ANTiMICROBLAL AGENTS AND CHEMOTHERAPY, Aug. 1993, p. 1649-1654 0066-4804/93/081649-06$00/0 Copyright 1993, American Society for Microbiology Vol. 37, No. 8 Susceptibilities of 428 Gram-Positive and -Negative Anaerobic Bacteria to Compared with Their Susceptibilities to,,,, -Tazobactam, and G. A. PANKUCH,1 M. R. JACOBS,2 AND P. C. APPELBAUM'* Departments ofpathology (Clinical Microbiology), Hershey Medical Center, Hershey, Pennsylvania 17033,1 and Case Western Reserve University, Cleveland, Ohio 441062 Received 8 March 1993/Accepted 26 May 1993 The susceptibilities of 428 gram-negative and gram-positive anaerobes (including selected cefoxitin-resistant strains) to (a new fluoroquinolone), ciprofloxacin, clindamycin, metronidazole, cefoxitin, piperacillin, and piperacillin-tazobactam were tested. Organisms comprised 115 BacteroidesfragUis group, 116 non-b. fragilis Bacteroides, PrevoteUa, and Porphyromonas spp., 40 fusobacteria, 58 peptostreptococci, 48 grampositive non-spore-forming rods, and 51 clostridia. P-Lactamase production was demonstrated in 87% of the gram-negative rods but in none of the gram-positive organisms. Overall, was the most active agent, with all organisms inhibited at an MIC of <,g/ml (MICs for 50%o [MIC5J1 and %o [MIC] of strains tested, and p,g/ml, respectively). By contrast, ciprofloxacin was much less active, with only 42% of strains susceptible at a breakpoint of,lg/ml (MIC50,,ug/ml; MIC,0,,ug/ml). was active against all gram-negative rods, but 7% of peptostreptococci, 83% of gram-positive non-spore-forming rods, and 4% of non-clostridium perfringens, non-clostridium dijfficile clostridia were resistant to this agent (MICs, > ig/ml). was active against 94% of Bacteroides, PrevoteUa, and Porphyromonas spp., 91% of peptostreptococci, and %6 of gram-positive non-spore-forming rods, but was active against only 701% of fusobacteria and 53% of clostridia. was active against.%o of all groups except the B. fragilis group and non-propionibacterium acnes gram-positive non-spore-forming rods (both 85%) and C. diffile (20%k). Significant enhancement of piperacillin by tazobactam was seen in all 13-lactamase-positive strains (991% susceptible; MICo,,,Lg/ml), and all 13-lactamase-negative strains were susceptible to piperacillin (MIC,,ug/ml). Clinical studies are required to delineate the role of in the treatment of anaerobic infections. Anaerobes are well-established human pathogens, especially when host defenses are lowered by processes such as trauma, malignancy, surgery, and malnutrition (1, 27). Although gram-negative rods (in particular, the Bacteroides fragilis group) represent the most important group of pathogenic anaerobes in humans, infections with other anaerobic gram-negative rods as well as with anaerobic gram-positive cocci and rods (spore formers as well as non-spore formers) are increasingly encountered (5, 27). The susceptibility patterns of clinically significant anaerobes are changing. 3-Lactamase production and resistance to 3-lactams are usually encountered in the B. fragilis group. However, both of the latter phenomena have been increasingly found in non-b. fragilis group Bacteroides, Prevotella, Porphyromonas, and Fusobacterium species (2-7, 9, 19). Although imipenem resistance in the B. fragilis group is exceedingly rare in the United States and Europe (5, 19), approximately 6% of B. fragilis and Bacteroides thetaiotaomicron strains in Japan are currently resistant to this antimicrobial agent; 29% of these strains produce imipenemase (9a). Not all anaerobes are susceptible to clindamycin, and metronidazole resistance has been encountered in anaerobic gram-positive cocci (7, 12). 1B-Lactamase production has also been described in some non-clostridium perfringens Clostridium species (8, 22) Ṫhe development of agents such as fluoroquinolones, new * Corresponding author. 