- Email: [email protected]
Clinical characteristics and antimicrobial susceptibilities of anaerobic bacteremia in an acute care hospital
Accepted Manuscript Clinical characteristics and antimicrobial susceptibilities of anaerobic bacteremia in an acute care hospital Thean Yen Tan, Lily Siew Yong Ng, Lee Ling Kwang, Suma Rao, Li Ching Eng PII:
S1075-9964(16)30155-X
DOI:
10.1016/j.anaerobe.2016.11.009
Reference:
YANAE 1645
To appear in:
Anaerobe
Received Date: 7 May 2014 Revised Date:
17 November 2016
Accepted Date: 24 November 2016
Please cite this article as: Tan TY, Ng LSY, Kwang LL, Rao S, Eng LC, Clinical characteristics and antimicrobial susceptibilities of anaerobic bacteremia in an acute care hospital, Anaerobe (2016), doi: 10.1016/j.anaerobe.2016.11.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Clinical characteristics and antimicrobial susceptibilities of anaerobic bacteremia in an acute care hospital Thean Yen Tan*1 Lily Siew Yong Ng1
Suma Rao2 Li Ching Eng1
Short title:
M AN U
Clinical features of anaerobic bacteraemia
Affiliation: 1
Department of Laboratory Medicine, Changi General Hospital, 2 Simei Street 3,
TE D
Singapore, 529889
Department of Medicine, Changi General Hospital, 2 Simei Street 3, Singapore, 529889
EP
2
SC
RI PT
Lee Ling Kwang1
Corresponding author:
AC C
Thean Yen Tan
Department of Laboratory Medicine Changi General Hospital 2 Simei Street 3 Singapore 529889 Tel: 065-68504934 Fax: 065-64269507 Page -1-
ACCEPTED MANUSCRIPT Email:
[email protected]
RI PT
Article word count:
SC
2442
Conflict of interest declaration
M AN U
None to declare
Acknowledgements
AC C
EP
TE D
This study was funded by an internal grant from Changi General Hospital.
Page -2-
ACCEPTED MANUSCRIPT SUMMARY PAGE Introduction This study investigated the clinical features of anaerobic bacteraemia in an acute-care hospital, and evaluated the antimicrobial susceptibility of these isolates to commonly
RI PT
available antibiotics. Materials & Methods
Microbiological and epidemiological data from 2009-2011were extracted from the
SC
laboratory information system and electronic medical records. One hundred and eleven unique patient episodes consisting of 116 anaerobic isolates were selected for clinical
M AN U
review and antibiotic susceptibility testing. Susceptibilities to amoxicillin-clavulanate, clindamycin, imipenem, metronidazole, moxifloxacin, penicillin and piperacillin-tazobactam were performed using Etest strips with categorical interpretations according to current CLSI breakpoints. Metronidazole-resistant and carbapenem-resistant anaerobic Gram-
TE D
negative bacilli were screened for the nim and cfiA genes. Clinical data was obtained retrospectively from electronic medical records. Results
EP
During the 3 year period, Bacteroides fragilis group (41%), Clostridium species (14%), Propionibacterium species (9%) and Fusobacterium species (6%) were the most
AC C
commonly isolated anaerobes. Patients with anaerobic bacteraemia that were included in the study were predominantly above 60 years of age, with community-acquired infections. The most commonly used empiric antibiotic therapies were beta-lactam/beta-lactamase inhibitor combinations (44%) and metronidazole (10%). The crude mortality was 25%, and appropriate initial antibiotic therapy was not significantly associated with improved survival. Intra-abdominal infections (39%) and soft-tissue infections (33%) accounted for nearly three-quarters of all bacteraemia.
Page -3-
ACCEPTED MANUSCRIPT Antibiotics with the best anaerobic activity were imipenem, piperacillin-tazobactam, amoxicillin-clavulanate and metronidazole, with in-vitro susceptibility rates of 95%, 95%, 94% and 92% respectively. Susceptibilities to penicillin (31%), clindamycin (60%) and moxifloxacin (84%) were more variable. Two multidrug-resistant isolates of Bacteroides
resistant Fusobacterium species were negative for cfiA genes. Conclusion
RI PT
species were positive for nim and cfiA genes respectively, while another two imipenem-
SC
This study demonstrated that anaerobic bacteraemia in our patient population was
predominantly associated with intra-abdominal and soft-tissue infections. Overall antibiotic
M AN U
resistance was high for penicillin and clindamycin, and the presence of emerging resistance to carbapenems and metronidazole warrants further monitoring.
TE D
Key words:
anaerobes; bacteremia; Bacteroides; antibiotic resistance
AC C
305
EP
Synopsis word count:
Page -4-
ACCEPTED MANUSCRIPT ARTICLE
1.0 INTRODUCTION Anaerobic bacteraemia continues to account for a small, but significant, proportion of
RI PT
bacteraemia (around 7-10% of all bacteraemic episodes) [1, 2]. Common sources of infection include the gastrointestinal tract, abscesses, gynaecologic sources and soft tissue infections [3], while reported risk factors for anaerobic bacteraemia include
SC
gastrointestinal surgery and haematologic malignancy [4].
Antibiotic therapy for anaerobes is usually empiric. Anaerobic susceptibility testing is
M AN U
not performed routinely, as it is expensive, methodologically demanding and time consuming. Based on published reports, the prevalence of antibiotic resistance in anaerobes is increasing [5, 6]. Clinical data suggest that antibiotic resistance impacts both microbiological cure and patient mortality for patients with anaerobic bacteraemia [7].
TE D
There is little susceptibility data for anaerobic infections in Singapore, and no data exists for the clinical characteristics of patients with anaerobic bacteraemia. This study was performed in an 800-bed hospital that provides medical, surgical,
EP
orthopaedic and geriatric medical care to the eastern region of Singapore. The aims of the study were to determine clinical data for patients with anaerobic bacteraemia over a three-
AC C
year period, and to determine antibiotic susceptibilities for these anaerobic isolates.
2.0 METHODS
2.1. Retrospective microbiological data extraction Bacterial information and epidemiological data were extracted from the laboratory information system for the period 2009-2011. Duplicate isolates from the same patient (defined as similar isolates within a 30 day period) were excluded from analysis. During Page -5-
ACCEPTED MANUSCRIPT the study period, routine aerobic and anaerobic blood cultures were performed using BACTEC™ Plus Aerobic/F and Plus Anaerobic/F vials (BD, USA), incubated for five days in a continuous monitoring blood culture system (Bactec 9000 series, BD, USA). Anaerobic vials with positive growth indices were sub-cultured on to trypticase soy agar
RI PT
with 5% sheep blood (BD, USA), MacConkey agar (Oxoid, UK), Chocolate agar (BD, USA) and CDC anaerobe agar with 5% sheep blood (BD, USA). Media were routinely incubated for 2 days, with additional incubation for up to 4 days if a slow-growing anaerobe was
SC
suspected. Gram-stain and aero-tolerance testing were performed for suspected
anaerobic isolates, with speciation by Vitek® ANC identification cards tested on the Vitek®
M AN U
Compact system (bioMérieux, France), supplemented with API 20A (bioMérieux, France) where necessary.
Detailed clinical and bacteriological information was retrospectively obtained for 111 unique patient episodes. These cases were randomly chosen to represent the most
TE D
common clinically significant anaerobic isolates from blood cultures. Propionibacterium species, anaerobic Gram-positive cocci, and nonsporulating anaerobic Gram-positive rods were excluded from detailed analysis, as previous studies have demonstrated that these
EP
genera are much less likely to be clinically significant [3, 8, 9]. 2.2. Antimicrobial susceptibility testing
AC C
For the 111 unique patient episodes, 116 anaerobic isolates were retrieved from cryo-storage for antibiotic susceptibility testing. Bacterial identification was re-confirmed by MALDI-TOF testing (Vitek MS database 2.0, bioMérieux, France) prior to susceptibility testing. Susceptibilities to amoxicillin-clavulanate, clindamycin, imipenem, metronidazole, moxifloxacin, penicillin and piperacillin-tazobactam were performed using Etest antimicrobial strips on Brucella agar with 5% sheep blood, vitamin K and hemin (BD, USA). Overnight pre-reduction of testing media was performed prior to testing, and following inoculation, plates were incubated in an anaerobic workstation system at 35°C Page -6-
ACCEPTED MANUSCRIPT (Forma Anaerobic System, ThermoFisher Scientific, USA) in an anaerobic atmosphere comprising 5% hydrogen, 5-10% carbon dioxide and 85-90% nitrogen. Concurrent quality control (QC) testing was performed using Bacteriodes fragilis ATCC 25285 according to manufacturer’s guidelines, and QC values were within acceptable limits. Etest results were
RI PT
read after 48-72 hours incubation, with all except one Fusobacterium isolate achieving satisfactory growth endpoints at 48 hours. Antibiotic susceptibilities were interpreted according to existing CLSI breakpoints [10]. Metronidazole-resistant and carbapenem-
SC
resistant anaerobic Gram-negative bacilli were screened for nim and cfiA genes encoding 5-nitroimidazole resistance and carbapenem resistance respectively, by conventional PCR
M AN U
[11, 12]. 2.3. Clinical data extraction and analysis
Clinical data collection for each patient episode of anaerobic bacteraemia was obtained retrospectively from electronic medical records. Data collected included the
TE D
admitting and final discharge diagnosis, empiric antibiotic prescription, results of clinical investigations and inpatient mortality. The presence of polymicrobial bacteraemia was defined as the presence of other bacterial isolates from blood cultures collected on the
EP
same day as the primary anaerobic isolate. The Charlson co-morbidity scores were calculated for each patient using an electronic spreadsheet [13], while statistical analysis
AC C
and odds ratios were calculated using OpenEpi (online open source statistical software) [14] supplemented with SPSS Statistics (IBM, USA).
