Abstract
A retrospective analysis of susceptibility data on 542 blood isolates of the Bacteroides fragilis group tested from 1987 to 1999 by the same NCCLS-recommended broth microdilution method throughout is presented. Metronidazole, β-lactam-β-lactamase inhibitor combinations, carbapenems, and trovafloxacin were the most active agents (susceptibility of ≥93%). Among the cephalosporin-cephamycins, the order of activity was cefoxitin > ceftizoxime > cefotetan = cefotaxime = cefmetazole > ceftriaxone. All isolates were resistant to penicillin G, and 22% were resistant to clindamycin. The susceptibility rates to piperacillin-tazobactam, imipenem, and meropenem were affected least among isolates resistant to cefoxitin or clindamycin. Except for piperacillin-tazobactam, imipenem, and meropenem, the B. fragilis species was more susceptible than were the non-B. fragilis species. These data underscore the importance of susceptibility testing of the B. fragilis group and can serve as a guide in the choice of empirical antimicrobial therapy.
The Bacteroides fragilis group comprises the most important anaerobic pathogens in human infections. These pathogens are often associated with polymicrobial infections such as intra-abdominal, obstetric-gynecologic, diabetic foot, and skin and skin structure infections. The virulence of these pathogens contributes to the spread of the infection, abscess formation, and the production of an antiphagocytic polysaccharide capsule (6).
Documented increases in antimicrobial resistance to β-lactams and clindamycin have been reported (5, 28). β-Lactam resistance is mediated predominantly by production of β-lactamase enzymes among both gram-negative and -positive bacilli, as well as Acidaminococcus (6, 14, 22). The incidence of β-lactamase production among the B. fragilis group has been reported to be ≥95% of isolates (6). These enzymes have been characterized primarily as cephalosporinases; however, penicillinases and carbapenemases have also been reported (8, 17, 18, 22, 23, 27, 30).
Anaerobic bacteremia is an infrequent infection; however, it can result in a high mortality rate. Two previous studies (9, 10) of Bacteroides bacteremia determined that the mortality rate varied according to the species isolated. Bacteremia due to B. fragilis had associated mortality rates of 24 and 31%, respectively, whereas the mortality rate for B. distasonis was 50% and that for B. thetaiotaomicron was as high as 100%. More recently Redondo et al. (25) compared a group of patients with B. fragilis group bacteremia to a demographically matched-pair group of patients without bacteremia. The mortality rate was significantly higher in the bacteremic group (28% versus 8.7%, P = 0.002) and had a mortality risk ratio of 4.9. Moreover, patients with bacteremia remained in the hospital 16 days longer. These authors also reported that the most common source of bacteremia was associated with bowel surgery or disease. Montravers et al. (19) studied 100 patients who developed postoperative peritonitis due to aerobes and anaerobes and compared their clinical outcome to the appropriateness of their antimicrobial therapy based on in vitro susceptibility data. Patients who were infected with resistant isolates had a significantly higher mortality rate than those infected with susceptible isolates (45% versus 16%, P < 0.05). These authors concluded that the choice of initial empirical therapy directly affected the clinical outcome and that the subsequent change of antibiotic therapy based on culture results did not affect clinical outcome when the initial therapy was judged as inadequate. In a similar study recently reported by Nguyen et al. (21), a multicenter, observational study of 128 patients with Bacteriodes bacteremia compared the clinical outcome of these patients with the in vitro susceptibility of their blood isolates. The mortality rate of patients who received appropriate therapy (susceptible in vitro) was 16% compared to patients who received inappropriate therapy (resistant in vitro), whose mortality rate was 45% (P = 0.04). Microbiological persistence was higher with inactive versus active therapy (42% versus 12%, P = 0.06). That study concluded that in vitro susceptibility results reliably predicted the clinical outcome of the patients with a specificity of 97% and positive predictive value of 82%.
We report here the results of a retrospective analysis of in vitro susceptibility of B. fragilis group blood isolates as a subset of >3,600 isolates tested from 1987 to 1999.
MATERIALS AND METHODS
Organisms.
A total of 542 B. fragilis group isolates cultured from blood and representing isolates from 12 different medical centers throughout the United States were analyzed retrospectively for their susceptibility to various antimicrobial agents. Of this 542 isolate group, 233 isolates were included in a 1992 report (3) on the susceptibility of body fluid isolates, but the results were not detailed in that report except for the overall resistance rate to ceftizoxime (8%) and the lack of resistance to metronidazole and imipenem. We have, however, included those data in the present study. Due to the availability of various antimicrobial agents and formulary considerations during the testing period, not all antimicrobial agents were tested against every isolate. Each clinical isolate was identified by using selective growth media, biochemical profiles, and gas-liquid chromatography (16, 29). The distribution of the various species of isolates is shown in Table 1. Production of β-lactamase was determined by using a nitrocephin test (Cefinase; BBL, Cockeysville, Md.).
TABLE 1.
