A 21-year survey of Escherichia coli from bloodstream infections (BSI) in a tertiary hospital reveals how community-hospital dynamics of B2 phylogroup clones influence local BSI rates

This is a longitudinal study comprising 649 Escherichia coli isolates representing all 7,165 E. coli bloodstream infection (BSI) episodes recorded in a hospital (1996 to 2016). Strain analysis included clonal identification (phylogenetic groups/subgroups, STc131 subclades, pulsed-field gel electrophoresis [PFGE], and whole-genome sequencing [WGS]), antibiotic susceptibility (13 antibiotics), and virulence-associated genes (VAGs; 29 genes). The incidence of E. coli BSI increased from 1996 to 2016 (5.5 to 10.8 BSI episodes/1,000 hospitalizations, average 7 to 8/1,000). B2 isolates predominate (53%), with subgroups B2-I (STc131), B2-II, B2-IX, and B2-VI representing 25%, 25%, 14%, and 9%, respectively. Intertwined waves of community-acquired (CA) plus health care-associated and community-onset health care-associated (HCA) and hospital-acquired (HA) episodes of both B2 and non-B2 phylogroups occurred. A remarkable increase was observed only for B2-I-STc131 (C1/C2 subclades), with oscillations for other B2 subgroups and phylogroups throughout the years. Epidemic and persistent clones (comprising isolates with highly similar/identical PFGE types and genomes differing in 6 to 173 single nucleotide polymorphisms [SNPs]) of B2-I (STc131), B2-II (STc73), B2-III (STc127), B2-IX (STc95), and B2-VI (STc12) were recovered from different patients, most at hospital admission, for long periods (2 to 17 years), and extended-spectrum beta-lactamase (ESBL) producers or resistance to ciprofloxacin in B2 isolates was almost restricted to B2-I (STc131) subclade C. STc131 contributed to increasing the B2 rates but only transiently altered the E. coli population structure. The increase of E. coli BSI was determined by waves of CA1HCA BSI episodes that predate the waves of HA BSI. Besides the risk of hospital transmission that led to temporal increases in BSI, this study suggests that E. coli populations/clones from community-based healthy individuals may occasionally have an epidemic structure and provide a source of transmissible strains influencing the HA BSI incidence. IMPORTANCE Sepsis is the third leading cause of mortality in Western countries and one of the Global Health Threats recognized by the WHO since 2017. Despite Escherichia coli constituting the most common cause of bloodstream infections (BSI), its epidemiology is not fully understood, in part due to the scarcity of local and longitudinal studies. Our work analyzes the long-term dynamics of E. coli causing bacteremia in a single institution and reveals waves of different clonal lineages that emerge periodically and successfully spread afterward in both the community and hospitals. Because the origin of E. coli bloodstream infections is the gut, the microbiota of healthy individuals might occasionally have an epidemic structure, providing a source of E. coli strains to influence the incidence of hospital BSI. The study complements previous fractionated observations focusing on specific E. coli lineages or antibiotic-resistant isolates in the last decades and helps to understand the epidemiology of E. coli BSI and the dynamics of pandemic clones.