Page -7-
ACCEPTED MANUSCRIPT 3.0 RESULTS 3.1. Anaerobic bacteraemia over the three year period During the 3 year period, 250 obligate anaerobes were recovered from blood cultures, representing 4.1% of all positive blood cultures. The most frequent anaerobic
RI PT
isolates were the Bacteroides fragilis group (n=102; 41%), Clostridium species (n=34; 14%), Propioniobacterium species (n=23; 9%) and Fusobacterium species (n=16; 6%). The remainder of the isolates included Prevotella spp., Veillonella spp., Eubacterium spp.,
SC
Actinomyces spp., Peptostreptococcus spp., Lactobacillus spp., Parvimonas and
Porphyromonas spp. Twenty three isolates could not be identified by phenotypic tests,
M AN U
consisting predominantly of anaerobic Gram-positive bacilli.
3.2. Detailed clinical information on subset of 111 patients For the detailed subset of 111 study patients with anaerobic bacteraemia, the average age was 73 years (range: 21-99 years). Patients older than 60 years of age
TE D
accounted for 84% of cases. Most patients (n=89, 80%) were admitted from home, with a futher 21 (19%) patients admitted from a long-term residential care facility. The majority (87%) of anaerobic bacteraemia were reported from blood cultures taken within two days
EP
of hospital admission. The majority of patients were reported as independently mobile (n=70, 63%), while 24 (22%) patients were reported to be bed-bound. The average
AC C
Charlson co-morbidity score was 2.6 (range 0-12), while the average age-adjusted score was 5.1 (range 0-15). Information on antibiotic therapy was available for 97 patient episodes. The most commonly used empiric antibiotic therapy were beta-lactam/betalactamase inhibitor combinations (amoxicillin/clavulanic acid or piperacillin/tazobactam) (n=49, 44%), metronidazole (n=11, 10%) and carbapenems (n=8, 8%). Eighteen patients (16%) were on antibiotics with no activity against anaerobic bacteria, and two patients had no antibiotic coverage. The majority of patients were started on initial appropriate antibiotic therapy (n=72, 65%). The crude mortality was 25%, with a higher mortality risk for patients Page -8-
ACCEPTED MANUSCRIPT with infections due to an extra-abdominal source (odds ratio 0.3, 95%CI 0.12-0.92), increasing age (increased risk of 3.7% per year, 95%CI 0.2- 7.3%) and increasing adjusted Charlson co-morbidity score (odds ratio 1.2, 95%CI 1.0-1.3) on univariate analysis. However, on multivariate analysis, the increased risk of mortality remained
RI PT
statistically significant only for the source of infection. Infection with either Bacteroides spp. or Clostridium spp. was not significantly associated with a higher mortality, while
appropriate initial antibiotic therapy was not associated with improved survival (odds ratio
SC
1.1, 95%CI 0.4-3.3). Intra-abdominal infections were the predominant source of infection (n=43, 39%), followed by soft-tissue infections (n=36, 32%) and lower respiratory tract
M AN U
infections (n=5, 5%). No obvious source of infection was identified for 18 patients (16%), 9 of which were judged to be likely contaminants due to the lack of clinical signs of infection and/or the presence of other skin commensals from corresponding blood cultures (Bacteroides spp. n=3, Clostridium spp. n=2, Prevotella spp. n=2, Veillonella spp. n=2).
TE D
The most common causes of intra-abdominal infections were hepatobiliary sepsis, appendicitis, perforated viscus and underlying colonic cancer. For soft-tissue infections, the most common sources of infection were decubitus pressure ulcers, diabetic foot
EP
infection and cellulitis with associated gangrene (Table 1). One case of Lemierre’s syndrome was reported, involving Prevotella melaninogenica and ‘Anginosus group’
AC C
streptococci. Fifty-seven percent of anaerobic bacteraemias were polymicrobial, with 19% of cases having three or more separate organisms in the same culture set. For episodes of polymicrobial bacteraemia, aerobic bacteria were also isolated in 70% of cases. The most common other bacterial pathogens isolated from polymicrobial bacteraemia were other anaerobes (n=29, 33% of all polymicrobial cases), Enterobacteriaceae (n=25, 29%), coagulase negative staphylococci (n=10, 11%), Enterococcus species (n= 6, 7%) and ‘Anginosus group’ streptococci (n=6, 7%). Patients with soft-tissue infections were twice as
Page -9-
ACCEPTED MANUSCRIPT likely to have polymicrobial bacteraemia (odds ratio 2.2, 95%CI 0.95-5.3) compared to patients with infections from all other sources. 3.3. Antimicrobial susceptibilities of anaerobic isolates from 111 patients Antibiotic susceptibilities were determined for 116 anaerobic isolates, consisting of
RI PT
Bacteroides spp. (n=68), Clostridium spp. (n=26), Fusobacterium spp. (n=9) and other species (n=13). Antibiotics with the best activity against anaerobic bacteria were imipenem, piperacillin-tazobactam, metronidazole, amoxicillin-clavulanate and
SC
moxifloxacin, with in-vitro susceptibility rates of 95%, 95%, 92%, 94% and 84%
respectively. Susceptibilities to penicillin (31%) and clindamycin (60%) were much lower.
M AN U
Bacteroides spp. were uniformly resistant to penicillin, with lower susceptibility to clindamycin (54%) and moxifloxacin (85%) (Table 2). Moxifloxacin resistance was higher in B. ovatus and B. uniformis, while Parabacteroides distasonis was more likely to be resistant to the beta-lactam/beta-lactamase inhibitors and clindamycin. One strain of P.
TE D
distasonis demonstrated high-level resistance to metronidazole (MIC>256 mg/L) and this isolate was nim-positive by PCR. A multidrug-resistant isolate of B. fragilis was also present with high-level resistance to all tested antibiotics except for metronidazole, and
EP
this isolate was positive for the cfiA gene (Table 3). Clostridium spp. demonstrated lowest susceptibility to clindamycin (65%) and
AC C
variable susceptibility to metronidazole (81%), moxifloxacin (77%) and penicillin (81%), but were almost uniformly susceptible to the beta-lactam/beta-lactamase inhibitors (96-100%). Fusobacterium spp. remained highly susceptible to metronidazole (100%), with lower susceptibilities to penicillin (89%) and the beta-lactam/beta-lactamase inhibitors (89%). Although the overall numbers tested were small, the lowest susceptibility to imipenem was seen in Fusobacterium sp. (78%), with two unusually multidrug-resistant isolates of Fusobacterium detected. One strain of F. mortiferum demonstrated high-level resistance to all tested beta-lactams but was susceptible to metronidazole, clindamycin and Page -10-
ACCEPTED MANUSCRIPT moxifloxacin, The other F. varium was resistant to imipenem, moxifloxacin and clindamycin. The cfiA gene was not detected for either of these two strains.
AC C
EP
TE D
M AN U
SC
RI PT
MIC distributions for all tested antibiotics are listed in Table 4.