Distribution of blood isolates of the B. fragilis group and β-lactamase production
| Organism | No. of isolates (% of total)a | % β-Lactamase+b |
|---|---|---|
| B. fragilis | 339 (62.5) | 98.5 |
| B. thetaiotaomicron | 95 (17.5) | 100 |
| B. ovatus | 54 (10) | 94.7 |
| B. vulgatus | 30 (5.5) | 100 |
| B. distasonis | 17 (3.2) | 75 |
| B. uniformis | 5 (0.9) | 100 |
| B. caccae | 2 (0.4) | 100 |
The number in parenthesis indicates the percentage of the total B. fragilis group isolates.
That is, the percentage of each species producing β-lactamase.
Antimicrobial agents.
Each of the following standard laboratory powders was provided by the manufacturer: penicillin G (Eli Lilly, Indianapolis, Ind.); clindamycin and cefmetazole (Pharmacia-Upjohn, Kalamazoo, Mich.); trovafloxacin, ampicillin, and sulbactam (Pfizer, Groton, Conn.); imipenem, ertapenem, and cefoxitin (Merck, West Point, Pa.); metronidazole (Searle, Skokie, Ill.); piperacillin and tazobactam (Wyeth-Ayerst, St. Davids, Pa.); meropenem and cefotetan (AstraZeneca, Wilmington, Del.); ticarcillin and clavulanate (Glaxo SmithKline, Philadelphia, Pa.); cefotaxime (Aventis, Bridgewater, N.J.); ceftriaxone (Roche, Nutley, N.J.); and ceftizoxime (Fujisawa, Deerfield, Ill.). All antimicrobial powders were stored at −20°C prior to use.
Susceptibility testing.
MICs were determined by a broth microdilution method based on recommendations of the National Committee for Clinical Laboratory Standards (NCCLS) (20). Antimicrobial agents were prepared in serial twofold dilutions in anaerobe broth MIC (Difco) within a dilution range of 0.008 to 256 μg/ml. Ampicillin was combined with sulbactam in a 2:1 ratio, while serial twofold dilutions of piperacillin and ticarcillin were combined with tazobactam and clavulanate at constant concentrations of 4 and 2 μg/ml, respectively. Final test volumes of 100 μl were dispensed into microdilution wells. The inoculum of each isolate was prepared by suspending colonies from an anaerobic sheep blood agar plate incubated for 18 to 24 h in 5 ml of prereduced anaerobe broth MIC to a density equal to that of a no. 1 McFarland standard. The suspension was further diluted to give a final inoculum size of 105 CFU per well (106 CFU/ml). All plates were incubated at 35°C anaerobically for 48 h and then read with a mirror reader. The MIC was defined as the lowest concentration of each antimicrobial agent that inhibited the visible growth of the test isolate. With each susceptibility test run, quality control was performed with B. fragilis ATCC 25285, B. thetaiotaomicron ATCC 29741, and Eubacterium lentum ATCC 43055.
Data management.
MICs were collated to determine the mode MIC, the MIC at which 50% of the isolates were inhibited (MIC50), the MIC90, and the percentage of isolates susceptible (%S) to each antimicrobial agent based on NCCLS recommendations (20) The statistical significance of differences in susceptibility rates for all pairwise comparisons was performed by using Pearson's chi-square test or the Fisher exact test. To control the familywise error rate for multiple comparisons, the P values were adjusted by the Bonferroni method. Statistical significance was achieved with a P value of ≤0.05 (12).
RESULTS AND DISCUSSION
The distribution of the 542 clinical isolates and β-lactamase production is shown in Table 1. B. fragilis isolates were the predominant species (62.5%), followed by B. thetaiotaomicron (17.5%), B. ovatus (10%), B. vulgatus (5.5%), and B. distasonis (3.2%). Only five and two isolates of B. uniformis and B. caccae, respectively, were isolated from blood. Nguyen et al. (21) recently reported a similar distribution of B. fragilis group blood isolates. That study showed essentially the same incidence of B. fragilis, B. distasonis, B. thetaiotamicron, and B. vulgatus, whereas we tested more B. ovatus and fewer B. uniformis and B. caccae. Overall, β-lactamase production for the B. fragilis group was 97.4%. Variation in β-lactamase production among the various species is shown in Table 1. Previous studies have reported a high rate of β-lactamase production among the B. fragilis group. In 1983 we found 87% of the B. fragilis group were cephalosporinase producers, whereas in 1988 and 2001 this rate was ≥97% (1, 2,7). Interestingly the B. distasonis isolates are often β-lactamase nonproducers but remain highly resistant to penicillin G or ampicillin (1, 13). Moreover, we previously reported (4) that of the B. fragilis group species, B. distasonis isolates were the most resistant to ampicillin-sulbactam, ticarcillin-clavulanate, and amoxicillin-clavulanate, whereas piperacillin-tazobactam was active against all of the same isolates.