1649 j-lactam-1-lactamase inhibitor combinations, new broadspectrum,b-lactams, and other compounds mandates susceptibility testing surveys by standardized methods for clinical strains isolated within a few years of the survey (5, 19). The currently marketed fluoroquinolones such as ciprofloxacin and ofloxacin are not active against most groups of anaerobic bacteria (16, 17). {8-chloro-1-cyclopro- pyl-7-[(s,s)-2,8-diazabicyclo[4.3.0.1-non-8-yl]-6-fluoro-1,4- dihydro-4-oxo-3-quinolone-carboxylic acid hydrochloride} is a new fluoroquinolone with potent antibacterial activity against both aerobic and anaerobic bacteria (10, 13, 20, 23, 28). In the survey described here, we used standard methods to compare the activity of with those of clindamycin, ciprofloxacin, metronidazole, piperacillin, piperacillintazobactam, and cefoxitin against 115 B. fragilis group, 116 non-b. fragilis group Bacteroides, Prevotella, and Porphyromonas species, 40 fusobacteria, 58 anaerobic cocci, 48 anaerobic gram-positive non-spore-forming rods, and 51 clostridia. MATERIALS AND METHODS Bacteria. Organisms were obtained from Hershey Medical Center, University Hospitals of Cleveland, and the centers listed in the Acknowledgments Ijn addition, 3 imipenemresistant B. fragilis strains (2 irtipenemase producers) were kindly provided by K. Ueno (Gifu University School of Medicine, Gifu, Japan), and 11 cefoxitin-resistant members of the B. fragilis group were obtained from D. Shungu
1650 PANKUCH ET AL. (Merck & Co., Rahway, N.J.). Strains other than the latter were all recent clinical strains that were isolated between 1988 and 19 and that were obtained from normally sterile sites, blood cultures, and material from patients with orofacial, pulmonary, intra-abdominal, upper genital tract, and other infections. Most strains were obtained prior to the institution of antimicrobial therapy. Gram-negative rods were selected such that f-lactamase-positive strains predominated. Strains were stored in double-strength skim milk (Difco Laboratories, Detroit, Mich.) at -70 C. Prior to testing, the organisms were checked for purity as described previously (5). Identification was by standard methods (1, 18, 27). 13-Lactamase and susceptibility testing. f-lactamase testing was done by the nitrocefin disk method (Cefinase disks; BBL Microbiology Systems, Cockeysville, Md.) as described previously (4). Testing of susceptibility to and ciprofloxacin (Miles, Inc., West Haven, Conn.), clindamycin (The Upjohn Company, Kalamazoo, Mich.), metronidazole (Searle, Inc., Skokie, Ill.), piperacillin and piperacillin-tazobactam (Lederle Laboratories, Pearl River, N.Y.), and cefoxitin (Merck Laboratories, West Point, Pa.) was performed by the agar dilution method recommended by the National Committee for Clinical Laboratory Standards (21) with Wilkins-Chalgren agar supplemented with 5% sterile defibrinated sheep blood for fastidious organisms. Plates were incubated in an anaerobe chamber (Coy Laboratory Products, Ann Arbor, Mich.) in an atmosphere of 80% N2-10% C02-10% H2. Standard anaerobic quality controls were included in each run (21). Tazobactam was added to piperacillin at a fixed concentration of,ug/ml (7). Definition of enhancement of piperacillin by tazobactam was as described previously (2). When available, National Committee for Clinical Laboratory Standards breakpoints (21) were used. The following susceptibility breakpoints were used:,ug/ml for and ciprofloxacin,,g/ml for clindamycin,,ug/ml for metronidazole, 3 p,g/ml for cefoxitin, and 6,g/ml for piperacillin and piperacillintazobactam. The breakpoint chosen for was empiric, because no standardized breakpoint is currently available for this compound. RESULTS P-Lactamase production was demonstrated in 97% of B. fragilis group strains, % of non-b. fragilis group Bacteroides, Prevotella, and Porphyromonas