Page -11-
ACCEPTED MANUSCRIPT 4.0 DISCUSSION This study reports on the clinical features of anaerobic bacteraemia, with associated antibiotic susceptibilities. There are some limitations to our study. We studied a specific subset of bacteraemic
RI PT
episodes by specifically excluding anaerobe species which were most likely to be contaminants from the analysis. Secondly, the spectrum of clinical infections may not be representative of infections seen in other healthcare institutions, which may serve different
SC
patient populations or with additional clinical specialties not present in our institution. Bacteroides spp. accounted for nearly half of all bacteraemia, followed by
M AN U
Clostridium spp. and propionibacteria. In our patient population, patients with anaerobic bacteraemia were mostly elderly patients admitted from the community, with several existing co-morbidities. Although intra-abdominal infections were the most common source of infection, in contrast with other published studies, nearly one-third of anaerobic
TE D
bacteraemia was associated with soft-tissue infections. In a review of 81 patients with anaerobic bacteraemia from Finland from the 1990s, soft-tissue infections were documented in 23% of patients [3], compared to 8% of patients from a more recent
EP
publication from New Zealand [15]. Of note, the aetiology of bacteraemic infections in our study is likely to differ from larger medical centres offering oncology, organ transplant and
AC C
gynaecologic services, which are absent in our hospital. Studies from hospitals with comprehensive medical services suggest that patients with haematological disorders have a higher underlying rate of anaerobic bacteraemia compared with other patient groups, and in one such study, the upper respiratory system constituted the second most common portal of infection [4]. The inpatient mortality of patients in our series was 25%, which is largely consistent with the data from other reported studies [16]. Our study did not demonstrate a reduction in mortality for patients receiving appropriate initial antibiotic therapy. Page -12-
ACCEPTED MANUSCRIPT The carbapenems, beta-lactam/beta-lactamase inhibitors and metronidazole demonstrated excellent in-vitro activity against most anaerobes. Conversely, resistance to penicillin and clindamycin were high, which suggests that these antibiotics are not suitable for empiric treatment of anaerobic bacteraemia. A prospective study in the United States
RI PT
demonstrated increasing clindamycin and moxifloxacin resistance over a 3 year period, with differentially higher rates of resistance in non-Bacteroides fragilis isolates [17]. In Taiwan, higher rates of carbapenem resistance (up to 12% in B. fragilis) were reported
SC
over a ten-year period, but conversely, lower rates of resistance to moxifloxacin [6]. Carbapenem resistance in Bacteroides species is primarily caused by the presence of a
M AN U
metallo-β-lactamase (cfiA) with gene expression mediated by a corresponding insertion sequence [18]. The mechanisms of carbapenem resistance in other Gram-negative anaerobes remain poorly elucidated. Metronidazole resistance in anaerobes is mediated by multiple mechanisms, which include resistance conferred by a family (nim) of
TE D
transmissible genes [19]. A recent review noted that metronidazole resistance remains very low (<1%) in most geographic regions [20]. To our knowledge, this is the first study to screen for and detect the presence of cfiA and nim genes in organisms isolated in
EP
Singapore, emphasizing the importance of further surveillance to determine the prevalence of these transmissible genes in anaerobes from the Asia-Pacific region. This study also
AC C
provides the first information on antibiotic susceptibilities of anaerobic isolates associated with bacteraemic infections in Singapore. In summary, this study demonstrated that anaerobic bacteraemia in our patient population was predominantly associated with intra-abdominal and soft-tissue infections, with a relatively high mortality rate. Overall antibiotic resistance was high for penicillin and clindamycin and, to a lesser extent, moxifloxacin. The presence of emerging resistance to carbapenems and metronidazole warrants further monitoring.
Page -13-
ACCEPTED MANUSCRIPT REFERENCES [1]
Fenner L, Widmer AF, Straub C, Frei R. Is the incidence of anaerobic bacteremia
decreasing? Analysis of 114,000 blood cultures over a ten-year period. J Clin Microbiol 2008;46:2432-4. Lassmann B, Gustafson DR, Wood CM, Rosenblatt JE. Reemergence of anaerobic
bacteremia. Clin Infect Dis 2007;44:895-900. [3]
Salonen JH, Eerola E, Meurman O. Clinical significance and outcome of anaerobic
SC
bacteremia. Clin Infect Dis 1998;26:1413-7. [4]
RI PT
[2]
Blairon L, De Gheldre Y, Delaere B, Sonet A, Bosly A, Glupczynski Y. A 62-month
Microbiol Infect 2006;12:527-32. [5]
M AN U
retrospective epidemiological survey of anaerobic bacteraemia in a university hospital. Clin
Goldstein EJC, Citron DM. Resistance Trends in Antimicrobial Susceptibility of
Anaerobic Bacteria, Part I. Clin Microbiol Newsletter 2011;33:1-8. Liu CY, Huang YT, Liao CH, Yen LC, Lin HY, Hsueh PR. Increasing trends in
TE D
[6]
antimicrobial resistance among clinically important anaerobes and Bacteroides fragilis isolates causing nosocomial infections: emerging resistance to carbapenems. Antimicrob
[7]
EP
Agents Chemother 2008;52:3161-8.
Nguyen MH, Yu VL, Morris AJ, McDermott L, Wagener MW, Harrell L, et al.
AC C
Antimicrobial resistance and clinical outcome of Bacteroides bacteremia: findings of a multicenter prospective observational trial. Clin Infect Dis 2000;30:870-6. [8]
Brook I, Frazier EH. Significant recovery of nonsporulating anaerobic rods from
clinical specimens. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 1993;16:476-80. [9]
Park HJ, Na S, Park SY, Moon SM, Cho OH, Park KH, et al. Clinical significance of
Propionibacterium acnes recovered from blood cultures: analysis of 524 episodes. J Clin Microbiol 2011;49:1598-601. Page -14-
ACCEPTED MANUSCRIPT [10] CLSI. Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria; Approved Standard - Eighth Edition. CLSI document M11-A8. Wayne, PA: Clinical and Laboratory Standards Institute; 2012. [11] Trinh S, Reysset G. Detection by PCR of the nim genes encoding 5-nitroimidazole
RI PT
resistance in Bacteroides spp. J Clin Microbiol 1996;34:2078-84. [12] Thompson JS, Malamy MH. Sequencing the gene for an imipenem-cefoxitin-
hydrolyzing enzyme (CfiA) from Bacteroides fragilis TAL2480 reveals strong similarity
SC
between CfiA and Bacillus cereus beta-lactamase II. J Bacteriol 1990;172:2584-93. [13] Hall WH, Ramachandran R, Narayan S, Jani AB, Vijayakumar S. An electronic
M AN U
application for rapidly calculating Charlson comorbidity score. BMC Cancer 2004;4:94. [14] Dean AG SK, Soe MM. Version 3.01. OpenEpi: Open Source Epidemiologic Statistics for Public Health, . 2013.
[15] Muttaiyah S, Paviour S, Buckwell L, Roberts SA. Anaerobic bacteraemia in patients
journal 2007;120:U2809.
TE D
admitted to Auckland City Hospital: its clinical significance. The New Zealand medical
[16] Brook I. The role of anaerobic bacteria in bacteremia. Anaerobe 2010;16:183-9.
EP
[17] Snydman DR, Jacobus NV, McDermott LA, Golan Y, Hecht DW, Goldstein EJ, et al. Lessons learned from the anaerobe survey: historical perspective and review of the most
AC C
recent data (2005-2007). Clin Infect Dis 2010;50 Suppl 1:S26-33. [18] Ang L, Brenwald NP, Walker RM, Andrews J, Fraise A. Carbapenem resistance in Bacteroides fragilis. J Antimicrob Chemother 2007;59:1042-4. [19] Lofmark S, Edlund C, Nord CE. Metronidazole is still the drug of choice for treatment of anaerobic infections. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2010;50 Suppl 1:S16-23. [20] Boyanova L, Kolarov R, Mitov I. Recent evolution of antibiotic resistance in the anaerobes as compared to previous decades. Anaerobe 2015;31:4-10. Page -15-
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Page -16-
ACCEPTED MANUSCRIPT
Table 1: Sources of anaerobic bacteraemia
36 26 3 3 4(1) 5
Pyelonephritis
12
8 6 5 2 2
1 3 1 2
RI PT
Soft-tissue Sacral sore(4) Diabetic foot infection Gangrene & cellulitis Others(1) Lower respiratory tract infection Aspiration pneumonia & infected sacral sores Empyema Pneumonia Aspiration pneumonia
29
1
SC
43 12 8 6 3 2 1 1 10
No. of bacterial isolates FusoPorphybacterium romonas spp. spp. 2 0
M AN U
Intra-abdominal Hepatobiliary sepsis Appendicitis(1) Perforated viscus(1) Colonic cancer(1) Ischemic bowel C. difficile colitis Unspecific Colitis Others
Bacteroides Clostridium spp. spp.
Prevotella spp.
Veillonella spp.
2
1
1
1
1
1 1 3
1
23
7
6
1
2
2
17 1 3 2 4
6 1
4
1
1 1
1
6
TE D
No. of patients
EP
Type of infection
AC C
1
1 1 2 1
1 1 1 1
2
2
2 1
1
Page -17-
1
ACCEPTED MANUSCRIPT
1 2
Total No. = number (n)
2 2 1 1 1 18
1
1 2
1 1 8 69
111
5 27
RI PT
Bacteroides Clostridium spp. spp.
No. of bacterial isolates FusoPorphybacterium romonas spp. spp.
SC
Genito-urinary Lower urinary tract infection Joint infection Lemierre's syndrome Osteomyelitis No source identified
No. of patients
M AN U
Type of infection
1 10
Prevotella spp.
1
1
2 7
numbers in superscript indicate the number of cases with more than one anaerobe isolated for the bacteraemic episode
AC C
EP
TE D
3
Page -18-
Veillonella spp.