The susceptibility results of the 542 blood isolates are shown in Table 2. Since the collection of the isolates ranged from 1987 to 1999 not all antimicrobials were tested against all isolates; therefore, the mix of various species tested for each antimicrobial agent varied. However, we feel these comparisons are valid since the reports that the empirical choice of antimicrobial therapy are important, regardless of the species, in the outcome of bacteremia and intra-abdominal infections involving the B. fragilis group (19, 21, 25).
TABLE 2.
Comparison of in vitro susceptibility rates of blood isolates of the B. fragilis group
| Antimicrobial agent | No. of isolates tested | MIC (μg/ml)
|
All isolates (%S)a | Cfx-I&Rb
|
Clin-I&Rc
|
|||||
|---|---|---|---|---|---|---|---|---|---|---|
| Range | Mode | MIC50 | MIC90 | %S | %SΔd | %S | %SΔ | |||
| Penicillin G | 160 | 0.06-128 | 8 | 8 | 128 | 0 | 0 | 0 | 0 | 0 |
| Piperacillin | 384 | 0.12-256 | 2 | 8 | 128 | 77 | 46 | −31 | 60 | −17 |
| Ticarcillin | 137 | 0.12-256 | 16 | 32 | 128 | 63 | 21 | −42 | 39 | −24 |
| Piperacillin-tazobactam | 142 | 0.06-64 | 2 | 1 | 8 | 99.3 | 96 | −3.3 | 96.1 | −3.2 |
| Ticarcillin-clavulanate | 191 | 0.25-256 | 1 | 0.5 | 8 | 96 | 83 | −13 | 86 | −10 |
| Ampicillin-sulbactam | 382 | 0.03-128 | 1 | 1 | 8 | 93 | 81 | −12 | 80 | −13 |
| Cefotaxime | 384 | 0.06-256 | 4 | 8 | 64 | 62 | 32 | −30 | 34 | −28 |
| Ceftriaxone | 138 | 1-128 | 2 | 32 | 128 | 49 | 7 | −42 | 7 | −42 |
| Cefoxitin | 515 | 0.12-128 | 8 | 8 | 32 | 84 | 0 | 73 | −11 | |
| Cefotetan | 473 | 0.06-256 | 8 | 8 | 64 | 64 | 8 | −56 | 26 | −38 |
| Cefmetazole | 84 | 0.25-128 | 8 | 16 | 64 | 61 | 0 | −61 | 35 | −26 |
| Ceftizoxime | 358 | 0.06-256 | 4 | 4 | 64 | 78 | 67 | −11 | 65 | −13 |
| Imipenem | 378 | 0.015-32 | 0.03 | 0.12 | 1 | 99.5 | 97 | −2.5 | 98 | −1.5 |
| Meropenem | 127 | 0.015-32 | 0.12 | 0.5 | 0.5 | 98 | 94 | −4 | 94 | −4 |
| Ertapenem | 92 | 0.06-32 | 0.25 | 0.25 | 2 | 94 | 64 | −30 | 86 | −8 |
| Trovafloxacin | 156 | 0.03-16 | 0.25 | 0.25 | 2 | 96 | 86 | −10 | 88 | −8 |
| Clindamycin | 542 | 0.015-64 | 0.25 | 0.5 | 16 | 78 | 63 | −15 | 0 | |
| Metronidazole | 542 | 0.06-8 | 1 | 1 | 2 | 100 | 100 | 0 | 100 | 0 |
The following MIC values (μg/ml) were used as susceptibility breakpoints as recommended by the NCCLS (14): penicillin G, ≤0.5; piperacillin, ≤32; ticarcillin, ≤32; piperacillin-tazobactam, ≤32; ticarcillin-clavulanate, ≤32; ampicillin-sulbactam, ≤8; cefotaxime, ≤16; ceftriaxone, ≤16; cefoxitin, ≤16; cefotetan, ≤16; cefmetazole, ≤16; ceftizoxime, ≤16; imipenem, ≤4; meropenem, ≤4; trovafloxacin, ≤2; clindamycin, ≤2; and metronidazole, ≤8. Ertapenem susceptibility was ≤4 μg/ml, as recommended by the manufacturer. Test isolates with higher than stated susceptible MICs were considered “not susceptible” and were comprised of isolates with MICs interpreted as intermediate or fully resistant.
Contains isolates with cefoxitin (Cfx) MICs of ≥32 μg/ml.
Contains isolates with clindamycin (Clin) MICs of ≥4 μg/ml.
Indicates the decrease in the percentage of isolates susceptible to isolates “not-susceptible” to cefoxitin or clindamycin.
