TEXT

Top Abstract Text References

Coryneform bacteria are increasingly recognized as a cause of human infection, particularly in hospitalized patients (6, 9). Molecular and chemotaxonomic methods have successfully been used in the last decade to define new genera and species and to reclassify previously recognized coryneform bacteria, but very few studies have been performed relative to their medical and clinical importance. Corynebacterium urealyticum, C. jeikeium, C. amycolatum, C. striatum, and C. minutissimum and Brevibacterium spp. are some of the most frequently isolated coryneform bacteria in clinical microbiology laboratories (6, 9, 12). Coryneform Centers for Disease Control and Prevention (CDC) groups I2 and F2 are now included in the newly described species C. amycolatum (2, 4). Most of the organisms previously identified as C. xerosis, some C. minutissimum strains, and a few C. striatum strains are also currently known to be C. amycolatum. Another clinically important gram-positive rod to be considered is Listeria monocytogenes.

C. jeikeium and C. urealyticum are commonly multi-drug resistant, although strains of both species are known to be susceptible to a variety of antimicrobial agents (3, 11-14), such as glycopeptides, and variably resistant to erythromycin, tetracycline, and fluoroquinolones. Multi-drug resistance is not found only in these two organisms; other species and groups of coryneform bacteria (including, for example, C. amycolatum or coryneform CDC group G) are frequently resistant to many of the antimicrobial agents available for therapeutic use (5, 9).

Ketolide HMR 3647 is a new semisynthetic 14-membered ring macrolide with a 3-keto group instead of the L-cladinose sugar (1). Ketolides show the same antibacterial spectrum as reference macrolides but also have good activity against erythromycin-resistant isolates among gram-positive cocci (7). The purpose of this study was to evaluate the in vitro activity of the new ketolide HMR 3647 against coryneform bacteria and L. monocytogenes isolated from human samples compared with those of azythromycin, 14- and 16-membered macrolides, and clindamycin.

(This work was presented at the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 27 September to 1 October 1997.)

The following test organisms (number of isolates) were isolated from clinical samples at the Department of Clinical Microbiology, University Hospital Virgen Macarena, Seville, Spain: C. amycolatum (35), C. jeikeium (20), C. urealyticum (20), C. striatum (20), C. minutissimum (20), L. monocytogenes (15), and Brevibacterium spp. (10). Organisms were identified by conventional phenotypic tests and the API Coryne System, according to Funke et al. (6). C. amycolatum strains included organisms identified as C. xerosis, C. minutissimum, and CDC coryneform group I2 by the CDC scheme for identification of coryneform bacteria (4). After isolation and identification, bacteria were maintained in 10% glycerol in tryptic soy broth at 70°C until used. The following reference strains were included: C. jeikeium ATCC 43734, C. urealyticum ATCC 43042, and C. striatum ATCC 6940. Staphylococcus aureus ATCC 29213 and Enterococcus faecalis ATCC 29212 were used as control strains.

HMR 3647 (Hoechst-Marion-Roussell, Romainville, France), erythromycin A (Sigma, St. Louis, Mo.), clarithromycin (Abbott Laboratories, Madrid, Spain), roxithromycin (Hoechst-Marion-Roussell), josamycin (Sigma), azithromycin (Pfizer S.A., Madrid, Spain), and clindamycin (Pharmacia-Upjohn, Kalamazoo, Mich.) were used as powders of known potency. Solutions of antimicrobial agents were prepared on the day of antimicrobial testing, according to the manufacturer's instructions.

A microdilution method was used for susceptibility testing. The guidelines provided by the National Committee for Clinical Laboratory Standards (NCCLS) for dilution susceptibility testing (10) were followed, with some exceptions, as coryneform bacteria are not discussed in this document. Cation-adjusted Mueller-Hinton broth (Difco, Detroit, Mich.) was supplemented with 0.5% Tween 80 in tests of the lypophilic species C. jeikeium and C. urealyticum. Twofold dilutions ranging from 0.03 to 64 µg/ml for all antimicrobial agents were tested. The inoculum was prepared from bacterial cultures grown on Columbia agar with 5% sheep blood agar for 20 to 24 h. Bacteria were suspended in Mueller-Hinton broth with (C. jeikeium and C. urealyticum) or without 0.5% Tween 80 to a concentration of about 10⁸ CFU/ml and diluted in the same medium to obtain a final concentration of 5 × 10⁴ CFU per well. The final volume in each well of the microtiter plate was 100 µl. Plates were inoculated and then incubated in air at 35°C for 20 to 22 h or 48 h if no growth was observed at 20 to 22 h or if the organism was C. striatum.

The MICs, MICs at which 50% of the isolates are inhibited (MIC₅₀s), and MIC₉₀s of the different antimicrobial agents evaluated are shown in Table 1. The macrolides and clindamycin had low activity against Corynebacterium spp. The ketolide HMR 3647 showed high activity against C. amycolatum, C. striatum, and C. minutissimum but low activity against the more-resistant species C. urealyticum and C. jeikeium. Similar activities of macrolides and clindamycin against Corynebacterium spp. have been reported in other studies (5, 8, 14), although Soriano et al. (14) reported the MIC₉₀s of erythromycin for C. minutissimum strains to be several times higher than those obtained here. It is possible that some of the C. minutissimum strains included in the study of Soriano et al. belong to the newly-recognized C. amycolatum species.

It is important that remarkable differences in the MICs of ketolides and macrolides can be observed for C. striatum if readings are made at 48 h instead of 20 h. At 48 h, all C. striatum strains showed a hazy growth, lower than that observed in control strains or in other coryneforms, in wells containing concentrations of 32 to 64 µg/ml. Since the meaning of this residual growth remains unclear, and C. striatum is a nonfastidious bacterium, we chose to establish MICs at 20 h following NCCLS guidelines for nonfastidious microorganisms. This phenomenon could partially explain the higher MICs of HMR 3647 and erythromycin for C. striatum isolates that were obtained by other authors (15).

At a concentration of 0.5 µg/ml, which is the breakpoint of the NCCLS for erythromycin and clindamycin, more than 90% of C. amycolatum, C. striatum, and C. minutissimum strains were inhibited by HMR 3647. These values decreased to 20 and 45% for C. urealyticum and C. jeikeium, respectively. At this concentration, none of the macrolides inhibited more than 20% of Corynebacterium spp., with the exception of C. minutissimum. Forty to 70% of strains of the latter species were inhibited by macrolides at 0.5 µg/ml.

The lowest MIC₉₀s for L. monocytogenes were those of clarithromycin (0.125 µg/ml) and those of HMR 3647 and erythromycin (0.25 µg/ml for both antimicrobial agents). At a concentration of 0.5 µg/ml, more than 90% of the strains were inhibited by HMR 3647, erythromycin, clarithromycin, and roxithromycin. Josamycin, azithromycin, and clindamycin showed very low activity against this microorganism.

HMR 3647 was the most-active agent against Brevibacterium spp. The MIC₉₀ of clarithromycin (2 µg/ml) was much lower than those of the other macrolides (16 µg/ml).