2 5
ACCEPTED MANUSCRIPT
Table 2: Antibiotic susceptibilities of selected anaerobes
TE D
Page -19-
Imipenem S (%) 67 (99%) 39 (98%) 12 (100%) 6 (100%) 4 (100%) 3 (100%) 2 (100%) 1 (100%) 23 (89%) 10 (100%) 4 (100%) 3 (100%) 3 (100%) 1 (50%) 0 (0%) 1 (100%) 0 (0%) 1 (100%) 7 (78%) 3 (75%) 1 (50%) 2 (100%) 1 (100%) 7 (100%)
RI PT
Clindamycin S (%) 37 (54%) 25 (63%) 6 (50%) 1 (17%) 3 (75%) 1 (33%) 1 (50%) 0 (0%) 17 (65%) 9 (90%) 3 (75%) 0 (0%) 2 (67%) 2 (100%) 0 (0%) 0 (0%) 1 (100%) 0 (0%) 6 (67%) 1 (25%) 2 (100%) 2 (100%) 1 (100%) 4 (57%)
SC
B. fragilis B. thetaiotaomicron P. distasonis B. vulgatus B. ovatus B. uniformis B. stercoris Clostridium species C. perfringens C. bifermentans Clostridium species C. ramosum C. clostridioforme C. difficile C. tertium C. glycolium C. histolyticum Fusobacterium species F. varium F. mortiferum F. necrophorum F. nucleatum Prevotella species
S (%) 1 (2%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 1 (33%) 0 (0%) 0 (0%) 21 (81%) 10 (100%) 1 (100%) 0 (0%) 1 (100%) 1 (100%) 8 (89%) 4 (100%) 1 (50%) 2 (100%) 8 (89%) 4 (100%) 1 (50%) 2 (100%) 1 (100%) 3 (43%)
EP
Bacteroides species
n 68 40 12 6 4 3 2 1 26 10 4 3 3 2 1 1 1 1 9 4 2 2 1 7
Augmentin S (%) 63 (93%) 37 (93%) 12 (100%) 4 (67%) 4 (100%) 3 (100%) 2 (100%) 1 (100%) 25 (96%) 10 (100%) 4 (100%) 2 (67%) 3 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 1 (100%) 8 (89%) 4 (100%) 1 (50%) 2 (100%) 1 (100%) 7 (100%)
M AN U
Penicillin
AC C
1
Moxifloxacin S (%) 58 (85%) 36 (90%) 10 (83%) 6 (100%) 4 (100%) 0 (0%) 1 (50%) 1 (100%) 20 (77%) 10 (100%) 4 (100%) 2 (67%) 1 (33%) 0 (0%) 0 (0%) 1 (100%) 1 (100%) 1 (100%) 7 (78%) 2 (50%) 2 (100%) 2 (100%) 1 (100%) 6 (86%)
Piperacillintazobactam S (%) 64 (94%) 39 (98%) 12 (100%) 4 (67%) 4 (100%) 3 (100%) 1 (50%) 1 (100%) 26 (100%) 10 (100%) 4 (100%) 3 (100%) 3 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 1 (100%) 8 (89%) 4 (100%) 1 (50%) 2 (100%) 1 (100%) 7 (100%)
Metronidazole S (%) 67 (99%) 40 (100%) 12 (100%) 5 (83%) 4 (100%) 3 (100%) 2 (100%) 1 (100%) 21 (81%) 10 (100%) 3 (75%) 1 (33%) 1 (33%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 1 (100%) 9 (100%) 4 (100%) 2 (100%) 2 (100%) 1 (100%) 5 (71%)
ACCEPTED MANUSCRIPT
S (%) 1 (50%) 1 (50%) 0 (0%) 1 (100%) 0 (0%) 3 (60%) 0 (0%)
All Gram-negative anaerobes
90
15 (17%)
84 (93%)
52 (58%)
87 (97%)
77 (86%)
84 (93%)
86 (96%)
26
21 (81%)
25 (96%)
17 (65%)
23 (89%)
20 (77%)
26 (100%)
21 (81%)
1
All Gram-positive anaerobes n = number of isolates
2
S = susceptible
Imipenem S (%) 2 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 5 (100%) 1 (100%)
Moxifloxacin S (%) 2 (100%) 2 (100%) 0 (0%) 1 (100%) 1 (100%) 5 (100%) 1 (100%)
Piperacillintazobactam S (%) 2 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 4 (80%) 1 (100%)
Metronidazole S (%) 2 (100%) 2 (100%) 0 (0%) 0 (0%) 1 (100%) 5 (100%) 0 (0%)
SC
M AN U
TE D
4
EP
3
AC C
Prevotella species P. melaninogenica P. bivia P. intermedia P. oralis
Clindamycin S (%) 0 (0%) 1 (50%) 1 (100%) 1 (100%) 1 (100%) 5 (100%) 0 (0%)
RI PT
Veillonella spp. Porphyromonas asaccharolytica
n 2 2 1 1 1 5 1
Augmentin S (%) 2 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 5 (100%) 1 (100%)
Penicillin
Page -20-
ACCEPTED MANUSCRIPT
Table 3: Minimum inhibitory concentration and resistance gene testing for metronidazole and imipenem resistant
2
anaerobes
RI PT
1
nim gene
cfiA gene
16
1
0.25
≥256
≥256
Detected
ND
>256
≥256
≥64
≥64
≥256
0.5
ND
Detected
≥256
>256
≤0.01
≥64
0.125
≥256
0.125
ND
ND
0.25
1
8
≥64
8
2
0.5
ND
ND
2
≤0.01
0.06
≥64
≤0.01
64
ND
ND
≤0.016
0.125
0.016
0.5
≤0.01
≥256
ND
ND
Fusobacterium mortiferum Fusobacterium varium
3
EP
AC C
Prevotella intermedia
8
TE D
≥256
≤0.01
SC
Metronidazole
ND
Bacteroides fragilis
ND = not detected
5 Page -21-
Piperacillin-
ND
32
tazobactam
≥256
≥256
Moxifloxacin
2
Parabacteroides distasonis
Imipenem
0.25
8
Clindamycin
0.125
≥256
Amoxycillin-
clavulanate
≥256
Porphyromonas asaccharolytica
Prevotella bivia
4
Penicillin
Organism
PCR
M AN U
Minimum inhibitory concentration (mg/L)
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
1
Page -22-
ACCEPTED MANUSCRIPT
Table 4: MIC distributions for tested antibiotics
Antibiotic Clindamycin
1
1 ≤0.01 0.03 0.06 0.125 0.25 1 25 7 5 3 1 1 1
1
1
2
1
RI PT
2
M AN U
2
2
SC
3
1
TE D
Antibiotic Amoxycillinclavulanate
Organism Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Organism Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Organism Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp.
1 ≤0.01 0.03 0.06 0.125 0.25 4 2 4 2 3 1 1
EP
Antibiotic Penicillin
Minimum inhibitory concentration (mg/L) ≤0.01 0.03 0.06 0.125 0.25 0.5 1 2 4 8 16 1 4 1 9 11 2 7 3 4 3 2 1
AC C
1
1 1
1 Page -23-
64 2
128 2 1
≥256 29 2
1 1
1 0.5 9 5
1 1 7
1 2 10 2
3
2
1
1 4 8 1
2
1
1
8 4 1
16 1
32 2
64
128
≥256 1
16 3
32
64
128
≥256 24 8
1 2 2 0.5 4 1
1 1 1 8 3
4
2 3
32 9 1
1
1 1 2 11 4
4 5 2
8 3
2
2
1 1
1 3
ACCEPTED MANUSCRIPT
1
1
1
1 2
2
1
1
TE D
1 1 1 1 2 ≤0.01 0.03 0.06 0.125 0.25 1 3 12 4 6 2 2 2
4
1
1
32 1
2
2 2 0.5 18 6
1 20 1
2 6 1
2
1
3
1 1 0.5 10 3
2 1 1 7 1
M AN U
3 ≤0.01 0.03 0.06 0.125 0.25 1 11 2 2 8
1
RI PT
1
SC
1
EP
Organism Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Antibiotic Organism Moxifloxacin Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Antibiotic Organism PiperacillinBacteroides spp. tazobactam Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Antibiotic Organism Metronidazole Bacteroides spp.
AC C
Antibiotic Imipenem
Minimum inhibitory concentration (mg/L) ≤0.01 0.03 0.06 0.125 0.25 0.5 1 2 4 8 16 8 17 14 8 11 6 1 2 2 6 8 4 2 2
1
4 3
≥64 1 1 1
8 1 2
16
32
1
≥64 8 4 1
1
1
2 9 3
4 6 1
8 5 2
16 7 1
1
3
1
3 4 2
1 16
32 3
64 1 1
128
≥256 4
1 1
≤0.01 0.03 0.06 0.125 0.25 1 3 4 10 Page -24-
1 0.5 26
1 17
2 3
8
32
64
128
1 ≥256 2
ACCEPTED MANUSCRIPT
1
2 1
1
1
1
2
4
3
RI PT
1
2
SC
Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp.
