Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, September 2008, p. 3161-3168, Vol. 52, No. 9
0066-4804/08/$08.00+0 doi:10.1128/AAC.00355-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Departments of Internal Medicine,1 Laboratory Medicine, National Taiwan University Hospital, National Taiwan University College of Medicine, Taipei,2 Department of Internal Medicine, Far Eastern Memorial Hospital, Taipei County, Taiwan3
Received 14 March 2008/ Returned for modification 16 April 2008/ Accepted 5 July 2008
|
|
|---|
|
|
|---|
The rates of resistance of aerobic pathogens have been increasing over the last decade in Taiwan (17, 33). However, surveillance data for anaerobic pathogens remain scarce (28). The antimicrobial resistance of anaerobic pathogens adversely affects clinical outcome, resulting in treatment failure and mortality (20). High incidences (28) of cefoxitin and clindamycin resistance in Taiwan were reported in 2002, but the same study disclosed a low rate of resistance to meropenem (<1%) at that time.
The aim of this study was to determine the in vitro susceptibilities of anaerobic blood isolates to several old and newer agents and to analyze the resistance trends of anaerobic clinical isolates and isolates associated with nosocomial infections at a tertiary care medical center in Taiwan.
|
|
|---|
All isolates of B. fragilis associated with nosocomial infections (from 1990 to 2006) were also evaluated. The Nosocomial Infection Control Committee of the hospital was established in 1980 to identify pathogens causing nosocomial infections by chart review and to obtain and analyze antimicrobial susceptibility results for these pathogens from the hospital's clinical microbiology laboratory. National Nosocomial Infections Surveillance System (NNIS) definitions for nosocomial infections were applied (8). These isolates were collected from patients treated at National Taiwan University Hospital (NTUH), a tertiary referral center with a capacity of 2,500 beds in northern Taiwan. The susceptibilities of these isolates were determined by the breakpoint agar dilution method (7). Susceptibility data for penicillin, ampicillin-sulbactam, cefmetazole, clindamycin, chloramphenicol, and metronidazole were analyzed.
Antimicrobial susceptibility testing. β-Lactamase production was determined for all gram-negative isolates other than B. fragilis isolates by using the Cefinase disc method (BBL Microbiology Systems, Cockeysville, MD) (7). The MICs of 207 nonduplicate anaerobic blood isolates to 14 antimicrobial agents were determined by the agar dilution method according to the guidelines recommended by the CLSI (7). These isolates were collected from patients with bloodstream infections at NTUH from January 2006 to December 2006. An inoculation of 105 CFU per well was applied with a Steers replicator onto supplemented Brucella blood agar. The plates were incubated in an anaerobic chamber for 48 h at 35°C. The MIC was defined as the lowest concentration of each antimicrobial agent that inhibited the growth of an organism. The antimicrobial agents used for susceptibility testing were obtained from their corresponding manufacturers.
B. fragilis ATCC 25285, Bacteroides thetaiotaomicron ATCC 29741, and Eubacterium lentum ATCC 43055 were used for quality control of the susceptibility tests for each run of susceptibility testing. The breakpoints for the carbapenems, β-lactam-β-lactamase inhibitor combinations, cefmetazole, moxifloxacin, chloramphenicol, and clindamycin were established by the CLSI for anaerobic bacteria (7). All isolates that were resistant to any carbapenem in the initial run were retested to confirm the accuracy of the results.
Statistical analysis. The trends of susceptibility over time were analyzed by using the Cochran-Armitage trend test. A P value of less than 0.05 was considered statistically significant.
|
|
|---|
![]() View larger version (42K): [in a new window] |
FIG. 1. Trends of susceptibility of clinical isolates of common anaerobic pathogens from 2000 to 2007 at National Taiwan University Hospital. Bacteroides fragilis (A), Bacteroides species other than B. fragilis (B), Prevotella species (C), Fusobacterium species (D), and gram-positive cocci other than Peptostreptococcus anaerobius (E). Susceptibilities of the isolates were determined by the breakpoint agar dilution method. An asterisk denotes a significant decreasing trend (P < 0.05) of susceptibility to the indicated antimicrobial agent from 2000 to 2007.
