Organic and conventional fruits and vegetables contain equivalent counts of Gram-negative bacteria expressing resistance to antibacterial agents.

Resistance to antibiotics is a major public health problem which might culminate in outbreaks caused by pathogenic bacteria untreatable by known antibiotics. Most of the genes conferring resistance are acquired horizontally from already resistant commensal or environmental bacteria. Food contamination by resistant bacteria might be a significant source of resistance genes for human bacteria but has never been precisely assessed, nor is it known whether organic products differ in this respect from conventionally produced products. We showed here, on a large year-long constructed sample set containing 399 products that, irrespective of their mode of production, raw fruits and vegetables are heavily contaminated by Gram-negative bacteria (GNB) resistant to multiple antibiotics. Most of these bacteria originate in the soil and environment. We focused on non-oxidative GNB resistant to third-generation cephalosporins, because of their potential impact on human health. Among them, species potentially pathogenic for immunocompetent hosts were rare. Of the products tested, 13% carried bacteria producing extended-spectrum beta-lactamases, all identified as Rahnella sp. which grouped into two phylotypes and all carrying the bla(RAHN) gene. Thus, both organic and conventional fruits and vegetables may constitute significant sources of resistant bacteria and of resistance genes.

Integrons and antibiotic resistance in phylogenetic group B2 Escherichia coli.

It has been reported that Escherichia coli B2 phylogenetic group strains are more susceptible to antibiotics, especially to quinolones, and tend to carry less integrons than other phylogenetic groups in commensal environments. To gain a better understanding of the relationships between antibiotic resistance, integrons, and phylogenetic groups in an environment with high antibiotic selective pressure, we compared these characteristics in three selected groups of urinary tract infection E. coli isolated in a university hospital (G1, G2, and G3). The isolates were fully susceptible to antibiotics, resistant to amoxicillin and cotrimoxazol, or resistant to amoxicillin, cotrimoxazol, and nalidixic acid in the G1, G2, and G3 group, respectively. The prevalence of B2 isolates was significantly lower in the most resistant G3 group (22.6%) than in susceptible G1 (57.8%, p < 0.001) and G2 groups (50%, p < 0.01). In contrast, the prevalence of B2 isolates was not significantly different between G1 and G2 groups. The prevalence of integrons was nil in G1 isolates but very high in G2 (94.3%) and G3 (87.5%) isolates, and integrons were equally distributed among the phylogenetic groups. We propose a step-by-step mechanism for the emergence of antibiotic resistance in E. coli. Under very low selective pressure, resistance emerges without integrons. When the antibiotic pressure increases, quinolone and integron-mediated resistance occurs outside phylogenetic group B2. With strong antibiotic selective pressure, integrons are highly prevalent and widespread regardless of the phylogenetic group.

Accuracy and potential usefulness of triplex real-time PCR for improving antibiotic treatment of patients with blood cultures showing clustered gram-positive cocci on direct smears.

Bacterial identification and antibiotic susceptibility testing currently require 48 h when a first blood culture (BC) is positive for clustered gram-positive cocci on direct smear examination (DSE). Meanwhile, antibiotic treatment is often inadequate, reducing the chances of effective treatment or creating unnecessary selective pressure. A new real-time PCR (RT-PCR) technique that differentiates Staphylococcus aureus from coagulase-negative staphylococci (CoNS) and detects methicillin resistance in 90 min in BC bottles could help solve these problems. BC bottles from 410 patients with gram-positive cocci on DSE were processed by current methods, and patients' treatments were prospectively recorded. The RT-PCR assay was performed on aliquots of these BCs, which had been kept frozen. For the 121 patients who had true bacteremia, we established whether the faster availability of RT-PCR results could have led to the initiation of treatments different from those actually given. RT-PCR sensitivity and specificity were 100% for differentiating between S. aureus and CoNS and detecting methicillin resistance with two manufacturers' BC bottles. For 31/86 (36%) of the S. aureus-infected patients and for 8/35 (23%) of the CoNS-infected patients who either had suboptimal or nonoptimal treatment or were untreated 48 h after positivity was detected, the early availability of RT-PCR results could have allowed more effective treatment. Unnecessary glycopeptide treatments could have been avoided for 28 additional patients. The use of RT-PCR would increase treatment effectiveness in patients with staphylococcal bacteremia and reduce the selective pressure created by glycopeptides.