In Vitro Activity of Cefotetan against ESBL-Producing Escherichia coli and Klebsiella pneumoniae Bloodstream Isolates from the MERINO Trial.

Extended-spectrum-beta-lactamase (ESBL)-producing Enterobacterales continue to pose a major threat to human health worldwide. Given the limited therapeutic options available to treat infections caused by these pathogens, identifying additional effective antimicrobials or revisiting existing drugs is important. Ceftriaxone-resistant Escherichia coli and Klebsiella pneumoniae containing CTX-M-type ESBLs or AmpC, in addition to narrow-spectrum OXA and SHV enzymes, were selected from blood culture isolates obtained from the MERINO trial. Isolates had previously undergone whole-genome sequencing (WGS) to identify antimicrobial resistance genes. Cefotetan MICs were determined by broth microdilution (BMD) testing with a concentration range of 0.125 to 64 mg/liter; CLSI breakpoints were used for susceptibility interpretation. BMD was performed using an automated digital antibiotic dispensing platform (Tecan D300e). One hundred ten E. coli and 40 K. pneumoniae isolates were used. CTX-M-15 and CTX-M-27 were the most common beta-lactamases present; only 7 isolates had coexistent ampC genes. Overall, 98.7% of isolates were susceptible, with MIC50s and MIC90s of 0.25 mg/liter and 2 mg/liter (range, ≤0.125 to 64 mg/liter), respectively. MICs appeared higher among isolates with ampC genes present, with an MIC50 of 16 mg/liter, than among those containing CTX-M-15, which had an MIC50 of only 0.5 mg/liter. Isolates with an ampC gene exhibited an overall susceptibility of 85%. Presence of a narrow-spectrum OXA beta-lactamase did not appear to alter the cefotetan MIC distribution. Cefotetan demonstrated favorable in vitro efficacy against ESBL-producing E. coli and K. pneumoniae bloodstream isolates. IMPORTANCE Carbapenem antibiotics remain the treatment of choice for severe infection due to ESBL- and AmpC-producing Enterobacterales. The use of carbapenems is a major driver of the emergence of carbapenem-resistant Gram-negative bacilli, which are often resistant to most available antimicrobials. Cefotetan is a cephamycin antibiotic developed in the 1980s that demonstrates enhanced resistance to beta-lactamases and has a broad spectrum of activity against Gram-negative bacteria. Cefotetan holds potential to be a carbapenem-sparing treatment option. Data on the in vitro activity of cefotetan against ESBL-producing Enterobacterales remain scarce. Our study assessed the in vitro activity of cefotetan against ceftriaxone-nonsusceptible blood culture isolates obtained from patients enrolled in the MERINO trial.

[THE APPLICATION OF SELECTIVE CHROMOGENIC AGAR FOR DETECTING ENTEROBACTERIA WITH PRODUCTION OF BETA-LACTAMASES].

The detection of enterobacteria with production of beta-lactamases of extended spectrum in selective chromogenic agar was analyzed The results ofdetection of beta-lactamases of extended spectrum was compared with "double disc" technique. The smears from mucous membrane of guttur and rectum from patients were analyzed in parallel on solid growth agar (Endo or Mac Conkey) and on selective agar CHROMagartm ESBL (CHROMagar France). The production of beta-lactamases of extended spectrum was confirmed using "double discs" technique. To exclude hyper-production of ampC beta-lactamases E-test was applied containing cefotetan and cefotetan with cloxacillin. The sampling consisted of 1552 samples from patients. The study permitted to isolate 1243 strains of enterobacteria on agar Endo or Mac Conkey and 409 strains of enterobacteria on selective agar CHROMagartm ESBL (Escherichia coli n = 226, Klebsiella pneumoniae n = 105, enterobacter spp. n = 35, Citrobacter spp. n = 21, others n = 22). The application of "double discs" technique confirmed production of beta-lactamases of extended spectrum in 386 (94%) out of 409 strains isolated on agar CHROMagartm ESBL. In 23 (6%) of strains no confirmation was established and hyper-production of ampC of beta-lactamases was established 15 out of total. Additionally, 8 were sensitive to cephalosporin of third generation. All enterobacteria isolated on agar Endo or Mac Conkey also were tested by "double discs" technique. Overall, 394 strains of enterobacteria with production of beta-lactamases of extended spectrum were obtained. On all agars (agar Endo or Mac Conkey and CHROMagartm ESBL)--263 (67%) strains; only on CHROMagartm ESBL--123 (31%) and only on agar Endo or Mac Conkey--8 (2%) (p < 0.0001). The sensitivity of selective agar CHROMagartm ESBL made up to 98% and specificity--97%. The resolution about detection of enterobacteria producing beta-lactamases of extended spectrum were submitted to clinic in 18-24 hours after arrival ofsamplesfrom patients in laboratory. The CHR OMagartm ESBL has higher sensitivity and specificity to detect enterobacteria with production of beta-lactamases of extended spectrum and can be applied in common laboratory practice.