Only metronidazole was active against all of the test isolates. All test isolates were resistant to penicillin G based on β-lactamase production or MICs. Both piperacillin and ticarcillin showed moderate activity with susceptibility rates (i.e., %S) of 77 and 63%, respectively. However, the piperacillin mode MIC was fourfold less than that of ticarcillin. Piperacillin-tazobactam, ticarcillin-clavulanate and ampicillin-sulbactam showed good activity (%S ≥ 96%) against the test isolates. Although no statistical differences were noted, more isolates were susceptible to piperacillin-tazobactam (%S = 99.1) than to ticarcillin-clavulanate (%S = 96%) and ampicillin-sulbactam (%S = 95%). Of 142 simultaneous comparisons of all three β-lactam-β-lactamase inhibitor combinations, 137 isolates were susceptible to all three agents. One B. fragilis isolate had an intermediate MIC for piperacillin-tazobactam but had resistant MICs for ticarcillin-clavulanate and ampicillin-sulbactam. Of the other four isolates (two B. fragilis and two B. ovatus), all were susceptible to piperacillin-tazobactam, three were intermediate, and one was susceptible to ticarcillin-clavulanate, and three were intermediate and one was resistant to ampicillin-sulbactam. The activity of the various cephalosporin-cephamycin agents varied widely. Cefoxitin was the most active of the group, with 83% of the isolates being susceptible, followed by ceftizoxime, with 78% being susceptible. Against the B. fragilis group overall, cefotaxime, cefotetan, and cefmetazole showed equal activity with susceptibility rates of 61 to 63%; however, distinct differences can be seen when analyzing individual species (see Table 3). Ceftriaxone had the poorest activity against the B. fragilis group, with only 49% of isolates being susceptible.
TABLE 3.
Comparison of susceptibility rates of various species of the B. fragilis group isolated from blood
| Antimicrobial agent | % S value(s)a
|
||||||
|---|---|---|---|---|---|---|---|
| B. fragilis | B. thetaiotaomicron | B. distasonis | B. ovatus | B. vulgatus | B. uniformis | B. caccae | |
| Penicillin G | 0 | 0 | 0 | 0 | 0 | 0 | 4, 32∗ |
| Piperacillin | 80 | 76 | 64 | 74 | 45 | 32, 32, 32∗ | NT |
| Ticarcillin | 69 | 68 | 16, 256 | 39 | 44 | 256 | NT |
| Piperacillin-tazobactam | 99.1 | 100 | 100 | 100 | 100 | 0.12, 0.25∗ | 0.5, 2∗ |
| Ticarcillin-clavulanate | 99.2 | 100 | 1, 32, 128, 128 | 77 | 100 | 8∗ | NT |
| Ampicillin-sulbactam | 95 | 98 | 77 | 84 | 88 | 1, 2, 2, 2∗ | 1, 16∗ |
| Cefotaxime | 74 | 39 | 36 | 46 | 50 | 32, 64, 64∗ | NT |
| Ceftriaxone | 69 | 8 | 17 | 11 | 0 | NT | NT |
| Cefoxitin | 90 | 70 | 53 | 88 | 69 | 1, 2, 16, 64∗ | 4, 32∗ |
| Cefotetan | 87 | 19 | 21 | 26 | 50 | 64, 128∗ | NT |
| Cefmetazole | 85 | 6 | 33 | 20 | 8, 16∗ | 128∗ | NT |
| Ceftizoxime | 81 | 76 | 64 | 81 | 56 | 32, 32∗ | NT |
| Imipenem | 99.2 | 100 | 100 | 100 | 100 | 100 | 0.03, 0.25∗ |
| Meropenem | 96 | 100 | 100 | 100 | 100 | 0.06, 0.12, 0.25∗ | 0.12, 2∗ |
| Ertapenem | 94 | 100 | 60 | 88 | 100 | 0.06, 0.12∗ | 0.25, 2∗ |
| Trovafloxacin | 100 | 92 | 100 | 90 | 92 | 0.03, 0.5∗ | 0.12, 2∗ |
| Clindamycin | 89 | 63 | 53 | 48 | 73 | 60 | 16, 16∗ |
| Metronidazole | 100 | 100 | 100 | 100 | 100 | 100 | 0.25, 0.5∗ |
∗, when four or fewer isolates were tested, the individual MICs are listed. NT, not tested.
All three carbapenems were highly active. Imipenem was the most active, with an isolate susceptibility of 99.5%. Meropenem and ertapenem had equal activity (%S ≥ 94%). Ertapenem is a newly Food and Drug Administration-approved carbapenem with pharmacokinetics compatible with once-a-day dosing (15). The mode MICs of imipenem, however, were four- and eightfold lower than those of meropenem and ertapenem, respectively. For one isolate of B. fragilis, the imipenem MIC was 8 μg/ml (intermediate) and had meropenem and ertapenem MICs of ≥32 μg/ml (resistant). Another isolate of B. fragilis was susceptible (MIC = 1 μg/ml) to imipenem but was resistant to meropenem and ertapenem (MICs of ≥32 μg/ml). Trovafloxacin was active against 96% of the isolates, whereas clindamycin was active against only 76% of the test isolates.
