Journal of Antimicrobial Chemotherapy (9), Supp. D, 9 6 Resistance pattern of anaerobic bacteria isolated in a general hospital during a two-year period J. F. Acar, F. W. Goldstein, M. D. Kitzis and M. T. Eyquem Hopital Saint Joseph, Universite Pierre & Marie Curie, 7 Rue Pierre Larousse, 767 Paris Cedex, France Summary During a -year period, 77 anaerobic bacteria were isolated from clinical specimens. Bacteroides sp. were the most common bacteria (9%) followed by Peptococcus and Peptostreptococcus (%) and Clostridium (6%). On the basis of disc tests, Bacteroides sp. have been found resistant to ampicillin (6%) tetracycline (6%), clindamycin (0%), chloramphenicol (6-%) and metronidazole (%). Some strains had joint resistance to several useful antibiotics. In some cases, resistance was confirmed by MIC determinations. Introduction Anaerobic bacteria are isolated from infected patients with increasing frequency, even in unspecialized laboratories. It has been generally assumed, since anaerobic infections are frequently polymicrobic and cultures require relatively long periods of time for growth and identification, that the routine susceptibility testing of isolates is unpractical. However, the in-vitro antimicrobial susceptibility pattern of anaerobic bacteria seems to undergo gradual change (Kislak, 97; Phillips, 97; Chow, Patten & Guze 97; Sutter & Finegold, 976; Appelbaum & Chatterton, 97; Dubois, Pechere & Turgeon 97; Percival & Cumberland 97; Sutter et al. 97; Brown & Waati, 90; King et al., 90; Tytgat, 90; Soussy, Thibault & Duval, 9). Periodic surveys to monitor changing patterns of resistance and to determine the possible effectiveness of new agents are thus very desirable. Such surveys can be made by centres having the capability to receive, to store and to test a large number of strains sent from different hospitals in a country. The results might help in the empirical selection of the right antibiotic at the bedside of the patient. We have developed in an unspecialized routine laboratory, a simple disc method for rapid estimation of the antibiotic susceptibility of anaerobic bacteria, both to help the making of therapeutic decisions in individual cases and to improve the sampling of strains which have to be saved and sent to the centre doing periodic surveys of resistance patterns in anaerobes. Downloaded from http://jac.oxfordjournals.org/ at Pennsylvania State University on May, 06 Material and Methods In a general hospital of 900 beds, the study reveiwed anaerobic infections from which was isolated at least one anaerobic bacterium, in a two-year period from January 979 00-7//DO0O9 + 0 $0.00/0 9 9 The British Society for Antimicrobial Chemotherapy
0 J. F. Acar et al Table I. Estimation of susceptibility by inhibition zone diameter measurements. Antibiotics Ampicillin Cefoxitin Tetracycline Clindamycin Chloramphenicol Metronidazolc Disc content G*g) 0 0 0 0 Susceptible > 7mm > mm > 6mm > mm > mm > 0mm to December 90. The procedure of isolation was based on plating the suspected specimen on Columbia ANC-blood agar, Chocolate Isovitalex agar and B.H.I, agar supplemented with vitamin K, ( mg/) Isovitalex (%), vancomycin (0 mg/) and kanamycin (0 mg/). The plates were incubated h in Gas Pak jars at 7 C. Identification was mainly performed by API 0 ANA. The antibiotic susceptibility testing was performed by the single disc method on Mueller Hinton agar with % horse blood and vitamin K, ( mg/). Suspensions of colonies taken from the plates were adjusted to the turbidity of No. MacFarland standard giving 0 7-0 cfu/ml. The inoculum was applied by flooding the agar plates. Zones of inhibition were measured after h incubation in Gas Pak jars. The antibiotics tested and the interpretation of susceptibility are summarized on Table I. Bacteroides fragilis and clostridia could be easily tested by this method. Agar dilution tests were performed by incorporating serial twofold dilutions of the antibiotics in Mueller-Hinton agar with % horse blood and vitamin K, ( mg/). The inoculum was applied by means of a Steers replicator, yielding 0* cfu/spot. The Minimum Inhibitory Concentration (MIC) was determined after -h incubation in Gas Pak jars. The agar dilution test was applied to organisms exhibiting