Conn.), ceftriaxone (Roche, Nutley, N.J.), ciprofloxacin (Miles Laboratories, West Haven, Conn.), clindamycin (Upjohn Co., Kalamazoo, Mich.), and metronidazole (McGaw, Irvine, Calif.).

The P. acnes strains in this study were the same strains as those in our previous study (7). All strains demonstrated high-level susceptibilities to moxifloxacin, ceftriaxone, ciprofloxacin, and clindamycin by the E test. The MICs of moxifloxacin were within 1 antibiotic dilution of each other (0.094 to 0.125 µg/ml). MICs clustered around 0.25 µg/ml for ciprofloxacin, with a range of 0.38 to 0.19 µg/ml. There was greater variability in P. acnes susceptibility to ceftriaxone, with MICs ranging from 0.064 to 0.38 µg/ml. For 50% of the organisms tested, the MIC was 0.38 µg/ml. For the majority of the isolates, the MIC of clindamycin was either 0.032 or 0.047 µg/ml. For two isolates, the MICs were somewhat higher: 0.125 and 0.25 µg/ml. All isolates were found to be resistant to metronidazole. Results are shown in Table 1.

P. acnes clinical isolates demonstrated high-level in vitro susceptibilities to moxifloxacin, ciprofloxacin, ceftriaxone, and clindamycin and resistance to metronidazole. Results obtained for moxifloxacin, ceftriaxone, ciprofloxacin, clindamycin, and metronidazole are in agreement with previously reported findings (1-3, 7).

Interestingly, the MICs of ciprofloxacin were lower for several P. acnes clinical isolates compared with data from our previous study. The MICs of ciprofloxacin were comparable to those of sparfloxacin and ofloxacin in that study. We did not test ofloxacin or sparfloxacin in the present study.

On the basis of our study, all the clinical eye isolates were susceptible to the new fluoroquinolone. This study supplements existing information regarding the in vitro susceptibility of P. acnes. The data presented suggest that moxifloxacin is a suitable antimicrobial agent for evaluation in a rabbit model of P. acnes endophthalmitis.