Table 2 also presents the changes in susceptibility rates to the various antimicrobial agents when subsets of isolates for which the MICs of cefoxitin or clindamycin are intermediate or fully resistant are compared to the entire test group. Except for penicillin or metronidazole, all of the remaining antimicrobial agents altered the susceptibility rates. Only piperacillin-tazobactam and imipenem were associated with decreases in susceptibility rates of <4% among both cefoxitin- and clindamycin-resistant isolates. Ticarcillin-clavulanate and ampicillin-sulbactam were associated with susceptibility decreases of from 10 to 13%, whereas meropenem and ertapenem were associated with reductions in susceptibility of 4 to 30%. With cefoxitin-resistant isolates, the susceptibility rates associated with other cephalosporin-cephamycin agents were reduced from 11 to 61%, and with clindamycin-resistant isolates the decreases in susceptibility rates ranged from 11 to 42%. Susceptibility to clindamycin decreased 15% in cefoxitin-resistant isolates. The susceptibility rate associated with trovafloxacin decreased 10% with cefoxitin-resistant isolates and 8% with clindamycin-resistant isolates.
Table 3 indicates the susceptibility rates among the B. fragilis group species to the test antimicrobial agents. Overall, the B. fragilis species remains the most susceptible of the B. fragilis group. Only metronidazole was active against all of the species; however, Rotimi et al. (26) recently reported from Kuwait the isolation of metronidazole-resistant B. fragilis, B. ovatus, and B. distasonis strains from patients with abdominal infections who were clinical failures while being treated with metronidazole. For ticarcillin-clavulanate and ampicillin-sulbactam, isolates of B. distasonis and B. ovatus were the least susceptible (%S = 50 and 77%, respectively). With piperacillin-tazobactam, reduced susceptibility (intermediate) was noted for a single isolate of B. fragilis.
Interestingly, resistance to the carbapenems was found only among B. fragilis species isolates, whereas resistance to ertapenem varied from 6 to 40% among isolates of B. fragilis, B. ovatus, and B. distasonis. The non-B. fragilis species varied widely in susceptibility rates to the cephalosporin-cephamycin agents (Table 3). Overall, cefoxitin remained the most active, with susceptibility rates ranging from 53 to 84% compared to cefotetan, with susceptibility rates ranging from 0 to 50%. Of the cephalosporins, ceftizoxime was the most active, with the exception of B. uniformis isolates, which were resistant to all cephalosporin-cephamycins except for cefoxitin. For cefotaxime, ceftriaxone, and cefmetazole, ≤50% of all isolates of the non-B. fragilis species were resistant to these agents. Patey et al. (24) studied 416 blood isolates of the B. fragilis group from patients in France over a 2-year period. These researchers obtained results similar to ours in that all species were highly susceptible (>98%) to metronidazole, β-lactam-β-lactamase inhibitors, and carbapenems. Compared to the B. fragilis species, the non-B. fragilis species showed reduced susceptibility to piperacillin (88%), clindamycin (86%), cefotetan (23%), and cefotaxime (69%), although these authors did not specify results for individual species. In the present study, only strains of B. ovatus and B. vulgatus were found to be resistant to trovafloxacin, whereas resistance to clindamycin occurred among all of the species, including the two isolates of B. caccae.
Table 4 summarizes the statistical relationships between all pairwise combinations of antimicrobial agents against the B. fragilis group isolates when susceptibility rates were compared. Metronidazole was statistically more active than all of the antimicrobial agents except for piperacillin-tazobactam, the carbapenems, and trovafloxacin. No statistical differences were found within the β-lactam-β-lactamase inhibitor group, but as a group they were statistically more active than the cephalosporin-cephamycins, piperacillin, ticarcillin, and clindamycin. Similarly, no statistical differences were noted among the carbapenem compounds compared to the β-lactam-β-lactamase inhibitor agents and trovafloxacin; however, significance was achieved with few exceptions between carbapenems and cephalosporin-cephamycins, piperacillin, ticarcillin, and clindamycin. Various degrees of significance were also achieved among isolated comparisons, particularly among the cephalosporin-cephamycin agents (Table 4).
TABLE 4.