Detection and occurrence of plasmid-mediated AmpC in highly resistant gram-negative rods.

OBJECTIVES: The aim of this study was to compare the current screening methods and to evaluate confirmation tests for phenotypic plasmidal AmpC (pAmpC) detection. METHODS: For this evaluation we used 503 Enterobacteriaceae from 18 Dutch hospitals and 21 isolates previously confirmed to be pAmpC positive. All isolates were divided into three groups: isolates with 1) reduced susceptibility to ceftazidime and/or cefotaxime; 2) reduced susceptibility to cefoxitin; 3) reduced susceptibility to ceftazidime and/or cefotaxime combined with reduced susceptibility to cefoxitin. Two disk-based tests, with cloxacillin or boronic acid as inhibitor, and Etest with cefotetan-cefotetan/cloxacillin were used for phenotypic AmpC confirmation. Finally, presence of pAmpC genes was tested by multiplex and singleplex PCR. RESULTS: We identified 13 pAmpC producing Enterobacteriaceae isolates among the 503 isolates (2.6%): 9 CMY-2, 3 DHA-1 and 1 ACC-1 type in E. coli isolates. The sensitivity and specificity of reduced susceptibility to ceftazidime and/or cefotaxime in combination with cefoxitin was 97% (33/34) and 90% (289/322) respectively. The disk-based test with cloxacillin showed the best performance as phenotypic confirmation method for AmpC production. CONCLUSIONS: For routine phenotypic detection of pAmpC the screening for reduced susceptibility to third generation cephalosporins combined with reduced susceptibility to cefoxitin is recommended. Confirmation via a combination disk diffusion test using cloxacillin is the best phenotypic option. The prevalence found is worrisome, since, due to their plasmidal location, pAmpC genes may spread further and increase in prevalence.

Comparison of 3 phenotypic-detection methods for identifying plasmid-mediated AmpC beta-lactamase-producing Escherichia coli, Klebsiella pneumoniae, and Proteus mirabilis strains.

BACKGROUND: Plasmid-mediated AmpC beta-lactamases (PABLs) have been detected in the strains of Escherichia coli, Klebsiella spp., Proteus mirabilis, and Salmonella spp. PABLs may be difficult to detect and might interfere in the therapeutic and infection-control processes. Although several PABL-detection methods based on phenotypes have been reported, the Clinical and Laboratory Standards Institute currently does not recommend a routine detection method for PABLs. The aim of this study is to compare the performances of 3 phenotypic PABL detection methods. METHODS: Total 276 non-duplicated clinical isolates of E. coli (N=97), K. pneumoniae (N=136), and P. mirabilis (N=43) were collected from 14 hospitals in Korea between April and June 2007 in a non-consecutive and non-random manner. Multiplex PCR was performed to detect the PABL genes. Further, 3 phenotypic detection methods-cephamycin-Hodge test, Tris-EDTA (TE) disk test, and combination-disk test with 3-aminophenylboronic acid (BA)-were performed using cefoxitin and cefotetan disks. RESULTS: PABL genes were detected by multiplex PCR in 122/276 isolates, including 14/97 E. coli, 105/136 K. pneumoniae, and 3/43 P. mirabilis isolates. The combination-disk test with BA showed higher sensitivity (98.4%), specificity (92.2%), and efficiency (96.3%) than the cephamycin-Hodge (76.2%, 96.1%, and 88.6%, respectively) and the TE-disk (80.3%, 91.6%, and 87.9%, respectively) tests. CONCLUSIONS: The combination-disk test with BA is a simple, efficient, and interpretable test that can be applicable in clinical laboratories involved in the detection of PABLs in clinical isolates of E. coli, K. pneumoniae, and P. mirabilis.

[Evaluation of the MicroScan NegCombo panel Type 44 for detection of extended-spectrum beta-lactamase among clinical isolates of Escherichia coli, Klebsiella species, and Proteus mirabilis].