species, and 52% of fusobacteria. By contrast, none of the gram-positive strains was shown to produce 1-lactamase. The susceptibilities of all organisms to the various antimicrobial agents tested are presented in Table 1. Because the majority of gram-negative rods produced j3-lactamase, results for enzyme-positive and -negative strains are presented together. With the exception of the fusobacteria (see below), the results for enzymenegative strains did not differ significantly from those obtained when,-lactamase-producing strains were analyzed separately. As can be seen from Table 1, was extremely active against all anaerobe groups at MICs of.,ug/ml (% susceptible at this provisional breakpoint), with an overall MIC for 50% of isolates tested (MIC50) of p,g/ml and an MIC of p,g/ml. For only five strains MICs were,ug/ml; these strains comprised four strains of Fusobacterium varium and one strain of Clostridium clostridioforme. Members of the B. fragilis groups were more susceptible than members of the other groups; the MIC50 of ANTIMICROB. AGENTS CHEMOTHER. was,ig/ml, and the MIC was,ug/ml. By contrast, only 42% of strains were susceptible to ciprofloxacin; the MIC50 was,g/ml, and the MIC. was,ug/ml. Only Porphyromonas asaccharolytica, Propionibactenum acnes, and C. perfringens strains were consistently susceptible to ciprofloxacin, with >% of strains inhibited at MICs of ',ug/ml. Although metronidazole was active against all gram-negative rods, 7% of peptostreptococci, 83% of grampositive non-spore-forming rods, and 4% of non-c. perfringens, non-clostridium difficile clostridia were resistant to this agent. was active against % of B. fragilis group strains and 96% of non-b. fragilis group Bacteroides, Prevotella, and Porphyromonas species, 91% of peptostreptococci, % of gram-positive non-spore-forming rods, and % of C. perfringens. By contrast, only 70% of fusobacteria (10% of F. varium), 53% of C. difficile, and 33% of miscellaneous clostridia were susceptible to this compound. was active against % of strains overall, but only 20% of the C. difficile strains were susceptible. Significant enhancement of piperacillin by tazobactam was seen in all,-lactamase-positive strains (99% susceptible to the combination; MICg,,ug/ml), and all P-lactamase-negative strains were susceptible to piperacillin (MICg, jig/ml).,3-lactamase-producing strains were all resistant to tazobactam alone (MIC, 2,ug/ml), whereas 55% of,-lactamasenegative, gram-negative and -positive strains were more susceptible to this compound (MICs, <,g/ml). MICs of the non-1-lactams for the 35,B-lactamase-negative, gram-negative rods did not differ from those seen for P-lactamase-positive strains. However, piperacillin MICs for enzyme-negative strains were lower than those seen for enzyme-positive strains (MICs, versus >6,ug/ml). With the exception of C. difficile and some strains of lactobacilli, cefoxitin MICs were also lower for 3-lactamasenegative strains (MICs, versus 3,ug/ml). When results for P-lactamase-positive Fusobacterium nucleatum and Fusobacterium necrophorum strains were analyzed separately, enhancement of piperacillin by tazobactam similar to that seen in other 3-lactamase-positive, gram-negative anaerobic rods was observed (MICs, 6,ug/ml without inhibitor and,ug/ml with inhibitor). DISCUSSION The existing fluoroquinolones, although exhibiting broadspectrum activity against aerobes, yield relatively high MICs for anaerobes (16, 17). Goldstein and Citron (16, 17) have found the MICs of ciprofloxacin and ofloxacin to be.,ug/ml for most anaerobic species, including the B. fragilis group; only the Bacteroides ureolyticus group, C. perfringens, and Propionibacterium species were susceptible (MIC.s,. jig/ml). In the current study, only P. asaccharolytica, P. acnes, and C. perffingens strains were susceptible to ciprofloxacin (MICs,.,ug/ml)., a new fluoroquinolone, has been found to exhibit potent activity against a broad spectrum of aerobic organisms, including methicillin-resistant Staphylococcus aureus, streptococci (including Enterococcus faecalis and Streptococcus pneumoniae), members of the family Enterobacteriaceae, and Pseudomonas aeruginosa (10, 20, 23, 28). In one preliminary study, Endermann and Bremm (13) have also shown that has potent activity against anaerobes in vitro, with all 32 strains tested (including 19 B. fragilis group, 8 Clostridium species, and 5 peptostreptococci) being inhibited by at <,g/ml. Endermann and Bremm (13) also reported the excellent activity of
VOL. 37, 1993 ANAEROBE SUSCEPTIBILITY TO 1651 TABLE 1. Antimicrobial susceptibilities of anaerobic strains Organism and MIC.5 MIC % Suscepantimicrobial agent (,g/ml) (Pg/ml) tibility Bacteroides fragilis (43/43b) >6 >6 Organism and MIC50 MICgo % Suscepantimicrobial agent (Ag/ml) (pg/ml) tibilitya Bacteroides ureolyticus (9/9) 2 91 84 72 93 33 89 89 Bacteroides thetaiotaomicron (31/31) Bacteroides ovatus (11/11) Piperacilin Bacteroides distasonis (10/8) Piperacilin-tazobactam Bacteroides vulgatus (11/9) Bacteroides uniformis (9/9) 3 6 > > 6 6 >6 6 >6 > > 3 6 16 16 >6 >6 13 94 Bacteroides capillosus (10/10) Porphyromonas asaccharolytica (10/8) 0 73 73 55 Piperacilin Piperacilin-tazobactam Prevotella bivia (35/34) 10 70 80 18 82 91 Prevotella disiens (11/9) Prevotella oris (Prevotella buccae) (10/5) 0 89 78 67 Piperacilin > 3 3 6 40 0 94 73 40 80 Bacteroides fragilis group (115/111) 6 >6 Prevotella melaninogenica (12/12) 7 85 77 99 Continued 3 25 Continued on following page
1652 PANKUCH ET AL. ANTIMICROB. AGENTS CHEMOTHER. Organism and MIC50 MIC % Suscepantimicrobial agent (pg/ml) (p1g/mi) tibilitya Prevotella intermedia (12/12) Miscellaneous strainsc (7/5) Chindamycin Non-Bacteroides fragilis Bacteroides, Prevotella, and Porphyromonas spp. (116/104) Fusobacterium nucleatum (11/6) Fusobacterium necrophorum (12/2) Fusobacterium mortiferum (7/6) Fusobacterinum varium (10/7) > 3 3 3 0.06 > 3 < 0.03 < 0.06 > > 3 33 57 34 96 98 97 Organism and MIC50 MIC % Suscepantimicrobial agent (pg/ml) (pg/ml) tibilitya All fusobacteria (40/21) Peptostreptococcid (58/0) Propionibacterium acnes (15/0) Other gram-positive anaerobic non-spore-forming rodse (33/0) 55 82 91 75 Clostridium perfringens (12/0) Clostridium difficile (15/0) 14 Other clostridiaf (24/0) 10 10 Continued 0.06 > 3 0.06 > > > > >6 > >6 >6 45 70 95 81 93 91 0 61 24 85 13 53 20 79 96 > 33 Continued on following page
VOL. 37, 1993 Organism and MIC50 MIC9( % Suscepantimicrobial agent (pg/ml) (ig/ml) tibility" All strains (428/236) 42 89 87 3 6 93 99 a See Materials and Methods for breakpoints. b Number of strains tested/number of strains,b-lactamase positive. c Three Prevotella loescheii, two Prevotella corpora, one Prevotella oralis, and one Porphyromonas gingivalis. d Fifteen P. anaerobius, 14 P. tetradius, 13 P. magnus, 11 P. asaccharolyticus, 2 P. micros, 1 P. prevotii, 1 P. productus, and 1 P. hydrogenalis. ' Eight Eubactenum lentum, one Eubacteriwn aerofaciens, four Lactobacillus casei, three Lactobacillus acidophilus, one Lactobacillus plantarum, three LactobaciUlus spp., one Bifidobacterium breve, four Bifidobacterium spp., three Actinomyces odontolyticus, two Actinomyces naeslundii, one Actinomyces israelii, one Actinomyces viscosus, and one Actinomyces meyeni ḟten C. tertium, five C innocuum, three C. clostridioforme, two C. ramosum, one C sporogenes, one C. butyricum, one C. paraputificum, and one C. cadaveris. against B. fragilis in a groin infection animal model, with a dose of mg/kg of body weight resulting in a 7-log-unit decrease in the mean bacterial count. The MICs in the current study were a little higher than those found in the German study (13); this can be explained by the lower