|
The rates of susceptibility to penicillin for Bacteroides species other than B. fragilis (n = 2,701) were higher than those for B. fragilis isolates (n = 2,754) (Fig. 1B). Cefmetazole and clindamycin were only moderately active, with susceptibility rates of less than 50% since 2001. Ampicillin-sulbactam had better activity than cefmetazole and clindamycin from 2000 to 2006 but became less potent than cefmetazole after 2006. Chloramphenicol and metronidazole were most active in the preceding 7 years, but resistant isolates were present.
During the preceding 7 years, chloramphenicol and metronidazole were almost fully active against Fusobacterium species (n = 758). The rates of nonsusceptibility to ampicillin-sulbactam, penicillin, and cefmetazole increased during this period. Ampicillin-sulbactam and cefmetazole were reliable before 2002 (susceptibility rate, >80%), but had moderate activity against only 84% and 81% of Fusobacterium species in the first half of 2007. Whether these rates of nonsusceptibility will continue to increase remains unclear, and Fusobacterium species should be monitored.
Penicillin and clindamycin had poor activity against Prevotella species (n = 1,075) from 2000 (Fig. 1C). Chloramphenicol (93% to 100%) and metronidazole (97% to 100%) remained the most-active agents. Penicillin (43% to 88%) and clindamycin (46% to 88%) were moderately active against Fusobacterium species (Fig. 1D). These isolates were almost all susceptible to chloramphenicol (91% to 100%) and metronidazole (97% to 100%).
Peptostreptococcus species and other gram-positive coccus (n = 4,106) isolates showed a stationary susceptibility trend over the study period (Fig. 1E). Clindamycin (54% to 61%) was the only agent to show a worse susceptibility rate in recent years than the other antimicrobial agents tested.
Figure 2 illustrates the trends of the rates of susceptibility of 606 isolates of B. fragilis associated with nosocomial infections in the preceding 16 years (1990 to 2007). For Bacteroides species, the susceptibility to penicillin was less than 20%. Excellent antimicrobial activity was noted for metronidazole and chloramphenicol. The rates of susceptibility to clindamycin were generally above 50% before 1999 but decreased thereafter. Ampicillin-sulbactam retained moderate antimicrobial activities ranging from 60% to 70% between 1999 and 2005 but had a decreased rate of susceptibility of 48.8% in 2006. In general, the rates of susceptibility of all clinical isolates of B. fragilis to clindamycin and cefmetazole were higher than those of isolates associated with nosocomial infections.
![]() View larger version (26K): [in a new window] |
FIG. 2. Trends of susceptibility of all B. fragilis isolates associated with nosocomial infections from 1990 to 2006 at National Taiwan University Hospital. An asterisk denotes a significant decreasing trend (P < 0.05) of susceptibility to the indicated antimicrobial agent from 1996 to 2006.
|
|
View this table: [in a new window] |
TABLE 1. Trend of decreased susceptibility in selected antimicrobial-microorganism combinations with time as analyzed by Cochran-Armitage trend test
|
All MICs of isolates of the three control strains to the antimicrobial agents tested (except penicillin and levofloxacin, for which MIC ranges were not provided by CLSI) for each run (total of seven runs) were within the recommended ranges (7). The MIC ranges, MIC50s, MIC90s, and percentages of the 207 anaerobic blood isolates that were susceptible intermediate, or resistant to the 14 antimicrobial agents are summarized in Table 2.
|
View this table: [in a new window] |
TABLE 2. In vitro susceptibilities of 207 anaerobic isolates associated with bloodstream infections in 2006 at NTUH
|
![]() View larger version (16K): [in a new window] |
FIG. 3. Differences in rates of nonsusceptibility to imipenem and meropenem for selected clinical isolates of anaerobes isolated between 2002 and 2006 at National Taiwan University Hospital. The numbers of isolates are shown above the bars.
|
Of B. fragilis isolates, only 63% were susceptible to ampicillin-sulbactam, compared with 95% to piperacillin-tazobactam. Three percent of B. fragilis isolates showed resistance to chloramphenicol, which was not noted in the 2002 study by our group (28).