BACKGROUND: Accurate and rapid detection of extended-spectrum beta-lactamases (ESBLs) is important in guiding proper antimicrobial therapy for infected patients. We evaluated the performance of MicroScan NegCombo Type 44 panel (Dade Behring, USA), which was developed to confirm ESBL-producing Enterobacteriaceae using ceftazidime/clavulanate and cefotaxime/clavulanate. METHODS: From August 30 to September 20, 2007, 206 non-duplicate clinical isolates, including 106 Escherichia coli, 81 Klebsiella pneumoniae, 11 Klebsiella oxytoca, and 8 Proteus mirabilis were subcultured and tested with Type 32 and Type 44 panels. The results were compared with those of the CLSI phenotypic confirmatory test (CLSI-PCT) and disk approximation test (DAT). Isolates not susceptible to cefotetan or flagged as "Possible ESBL, unable to interpret confirm test (Possible ESBL)" on Type 44 panel were tested with boronic acid disks to confirm AmpC beta-lactamases (AmpC) production. RESULTS: Of the 206 isolates tested, 44 (21.4%) produced ESBL by CLSI-PCT or DAT, including 27 E. coli, 14 K. pneumoniae, 2 K. oxytoca, and 1 P. mirabilis. Thirty-eight isolates flagged as "Confirmed ESBL" on Type 44 panel were all confirmed as ESBL-producers. Of 14 K. pneumoniae flagged as "Possible ESBL", 6 were confirmed as ESBL and AmpC co-producers and 8 as AmpC-producers. CONCLUSIONS: Type 44 panel showed an excellent performance in detecting ESBL-producing E. coli, Klebsiella spp., and P. mirabilis. When flagged as "Confirmed ESBL", no other confirmatory test was necessary to report as ESBL; however, "Possible ESBL" required a differential test for AmpC production.

Infection after elective colorectal surgery: bacteriological analysis of failures in a randomized trial of cefotetan vs. ertapenem prophylaxis.

BACKGROUND: A randomized study comparing single-dose cefotetan and ertapenem prophylaxis for elective colorectal surgery in 1,002 patients found ertapenem to be significantly more effective (p < 0.001). Failures of prophylaxis were thought to involve organisms resistant to both antimicrobial agents, isolated most often from deep or superficial incision sites. METHODS: Further testing and analysis of the microbial data was performed. Susceptibility results were correlated with the clinical outcomes reported previously. RESULTS: Of the 216 aerobes tested, 62.6% were resistant to cefotetan and 44% to ertapenem. Enterococci and methicillin-resistant Staphylococcus epidermidis were the aerobes recovered most frequently, and Bacteroides thetaiotaomicron, Clostridium innocuum, and Eubacterium lentum were the most frequent anaerobes. Enterococcus faecalis usually was associated in mixed culture with Bacteroides fragilis group species. Approximately one-half of the 158 anaerobes (50.7%), including all the species above, were resistant to cefotetan; most of these (61.4%) came from superficial incision sites. Only one anaerobe (Desulfovibrio fairfieldensis), found in a superficial incisional infection, was resistant to ertapenem, and no ertapenem-resistant enteric bacteria were recovered. In vitro resistance was associated with therapeutic failure. CONCLUSIONS: The in vitro activity of ertapenem was superior to that of cefotetan against all anaerobic and many aerobic bacteria isolated from postoperative cultures of patients who failed prophylaxis with these agents. Our findings help to elucidate the results of the clinical trial.

Evaluation of the MicroScan NegCombo Panel Type 44 for Detection of Extended-Spectrum beta-Lactamase among Clinical Isolates of Escherichia coli, Klebsiella species, and Proteus mirabilis / 대한진단검사의학회지