inoculum size (104 bacteria per ml) used by the latter workers (13). Both studies, however, showed that has superior activity against B. fragilis compared with that against other anaerobes, with MICs for most organisms being < p,g/ml; the in vitro activity of was superior to those of ciprofloxacin, clindamycin, cefoxitin, and metronidazole (13). In the current study, for only five strains were MICs 2 p,g/ml. Of these, four were F. vanium. Other reports have documented the relative resistance of fusobacteria to quinolones compared with the resistances of other groups of anaerobes (6, 9, 17). The uniform susceptibility of the three imipenem-resistant strains from Japan to in the current study mirrors the findings of Ueno (29) that imipenem-resistant strains from Japan are susceptible to but are resistant to the other new quinolones tested. The results of the present study also confirm the excellent activity of piperacillin against 3-lactamase-negative anaerobes and of piperacillin-tazobactam against P-lactamasepositive anaerobes (2, 6, 7). Lower f-lactam MICs for P-lactamase-negative anaerobic strains and the lack of enhancement of 1-lactams by P-lactamase inhibitors in 1- lactamase-positive F. varium strains have been described before (5-7). As expected, all gram-negative strains tested were susceptible to metronidazole (5, 12, 19). The resistance of anaerobic, gram-positive non-spore-forming rods to metronidazole is expected, and the resistance of peptostreptococci to metronidazole has been described previously (7, 11, 12, 24). However, the relatively high metronidazole MICs for non-c. perfringens Clostridium species were unexpected. Chow and coworkers (11) have found that for 4 of 13 (31%) non-c. perfringens clostridia tested, metronidazole MICs were >25.0,ug/ml, compared with % susceptibility at MICs of <1.6,g/ml for C. perfringens (11). Rolfe and Finegold (24) ANAEROBE SUSCEPTIBILITY TO 1653 have reported that two of four strains of C. clostnidioforme tested resistant to metronidazole (MICs, >256.0,ug/ml). Additionally, clinical failures with metronidazole therapy have been reported in patients with systemic Clostridium sordellii infections (26). More studies are necessary in order to determine whether metronidazole resistance in non-c. perfringens clostridia is more common than previously suspected. Because of the relatively high number of cefoxitin-resistant members of the B. fragilis group included in the current study, MICs were higher than those reported by other workers (5, 19). Lack of cefoxitin activity against C. difficile has been described before (24). Some members of all groups of anaerobes, especially F. vanium and clostridia, were resistant to clindamycin. F. varium is inherently resistant to a range of 1-lactam and non-13-lactam agents; George and coworkers (15) have reported that none of five F. varium strains tested susceptible to clindamycin. However, with 87% of strains being susceptible in the current study, clindamycin remains active against most clinically significant anaerobes. It is recognized that the results of in vitro anaerobe susceptibility testing do not always correlate with in vivo findings. Reasons for the latter phenomenon may include methodological problems in susceptibility testing (e.g., medium, inoculum, susceptibility breakpoint), the type and spectrum of anaerobes tested, and pharmacokinetic factors. Most clinicians feel that penicillin is the preferred drug for therapy of fusobacterial infections (14), despite the propensity of fusobacteria to produce 1-lactamase (5-7). The reason for the latter discrepancy may be the relatively low 1-lactam MICs for 3-lactamase-producing fusobacteria (5-7) compared with those for 3-lactamase-producing species such as members of the B. fragilis group (19), together with the high therapeutic doses of 3-lactam used (14). is almost