Minimum inhibitory concentration (mg/L) 2 2 9 2 2 1 1
AC C
EP
TE D
M AN U
1
Page -25-
3
1
1
3
1 1
ACCEPTED MANUSCRIPT
1
Highlights
2
•
Bacteroides species is the most commonly isolated anaerobe from blood cultures
3
•
The primary sources of infection were intra-abdominal and soft-tissue infections
4
•
Metronidazole and imipenem resistance was uncommon in Gram-negative isolates
EP
TE D
M AN U
SC
RI PT
(<5%)
AC C
5
Page -1-
S1075-9964(16)30155-X
DOI:
10.1016/j.anaerobe.2016.11.009
Reference:
YANAE 1645
To appear in:
Anaerobe
Received Date: 7 May 2014 Revised Date:
17 November 2016
Accepted Date: 24 November 2016
Please cite this article as: Tan TY, Ng LSY, Kwang LL, Rao S, Eng LC, Clinical characteristics and antimicrobial susceptibilities of anaerobic bacteremia in an acute care hospital, Anaerobe (2016), doi: 10.1016/j.anaerobe.2016.11.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Clinical characteristics and antimicrobial susceptibilities of anaerobic bacteremia in an acute care hospital Thean Yen Tan*1 Lily Siew Yong Ng1
Suma Rao2 Li Ching Eng1
Short title:
M AN U
Clinical features of anaerobic bacteraemia
Affiliation: 1
Department of Laboratory Medicine, Changi General Hospital, 2 Simei Street 3,
TE D
Singapore, 529889
Department of Medicine, Changi General Hospital, 2 Simei Street 3, Singapore, 529889
EP
2
SC
RI PT
Lee Ling Kwang1
Corresponding author:
AC C
Thean Yen Tan
Department of Laboratory Medicine Changi General Hospital 2 Simei Street 3 Singapore 529889 Tel: 065-68504934 Fax: 065-64269507 Page -1-
ACCEPTED MANUSCRIPT Email:
[email protected]
RI PT
Article word count:
SC
2442
Conflict of interest declaration
M AN U
None to declare
Acknowledgements
AC C
EP
TE D
This study was funded by an internal grant from Changi General Hospital.
Page -2-
ACCEPTED MANUSCRIPT SUMMARY PAGE Introduction This study investigated the clinical features of anaerobic bacteraemia in an acute-care hospital, and evaluated the antimicrobial susceptibility of these isolates to commonly
RI PT
available antibiotics. Materials & Methods
Microbiological and epidemiological data from 2009-2011were extracted from the
SC
laboratory information system and electronic medical records. One hundred and eleven unique patient episodes consisting of 116 anaerobic isolates were selected for clinical
M AN U
review and antibiotic susceptibility testing. Susceptibilities to amoxicillin-clavulanate, clindamycin, imipenem, metronidazole, moxifloxacin, penicillin and piperacillin-tazobactam were performed using Etest strips with categorical interpretations according to current CLSI breakpoints. Metronidazole-resistant and carbapenem-resistant anaerobic Gram-
TE D
negative bacilli were screened for the nim and cfiA genes. Clinical data was obtained retrospectively from electronic medical records. Results
EP
During the 3 year period, Bacteroides fragilis group (41%), Clostridium species (14%), Propionibacterium species (9%) and Fusobacterium species (6%) were the most
AC C
commonly isolated anaerobes. Patients with anaerobic bacteraemia that were included in the study were predominantly above 60 years of age, with community-acquired infections. The most commonly used empiric antibiotic therapies were beta-lactam/beta-lactamase inhibitor combinations (44%) and metronidazole (10%). The crude mortality was 25%, and appropriate initial antibiotic therapy was not significantly associated with improved survival. Intra-abdominal infections (39%) and soft-tissue infections (33%) accounted for nearly three-quarters of all bacteraemia.
Page -3-
ACCEPTED MANUSCRIPT Antibiotics with the best anaerobic activity were imipenem, piperacillin-tazobactam, amoxicillin-clavulanate and metronidazole, with in-vitro susceptibility rates of 95%, 95%, 94% and 92% respectively. Susceptibilities to penicillin (31%), clindamycin (60%) and moxifloxacin (84%) were more variable. Two multidrug-resistant isolates of Bacteroides
resistant Fusobacterium species were negative for cfiA genes. Conclusion
RI PT
species were positive for nim and cfiA genes respectively, while another two imipenem-
SC
This study demonstrated that anaerobic bacteraemia in our patient population was
predominantly associated with intra-abdominal and soft-tissue infections. Overall antibiotic
M AN U
resistance was high for penicillin and clindamycin, and the presence of emerging resistance to carbapenems and metronidazole warrants further monitoring.
TE D
Key words:
anaerobes; bacteremia; Bacteroides; antibiotic resistance
AC C
305
EP
Synopsis word count:
Page -4-
ACCEPTED MANUSCRIPT ARTICLE
1.0 INTRODUCTION Anaerobic bacteraemia continues to account for a small, but significant, proportion of
RI PT
bacteraemia (around 7-10% of all bacteraemic episodes) [1, 2]. Common sources of infection include the gastrointestinal tract, abscesses, gynaecologic sources and soft tissue infections [3], while reported risk factors for anaerobic bacteraemia include
SC
gastrointestinal surgery and haematologic malignancy [4].
Antibiotic therapy for anaerobes is usually empiric. Anaerobic susceptibility testing is
M AN U
not performed routinely, as it is expensive, methodologically demanding and time consuming. Based on published reports, the prevalence of antibiotic resistance in anaerobes is increasing [5, 6]. Clinical data suggest that antibiotic resistance impacts both microbiological cure and patient mortality for patients with anaerobic bacteraemia [7].
TE D
There is little susceptibility data for anaerobic infections in Singapore, and no data exists for the clinical characteristics of patients with anaerobic bacteraemia. This study was performed in an 800-bed hospital that provides medical, surgical,
EP
orthopaedic and geriatric medical care to the eastern region of Singapore. The aims of the study were to determine clinical data for patients with anaerobic bacteraemia over a three-
AC C
year period, and to determine antibiotic susceptibilities for these anaerobic isolates.
2.0 METHODS
2.1. Retrospective microbiological data extraction Bacterial information and epidemiological data were extracted from the laboratory information system for the period 2009-2011. Duplicate isolates from the same patient (defined as similar isolates within a 30 day period) were excluded from analysis. During Page -5-
ACCEPTED MANUSCRIPT the study period, routine aerobic and anaerobic blood cultures were performed using BACTEC™ Plus Aerobic/F and Plus Anaerobic/F vials (BD, USA), incubated for five days in a continuous monitoring blood culture system (Bactec 9000 series, BD, USA). Anaerobic vials with positive growth indices were sub-cultured on to trypticase soy agar
RI PT
with 5% sheep blood (BD, USA), MacConkey agar (Oxoid, UK), Chocolate agar (BD, USA) and CDC anaerobe agar with 5% sheep blood (BD, USA). Media were routinely incubated for 2 days, with additional incubation for up to 4 days if a slow-growing anaerobe was
SC
suspected. Gram-stain and aero-tolerance testing were performed for suspected
anaerobic isolates, with speciation by Vitek® ANC identification cards tested on the Vitek®
M AN U
Compact system (bioMérieux, France), supplemented with API 20A (bioMérieux, France) where necessary.
Detailed clinical and bacteriological information was retrospectively obtained for 111 unique patient episodes. These cases were randomly chosen to represent the most
TE D
common clinically significant anaerobic isolates from blood cultures. Propionibacterium species, anaerobic Gram-positive cocci, and nonsporulating anaerobic Gram-positive rods were excluded from detailed analysis, as previous studies have demonstrated that these
EP
genera are much less likely to be clinically significant [3, 8, 9]. 2.2. Antimicrobial susceptibility testing
AC C
For the 111 unique patient episodes, 116 anaerobic isolates were retrieved from cryo-storage for antibiotic susceptibility testing. Bacterial identification was re-confirmed by MALDI-TOF testing (Vitek MS database 2.0, bioMérieux, France) prior to susceptibility testing. Susceptibilities to amoxicillin-clavulanate, clindamycin, imipenem, metronidazole, moxifloxacin, penicillin and piperacillin-tazobactam were performed using Etest antimicrobial strips on Brucella agar with 5% sheep blood, vitamin K and hemin (BD, USA). Overnight pre-reduction of testing media was performed prior to testing, and following inoculation, plates were incubated in an anaerobic workstation system at 35°C Page -6-
ACCEPTED MANUSCRIPT (Forma Anaerobic System, ThermoFisher Scientific, USA) in an anaerobic atmosphere comprising 5% hydrogen, 5-10% carbon dioxide and 85-90% nitrogen. Concurrent quality control (QC) testing was performed using Bacteriodes fragilis ATCC 25285 according to manufacturer’s guidelines, and QC values were within acceptable limits. Etest results were
RI PT
read after 48-72 hours incubation, with all except one Fusobacterium isolate achieving satisfactory growth endpoints at 48 hours. Antibiotic susceptibilities were interpreted according to existing CLSI breakpoints [10]. Metronidazole-resistant and carbapenem-
SC
resistant anaerobic Gram-negative bacilli were screened for nim and cfiA genes encoding 5-nitroimidazole resistance and carbapenem resistance respectively, by conventional PCR
M AN U
[11, 12]. 2.3. Clinical data extraction and analysis
Clinical data collection for each patient episode of anaerobic bacteraemia was obtained retrospectively from electronic medical records. Data collected included the
TE D
admitting and final discharge diagnosis, empiric antibiotic prescription, results of clinical investigations and inpatient mortality. The presence of polymicrobial bacteraemia was defined as the presence of other bacterial isolates from blood cultures collected on the
EP
same day as the primary anaerobic isolate. The Charlson co-morbidity scores were calculated for each patient using an electronic spreadsheet [13], while statistical analysis
AC C
and odds ratios were calculated using OpenEpi (online open source statistical software) [14] supplemented with SPSS Statistics (IBM, USA).
