We aimed to determine the duration of faecal carriage of extended-spectrum β-lactamase (ESBL) -producing Enterobacteriaceae (EPE) in patients with clinical infection caused by an EPE, to study host strains during carriage, and to identify factors associated with prolonged carriage. Patients (n = 61) were followed with faecal samples and questionnaires about antimicrobial treatment and risk factors for EPE, 1, 3, 6 and 12 months after EPE infection. The EPE isolates were subjected to ESBL genotyping, epidemiological typing with pulsed-field gel electrophoresis and PCR-based replicon typing. Escherichia coli isolates were analysed with PCR for phylogrouping, detection of pabB (ST131) and virulence content. Patient-related and strain-related variables were compared for carriers and non-carriers at 12 months. Carriage of EPE was observed in 51 of 61 (84%) patients after 1 month, 36 of 61 (66%) after 3 months, 31 of 61 (55%) after 6 months and 26 of 61 (43%) after 12 months. Of the 26 carriers at 12 months, five had previous negative samples. In 17 of 61 patients, ESBL was found in a new bacterial species and/or strain during carriage. Among E. coli, 14 of 49 belonged to the international clone ST131. Phylogroup B2 and CTX-M-gr.-9 were associated with being carriers at 12 months (OR 4.3, 95% CI 1.1-16.3 and OR 6.4, 95% CI 1.3-30.9, respectively). In conclusion, EPE carriage is common 12 months after infection and persisting carriage may be associated with E. coli phylogroup B2 and CTX-M-gr.-9. The host strain frequently changes throughout carriage and negative samples do not imply eliminated carriage.

No abstract

Activity of oral and parenteral antimicrobials against consecutively isolated extended-spectrum β-lactamase (ESBL)-producing Escherichia coli (n = 149) and Klebsiella pneumoniae (n = 20) was determined, and susceptibility test methods were compared for parenteral β-lactams. Polymerase chain reaction (PCR) targeting bla(CTX-M), bla(SHV) and bla(TEM), and DNA sequencing and epidemiological typing with pulsed-field gel electrophoresis were performed. PCR targeting pabB was screened for E. coli O25b-ST131. Minimum inhibitory concentrations (MICs) were determined using Etest and broth microdilution. Disc diffusion was performed according to European Committee on Antimicrobial Susceptibility Testing (EUCAST). Dominating genotypes were bla(CTX-M-15) (75%) and bla(CTX-M-14) (23%). Four E. coli clusters (7-18 isolates) were found. Forty-two per cent of E. coli belonged to O25b-ST131. Ciprofloxacin resistance was 72%, trimethoprim resistance was 70%. Among E. coli, resistance to mecillinam (13%), nitrofurantoin (7%) and fosfomycin (3%) was low, although resistance was high in K. pneumoniae (25%, 60%, 85%). Susceptibility to ertapenem was 99%, piperacillin-tazobactam 91%, tigecycline 96% and temocillin 76%. Susceptibility rates obtained with broth microdilution and Etest were in agreement for cefotaxime (2 vs 1%) and ceftazidime (9 vs 11%), but not for piperacillin-tazobactam (59 vs 91%). With disc diffusion major errors occurred with piperacillin-tazobactam (18/169). Several therapeutic alternatives exist for ESBL-producing E. coli, but few exist for K. pneumoniae. Disc diffusion and Etest can accurately predict susceptibility to cefotaxime and ceftazidime, but not to piperacillin-tazobactam with the present breakpoints.