Article

Brief Communication

Ann Lab Med 2025; 45(2): 223-227

Published online October 31, 2024 https://doi.org/10.3343/alm.2024.0369

Copyright © Korean Society for Laboratory Medicine.

Clinical Outcomes and Molecular Characteristics of Bacteroides fragilis Infections

Bongyoung Kim , M.D., Ph.D.1, Myungsook Kim , Ph.D.2, Kyungwon Lee , M.D., Ph.D.2,3, and Yangsoon Lee, M.D., Ph.D.4

1Department of Internal Medicine, Hanyang University College of Medicine, Seoul, Korea; 2Seoul Clinical Laboratory, Yongin, Korea; 3Department of Laboratory Medicine, Research Institute of Bacterial Resistance, Yonsei University College of Medicine, Seoul, Korea; 4Department of Laboratory Medicine, Hanyang University College of Medicine, Seoul, Korea

Correspondence to: Yangsoon Lee, M.D., Ph.D.
Department of Laboratory Medicine, Hanyang University College of Medicine, 222 Wangsimni-ro, Seongdong-gu, Seoul 04763, Korea
E-mail: yangsoon@hanyang.ac.kr

Received: July 17, 2024; Revised: August 26, 2024; Accepted: October 24, 2024

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Bacteroides fragilis is the most common opportunistic anaerobic pathogen. In the absence of appropriate antimicrobial therapy, mortality rates associated with B. fragilis group infections can reach as high as 50%. Therefore, we aimed to elucidate the clinical characteristics and outcomes of B. fragilis infections and the molecular genetic characteristics of B. fragilis isolates. Forty B. fragilis clinical isolates were collected at Hanyang University Hospital between January 2022 and December 2023. Antimicrobial susceptibility was tested using the agar dilution method. Whole-genome sequencing was conducted using the Illumina platform (Illumina, San Diego, CA, USA). Various multilocus sequence types of B. fragilis were identified, including ST149 (N=4), ST11 (N=4), ST1 (N=3), ST21 (N=2), and ST157 (N=1). The insertion sequence (IS) IS1187, located upstream of cfiA, was associated with high-level carbapenem resistance in the ST157 isolate. B. fragilis toxin genes (bft ) were identified in 30% of isolates. The most common comorbidities were diabetes mellitus (26.5%) and non-metastatic cancer (23.5%). Five patients (14.7%) died within 30 days, and two (5.9%) deaths were directly attributable to B. fragilis infection. The emergence of high-level MIC carbapenem-resistant B. fragilis ST157 has led to caution in the presence of B. fragilis infections.

Keywords: Bacteremia, Bacteroides fragilis, cfiA, Intraabdominal infection, ST157

Bacteroides fragilis group (BFG) are residents of the healthy intestinal microbiota, but are frequently isolated from anaerobic infections [1]. In the absence of appropriate antimicrobial therapy, mortality rates associated with BFG infections can reach as high as 50% [2]. Antimicrobial resistance (AMR) profiles vary across BFG members [3, 4]. Among the BFG, B. fragilis sensu stricto (BFSS) is the most prevalent anaerobic microorganism in clinical specimens. Genetic divergence at the species level has been observed between division I and division II BFSS bacteria [5]. Division I BFSS harbor the cephalosporinase gene cepA and generally are susceptible to frontline β-lactam regimens. Division II BFSS can achieve high-level carbapenem resistance through insertion sequence (IS)-mediated activation of the chromosomal carbapenemase gene cfiA, which is absent in division I BFSS [1, 6]. Numerous virulence factors contribute to the pathogenic potential of BFSS, including adhesin and capsule synthesis, relative oxygen tolerance, and antigenic variation in surface structures [7]. In-vitro studies have indicated that B. fragilis toxin (bft) may induce intestinal colonic pathology, including inflammation and carcinogenesis [1, 8]. We aimed to elucidate the clinical characteristics and outcomes of BFSS infections and the molecular genetic characteristics of BFSS isolates for the first time in Korea.