Page -7-
ACCEPTED MANUSCRIPT 3.0 RESULTS 3.1. Anaerobic bacteraemia over the three year period During the 3 year period, 250 obligate anaerobes were recovered from blood cultures, representing 4.1% of all positive blood cultures. The most frequent anaerobic
RI PT
isolates were the Bacteroides fragilis group (n=102; 41%), Clostridium species (n=34; 14%), Propioniobacterium species (n=23; 9%) and Fusobacterium species (n=16; 6%). The remainder of the isolates included Prevotella spp., Veillonella spp., Eubacterium spp.,
SC
Actinomyces spp., Peptostreptococcus spp., Lactobacillus spp., Parvimonas and
Porphyromonas spp. Twenty three isolates could not be identified by phenotypic tests,
M AN U
consisting predominantly of anaerobic Gram-positive bacilli.
3.2. Detailed clinical information on subset of 111 patients For the detailed subset of 111 study patients with anaerobic bacteraemia, the average age was 73 years (range: 21-99 years). Patients older than 60 years of age
TE D
accounted for 84% of cases. Most patients (n=89, 80%) were admitted from home, with a futher 21 (19%) patients admitted from a long-term residential care facility. The majority (87%) of anaerobic bacteraemia were reported from blood cultures taken within two days
EP
of hospital admission. The majority of patients were reported as independently mobile (n=70, 63%), while 24 (22%) patients were reported to be bed-bound. The average
AC C
Charlson co-morbidity score was 2.6 (range 0-12), while the average age-adjusted score was 5.1 (range 0-15). Information on antibiotic therapy was available for 97 patient episodes. The most commonly used empiric antibiotic therapy were beta-lactam/betalactamase inhibitor combinations (amoxicillin/clavulanic acid or piperacillin/tazobactam) (n=49, 44%), metronidazole (n=11, 10%) and carbapenems (n=8, 8%). Eighteen patients (16%) were on antibiotics with no activity against anaerobic bacteria, and two patients had no antibiotic coverage. The majority of patients were started on initial appropriate antibiotic therapy (n=72, 65%). The crude mortality was 25%, with a higher mortality risk for patients Page -8-
ACCEPTED MANUSCRIPT with infections due to an extra-abdominal source (odds ratio 0.3, 95%CI 0.12-0.92), increasing age (increased risk of 3.7% per year, 95%CI 0.2- 7.3%) and increasing adjusted Charlson co-morbidity score (odds ratio 1.2, 95%CI 1.0-1.3) on univariate analysis. However, on multivariate analysis, the increased risk of mortality remained
RI PT
statistically significant only for the source of infection. Infection with either Bacteroides spp. or Clostridium spp. was not significantly associated with a higher mortality, while
appropriate initial antibiotic therapy was not associated with improved survival (odds ratio
SC
1.1, 95%CI 0.4-3.3). Intra-abdominal infections were the predominant source of infection (n=43, 39%), followed by soft-tissue infections (n=36, 32%) and lower respiratory tract
M AN U
infections (n=5, 5%). No obvious source of infection was identified for 18 patients (16%), 9 of which were judged to be likely contaminants due to the lack of clinical signs of infection and/or the presence of other skin commensals from corresponding blood cultures (Bacteroides spp. n=3, Clostridium spp. n=2, Prevotella spp. n=2, Veillonella spp. n=2).
TE D
The most common causes of intra-abdominal infections were hepatobiliary sepsis, appendicitis, perforated viscus and underlying colonic cancer. For soft-tissue infections, the most common sources of infection were decubitus pressure ulcers, diabetic foot
EP
infection and cellulitis with associated gangrene (Table 1). One case of Lemierre’s syndrome was reported, involving Prevotella melaninogenica and ‘Anginosus group’
AC C
streptococci. Fifty-seven percent of anaerobic bacteraemias were polymicrobial, with 19% of cases having three or more separate organisms in the same culture set. For episodes of polymicrobial bacteraemia, aerobic bacteria were also isolated in 70% of cases. The most common other bacterial pathogens isolated from polymicrobial bacteraemia were other anaerobes (n=29, 33% of all polymicrobial cases), Enterobacteriaceae (n=25, 29%), coagulase negative staphylococci (n=10, 11%), Enterococcus species (n= 6, 7%) and ‘Anginosus group’ streptococci (n=6, 7%). Patients with soft-tissue infections were twice as
Page -9-
ACCEPTED MANUSCRIPT likely to have polymicrobial bacteraemia (odds ratio 2.2, 95%CI 0.95-5.3) compared to patients with infections from all other sources. 3.3. Antimicrobial susceptibilities of anaerobic isolates from 111 patients Antibiotic susceptibilities were determined for 116 anaerobic isolates, consisting of
RI PT
Bacteroides spp. (n=68), Clostridium spp. (n=26), Fusobacterium spp. (n=9) and other species (n=13). Antibiotics with the best activity against anaerobic bacteria were imipenem, piperacillin-tazobactam, metronidazole, amoxicillin-clavulanate and
SC
moxifloxacin, with in-vitro susceptibility rates of 95%, 95%, 92%, 94% and 84%
respectively. Susceptibilities to penicillin (31%) and clindamycin (60%) were much lower.
M AN U
Bacteroides spp. were uniformly resistant to penicillin, with lower susceptibility to clindamycin (54%) and moxifloxacin (85%) (Table 2). Moxifloxacin resistance was higher in B. ovatus and B. uniformis, while Parabacteroides distasonis was more likely to be resistant to the beta-lactam/beta-lactamase inhibitors and clindamycin. One strain of P.
TE D
distasonis demonstrated high-level resistance to metronidazole (MIC>256 mg/L) and this isolate was nim-positive by PCR. A multidrug-resistant isolate of B. fragilis was also present with high-level resistance to all tested antibiotics except for metronidazole, and
EP
this isolate was positive for the cfiA gene (Table 3). Clostridium spp. demonstrated lowest susceptibility to clindamycin (65%) and
AC C
variable susceptibility to metronidazole (81%), moxifloxacin (77%) and penicillin (81%), but were almost uniformly susceptible to the beta-lactam/beta-lactamase inhibitors (96-100%). Fusobacterium spp. remained highly susceptible to metronidazole (100%), with lower susceptibilities to penicillin (89%) and the beta-lactam/beta-lactamase inhibitors (89%). Although the overall numbers tested were small, the lowest susceptibility to imipenem was seen in Fusobacterium sp. (78%), with two unusually multidrug-resistant isolates of Fusobacterium detected. One strain of F. mortiferum demonstrated high-level resistance to all tested beta-lactams but was susceptible to metronidazole, clindamycin and Page -10-
ACCEPTED MANUSCRIPT moxifloxacin, The other F. varium was resistant to imipenem, moxifloxacin and clindamycin. The cfiA gene was not detected for either of these two strains.
AC C
EP
TE D
M AN U
SC
RI PT
MIC distributions for all tested antibiotics are listed in Table 4.