an unusual pattern of resistance on disc testing and to organisms isolated from severe or poorly-controlled infections. The antibiotics tested and the critical MIC for considering a strain resistant were as follows: ampicillin (^ mg/), carbenicillin (^ mg/), cefoxitin (;> mg/), moxalactam (^ mg/), thienamycin (^ mg/), chloramphenicol (> 6 mg/), tetracycline (S; mg/), metronidazole (>mg/l), clindamycin (^ mg/). Downloaded from http://jac.oxfordjournals.org/ at Pennsylvania State University on May, 06 Results The specific identification of the 77 bacteria used in this study is given in Table II. Gram-positive cocci include peptococci and peptostreptocci. Table III gives the origin of the strains studied: the 77 strains were isolated from 9 patients. The overall proportion of B. fragilis, Bacteroides sp. and clostridia resistant to the antibiotics tested by disc method is given in Table IV. Seventy-five per cent of the B. fragilis resistant to clindamycin were also resistant to tetracycline. Tables V to VII summarize the results of MIC determinations. Nine out of 7 Bacteroides spp. tested were resistant to clindamycin with MIC ^6 mg/, and strains were highly resistant to metronidazole with MIC;> mg/. The resistance pattern of the strains tested by MIC is given in Table VIII.
Resistance pattern of anaerobic bacteria Table n. Identification of the anaerobic species isolated Gram-negative bacilli Pus* Genital infections Blood cultures Pulmonary infections Miscellanous Total Bacteroides fragilis Other Bacteroides Bacteroides spp. Fusobacterium sp. Unidentified GNB Gram-positive bacilli Clostridium perfringens Clostridium difficile Actinomyces sp. Bifidobacterium sp. Propionibacterium sp. Other bacilli Gram-positive cocci 0 9 7 9 9 Table HI. Origin of the anaerobic bacteria Gram-negative Gram-positive Gram-positive bacilli bacilli cocci Total 6 6 0 (9%) 'Including peritoneal drainage. 0 9 (6%) Discussion 76 9 (%) B. fragilis and clostndia can be easily isolated even in ill-equipped routine laboratories. Several studies have demonstrated that for these rapidly growing bacteria, the prediction of susceptibility based on zone diameter measurements is generally accurate. (Wilkins et al, 97; Sutter, Kwok & Finegold, 97; Kwok et al, 97). However, this method may overestimate resistance to some antibiotics if the media, the inoculum and the incubation conditions are not carefully respected (Rahimi, Martin & Washington, 97). This concerns especially metronidazole and clindamycin. With a too heavy inoculum or when the anaerobic conditions are not satisfactory, growth of anaerobic bacteria may still occur but a sharp decrease of the inhibition zone around the metronidazole disc will be observed. If the concentration of CO, is too high, as often observed in the Gas Pak jars there may be a decrease in the ph of unbuffered media which also results in a decrease of the inhibition zone around the clindamycin disc, as reported for other macrolides (Ingham et al, 970). A too low disc content, as in the 7. 77 Downloaded from http://jac.oxfordjournals.org/ at Pennsylvania State University on May, 06
Table IV. Resistance of anaerobic bacteria to antibiotics when tested by the disc method. No. of isolates Ampicillin Cefoxitin % Isolates iresistant to Chloramphenicol Tetracycline Clindamycin Metronidazole Bacteroides fragilis Bacteroides sp. Clostridium perfringens 9 96 70 0 6-0 -9 60-7 60-9 6 9-6 0-7- -
Resistance pattern of anaerobic bacteria Table V. MICs for strains of Bacteroides fragilis 0 0-0- Number of isolates with MIC 6 (rag/) of 6 :6 Ampicillin Carbenicillin Cefoxitin Moxalactam Thienamycin 9 7 7 9 6 Chloramphenicol Tetracycline Clindamycin Metronidazote npicillin irbcnicillin :foxitin oxalactam lienamycin lloramphenicol itracycline indamycin etronidazole 00 6 TaWe VI. MICs for strains of Bacteroides sp. Number of isolates with MIC (mg/) of 006 0 0-0- 6 Table VII. MICs for 0 strains of Clostridium sp. 6 6 >6 Downloaded from http://jac.oxfordjournals.org/ at Pennsylvania State University on May, 06 00 006 Number of isolates with 0 0-0- MIC (mg/) of: 6 6 6 jnpicillin 'arbenicillin efoxitin foxalactam hienamycin 7 Woramphenicol etracycline lindamycin letronidazole