BACKGROUND: Accurate and rapid detection of extended-spectrum beta-lactamases (ESBLs) is important in guiding proper antimicrobial therapy for infected patients. We evaluated the performance of MicroScan NegCombo Type 44 panel (Dade Behring, USA), which was developed to confirm ESBL-producing Enterobacteriaceae using ceftazidime/clavulanate and cefotaxime/clavulanate. METHODS: From August 30 to September 20, 2007, 206 non-duplicate clinical isolates, including 106 Escherichia coli, 81 Klebsiella pneumoniae, 11 Klebsiella oxytoca, and 8 Proteus mirabilis were subcultured and tested with Type 32 and Type 44 panels. The results were compared with those of the CLSI phenotypic confirmatory test (CLSI-PCT) and disk approximation test (DAT). Isolates not susceptible to cefotetan or flagged as "Possible ESBL, unable to interpret confirm test (Possible ESBL)" on Type 44 panel were tested with boronic acid disks to confirm AmpC beta-lactamases (AmpC) production. RESULTS: Of the 206 isolates tested, 44 (21.4%) produced ESBL by CLSI-PCT or DAT, including 27 E. coli, 14 K. pneumoniae, 2 K. oxytoca, and 1 P. mirabilis. Thirty-eight isolates flagged as "Confirmed ESBL" on Type 44 panel were all confirmed as ESBL-producers. Of 14 K. pneumoniae flagged as "Possible ESBL", 6 were confirmed as ESBL and AmpC co-producers and 8 as AmpC-producers. CONCLUSIONS: Type 44 panel showed an excellent performance in detecting ESBL-producing E. coli, Klebsiella spp., and P. mirabilis. When flagged as "Confirmed ESBL", no other confirmatory test was necessary to report as ESBL; however, "Possible ESBL" required a differential test for AmpC production.

Comparison of 3 Phenotypic-detection Methods for Identifying Plasmid-mediated AmpC beta-lactamase-producing Escherichia coli, Klebsiella pneumoniae, and Proteus mirabilis Strains / 대한진단검사의학회지

BACKGROUND: Plasmid-mediated AmpC beta-lactamases (PABLs) have been detected in the strains of Escherichia coli, Klebsiella spp., Proteus mirabilis, and Salmonella spp. PABLs may be difficult to detect and might interfere in the therapeutic and infection-control processes. Although several PABL-detection methods based on phenotypes have been reported, the Clinical and Laboratory Standards Institute currently does not recommend a routine detection method for PABLs. The aim of this study is to compare the performances of 3 phenotypic PABL detection methods. METHODS: Total 276 non-duplicated clinical isolates of E. coli (N=97), K. pneumoniae (N=136), and P. mirabilis (N=43) were collected from 14 hospitals in Korea between April and June 2007 in a non-consecutive and non-random manner. Multiplex PCR was performed to detect the PABL genes. Further, 3 phenotypic detection methods-cephamycin-Hodge test, Tris-EDTA (TE) disk test, and combination-disk test with 3-aminophenylboronic acid (BA)-were performed using cefoxitin and cefotetan disks. RESULTS: PABL genes were detected by multiplex PCR in 122/276 isolates, including 14/97 E. coli, 105/136 K. pneumoniae, and 3/43 P. mirabilis isolates. The combination-disk test with BA showed higher sensitivity (98.4%), specificity (92.2%), and efficiency (96.3%) than the cephamycin-Hodge (76.2%, 96.1%, and 88.6%, respectively) and the TE-disk (80.3%, 91.6%, and 87.9%, respectively) tests. CONCLUSIONS: The combination-disk test with BA is a simple, efficient, and interpretable test that can be applicable in clinical laboratories involved in the detection of PABLs in clinical isolates of E. coli, K. pneumoniae, and P. mirabilis.

False susceptibility to cefotetan reported by MicroScan for DHA-type AmpC beta-lactamase-producing Klebsiella pneumoniae.

This study evaluated the accuracy of cefotetan susceptibility determination using the MicroScan WalkAway system for AmpC-producing Klebsiella pneumoniae. In total, 57 K. pneumoniae isolates that showed a D-shape flattening in a double-disk synergy test were studied. Cefotetan MICs were determined by the agar dilution method. The bla(DHA) gene was detected in all 57 isolates, one of which co-harboured bla(CMY-1). According to the MicroScan system, 28 isolates were susceptible, 18 were intermediately-resistant, and 11 were resistant to cefotetan. Compared with the agar dilution method, very major, minor and major error rates were 28.1% (16/57), 47.4% (27/57) and 1.8% (1/57), respectively.

Evaluation of beta-lactamase inhibitors in disk tests for detection of plasmid-mediated AmpC beta-lactamases in well-characterized clinical strains of Klebsiella spp.

The diagnostic utility of the AmpC beta-lactamase inhibitors LN-2-128, 48-1220, and Syn 2190 in combination with cefotetan (CTT) or cefoxitin in a disk test for the detection of clinical isolates of Klebsiella spp. producing plasmid-mediated AmpC beta-lactamases (pAmpCs) was evaluated. The combination of Syn 2190 and CTT had a sensitivity of 91%, a specificity of 100%, and a reproducibility of 100% and showed the best potential of using an inhibitor for detection of Klebsiella spp. producing pAmpCs.

Bacteroides fragilis-induced intra-abdominal abscess in an experimental model treated with telithromycin (HMR 3647).