Bacterial isolates were consecutively collected at Hanyang University Hospital (Seoul, Korea) between January 2022 and December 2023. In total, 40 BFSS isolates were obtained from 40 patients. The most common specimen was blood (N=25, 62.5%), followed by pus (N=10), tissue (N=4), and pleural fluid (N=1). Demographic data included age and sex, and underlying comorbidities were recorded using the Charlson comorbidity index [9]. Additional medical conditions noted included the use of chemotherapeutic and immunosuppressive agents, bedridden status, history of previous admissions, and anaerobic antimicrobial use prior to admission. Clinical features associated with severity included the presence of hypotension and the use of inotropics. Laboratory findings at the time of bacterial culture included white blood cell count and C-reactive protein, procalcitonin, blood urea nitrogen, and creatinine levels. Therapeutic interventions included the use of anaerobic antimicrobials within 48 hrs of bacterial culture and any surgical intervention to control the infection source. Clinical outcomes were assessed based on 30-day mortality. To evaluate clinical severity, we calculated Pitt’s bacteremia score [10]. Anaerobic antibiotics included carbapenems, cefotetan, ampicillin-sulbactam, amoxicillin-clavulanate, piperacillin-tazobactam, cefoperazone-tazobactam, clindamycin, moxifloxacin, and metronidazole. The type of infection and infection-related mortality were determined by the attending physicians and confirmed by an infectious disease specialist through a medical record review using the CDC/NHSN surveillance definition [11]. The Institutional Review Board of Hanyang University Hospital approved the study protocol (IRB No. 2022-07-037). Informed consent was waived for this study.

Bacterial species were identified using a matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MS) system (Bruker Biotyper MS; Bruker Daltonics, Bremen, Germany). Antimicrobial susceptibility was tested using the agar dilution method according to CLSI guidelines [12, 13]. The culture medium was Brucella agar supplemented with hemin and vitamin K1 (Sigma-Aldrich, Seoul, Korea) and 5% laked sheep blood. The antimicrobial powders used included penicillin, piperacillin (Sigma-Aldrich), tazobactam (Yuhan, Seoul, Korea), cefoxitin (Merck Sharp & Dohme, West Point, PA, USA), cefotetan (Daiichi Pharmaceutical, Tokyo, Japan), clindamycin, imipenem, meropenem, moxifloxacin and metronidazole (Sigma-Aldrich). An inoculum of 105 colony-forming units was streaked onto a culture plate using a Steers replicator (CMI-Promex, Pedricktown, NJ, USA), and the plates were incubated in an anaerobic chamber (Bactron; Sheldon Manufacturing, Cornelius, OR, USA) at 35°C for 48 hrs. The minimum inhibitory concentration (MIC) was determined as the concentration at which a marked reduction in growth occurred, such as the transition from confluent colonies to a haze, <10 tiny colonies, or several normal-sized colonies [12]. B. fragilis ATCC 25285 and Bacteroides thetaiotaomicron ATCC 29741 were used as quality control strains. Bacterial isolates cultured on Brucella blood agar plates were sent to Macrogen (Seoul, Korea) for whole-genome sequencing (WGS) using the Illumina platform (Illumina, San Diego, CA, USA). De-novo assembly and validation were performed for all 40 isolates. The WGS data were analyzed using Center for Genomic Epidemiology resources (http://www.genomicepidemiology.org/) and PubMLST (https://pubmlst.org/). Multilocus sequence types (MLSTs), AMR gene repertoires, and toxin genes were determined. The IS element was identified using ISfinder (https://www-is.biotoul.fr/index.php).

Six cases (15.0%) were classified as colonization rather than infection. In all colonization cases, the bacteria were isolated from a wound site near the anus. The clinical characteristics and outcomes of patients with B. fragilis infection (N=34) are provided in Table 1. The mean age was 65.5±18.7 yrs, and 32.5% of patients were female. The mean Charlson comorbidity index was 2.6±2.8, and the most common comorbidities were diabetes mellitus (26.5%) and non-metastatic cancer (23.5%). Five patients (14.7%) were bedridden, 12 (35.3%) had a history of admission within the past 3 months, and 10 (29.4%) used anaerobic antimicrobials within 1 yr prior to admission. The most common type of infection was intra-abdominal infection (58.8%), followed by skin and soft-tissue infection (14.7%) and empyema (8.8%). Among patients with intra-abdominal infection, 90.0% (18/20) exhibited bacteremia caused by B. fragilis, and 25.0% (5/20) had a history of major gastrointestinal tract surgery.