The other Bacteroides species included in this study were B. vulgatus (13 isolates), B. uniformis (eight isolates), and B. caccae (two isolates). The rates of resistance to clindamycin (39%), cefmetazole (30%), ampicillin-sulbactam (30%), and amoxicillin-clavulanic acid (22%) were also high. One isolate of B. vulgatus was resistant to piperacillin-tazobactam, doripenem, imipenem, ertapenem, and meropenem.
All 27 isolates of Fusobacterium species were susceptible to metronidazole, piperacillin-tazobactam, cefmetazole, doripenem, and ertapenem. One isolate of Fusobacterium varium was resistant to both imipenem and meropenem but susceptible to ertapenem (MIC, 0.12 µg/ml). Of the 16 isolates of Prevotella species identified,
25% were not susceptible to any fluoroquinolones or clindamycin, and 94% were resistant to penicillin.
Isolates of Clostridium species had various levels of nonsusceptibility to the antimicrobial agents tested. All isolates were susceptible to piperacillin-tazobactam, meropenem, and doripenem. Of the Peptostreptococcus species isolates, >20% were not susceptible to any fluoroquinolones.
|
|
|---|
The rates of resistance of B. fragilis isolates to imipenem or meropenem found in the present study were higher than those reported previously (24, 25). In Japan, imipenem-resistant Bacteroides species in the range of 2% to 4% 15 years ago were reported, and these resistance rates have not changed recently (29). In a U.S. national survey from 1997 to 2004, the overall rates of resistance to imipenem and meropenem were 0.5% and 1.0%, respectively (24). In a European study published in 2006, the rates of resistance of B. fragilis isolates to imipenem (0.8%) and meropenem (1.3%) were similar to those found in the survey in the United States (25).
In addition, a B. fragilis isolate in this study exhibited a higher MIC90 (4 µg/ml) to ertapenem than those reported in the United States in 2004 (1 to 2 µg/ml) (32). Furthermore, one isolate of B. fragilis exhibited high-level resistance to all four carbapenems (>32 µg/ml) tested in this study. Such a high rate of resistance to carbapenems in Taiwan is alarming, since empirical use of an antimicrobial agent to which B. fragilis is nonsusceptible may lead to an adverse outcome (21).
The analysis of data for nosocomial isolates and all clinical isolates of B. fragilis in this study revealed a continually decreasing trend in rates of susceptibility of nosocomial isolates of B. fragilis to cefmetazole, from 72% in 1999 to 14% in 2006, while the trend for clinical isolates of B. fragilis and other Bacteroides species remained stationary. This cephamycin is widely used in the treatment of intra-abdominal infections (IAIs) and mixed infections in Taiwan and at our hospital (data not shown). Whether this widespread use induces acquired resistance during treatment or the increased resistance trend is leading to adverse clinical outcomes needs further study.
In the B. thetaiotaomicron and other Bacteroides species isolates collected during 2006, the rates of resistance to carbapenems were rather low. Compared with the results of our previous study in 2002, dramatic increases in the rates of nonsusceptibility of B. thetaiotaomicron isolates to cefmetazole (80% in 2006 versus 17% in 2002), imipenem (8% versus 0%), and clindamycin (70% versus 43%) were noted (28). A U.S. national survey conducted during 1997 to 2004 found that the rate of resistance of B. thetaiotaomicron isolates to ampicillin-sulbactam (2.1%) was much lower than the rate of resistance of the more-recent isolates in our study in 2006 (33%), while the rates of resistance for ertapenem (0.4%), imipenem (0.2%), meropenem (0.1%), piperacillin-tazobactam (0.4%), and clindamycin (33.3%) were comparable (24).