BACKGROUND: Telithromycin (HMR 3647), a novel ketolide, is known to have activity against Bacteroides fragilis in vitro. METHODS: We tested this ketolide in an animal model of intra-abdominal abscess produced by intraperitoneal injection of B. fragilis with sterile feces and BaSO(4) mixture. Telithromycin was tested at two doses, 1. 25 and 2.0 mg/dose twice daily, and compared with clindamycin, cefotetan or metronidazole, all given at 2.0 mg/dose twice daily for 10 days. Absence of bacteria at the infected site was considered a cure and a positive culture considered a failure. RESULTS: The cure rate was 18% (5/28) on saline therapy, 74% (20/27) on telithromycin and 82% (23/28) on clindamycin, whereas it was 61% (17/28) on metronidazole and 59% (16/27) on cefotetan therapy. A high tissue antibiotic concentration (3-5 times the MIC) of telithromycin was found and this is presumably related to its superior efficacy. Delayed therapy initiated 7 days after infection instead of immediate therapy cured only 32% of the animals treated. The lower dose of telithromycin (1.25 mg/dose twice daily) was as effective as the higher dose (2 mg/dose twice daily). CONCLUSIONS: We found that telithromycin is as effective as clindamycin and more effective than metronidazole and cefotetan in this experimental model. These results suggest that telithromycin may be tested in future for the treatment of B. fragilis infections in humans.

Interaction of cefotetan and the metallo-beta-lactamases produced in Aeromonas spp. and in vitro activity.

Aeromonas spp. are increasingly being recognized as human pathogens. The presence of metallo-beta-lactamases in these organisms represents a potential problem in antimicrobial therapy. Mechanism-based inactivators of beta-lactamases are used to overcome the resistance of clinical pathogens to beta-lactam antibiotics, but no clinical useful inhibitors of the metallo-beta-lactamases are presently known. Studying the interaction between cefotetan and Aeromonas spp. producing metallo-beta-lactamase activity, we observed that cefotetan behaved as a transient inactivator for both the crude extracts of Aeromonas strains and the purified enzymes from Aeromonas hydrophila AE036 and Aeromonas schubertii MNSA20. The direct hydrolysis of cefotetan showed that it was a poor substrate for both purified enzymes. In view of the minimum inhibitory concentrations, cefotetan shows to be a useful antimicrobial agent against Aeromonas spp.

Longitudinal study of susceptibilities of species of the Bacteroides fragilis group to five antimicrobial agents in three medical centers.

A total of 579 clinical isolates of the Bacteroides fragilis group collected from three Canadian hospitals were tested for susceptibility to five antimicrobial agents by using an agar dilution method. During the 4-year survey, isolates from intra-abdominal infections were collected from the following sites: abdominal abscesses (48%), peritoneal fluid (39%), blood (10%), and bile (3%). B. fragilis was the most prevalent species (35.4%), followed by B. thetaiotaomicron (19.2%), B. ovatus (15.9%), and B. vulgatus (11%). No metronidazole- or imipenem-resistant strains were found during the survey. Resistance profiles varied among the different species tested: 7.8, 2.9, and 7.3% of B. fragilis strains (n = 205) and 68.1, 17.2, and 9.4% of non-B. fragilis strains (n = 373) were resistant to cefotetan, cefoxitin, and clindamycin, respectively. B. fragilis and B. vulgatus demonstrated lower resistance rates than B. thetaiotaomicron, B. ovatus, B. distasonis, and B. caccae. During the study, rates of resistance to cefotetan and clindamycin fluctuated but rates of resistance to cefoxitin increased, particularly at one center. These data indicate a need to determine the susceptibility patterns of the B. fragilis group periodically at each hospital.

Evaluation of activity of temafloxacin against Bacteroides fragilis by an in vitro pharmacodynamic system.