Clinical characteristics of 34 patients with Bacteroides fragilis infection
CharacteristicValue
Age, yrs65.5±18.7
Female sex13 (32.5)
Underlying comorbidities
Diabetes mellitus9 (26.5)
Non-metastatic cancer8 (23.5)
Chronic liver disease6 (17.6)
Chronic renal disease5 (14.7)
Metastatic solid tumor4 (11.8)
Dementia3 (8.8)
Cerebrovascular accident2 (5.9)
Chronic pulmonary disease2 (5.9)
Myocardial infarction2 (5.9)
Congestive heart failure2 (5.9)
Peripheral vascular disease1 (2.9)
Hemiplegia1 (2.9)
Other medical conditions
Chemotherapeutic agents1 (2.9)
Immunosuppressants3 (8.8)
Bedridden status5 (14.7)
History of admission within the past 3 months12 (35.3)
History of admission within the past 1 yr18 (52.9)
History of anaerobic antibiotic use within the past 3 months5 (14.7)
History of anaerobic antibiotic use within the past 1 yr10 (29.4)
Hypotension8/33 (24.2)
Inotropics4/33 (12.1)
Type of infection
Intra-abdominal20 (58.8)
Skin and soft tissue5 (14.7)
Empyema3 (8.8)
Others6* (17.6)
Clinical severity
Charlson comorbidity index2.6±2.8
Pitt’s bacteremia score0.9±1.4
Therapeutic interventions
Use of anaerobic coverage antimicrobials within 48 hrs after culture28 (82.4)
Surgical intervention7 (20.6)
Outcomes
30-day mortality, all-cause5 (14.7)
30-day mortality, infection-related2 (5.9)
Laboratory findings
White blood cells (×109/L)16.4±7.1
C-reactive protein (mg/L)221±114
Blood urea nitrogen (mmol/L)14.75±14.78
Creatinine (µmol/L)194.52±254.42

Continuous variables are presented as mean±standard deviation.

Categorical variables are presented as N (%).

*These included complicated urinary tract infection (N=2), pyogenic arthritis (N=1), osteomyelitis (N=1), acute cholangitis (N=1), and primary bacteremia (N=1).



Polymicrobial pathogens in the same specimen were identified in 23 patients (57.5%). Low blood pressure was measured in 24.2% (8/33) of patients and 12.1% (4/33) of patients who used inotropics at the time of culture. The mean white blood cell count and C-reactive protein level were 16.4×109±7.1×109/L and 221±114 mg/L, respectively. More than 80% of patients (82.4%) received anaerobic antimicrobials within 48 hrs after culture, and seven patients (20.6%) underwent surgery to treat B. fragilis infection. Five patients (14.7%) died within 30 days after culture, and two deaths (5.9%) were directly attributable to B. fragilis infection.

The resistance rates in our study were comparable to those reported in previous studies in Korea (Table 2) [3, 14]. All isolates were penicillin-resistant. The resistance rate to cefotetan (12.5%) was higher than that to cefoxitin (5.0%). The resistance rates for clindamycin and moxifloxacin were 37.5% and 15.0%, respectively. Piperacillin-tazobactam, imipenem, and meropenem were active, with 2.5% resistance rates. None of the isolates were resistant to metronidazole. Among β-lactam resistance genes in BFG, cepA, cfxA, and cfiA are associated with resistance to penicillin, cephamycin, and carbapenem, respectively. In our study, 38 isolates (95%) were division I BFSS carrying cepA and showed penicillin resistance. cfxA was detected in 13 isolates (32.5%), among which nine were non-susceptible to cefotetan. Two isolates (5%) (Nos. 20 and 25) were division II BFSS carrying cfiA. The Bruker Biotyper MS system identified these as cfiA-positive B. fragilis, as previously reported [5, 15]. Isolate No. 20 exhibited high-level resistance to both imipenem (MIC, 64 µg/mL) and meropenem (MIC, 128 µg/mL), whereas isolate No. 25 was susceptible to imipenem (MIC, 4 µg/mL) and meropenem (MIC, 8 µg/mL). The presence of the cfiA gene does not necessarily imply resistance, as imipenem MICs in division II BFSS ranged from 0.06 to >32 mg/L in a previous study [15]. Silent chromosomal cfiA can be overexpressed upon insertion of a single mobile genetic element, leading to phenotypic carbapenem resistance. The conserved ISs IS1186, IS1187, and IS613 located upstream of cfiA have been associated with high-level carbapenem resistance [6, 16]. In our study, IS1187 was detected upstream of cfiA in isolate No. 20 but not in isolate No. 25.