Table VHI. Resistance pattern of anaerobes tested by agar dilution method. o a -a a. S u e g III II <U S P u H u S. fragilis + + B. fragilis + + B. fragilis + + + fi. fragilis + + - +. fragilis + + + B. fragilis + + + B. fragilis + + + B. fragilis B. fragilis + + + + + + B. fragilis + + + B. fragilis + + + + B. fragilis + + + + B. fragilis + + + + + B. distasonis + + B. distasonis + + + + B. distasonis + + + + + + + + Bacteroides sp. + + Bacteroides sp. + + + Cl. perfringens + + Cl. perfringens + + tp://jac.oxfordjournals.org/ at Pennsylvania State University on May, 06
Resistance pattern of anaerobic bacteria fig clindamycin disc, may also result in false reports of resistant strains. Thus, suspected resistance to these antibiotics must always be checked by agar dilution methods in presence of control strains with known MIC (Sutter et al., 979). In our study, about % of the bacteroides are resistant to mg/ of metronidazole, which is higher than some previous reports from France and U.S.A. (Sutter & Finegold, 976; Henderson, Chow & Guze, 977; Dubois el al., 97; Tytgat 90; Dublanchet & Durieux, 90). In South Africa, % of B. fragilis have been reported resistant to 6 mg/ metronidazole (Appelbaum & Chatterton, 97). In another U.S. report, 7% of the Bacteroides sp. studied were resistant to 00 mg/ metronidazole (Chow et al., 97). However, it must be emphasized that these findings do not appear to have been independently corroborated. The widespread use of metronidazole as a therapeutic and prophylactic agent in some areas may explain some of these discrepancies. Resistance to clindamycin was uncommon a few years ago and remains low even in recent reports (Sutter & Finegold, 976; Appelbaum & Chatterton 97; Kwok et al., 979; Tytgat, 90; Brown & Waati, 90). The high level of 0% clindamycin-resistant Bacteroides encountered in this study cannot be easily explained: when tested by agar dilution methods, 9 out 7 Bacteroides sp. were resistant to more than 6 mg/ clindamycin. Streptogramin antibiotics, not used elsewhere, are widely used in France in animal and human infections. These antibiotics are able to select strains resistant to macrolides and clindamycin. Finally, the incidence of plasmid-mediated clindamycin resistance has to be monitored (Privitera et al., 979). Results for the other antibiotics are in agreement with the findings in previous studies. Almost all the B. fragilis are resistant to ampicillin. More than 60% of Clostridium sp. and 70% of Bacteroides sp. are resistant to tetracycline. In contrast, all the Clostridium sp. were sensitive to penicillins and cephalosoporins. Resistance of Bacteroides sp. to cefoxitin and moxalactam is still rare, and resistance to chloramphenicol is very uncommon. Of special interest are strains exhibiting joint resistance to the clinically most useful antibiotics cefoxitin, clindamycin and metronidazole. In our survey, % of the Bacteroides sp. are resistant to at least of these antibiotics. One strain of Bacteroides distasonis isolated from a blood culture was resistant to all the antibiotics tested. Antibiotic resistant anaerobic bacteria are encountered with an increasing frequency but with few exceptions the percentage of resistant strains is still moderate and in most cases at least two major products remain effective. However, periodic surveys should be done regularly to detect and to follow the spread of anaerobic bacteria resistant to antibiotics. Downloaded from http://jac.oxfordjournals.org/ at Pennsylvania State University on May, 06 References Appelbaum, P. C. & Chatterton, S. A. (97). Susceptibility of anaerobic bacteria to ten antimicrobial agents. Antimicrobial Agents and Chemotherapy, 7-6. Brown, W. J. & Waati, P. E. (90). Susceptibility testing of clinically isolated anaerobic bacteria by an agar dilution technique. Antimicrobial Agents and Chemotherapy 7, 69-. Chow, A. W., Patten, V. & Guze, L. B. (97). Susceptibility of anaerobic bacteria to metronidazole: relative resistance of non-spore-forming gram positive bacilli. Journal of Infectious Diseases, -. Dublanchet, A. & Durieux, R. (90). Sensibilite comparee in vitro de quelques bacteries aerobics et anaerobies a trois derives des -nitro-imidazoles: le metronidazole, l'ornidazole et le tinidazole. Annales de Microbiologie A, -9.