P values for pairwise comparisons of susceptibility rates among the various antimicrobial agents
| Antibiotic | n | %S |
P values for pairwise %S comparisonsa
|
||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PIP (384, 76.6) | TIC (137, 63) | TZP (142, 99.3) | TIM (191, 95.8) | SAM (382, 92.9) | CTX (384, 62) | CRO (138, 49.3) | FOX (515, 84.1) | CTT (473, 64.5) | CMZ (84, 61) | ZOX (358, 78) | IPM (378, 99.5) | MEM (127, 97.6) | ERT (92, 93.5) | TVA (156, 95.5) | CLI (542, 77.7) | MTR (542, 100) | |||
| PEN G | 160 | 0 | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ |
| PIP | 384 | 76.6 | - | 0.278 | <0.001∗ | <0.001∗ | <0.001∗ | 0.002∗ | <0.001∗ | 0.704 | 0.019∗ | 0.428 | 1.000 | <0.001∗ | <0.001∗ | 0.043∗ | <0.001∗ | 1.000 | <0.001∗ |
| TIC | 137 | 63 | - | - | <0.001∗ | <0.001∗ | <0.001∗ | 1.000 | 1.000 | <0.001∗ | 1.000 | 1.000 | 0.093 | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | 0.052 | <0.001∗ |
| TZP | 142 | 99.3 | - | - | - | 1.000 | 0.610 | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | 1.000 | 1.000 | 1.000 | 1.000 | <0.001∗ | 1.000 |
| TIM | 191 | 95.8 | - | - | - | - | 1.000 | <0.001∗ | <0.001∗ | 0.005∗ | <0.001∗ | <0.001∗ | <0.001∗ | 0.507 | 1.000 | 1.000 | 1.000 | <0.001∗ | 0.003∗ |
| SAM | 382 | 92.9 | - | - | - | - | - | <0.001∗ | <0.001∗ | 0.009∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | 1.000 | 1.000 | 1.000 | <0.001∗ | <0.001∗ |
| CTX | 384 | 62 | - | - | - | - | - | - | 1.000 | <0.001∗ | 1.000 | 1.000 | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ |
| CRO | 138 | 49.3 | - | - | - | - | - | - | - | <0.001∗ | 0.194 | 1.000 | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ |
| FOX | 515 | 84.1 | - | - | - | - | - | - | - | - | <0.001∗ | <0.001∗ | 1.000 | <0.001∗ | 0.008∗ | 1.000 | 0.034∗ | 1.000 | <0.001∗ |
| CTT | 473 | 64.5 | - | - | - | - | - | - | - | - | - | 1.000 | 0.004∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ |
| CMZ | 84 | 61 | - | - | - | - | - | - | - | - | - | - | 0.167 | <0.001∗ | <0.001∗ | <0.001∗ | <0.001∗ | 0.120 | <0.001∗ |
| ZOX | 358 | 78 | - | - | - | - | - | - | - | - | - | - | - | <0.001∗ | <0.001∗ | 0.104 | <0.001∗ | 1.000 | <0.001∗ |
| IPM | 378 | 99.5 | - | - | - | - | - | - | - | - | - | - | - | - | 1.000 | 0.149 | 0.515 | <0.001∗ | 1.000 |
| MEM | 127 | 97.6 | - | - | - | - | - | - | - | - | - | - | - | - | - | 1.000 | 1.000 | <0.001∗ | 1.000 |
| ERT | 92 | 93.5 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | 1.000 | 0.071 | 0.001∗ |
| TVA | 156 | 95.5 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | <0.001∗ | 0.004∗ |
| CLI | 542 | 77.7 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - | <0.001∗ |
P values are derived from Fisher's exact test or Pearson's chi-square test, adjusted for multiple comparisons with the Bonferroni correction. The asterisks indicate values that were statistically significant at an alpha level of 0.05. Antibiotics: PEN G, penicillin G; PIP, piperacillin; TIC, ticarcillin; TZP, piperacillin-tazobactam; TIM, ticarcillin-clavulanate; SAM, ampicillin-sulbactam; TAX, cefotaxime; CRO, ceftriaxone; FOX, cefoxitin; CTT, cefotetan; CMZ, cefmetazole; ZOX, ceftizoxime; IPM, imipenem; MEM, meropenem; ERT, ertapenem; TVA, trovafloxacin; CLI, clindamycin; MTR, metronidazole. Parenthetic values indicate the following: n, %S.
Many in vitro susceptibility studies have included B. fragilis group isolates, but few of these studies have selected out blood isolates for discussion. In 1974 Chow and Guze (10) detailed the clinical and demographic factors from 112 patients with Bacteroidaceae bacteremia.These authors found that, of 36 isolates of the B. fragilis group, 97, 94, 47, and 36% were susceptible to chloramphenicol, clindamycin, lincomycin, and tetracycline, respectively, by using higher breakpoints than are currently recommended by the NCCLS (14). Although no β-lactamase results were reported, 25% of the isolates were susceptible to penicillin at ≤5 μg/ml, whereas 75% (21 of 28) were susceptible to erythromycin at 2.5 μg/ml. By comparison, this report indicates that for blood isolates susceptibility to clindamycin has fallen to 76% and for cefoxitin it fell to 83%, whereas metronidazole remained active against all B. fragilis group isolates. Cuchural et al. (11) reported that, in a susceptibility study of B. fragilis group isolates done in 1984 and 1985, blood isolates had a resistance rate to cefoxitin of 10.7% at a breakpoint of ≤16 μg/ml, which is lower than the 16% we found in our analysis with the same breakpoint. This finding may be related to the test isolate population of Cuchural et al., 71.4% of which were B. fragilis species isolates compared to the 62.5% B. fragilis isolates in our study, which traditionally tend to be more susceptible to a variety of antimicrobial agents than other species. Moreover, these authors reported a susceptibility rate to clindamycin of 90% compared to our finding of 78%, which is probably due to their use of a higher breakpoint (4 μg/ml). In a similar report from that laboratory (28) comparing the susceptibilities of the B. fragilis group from 1995 and 1996 from all sources, the resistances to cefoxitin were 8 and 5%, respectively, whereas the resistance rate to clindamycin was 16% for both years. By comparison, in the latter study the percentage of B. fragilis species tested decreased to 53.4%. It should be pointed out that in both of their studies the authors chose to include intermediate MICs for both agents in the susceptible category, whereas we included them in the resistant category.