always active against fusobacteria in vivo (14), despite the relatively poor in vitro results obtained in the current study. It is, however, recognized that human infections with F. vanium, the most clindamycin-resistant Fusobacterium species tested in our study, are rather rare (15). Data on the efficacy of cefoxitin in the treatment of anaerobic infections correlate with results from unselected in vitro surveys (unlike the current survey, in which selected 1-lactamase-producing and cefoxitin-resistant strains were tested), in which >95% of strains were susceptible at a breakpoint of 3,ug/ml (5, 14, 19). Although 42% of all strains were susceptible to ciprofloxacin in the current study, clustering around the breakpoint occurred, such that ciprofloxacin is not recommended for use in the treatment of anaerobic infections. Indeed, anaerobic flora in human volunteers have been found to be unchanged after oral ciprofloxacin administration (25). In summary, showed excellent activity against all groups of anaerobes tested and compared favorably to the existing compounds used for the treatment of anaerobic infections. In particular, excellent activity against the B. fragilis group was found, in contrast to the greater resistance of this group to most antimicrobial agents. Clinical studies are required to confirm these in vitro findings. ACKNOWLEDGMENTS This study was supported by a grant from Miles, Inc. We thank P. Schreckenberger (University of Illinois Health Sciences Center, Chicago, Ill.), D. Citron (St. John and Santa Monica Hospitals, Santa Monica, Calif.), W. Brown (Hutzel Hospital, Detroit, Mich.), and J. Rosenblatt (May Clinic, Rochester,
1654 PANKUCH ET AL. Minn.) for provision of some cultures used in the present study and D. Citron for helpful advice. REFERENCES 1. Appelbaum, P. C. 1987. Anaerobic infections: nonsporeformers, p. 45-109. In B. B. Wentworth (coordinating ed.), Diagnostic procedures for bacterial infections. American Public Health Association, Washington, D.C. 2. Appelbaum, P. C., M. R. Jacobs, S. K. Spangler, and S. Yamabe. 1986. Comparative activity of P-lactamase inhibitors YTR 830, clavulanate, and sulbactam combined with,-lactams against 3-lactamase-producing anaerobes. Antimicrob. Agents Chemother. 30:789-791. 3. Appelbaum, P. C., A. Philippon, M. R. Jacobs, S. K. Spangler, and L. Gutmann. 19. Characterization of P-lactamases from non-bacteroides fragilis group Bacteroides spp. belonging to seven species and their role in 1-lactam resistance. Antimicrob. Agents Chemother. 34:2169-2176. 4. Appelbaum, P. C., S. K. Spangler, and M. R. Jacobs. 19. Evaluation of two methods for rapid testing for beta-lactamase production in Bacteroides and Fusobactenum. Eur. J. Clin. Microbiol. Infect. Dis. 9:47-50. 5. Appelbaum, P. C., S. K. Spangler, and M. R. Jacobs. 19. 13-Lactamase production and susceptibilities to amoxicillin, amoxicillin-clavulanate, ticarcillin, ticarcillin-clavulanate, cefoxitin, imipenem, and metronidazole of 320 non-bacteroides fragilis Bacteroides and 129 fusobacteria from 28 U.S. centers. Antimicrob. Agents Chemother. 34:1546-1550. 6. Appelbaum, P. C., S. K. Spangler, and M. R. Jacobs. 1991. Susceptibilities of 394 Bacteroidesfragilis, non-b. fragilis group Bacteroides species, and Fusobacterium species to newer antimicrobial agents. Antimicrob. Agents Chemother. 35:1214-1218. 7. Appelbaum, P. C., S. K. Spangler, and M. R. Jacobs. Susceptibility of 539 gram-positive and -negative anaerobes to new agents, including RP 59500, biapenem, trospectomycin and piperacillin/tazobactam. J. Antimicrob. Chemother., in press. 8. Appelbaum, P. C., S. K. Spangler, G. A. Pankuch, A. Philippon, M. R. Jacobs, R. Shiman, E. J. C. Goldstein, and D. Citron. Submitted for publication. 9. Appelbaum, P. C., S. K. Spangler, R. Shiman, and M. R. Jacobs. 