An in vitro pharmacodynamic system has been successfully adapted to simulate in vivo antimicrobial pharmacokinetics under anaerobic conditions. This system was used to perform time-kill kinetic studies which were designed to compare the activity of temafloxacin to ciprofloxacin and cefotetan against two strains of Bacteroides fragilis (ATCC 25285 and ATCC 23745). All experiments were performed as single-dose, 24-h, duplicate runs. Starting bacterial inocula of 10(7) CFU/ml were exposed to starting antimicrobial concentrations of 5 micrograms of temafloxacin per ml, 5 micrograms of ciprofloxacin per ml, and 100 micrograms of cefotetan per ml. Terminal half-lives of 8, 4, and 4 h were simulated for each antimicrobial agent. Temafloxacin was rapidly bactericidal against B. fragilis. Ciprofloxacin was not bactericidal (< 3 log10 unit decline in bacterial numbers) to either strain of B. fragilis. Cefotetan was bactericidal (> or = 3 log10 unit decline in bacterial numbers) to each strain but killed at a slower rate than temafloxacin. Times to 3 log10 unit declines of strain ATCC 25285 were 2, 4, and > 24 h, whereas those of strain ATCC 23745 were 4, 4, and > 24 h for temafloxacin, cefotetan, and ciprofloxacin, respectively. Total logarithmic declines of strain ATCC 25285 were > 4.5, > 4.5, and 2.9 log10 CFU/ml, whereas those of strain ATCC 23745 were 4.1, > 4.5, and 1.2 log10 CFU/ml for each drug, respectively. These and other studies demonstrated that temafloxacin showed potential as an agent that could have been further developed for use in the treatment of anaerobic infections. However, the drug was removed from the market by its manufacturer because of toxicity issues. Although the release of newer fluoroquinolones that possess significant activity against anaerobic bacteria does not appear imminent, the time-kill studies performed in this study demonstrate that further research is warranted in the development of fluoroquinolones which possess significant antianaerobic activity.

In vitro antibacterial activity of cefoxitin and cefotetan and pharmacokinetics in dogs.

The susceptibility of 50 clinical Escherichia coli isolates to various antibacterials, including cefoxitin and cefotetan was ascertained, and the minimal inhibitory concentration (MIC) of cefoxitin and cefotetan for each of these isolates was determined. The pharmacokinetics of cefoxitin and cefotetan after a single i.v. or SC injection (30 mg/kg of body weight) were determined in 4 dogs. Of the 50 E coli isolates, 98% were susceptible in vitro to cefotetan, 90% were susceptible to cefoxitin, and 88% were susceptible to gentamicin. The MIC that would inhibit the growth of 90% of the E coli isolates (MIC90) was 0.25 microgram/ml for cefotetan and 4 micrograms/ml for cefoxitin. Plasma cefotetan concentrations remained above MIC90 for (mean +/- SD) 8.2 +/- 1.72 hours and 13.52 +/- 0.28 hours after i.v. and SC administration, respectively. Plasma cefoxitin concentrations remained above MIC90 for (mean +/- SD) 5.37 +/- 1.18 hours and 7.95 +/- 0.71 hours after i.v. and SC administration, respectively. We concluded that cefotetan was superior to cefoxitin in activity against E coli in vitro. We recommend that cefotetan be given at a dosage of 30 mg/kg, i.v., every 8 hours, or SC, every 12 hours.

Use of cephalosporins for prophylaxis and therapy of polymicrobial infection in mice.

Cefoxitin, cefotetan, and cefmetazole were compared in 10-day therapy of intra-abdominal and subcutaneous infections caused by three organisms: Bacteroides fragilis and Bacteroides thetaiotaomicron combined with either Escherichia coli or Staphylococcus aureus. Intra-abdominal infection was caused by B. fragilis plus B. thetaiotaomicron plus E. coli. Therapy was initiated immediately before inoculation or was delayed for 8 h. Mortality was 14 of 30 (47%) for saline-treated mice, and all survivors developed abscesses. Immediate therapy reduced mortality and the percentage of mice with abscesses (in survivors), respectively, to 17 and 20% with cefoxitin, 0 and 13% with cefotetan, and 0 and 17% with cefmetazole, and the numbers of all bacteria were reduced by all the cephalosporins. Delayed therapy reduced mortality and abscess formation, respectively, to 20 and 8% of mice with cefoxitin, 10 and 93% with cefotetan, and 7 and 96% with cefmetazole. B. thetaiotaomicron survived in all abscesses treated with cefotetan and cefmetazole. Subcutaneous abscesses were caused by each organism alone or in combinations of one aerobe (S. aureus or E. coli) and one or two Bacteroides species. Early therapy reduced the numbers of all bacteria independent of their in vitro susceptibility. All agents reduced the number of each Bacteroides species with either E. coli or S. aureus. However, when therapy was delayed, cefotetan and cefmetazole were less effective than cefoxitin against B. thetaiotaomicron. Cefotetan was the most active agent against E. coli, and cefmetazole was the most effective against S. aureus. These data illustrate the efficacy of all tested cephalosporins in the prophylaxis of polymicrobial infections.