Resistance to carbapenems has not been previously documented among the 27 reported strains of Fusobacterium species isolated (1, 12, 14, 22, 28). In this survey, however, we identified an isolate of F. varium which was resistant to both imipenem (MIC, 32 µg/ml) and meropenem (MIC, 32 µg/ml). However, this isolate was susceptible to ertapenem and doripenem with low MIC levels.
Our results showed that the rate of resistance to penicillin among Prevotella species isolates increased from 62% to 94% from 2002 to 2006. Prevotella species resistant to carbapenem have not been previously reported (1, 12, 14, 28). In this study, one isolate of Prevotella species was resistant to imipenem (MIC, 16 µg/ml) but susceptible to ertapenem (MIC, 4 µg/ml), doripenem (MIC, 2 µg/ml), and meropenem (MIC, 2 µg/ml). In this study, isolates of Prevotella and Fusobacterium species that were susceptible to piperacillin-tazobactam but resistant to one or two carbapenems were noted, and this rare susceptibility pattern and the mechanisms that conferred resistance need further investigation.
In a previous study from Taiwan in 2002, all isolates of Clostridium perfringens were susceptible to all antimicrobials tested (28). In the present study, however, three isolates of C. perfringens were resistant to metronidazole. A previous study in New Zealand from 1999 to 2003 (23) found no resistance to metronidazole among all the Clostridium species identified. By contrast, the rate of resistance was as high as 9.4% in a series from Thailand (27). Emerging resistance to metronidazole in Southeast Asia, including Taiwan, is cause for concern (27, 28). In the present 2006 survey, three isolates of Clostridium species other than C. perfringens showed a concurrent decrease in susceptibility to cefmetazole, clindamycin, levofloxacin, and imipenem. Whether a transposable multidrug resistance gene was present in these species warrants further investigation.
Consistent with the results of previous studies in Taiwan and other countries (1, 12, 14, 23, 27, 28), metronidazole, chloramphenicol, piperacillin-tazobactam, imipenem, ertapenem, and meropenem remained effective against the common isolates of anaerobic pathogens, including gram-positive and gram-negative anaerobes. However, comparison with the findings of our previous study in 2002 revealed trends of emerging resistance (28). Except for a U.S. national survey of Bacteroides species isolates from 1997 to 2004 and the Belgian third multicenter survey (34) published in 2007, there have been no recent large-scale studies of anaerobic resistance in recent years. Surveys in Belgium, Brazil, England, Estonia, Korea, Spain, Turkey, the United States, and Europe did not report such a high prevalence of resistance to carbapenems (6, 9, 10, 13, 16, 19, 22, 30-32, 34). Newer fluoroquinolones, such as moxifloxacin, exerted antimicrobial activities of 75% to 93% against different species of anaerobic isolates associated with bacteremia in this study. Thus, a fluoroquinolone might not be a suitable substitute for carbapenems. Behra-Miellet et al. reported that linezolid is likely to be active against most anaerobes (5). It may be a potential candidate for antimicrobial therapy if resistance to carbapenems develops.
Ampicillin-sulbactam is a recommended treatment option for IAIs of mild-to-moderate severity in the current guidelines for treatment and the proposed Asian consensus statement on the treatment of IAIs (11, 26). However, the high rates of resistance of Bacteroides spp. to ampicillin-sulbactam found in this study suggest that ampicillin-sulbactam is not the drug of choice for IAIs in Taiwan. In addition, the possibility of anaerobes with less susceptibility to carbapenems should be taken into consideration when clinical failure occurs after carbapenem treatment for IAIs, particularly those associated with anaerobic bacteremia.
In conclusion, our findings of high rates of resistance to carbapenems among Bacteroides species, Fusobacterium species, and Prevotella species isolates, as well as of resistance to metronidazole among C. perfringens isolates, stress the need for continuous monitoring of susceptibility trends. The extent of the association between the use of antibiotics against a resistant isolate and adverse clinical outcome also requires further investigation. The emerging resistance to carbapenems appears to be mounting a challenge to their role in empirical use in complicated mixed infections in Taiwan.
Published ahead of print on 14 July 2008. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»