Page -11-
ACCEPTED MANUSCRIPT 4.0 DISCUSSION This study reports on the clinical features of anaerobic bacteraemia, with associated antibiotic susceptibilities. There are some limitations to our study. We studied a specific subset of bacteraemic
RI PT
episodes by specifically excluding anaerobe species which were most likely to be contaminants from the analysis. Secondly, the spectrum of clinical infections may not be representative of infections seen in other healthcare institutions, which may serve different
SC
patient populations or with additional clinical specialties not present in our institution. Bacteroides spp. accounted for nearly half of all bacteraemia, followed by
M AN U
Clostridium spp. and propionibacteria. In our patient population, patients with anaerobic bacteraemia were mostly elderly patients admitted from the community, with several existing co-morbidities. Although intra-abdominal infections were the most common source of infection, in contrast with other published studies, nearly one-third of anaerobic
TE D
bacteraemia was associated with soft-tissue infections. In a review of 81 patients with anaerobic bacteraemia from Finland from the 1990s, soft-tissue infections were documented in 23% of patients [3], compared to 8% of patients from a more recent
EP
publication from New Zealand [15]. Of note, the aetiology of bacteraemic infections in our study is likely to differ from larger medical centres offering oncology, organ transplant and
AC C
gynaecologic services, which are absent in our hospital. Studies from hospitals with comprehensive medical services suggest that patients with haematological disorders have a higher underlying rate of anaerobic bacteraemia compared with other patient groups, and in one such study, the upper respiratory system constituted the second most common portal of infection [4]. The inpatient mortality of patients in our series was 25%, which is largely consistent with the data from other reported studies [16]. Our study did not demonstrate a reduction in mortality for patients receiving appropriate initial antibiotic therapy. Page -12-
ACCEPTED MANUSCRIPT The carbapenems, beta-lactam/beta-lactamase inhibitors and metronidazole demonstrated excellent in-vitro activity against most anaerobes. Conversely, resistance to penicillin and clindamycin were high, which suggests that these antibiotics are not suitable for empiric treatment of anaerobic bacteraemia. A prospective study in the United States
RI PT
demonstrated increasing clindamycin and moxifloxacin resistance over a 3 year period, with differentially higher rates of resistance in non-Bacteroides fragilis isolates [17]. In Taiwan, higher rates of carbapenem resistance (up to 12% in B. fragilis) were reported
SC
over a ten-year period, but conversely, lower rates of resistance to moxifloxacin [6]. Carbapenem resistance in Bacteroides species is primarily caused by the presence of a
M AN U
metallo-β-lactamase (cfiA) with gene expression mediated by a corresponding insertion sequence [18]. The mechanisms of carbapenem resistance in other Gram-negative anaerobes remain poorly elucidated. Metronidazole resistance in anaerobes is mediated by multiple mechanisms, which include resistance conferred by a family (nim) of
TE D
transmissible genes [19]. A recent review noted that metronidazole resistance remains very low (<1%) in most geographic regions [20]. To our knowledge, this is the first study to screen for and detect the presence of cfiA and nim genes in organisms isolated in
EP
Singapore, emphasizing the importance of further surveillance to determine the prevalence of these transmissible genes in anaerobes from the Asia-Pacific region. This study also
AC C
provides the first information on antibiotic susceptibilities of anaerobic isolates associated with bacteraemic infections in Singapore. In summary, this study demonstrated that anaerobic bacteraemia in our patient population was predominantly associated with intra-abdominal and soft-tissue infections, with a relatively high mortality rate. Overall antibiotic resistance was high for penicillin and clindamycin and, to a lesser extent, moxifloxacin. The presence of emerging resistance to carbapenems and metronidazole warrants further monitoring.
Page -13-
ACCEPTED MANUSCRIPT REFERENCES [1]
Fenner L, Widmer AF, Straub C, Frei R. Is the incidence of anaerobic bacteremia
decreasing? Analysis of 114,000 blood cultures over a ten-year period. J Clin Microbiol 2008;46:2432-4. Lassmann B, Gustafson DR, Wood CM, Rosenblatt JE. Reemergence of anaerobic
bacteremia. Clin Infect Dis 2007;44:895-900. [3]
Salonen JH, Eerola E, Meurman O. Clinical significance and outcome of anaerobic
SC
bacteremia. Clin Infect Dis 1998;26:1413-7. [4]
RI PT
[2]
Blairon L, De Gheldre Y, Delaere B, Sonet A, Bosly A, Glupczynski Y. A 62-month
Microbiol Infect 2006;12:527-32. [5]
M AN U
retrospective epidemiological survey of anaerobic bacteraemia in a university hospital. Clin
Goldstein EJC, Citron DM. Resistance Trends in Antimicrobial Susceptibility of
Anaerobic Bacteria, Part I. Clin Microbiol Newsletter 2011;33:1-8. Liu CY, Huang YT, Liao CH, Yen LC, Lin HY, Hsueh PR. Increasing trends in
TE D
[6]
antimicrobial resistance among clinically important anaerobes and Bacteroides fragilis isolates causing nosocomial infections: emerging resistance to carbapenems. Antimicrob
[7]
EP
Agents Chemother 2008;52:3161-8.
Nguyen MH, Yu VL, Morris AJ, McDermott L, Wagener MW, Harrell L, et al.
AC C
Antimicrobial resistance and clinical outcome of Bacteroides bacteremia: findings of a multicenter prospective observational trial. Clin Infect Dis 2000;30:870-6. [8]
Brook I, Frazier EH. Significant recovery of nonsporulating anaerobic rods from
clinical specimens. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 1993;16:476-80. [9]
Park HJ, Na S, Park SY, Moon SM, Cho OH, Park KH, et al. Clinical significance of
Propionibacterium acnes recovered from blood cultures: analysis of 524 episodes. J Clin Microbiol 2011;49:1598-601. Page -14-
ACCEPTED MANUSCRIPT [10] CLSI. Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria; Approved Standard - Eighth Edition. CLSI document M11-A8. Wayne, PA: Clinical and Laboratory Standards Institute; 2012. [11] Trinh S, Reysset G. Detection by PCR of the nim genes encoding 5-nitroimidazole
RI PT
resistance in Bacteroides spp. J Clin Microbiol 1996;34:2078-84. [12] Thompson JS, Malamy MH. Sequencing the gene for an imipenem-cefoxitin-
hydrolyzing enzyme (CfiA) from Bacteroides fragilis TAL2480 reveals strong similarity
SC
between CfiA and Bacillus cereus beta-lactamase II. J Bacteriol 1990;172:2584-93. [13] Hall WH, Ramachandran R, Narayan S, Jani AB, Vijayakumar S. An electronic
M AN U
application for rapidly calculating Charlson comorbidity score. BMC Cancer 2004;4:94. [14] Dean AG SK, Soe MM. Version 3.01. OpenEpi: Open Source Epidemiologic Statistics for Public Health, . 2013.
[15] Muttaiyah S, Paviour S, Buckwell L, Roberts SA. Anaerobic bacteraemia in patients
journal 2007;120:U2809.
TE D
admitted to Auckland City Hospital: its clinical significance. The New Zealand medical
[16] Brook I. The role of anaerobic bacteria in bacteremia. Anaerobe 2010;16:183-9.
EP
[17] Snydman DR, Jacobus NV, McDermott LA, Golan Y, Hecht DW, Goldstein EJ, et al. Lessons learned from the anaerobe survey: historical perspective and review of the most
AC C
recent data (2005-2007). Clin Infect Dis 2010;50 Suppl 1:S26-33. [18] Ang L, Brenwald NP, Walker RM, Andrews J, Fraise A. Carbapenem resistance in Bacteroides fragilis. J Antimicrob Chemother 2007;59:1042-4. [19] Lofmark S, Edlund C, Nord CE. Metronidazole is still the drug of choice for treatment of anaerobic infections. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2010;50 Suppl 1:S16-23. [20] Boyanova L, Kolarov R, Mitov I. Recent evolution of antibiotic resistance in the anaerobes as compared to previous decades. Anaerobe 2015;31:4-10. Page -15-
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Page -16-
ACCEPTED MANUSCRIPT
Table 1: Sources of anaerobic bacteraemia
36 26 3 3 4(1) 5
Pyelonephritis
12
8 6 5 2 2
1 3 1 2
RI PT
Soft-tissue Sacral sore(4) Diabetic foot infection Gangrene & cellulitis Others(1) Lower respiratory tract infection Aspiration pneumonia & infected sacral sores Empyema Pneumonia Aspiration pneumonia
29
1
SC
43 12 8 6 3 2 1 1 10
No. of bacterial isolates FusoPorphybacterium romonas spp. spp. 2 0
M AN U
Intra-abdominal Hepatobiliary sepsis Appendicitis(1) Perforated viscus(1) Colonic cancer(1) Ischemic bowel C. difficile colitis Unspecific Colitis Others
Bacteroides Clostridium spp. spp.
Prevotella spp.
Veillonella spp.
2
1
1
1
1
1 1 3
1
23
7
6
1
2
2
17 1 3 2 4
6 1
4
1
1 1
1
6
TE D
No. of patients
EP
Type of infection
AC C
1
1 1 2 1
1 1 1 1
2
2
2 1
1
Page -17-
1
ACCEPTED MANUSCRIPT
1 2
Total No. = number (n)
2 2 1 1 1 18
1
1 2
1 1 8 69
111
5 27
RI PT
Bacteroides Clostridium spp. spp.
No. of bacterial isolates FusoPorphybacterium romonas spp. spp.
SC
Genito-urinary Lower urinary tract infection Joint infection Lemierre's syndrome Osteomyelitis No source identified
No. of patients
M AN U
Type of infection
1 10
Prevotella spp.
1
1
2 7
numbers in superscript indicate the number of cases with more than one anaerobe isolated for the bacteraemic episode
AC C
EP
TE D
3
Page -18-
Veillonella spp.