6 J. F. Acar et al. Dubois, J., Pechere, J. C. & Turgeon, P. (97). Activity of ten antimicrobial agents against anaerobic bacteria. Journal of Antimicrobial Chemotherapy., 9-. Henderson, D. K., Chow, A. W. & Guze, L. B. (977). Comparative susceptibility of anaerobic bacteria to ticarcillin, cefoxitin metronidazole and related antimicrobial agents. Antimicrobial Agents and Chemotherapy, 679-. Ingham, H. R., Selkon, J. B., Codd, A. A. & Hale, J. H. (970). The effect of carbon dioxide on the sensitivity of Bacteroides fragilis to certain antibiotics in vitro. Journal of Clinical Pathology, -. King, A., Warren, C, Shannon, K. & Phillips, I. (90). The in-vitro antibacterial activity of cefotaxime compared with that of cefuroxime and cefoxitin. Journal of Antimicrobial Chemotherapy 6, 79-. Kislak, J. W. (97). The susceptibility of Bacteroides fragilis to antibiotics. Journal of Infectious Diseases, 9-9. Kwok, Y. Y., Tally, F. P., Sutter, V. L. & Finegold, S. M. (97). Disc susceptibility testing of slow-growing anaerobic bacteria. Antimicrobial Agents and Chemotherapy 7, -7. Percival, A. & Cumberland, N. (97). Antimicrobial susceptibilities of Gram-negative anaerobes. Journal of Antimicrobial Chemotherapy, Suppl. C, -. Phillips, I. (97). Antibiotic sensitivity of non-sporing anaerobes. In Infection with Non-sporing Anaerobic Bacteria (Phillips, I. & Sussman, M., Eds). Churchill Livingstone, Edinburgh, London & New York. Privitera, G., Dublanchet, A. & Sebald, M. (979). Transfer of multiple antibiotic resistance between subspecies of Bacteriodes fragilis. Journal of Infectious Diseases 9, 97-0. Rahimi, A., Martin, W. J. & Washington, II J. A. (97). Effect of medium and inoculum on antimicrobial susceptibility of anaerobic bacteria. American Journal of Clinical Pathology 6, -7. Soussy, C. J., Thibault, M. & Duval, J. (9). Activite antibacterienne in vitro du cefotaxime sur les bacteries anaerobies. La Nouvelle Presse Medicate 0, 6-. Sutter, V. L., Kwok, Y. Y. & Finegold, S. M. (97). Susceptibility of Bacteroides fragilis to 6 antibiotics determined by standardized antimicrobial disc susceptibility testing. Antimicrobial Agents and Chemotherapy, -9. Sutter, V. L. & Finegold, S. M. (976). Susceptibility of anaerobic bacteria to antimicrobial agents. Antimicrobial Agents and Chemotherapy 0, 76-7. Sutter, V. L., Oberhammer, I., Kwok, Y. Y. & Finegold, S. M. (97). Susceptibility of anaerobes to cefoxitin and cephalotin. Journal of Antimicrobial Chemotherapy, -6. Sutter, V. L., Barry, A. L., Wilkins, T. D. & Zabransky, R. J. (979). Collaborative evaluation of a proposed reference dilution method of susceptibility testing of anaerobic bacteria. Antimicrobial Agents and Chemotherapy 6, 9-0. Tytgat, F. (90). Activite de sept antibiotiques sur Bacteroides fragilis. Annules de Microbiologie (lnstitut Pasteur), A: 9-. Wilkins, T. D., Holdeman, L. V., Abramson, I. J. & Moore, W. E. C. (97). Standardized singledisc method for antibiotic susceptibility testing of anaerobic bacteria. Antimicrobial Agents and Chemotherapy, -9. Downloaded from http://jac.oxfordjournals.org/ at Pennsylvania State University on May, 06