19. Susceptibilities of 540 anaerobic gram-negative bacilli to amoxicillin, amoxicillin-brl 42715, amoxicillin-clavulanate, temafloxacin, and clindamycin. Antimicrob. Agents Chemother. 36:1140-1143. 9a.Bandoh, K., K. Ueno, K. Watanabe, and N. Kato. 1993. Susceptibility patterns and resistance to imipenem in the Bacteroides fragilis group species in Japan: 4-year study. Clin. Infect. Dis. 16(Suppl. 4):S382-S386. 10. Bremm, K.-D., R. Endermann, and K. D. Metzger. 19. Program Abstr. 32nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. 649. 11. Chow, A. W., V. Patten, and L. B. Guze. 1975. Susceptibility of anaerobic bacteria to metronidazole: relative resistance of nonspore-forming gram-positive bacilli. J. Infect. Dis. 131:182-185. 12. Edson, R. S., J. E. Rosenblatt, D. T. Lee, and E. A. McVey. ANTIMICROB. AGENTS CHEMOTHER. 1982. Recent experience with antimicrobial susceptibility of anaerobic bacteria. Increasing resistance to penicillin. Mayo Clin. Proc. 57:737-741. 13. Endermann, R., and K.-D. Bremm. 19. Program Abstr. 32nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. 645. 14. Finegold, S. M., and W. L. George (ed.). 1989. Anaerobic infections in humans. Academic Press, Inc., New York. 15. George, W. L., B. D. Kirby, V. L. Sutter, D. M. Citron, and S. M. Finegold. 1981. Gram-negative anaerobic bacilli: their role in infection and patterns of susceptibility to antimicrobial agents. II. Little-known Fusobacterium species and miscellaneous genera. Rev. Infect. Dis. 3:599-626. 16. Goldstein, E. J. C., and D. Citron. 1985. Comparative activity of the quinolones against anaerobic bacteria isolated at community hospitals. Antimicrob. Agents Chemother. 27:657-659. 17. Goldstein, E. J. C., and D. Citron. 19. Comparative activity of ciprofloxacin, ofloxacin, sparfloxacin, temafloxacin, CI-960, CI- 9, and Win 57273 against anaerobic bacteria. Antimicrob. Agents Chemother. 36:1158-1162. 18. Holdeman, L. V., and W. E. C. Moore (ed.). 1977. Anaerobic laboratory manual, 4th ed. Virginia Polytechnic Institute and State University, Blacksburg. 19. Jacobs, M. R., S. K. Spangler, and P. C. Appelbaum. 19. 3-Lactamase production, j3-lactam sensitivity and resistance to synergy with clavulanate of 737 Bacteroides fragilis group organisms from thirty-three US centres. J. Antimicrob. Chemother. 26:361-370. 20. Krasemann, C. 19. Program Abstr. 32nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. 644. 21. National Committee for Clinical Laboratory Standards. 19. Methods for antimicrobial susceptibility testing of anaerobic bacteria, 2nd ed. Approved standard. NCCLS publication no. M11-A2. National Committee for Clinical Laboratory Standards, Villanova, Pa. 22. Nord, C. E. 1986. Mechanisms of P-lactam resistance in anaerobic bacteria. Rev. Infect. Dis. 8(Suppl. 5):S543-S548. 23. Petersen, U., K.-D. Bremm, A. Krebs, K. G. Metzger, T. Philipps, and T. Schenke. 19. Program Abstr. 32nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. 642. 24. Rolfe, R. D., and S. M. Finegold. 1982. Comparative in vitro activity of ceftriaxone against anaerobic bacteria. Antimicrob. Agents Chemother. 22:338-341. 25. Shah, P. M., R. Enzensberger, 0. Glogau, and H. Knothe. 1987. Influence of oral ciprofloxacin or ofloxacin on the fecal flora of healthy volunteers. Am. J. Med. 82(Suppl. 4A):333-335. 26. Spera, R. V., Jr., M. H. Kaplan, and S. L. Allen. 19. Clostridium sordellii bacteremia: case report and review. Clin. Infect. Dis. 15:950-954. 27. Summanen, P., E. J. Baron, D. M. Citron, C. A. Strong, H. M. Wexler, and S. M. Finegold. 1993. Wadsworth anaerobic bacteriology manual, 5th ed. Star Publishing Co., Belmont, Calif. 28. Thurberg, B. E., B. G. Painter, J. A. Herrington, J. A. Federici, J. M. Remy, and M. L. Barbiero. 19. Program Abstr. 32nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. 643. 29. Ueno, K. (Gifu University). 19. Personal communication.