Infections due to Escherichia coli producing extended-spectrum beta-lactamase (ESBL) or CMY-type beta-lactamase (CMY) are increasingly observed in non-hospitalized patients. The origin of these organisms is uncertain, but retail meat contaminated with E. coli may be a source. In the present study, clinical information and strains collected from patients infected or colonized with ESBL-producing and CMY-producing E. coli at hospitals in Pittsburgh, USA and Seville, Spain were investigated. Retail meat purchased in these cities was also studied for the presence of these organisms. Twenty-five and 79 clinical cases with ESBL-producing E. coli and 22 cases and one case with CMY-producing E. coli were identified in Pittsburgh and Seville, respectively. Among them all, community-acquired and healthcare-associated cases together constituted 60% of the cases in Pittsburgh and 73% in Seville. Community-acquired cases were more common in Seville than in Pittsburgh (49% vs. 13%; p <0.001). ESBL-producing and CMY-producing E. coli isolates were commonly recovered from the local retail meat. In particular, 67% (8/12) of retail chickens in Seville and 85% (17/20) of those in Pittsburgh contained ESBL-producing and CMY-producing E. coli isolates, respectively. Among the ESBL-producing isolates, CTX-M and SHV were the most common ESBL types in both clinical and meat isolates. Approximately half of the ESBL-producing and CMY-producing E. coli isolates from meat belonged to phylogenetic groups associated with virulent extra-intestinal infections in humans. Community and healthcare environments are now significant reservoirs of ESBL-producing and CMY-producing E. coli. Retail meat is a potential source of these organisms.

No abstract

Escherichia coli is a common etiologic agent of intra-abdominal infections (IAI) (1) associated with a heavy bacterial burden (2) and of urinary tract infections and bacteremia (3-5), among others. The recommendations for empirical antimicrobial therapy for high-risk, severe, community-acquired, extrabiliary complicated IAI and health care-associated complicated IAI include piperacillin-tazobactam or carbapenems (6). Recently, data from an observational study suggested that amoxicillin-clavulanate and piperacillin-tazobactam are as effective as carbapenems in patients with bacteremia due to extended-spectrum ␤-lactamase (ESBL)-producing E. coli; however, most of the patients in that study had urinary or biliary tract infections and the number of patients with heavy bacterial loads was probably low (7). In this context, it is important to note that, in some areas, many ESBLproducing and non-ESBL-producing isolates of Klebsiella pneumoniae and E. coli are susceptible in vitro to piperacillin-tazobactam and imipenem, but the MICs may increase substantially if the susceptibility tests are performed with an inoculum concentration higher than the standard one (8-11). Recently, we have demonstrated, using a microdilution method and time-kill assays, that the reduction of in vitro activity at a high inoculum concentration does not occur with the amoxicillin-clavulanate combination against clinical non-ESBL-producing and ESBL-producing E. coli strains (9). Previously, when several Enterobacteriaceae species produced complex beta-lactamase patterns or ESBL (12, 13), both piperacillin-tazobactam and carbapenem compounds showed increased MICs when inoculum concentrations higher than the standard one are used. On the other hand, some authors have raised doubts about the existence of an inoculum effect (14), but in fact, little is known about the real in vivo impact of this effect. Thus, the objective of this study was to compare the in vivo efficacies of amoxicillin-clavulanate, piperacillin-tazobactam, and imipenem against different tissue bacterial concentrations of two clinical E. coli strains in an experimental murine sepsis model.

No abstract