The present analysis, to our knowledge the largest yet reported, indicates that blood isolates collected from several geographical areas in the United States over a 12-year period have shown decreased susceptibility to certain antimicrobial agents, such as clindamycin, but remain susceptible to numerous other agents. Such data are important in the choice of empirical therapy of infections involving the B. fragilis group, particularly for bacteremia due to these organisms, which has been shown to be an independent risk factor for mortality (21, 25). Moreover, these data did not demonstrate any metronidazole-resistant B. fragilis group isolates as reported in Kuwait nor high resistance rates to imipenem as reported in Japan. However, we must remain vigilant through additional studies such as this to detect any future significant changes in antimicrobial resistance.
REFERENCES
- 1.Aldridge, K. E., C. V. Sanders, A. C. Lewis, and R. L. Marier. 1983. Susceptibility of anaerobic bacteria to beta-lactam antibiotics and beta-lactamase production. J. Med. Microbiol. 16:75-82. [DOI] [PubMed] [Google Scholar]
- 2.Aldridge, K. E., A. Henderberg, D. D. Schiro, and C. V. Sanders. 1988. Susceptibility of Bacteroides fragilis group isolates to broad-spectrum β-lactams, clindamycin, and metronidazole: rates of resistance, cross-resistance, and importance of β-lactamase production. Adv. Ther. 5:273-282. [Google Scholar]
- 3.Aldridge, K. E. 1992. In vitro susceptibility of Bacteroides fragilis group strains from abscesses, body fluids, and wound/tissue sources. Drug Investig. 4(Suppl. 1):1-6. [Google Scholar]
- 4.Aldridge, K. E. 1993. Activity of β-lactam-β-lactamase inhibitor combinations against the Bacteroides fragilis group: increased resistance among B. distasonis strains, p. 478-479. In Proceedings of the 18th International Congress of Chemotherapy. International Congress of Chemotherapy, Stockholm, Sweden.
- 5.Aldridge, K. E., M. Gelfand, L. B. Reller, L. W. Ayers, C. L. Pierson, F. Schoenknect, R. C. Tilton, J. Wilkins, A. Henderberg, D. D. Schiro, M. Johnson, A. Janney, and C. V. Sanders. 1994. A five-year multicenter study of the susceptibility of the Bacteroides fragilis group isolates to cephalosporins, cephamycins, penicillins, clindamycin, and metronidazole in the United States. Diagn. Microbiol. Infect. Dis. 18:235-241. [DOI] [PubMed] [Google Scholar]
- 6.Aldridge, K. E. 1995. The occurrence, virulence, and antimicrobial resistance of anaerobes in polymicrobial infections. Am. J. Surg. 169(Suppl. 5A):2S-7S. [PubMed] [Google Scholar]
- 7.Aldridge, K. E., D. Ashcraft, K. Cambre, C. L. Pierson, S. G. Jenkins, and J. E. Rosenblatt. 2001. Multicenter survey of the changing in vitro antimicrobial susceptibilities of clinical isolates of Bacteroides fragilis group, Prevotella, Fusobacterium, Porphyromonas, and Peptostreptococcus species. Antimicrob. Agents Chemother. 45:1238-1243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bandoh, K., K. Ueno, K. Watanabe, and N. Kato. 1993. Susceptibility patterns and resistance to imipenem in the B. fragilis group species in Japan: a 4-year study. Clin. Infect. Dis. 16(Suppl. 4):S382-S386. [DOI] [PubMed] [Google Scholar]
- 9.Brook, I. 1990. The clinical importance of all members of the Bacteroides fragilis group. J. Antimicrob. Chemother. 25:473-474. [DOI] [PubMed] [Google Scholar]
- 10.Chow, A. W., and L. B. Guze. 1974. Bacteroidaceae bacteremia: clinical experience with 112 patients. Medicine 53:93-126. [PubMed] [Google Scholar]
- 11.Cuchural, G. J., Jr., F. P. Tally, N. V. Jacobus, K. Aldridge, T. Cleary, S. M. Finegold, G. Hill, P. Iannini, J. P. O'Keefe, C. Pierson, D. Crook, T. Russo, and D. Hecht. 1988. Susceptibility of the Bacteroides fragilis group in the United States: analysis by site of isolation. Antimicrob. Agents Chemother. 32:717-722. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Daniel, W. W. 1995. Biostatistics: a foundation for analysis in the health sciences, 6th ed. John Wiley & Sons, Inc., New York, N.Y.