2 5
ACCEPTED MANUSCRIPT
Table 2: Antibiotic susceptibilities of selected anaerobes
TE D
Page -19-
Imipenem S (%) 67 (99%) 39 (98%) 12 (100%) 6 (100%) 4 (100%) 3 (100%) 2 (100%) 1 (100%) 23 (89%) 10 (100%) 4 (100%) 3 (100%) 3 (100%) 1 (50%) 0 (0%) 1 (100%) 0 (0%) 1 (100%) 7 (78%) 3 (75%) 1 (50%) 2 (100%) 1 (100%) 7 (100%)
RI PT
Clindamycin S (%) 37 (54%) 25 (63%) 6 (50%) 1 (17%) 3 (75%) 1 (33%) 1 (50%) 0 (0%) 17 (65%) 9 (90%) 3 (75%) 0 (0%) 2 (67%) 2 (100%) 0 (0%) 0 (0%) 1 (100%) 0 (0%) 6 (67%) 1 (25%) 2 (100%) 2 (100%) 1 (100%) 4 (57%)
SC
B. fragilis B. thetaiotaomicron P. distasonis B. vulgatus B. ovatus B. uniformis B. stercoris Clostridium species C. perfringens C. bifermentans Clostridium species C. ramosum C. clostridioforme C. difficile C. tertium C. glycolium C. histolyticum Fusobacterium species F. varium F. mortiferum F. necrophorum F. nucleatum Prevotella species
S (%) 1 (2%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 1 (33%) 0 (0%) 0 (0%) 21 (81%) 10 (100%) 1 (100%) 0 (0%) 1 (100%) 1 (100%) 8 (89%) 4 (100%) 1 (50%) 2 (100%) 8 (89%) 4 (100%) 1 (50%) 2 (100%) 1 (100%) 3 (43%)
EP
Bacteroides species
n 68 40 12 6 4 3 2 1 26 10 4 3 3 2 1 1 1 1 9 4 2 2 1 7
Augmentin S (%) 63 (93%) 37 (93%) 12 (100%) 4 (67%) 4 (100%) 3 (100%) 2 (100%) 1 (100%) 25 (96%) 10 (100%) 4 (100%) 2 (67%) 3 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 1 (100%) 8 (89%) 4 (100%) 1 (50%) 2 (100%) 1 (100%) 7 (100%)
M AN U
Penicillin
AC C
1
Moxifloxacin S (%) 58 (85%) 36 (90%) 10 (83%) 6 (100%) 4 (100%) 0 (0%) 1 (50%) 1 (100%) 20 (77%) 10 (100%) 4 (100%) 2 (67%) 1 (33%) 0 (0%) 0 (0%) 1 (100%) 1 (100%) 1 (100%) 7 (78%) 2 (50%) 2 (100%) 2 (100%) 1 (100%) 6 (86%)
Piperacillintazobactam S (%) 64 (94%) 39 (98%) 12 (100%) 4 (67%) 4 (100%) 3 (100%) 1 (50%) 1 (100%) 26 (100%) 10 (100%) 4 (100%) 3 (100%) 3 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 1 (100%) 8 (89%) 4 (100%) 1 (50%) 2 (100%) 1 (100%) 7 (100%)
Metronidazole S (%) 67 (99%) 40 (100%) 12 (100%) 5 (83%) 4 (100%) 3 (100%) 2 (100%) 1 (100%) 21 (81%) 10 (100%) 3 (75%) 1 (33%) 1 (33%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 1 (100%) 9 (100%) 4 (100%) 2 (100%) 2 (100%) 1 (100%) 5 (71%)
ACCEPTED MANUSCRIPT
S (%) 1 (50%) 1 (50%) 0 (0%) 1 (100%) 0 (0%) 3 (60%) 0 (0%)
All Gram-negative anaerobes
90
15 (17%)
84 (93%)
52 (58%)
87 (97%)
77 (86%)
84 (93%)
86 (96%)
26
21 (81%)
25 (96%)
17 (65%)
23 (89%)
20 (77%)
26 (100%)
21 (81%)
1
All Gram-positive anaerobes n = number of isolates
2
S = susceptible
Imipenem S (%) 2 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 5 (100%) 1 (100%)
Moxifloxacin S (%) 2 (100%) 2 (100%) 0 (0%) 1 (100%) 1 (100%) 5 (100%) 1 (100%)
Piperacillintazobactam S (%) 2 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 4 (80%) 1 (100%)
Metronidazole S (%) 2 (100%) 2 (100%) 0 (0%) 0 (0%) 1 (100%) 5 (100%) 0 (0%)
SC
M AN U
TE D
4
EP
3
AC C
Prevotella species P. melaninogenica P. bivia P. intermedia P. oralis
Clindamycin S (%) 0 (0%) 1 (50%) 1 (100%) 1 (100%) 1 (100%) 5 (100%) 0 (0%)
RI PT
Veillonella spp. Porphyromonas asaccharolytica
n 2 2 1 1 1 5 1
Augmentin S (%) 2 (100%) 2 (100%) 1 (100%) 1 (100%) 1 (100%) 5 (100%) 1 (100%)
Penicillin
Page -20-
ACCEPTED MANUSCRIPT
Table 3: Minimum inhibitory concentration and resistance gene testing for metronidazole and imipenem resistant
2
anaerobes
RI PT
1
nim gene
cfiA gene
16
1
0.25
≥256
≥256
Detected
ND
>256
≥256
≥64
≥64
≥256
0.5
ND
Detected
≥256
>256
≤0.01
≥64
0.125
≥256
0.125
ND
ND
0.25
1
8
≥64
8
2
0.5
ND
ND
2
≤0.01
0.06
≥64
≤0.01
64
ND
ND
≤0.016
0.125
0.016
0.5
≤0.01
≥256
ND
ND
Fusobacterium mortiferum Fusobacterium varium
3
EP
AC C
Prevotella intermedia
8
TE D
≥256
≤0.01
SC
Metronidazole
ND
Bacteroides fragilis
ND = not detected
5 Page -21-
Piperacillin-
ND
32
tazobactam
≥256
≥256
Moxifloxacin
2
Parabacteroides distasonis
Imipenem
0.25
8
Clindamycin
0.125
≥256
Amoxycillin-
clavulanate
≥256
Porphyromonas asaccharolytica
Prevotella bivia
4
Penicillin
Organism
PCR
M AN U
Minimum inhibitory concentration (mg/L)
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
1
Page -22-
ACCEPTED MANUSCRIPT
Table 4: MIC distributions for tested antibiotics
Antibiotic Clindamycin
1
1 ≤0.01 0.03 0.06 0.125 0.25 1 25 7 5 3 1 1 1
1
1
2
1
RI PT
2
M AN U
2
2
SC
3
1
TE D
Antibiotic Amoxycillinclavulanate
Organism Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Organism Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Organism Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp.
1 ≤0.01 0.03 0.06 0.125 0.25 4 2 4 2 3 1 1
EP
Antibiotic Penicillin
Minimum inhibitory concentration (mg/L) ≤0.01 0.03 0.06 0.125 0.25 0.5 1 2 4 8 16 1 4 1 9 11 2 7 3 4 3 2 1
AC C
1
1 1
1 Page -23-
64 2
128 2 1
≥256 29 2
1 1
1 0.5 9 5
1 1 7
1 2 10 2
3
2
1
1 4 8 1
2
1
1
8 4 1
16 1
32 2
64
128
≥256 1
16 3
32
64
128
≥256 24 8
1 2 2 0.5 4 1
1 1 1 8 3
4
2 3
32 9 1
1
1 1 2 11 4
4 5 2
8 3
2
2
1 1
1 3
ACCEPTED MANUSCRIPT
1
1
1
1 2
2
1
1
TE D
1 1 1 1 2 ≤0.01 0.03 0.06 0.125 0.25 1 3 12 4 6 2 2 2
4
1
1
32 1
2
2 2 0.5 18 6
1 20 1
2 6 1
2
1
3
1 1 0.5 10 3
2 1 1 7 1
M AN U
3 ≤0.01 0.03 0.06 0.125 0.25 1 11 2 2 8
1
RI PT
1
SC
1
EP
Organism Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Antibiotic Organism Moxifloxacin Bacteroides spp. Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Antibiotic Organism PiperacillinBacteroides spp. tazobactam Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp. Antibiotic Organism Metronidazole Bacteroides spp.
AC C
Antibiotic Imipenem
Minimum inhibitory concentration (mg/L) ≤0.01 0.03 0.06 0.125 0.25 0.5 1 2 4 8 16 8 17 14 8 11 6 1 2 2 6 8 4 2 2
1
4 3
≥64 1 1 1
8 1 2
16
32
1
≥64 8 4 1
1
1
2 9 3
4 6 1
8 5 2
16 7 1
1
3
1
3 4 2
1 16
32 3
64 1 1
128
≥256 4
1 1
≤0.01 0.03 0.06 0.125 0.25 1 3 4 10 Page -24-
1 0.5 26
1 17
2 3
8
32
64
128
1 ≥256 2
ACCEPTED MANUSCRIPT
1
2 1
1
1
1
2
4
3
RI PT
1
2
SC
Clostridium spp. Fusobacterium spp. Porphyromonas spp. Prevotella spp. Veillonella spp.
Minimum inhibitory concentration (mg/L) 2 2 9 2 2 1 1
AC C
EP
TE D
M AN U
1
Page -25-
3
1
1
3
1 1
ACCEPTED MANUSCRIPT
1
Highlights
2
•
Bacteroides species is the most commonly isolated anaerobe from blood cultures
3
•
The primary sources of infection were intra-abdominal and soft-tissue infections
4
•
Metronidazole and imipenem resistance was uncommon in Gram-negative isolates
EP
TE D
M AN U
SC
RI PT
(<5%)
AC C
5
Page -1-