Antimicrobial susceptibility patterns and molecular characteristics of Bacteroides fragilis clinical isolates
StrainSpecimenSTMIC (mg/L)β-lactam resistance genesToxin gene
PENFOXCTTPIP/TAZIMPMEMCLNMOXMETcepAcfxAcfiAbft
1Tissue6932440.250.250.120.50.061+
2Pus10564240.250.250.1210.51++
3*Pus125683210.50.5180.5++bft-2
4Blood2716440.250.250.121280.060.5+bft-1
5Blood21256816110.50.580.5++
6Blood1716440.250.250.122560.061+
7Blood10632440.250.250.120.50.121+
8Pus56256480.50.250.25256160.5+bft-2
9*Blood98440.250.250.1220.061+
10*,Blood116440.120.120.1220.060.5+bft-2
11Pus432440.250.250.120.060.121+
12Blood1032840.50.250.120.50.121+
13Pus132840.250.120.1210.51+bft-1
14Pus2125616640.250.250.252560.121++
15Blood39256880.250.50.252560.060.5+bft-3
16*Blood4816480.250.250.120.541+
17Pus14916440.120.250.122560.060.25+bft-3
18*Blood1461288320.060.250.1225611++
19Pleural fluid1116880.120.250.0640.061+
20Blood1572566464256641282560.062+, IS1187
21Blood732440.250.120.120.50.060.5+
22Blood9816440.250.120.120.50.061+
23*Blood13516440.250.120.1220.120.5+bft-1
24*Blood120256880.250.250.2510.061++
25Blood12232168848140.5+
26Pus1492568640.250.50.250.0680.25++bft-1
27Tissue13632840.250.250.12141+
28Blood14116480.250.250.122560.120.5+
29Blood14332440.250.250.120.060.061+
30Pus112568640.512120.5++
31*Blood7916161620.250.250.521+bft-1
32Blood1492568320.250.250.252560.060.25++bft-1
33Blood11256880.50.50.252560.061++
34*Pus4025683210.50.50.540.5++
35Blood118420.120.250.1210.061+
36*Blood14216820.250.250.120.520.5+
37Tissue1225612825622425640.25++
38Blood1498440.120.250.1225680.5+bft-1
39Blood2316480.250.250.1225640.5+
40Tissue872564160.50.250.2525680.5++
Resistant/positive rate (%)1005.012.52.52.52.537.515.0010032.55.030.0

*Patients with a history of anaerobic antibiotic use within 1 yr prior to admission

Patients who died due to infection with B. fragilis.

Abbreviations: MIC, minimum inhibitory concentration; PEN, penicillin; FOX, cefoxitin; CTT, cefotetan; PIP/TAZ, piperacillin-tazobactam; IMP, imipenem; CLN, clindamycin; MOX, moxifloxacin; MET, metronidazole.



Nielsen, et al. [17] constructed an MLST scheme for B. fragilis in 2023. They found that isolates with the same ST always belonged to the same division. All ST157 isolates belonged to the division II BFSS cluster CC7 and originated from Asia. ST119 and ST140 were dominant in North America and Europe, respectively. In our study, WGS analysis revealed highly diverse MLST patterns among B. fragilis clinical isolates (Table 2). The 40 isolates were distributed into 31 unique STs, including ST149 (N=4), ST11 (N=4), ST1 (N=3), ST21 (N=2), and ST157 (N=1). ST157 and ST122 were associated with division II BFSS. Notably, the ST157 isolate was highly resistant to carbapenem. The patient infected with this isolate was an 88-yr-old woman with an intra-abdominal infection and bacteremia. She improved with metronidazole treatment for 9 days. She had been hospitalized within the past 3 months but had not used anaerobic antimicrobials within the past year. To the best of our knowledge, this was the first study to conduct an ST analysis of B. fragilis clinical isolates in Korea. Twelve isolates (30%) carried bft. Among the patients associated with these isolates, two were diagnosed as having colorectal malignancy; however, the influence of bft on colorectal carcinogenesis in these cases remains unclear. Continuous monitoring is necessary to detect resistant and virulent BFG clinical isolates.

In conclusion, recently isolated B. fragilis strains in Korea are of diverse STs. The emergence of highly carbapenem-resistant B. fragilis ST157 warrants caution. Understanding the genetic diversity and virulence profiles of B. fragilis can aid in developing targeted antimicrobial therapies and infection control measures, potentially improving patient outcomes and reducing mortality associated with these infections.

Lee Y, Kim B, and Lee K designed the study. Lee Y and Kim M collected and identified the clinical isolates. Lee Y, Kim B, and Lee K performed the experiments and analyzed the data. Lee Y, Kim B, and Lee K contributed to the writing, editing, and review of the manuscript. All authors read and approved the final manuscript.

