Phenotypic and Genotypic Correlation of Antimicrobial Susceptibility of Bacteroides fragilis: Lessons Learnt

Background Bacteroides fragilis is an opportunistic pathogen causing severe infections, including bacteremia. There have been increased reports of antimicrobial resistance in B. fragilis. However, phenotypic testing of susceptibility is time consuming and not cost effective for anaerobes. The present study investigates the correlation of phenotypic susceptibility with genotypic markers; to determine if these could be considered for deciding empirical therapy for B. fragilis. Material and methods Bacteroides fragilis isolates from various clinical samples including exudates, tissue, and body fluids were collected between November 2018 and January 2020 in the Department of Clinical Microbiology, Christian Medical College (CMC) Vellore. Species identification was done by Matrix Assisted Laser Desorption Ionization time of flight mass spectrometry (MALDI TOF) according to the manufacturer's instructions. A total number of 51 B. fragilis isolates were tested against metronidazole, clindamycin, piperacillin/tazobactam, and meropenem phenotypically by agar dilution method using Clinical & Laboratory Standards Institute (CLSI) 2019 guidelines and minimum inhibitory concentrations (MIC) were interpretated. The genotypic markers for antimicrobial resistance genes (nim, emrF, and cfiA) were studied by polymerase chain reaction (PCR) assay as per the standard protocol on all isolates to detect resistance genes. Results B. fragilis isolates in this study expressed 45%, 41%, and 16% phenotypic resistance to clindamycin, metronidazole, and meropenem, respectively, with least resistance to piperacillin/tazobactam (6%). Among the metronidazole resistant isolates, 52% harbored nim gene. Nim gene was also present in 76% (23/30) of the metronidazole susceptible isolates. Similarly, cfiA was present in all eight meropenem resistant isolates in addition to 22% (9/41) of the susceptible isolates. All cfiA negative isolates were phenotypically susceptible. Interestingly, 74% (17/23) of the clindamycin resistant isolates were positive for ermF. Conclusions Detection of a limited set of genes does not always correlate with phenotypic resistance to metronidazole and clindamycin due to the reported influence of insertion sequence (IS) elements, efflux, and other genetic determinants. Certainly, the absence of the cfiA gene can be employed to rule out meropenem resistance. However, redundant use of antibiotics such as meropenem along with metronidazole could be avoided for B. fragilis, which might otherwise elevate meropenem resistance. Recommendation of metronidazole requires prior phenotypic testing due to the reported 41% resistance.