- 13.Ednie, L. M., M. R. Jacobs, and P. C. Appelbaum. 1998. Activities of gatifloxacin compared to those of seven other agents against anaerobic organisms. Antimicrob. Agents Chemother. 42:2459-2462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Galán, J. G., M. Reig, A. Navas, F. Baquero, and J. Blázquez. 2000. ACI-1 from Acidaminococcus fermentans: characterization of the first β-lactamase in anaerobic cocci. Antimicrob. Agents Chemother. 44:3144-3149. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Gill, C. J., J. J. Jackson, L. S. Gerckens, B. A. Pelak, R. K. Thompson, J. G. Sundelof, H. Kropp, and H. Rosen. 1998. In vitro activity and pharmacokinetic evaluation of a novel long-acting carbapenem antibiotic, MK-826 (L-748-345). Antimicrob. Agents Chemother. 42:1996-2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Holdeman, L. V., E. P. Cato, and W. E. C. Moore. 1977. Anaerobe laboratory manual, 4th ed. Virginia Polytechnic Institute and State University, Blacksburg, Va.
- 17.Kesado, T. L. Lindqvist, M. Hedberg, K. Tunér, and C. E. Nord. 1989. Purification and characterization of a new β-lactamase from Clostridium butyricum. Antimicrob. Agents Chemother. 33:1302-1307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Könöen, E., S. Nyfors, J. Mätö, S. Asikainen, and H. Joursimies-Somer. 1997. β-Lactamase production by oral pigmented Prevotella species isolated from young children. Clin. Infect. Dis. 25:S272-S274. [DOI] [PubMed] [Google Scholar]
- 19.Montravers, P., R. Gauzit, C. Muller, J. P. Marmuse, A. Fichelle, and J. M. Desmonts. 1996. Emergence of antibiotic-resistant bacteria in cases of peritonitis after intra-abdominal surgery affects the efficacy of empirical antimicrobial therapy. Clin. Infect. Dis. 23:486-494. [DOI] [PubMed]
- 20.National Committee for Clinical Laboratory Standards. 2001. Approved standard M11-A5. Methods for antimicrobial susceptibility testing of anaerobic bacteria. National Committee for Clinical Laboratory Standards, Villanova, Pa.
- 21.Nguyen, M. H., V. L. Yu, A. J. Morris, L. McDermott, M. W. Wagener, L. Harrell, and D. R. Snydman. 2000. Antimicrobial resistance and clinical outcome of Bacteroides bacteremia: findings of a multicenter prospective observational trial. Clin. Infect. Dis. 30:870-876. [DOI] [PubMed] [Google Scholar]
- 22.Nord, C. E., L. Lindqvist, B. Olsson-Liljequist, and K. Tuner. 1985. Beta-lactamases in anaerobia bacteria. Scand. J. Infect. Dis. 46(Suppl.):57-63. [PubMed] [Google Scholar]
- 23.Parker, A. C., and C. J. Smith. 1993. Genetic and biochemical analysis of a novel Ambler class A β-lactamase responsible for cefoxitin resistance in Bacteroides species. Antimicrob. Agents Chemother. 37:1028-1036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Patey, O., E. Varon, T. Prazuck, I. Podglajen, A. Dublanchet, L. Dubreuil, and J. Breuil. 1994. Multicentre survey in France of the antimicrobial susceptibilities of 416 blood culture isolates of the Bacteroides fragilis group. J. Antimicrob. Chemother. 33:1029-1034. [DOI] [PubMed]
- 25.Redondo, M. C., M. D. J. Arbo, J. Grindlinger, and D. R. Snydman. 1995. Attributable mortality of bacteremia associated with the Bacteroides fragilis group. Clin. Infect. Dis. 20:1492-1946. [DOI] [PubMed] [Google Scholar]
- 26.Rotimi, V. O., M. Khoursheed, J. S. Brazier, W. Y. Jamal, and F. B. Khodakhast. 1999. Bacteroides species highly resistant to metronidazole: an emerging clinical problem? Clin. Microbiol. Infect. 5:166-169. [DOI] [PubMed]
- 27.Smith, C. J., T. K. Bennett, and A. C. Parker. 1994. Molecular and genetic analysis of the Bacteroides uniformis cephalosporinase gene, cblA, encoding the species-specific β-lactamase. Antimicrob. Agents Chemother. 38:1711-1715. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Snydman, D. R., N. J. Jacobus, L. A. McDermott, S. Supran, G. J. Cuchural, Jr., S. Finegold, L. Harrell, D. W. Hecht, P. Iannini, S. Jenkins, C. Pierson, J. Rihs, and S. L. Gorbach. 1999. Multicenter study of in vitro susceptibility of the Bacteroides fragilis group, 1995 to 1996, with comparison of resistance trends from 1990 to 1996. Antimicrob. Agents Chemother. 43:2417-2422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Summanen, P., E. J. Baron, D. Citron, C. Strong, H. M. Wexler, and S. M. Finegold. 1993. Wadsworth anaerobic bacteriology manual, 5th ed. Star Publishing Company, Belmont, Calif.
- 30.Tunér, K., L. Lindqvist, and C. E. Nord. 1985. Purification and properties of a novel β-lactamase from Fusobacterium nucleatum. Antimicrob. Agents Chemother. 27:943-947. [DOI] [PMC free article] [PubMed] [Google Scholar]