This study was supported by a National Research Foundation of Korea grant funded by the Korean government (MSIT) (2022 R1F1A1063113).

  1. Jean S, Wallace MJ, Dantas G, Burnham CD. Time for some group therapy: update on identification, antimicrobial resistance, taxonomy, and clinical significance of the Bacteroides fragilis group. J Clin Microbiol 2022;60:e0236120.
    Pubmed KoreaMed CrossRef
  2. Cheng CW, Lin HS, Ye JJ, Yang CC, Chiang PC, Wu TS, et al. Clinical significance of and outcomes for Bacteroides fragilis bacteremia. J Microbiol Immunol Infect 2009;42:243-50.
  3. Byun JH, Kim M, Lee Y, Lee K, Chong Y. Antimicrobial susceptibility patterns of anaerobic bacterial clinical isolates from 2014 to 2016, including recently named or renamed species. Ann Lab Med 2019;39:190-9.
    Pubmed KoreaMed CrossRef
  4. Fang H, Li X, Yan MK, Tong MK, Chow KH, Cheng VC, et al. Antimicrobial susceptibility of Bacteroides fragilis group organisms in Hong Kong, 2020-2021. Anaerobe 2023;82:102756.
    Pubmed CrossRef
  5. Wallace MJ, Jean S, Wallace MA, Burnham CD, Dantas G. Comparative genomics of Bacteroides fragilis group isolates reveals species-dependent resistance mechanisms and validates clinical tools for resistance prediction. mBio 2022;13:e0360321.
    Pubmed KoreaMed CrossRef
  6. Gao Q, Wu S, Xu T, Zhao X, Huang H, Hu F. Emergence of carbapenem resistance in Bacteroides fragilis in China. Int J Antimicrob Agents 2019;53:859-63.
    Pubmed CrossRef
  7. Yekani M, Baghi HB, Naghili B, Vahed SZ, Sóki J, Memar MY. To resist and persist: important factors in the pathogenesis of Bacteroides fragilis. Microb Pathog 2020;149:104506.
    Pubmed CrossRef
  8. Garrett WS. The gut microbiota and colon cancer. Science 2019;364:1133-5.
    Pubmed CrossRef
  9. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373-83.
    Pubmed CrossRef
  10. Rhee JY, Kwon KT, Ki HK, Shin SY, Jung DS, Chung DR, et al. Scoring systems for prediction of mortality in patients with intensive care unit-acquired sepsis: a comparison of the Pitt bacteremia score and the Acute Physiology and Chronic Health Evaluation II scoring systems. Shock 2009;31:146-50.
    Pubmed CrossRef
  11. CDC/NHSN Surveillance definitions for specific types of infections. https://www.cdc.gov/nhsn/pdfs/pscmanual/17pscnosinfdef_current.pdf (Updated on Jan 2024).
  12. CLSI. Methods for antimicrobial susceptibility testing of anaerobic bacteria. 8th ed. M11-A8. Wayne, PA: Clinical and Laboratory Standards Institute, 2012.
    CrossRef
  13. CLSI. Performance standards for antimicrobial susceptibility testing. 33rd ed. M100. Wayne, PA: Clinical and Laboratory Standards Institute, 2023.
    CrossRef
  14. Yim J, Lee Y, Kim M, Seo YH, Kim WH, Yong D, et al. Antimicrobial susceptibility of clinical isolates of Bacteroides fragilis group organisms recovered from 2009 to 2012 in a Korean hospital. Ann Lab Med 2015;35:94-8.
    Pubmed KoreaMed CrossRef
  15. Kawamoto Y, Kosai K, Ota K, Uno N, Sakamoto K, Hasegawa H, et al. Rapid detection and surveillance of cfiA-positive Bacteroides fragilis using matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Anaerobe 2021;72:102448.
    Pubmed CrossRef
  16. Jeverica S, Sóki J, Premru MM, Nagy E, Papst L. High prevalence of division II (cfiA positive) isolates among blood stream Bacteroides fragilis in Slovenia as determined by MALDI-TOF MS. Anaerobe 2019;58:30-4.
    Pubmed CrossRef
  17. Nielsen FD, Skov MN, Sydenham TV, Justesen US. Development and clinical application of a multilocus sequence typing scheme for Bacteroides fragilis based on whole-genome sequencing data. Microbiol Spectr 2023;11:e0511122.
    Pubmed KoreaMed CrossRef