A large genomic epidemiology study by scientists in the United Kingdom has found that most bloodstream infections caused by drug-resistant Escherichia coli involve human-associated strains of the pathogen, with little contribution from the food chain.
The study, published yesterday in The Lancet Infectious Diseases, found that the extended-spectrum beta-lactamase-producing (ESBL) E coli sequence type (ST) 131 was the predominant strain found in bloodstream isolates, as well as in samples collected from human feces and sewage, while isolates from meat, veterinary diagnostic samples, and farm runoff were dominated by other ESBL E coli sequence types. Few drug-resistant E coli strains were shared among the animal and human isolates.
The authors of the study say the findings suggest that while ESBL E coli strains are widespread in humans, animals, and the environment, there's little crossover between these strains, and efforts to reduce invasive ESBL E coli infections should focus on limiting human transmission.
Comparison of resistant E coli reservoirs
The aim of the study was to examine the role of the food chain in transmitting drug-resistant E coli to humans, an issue that is being increasingly scrutinized with the emergence of antibiotic resistance as a global health threat. While the connection between E coli strains that cause diarrhea and other intestinal illness and food products is well established, in recent decades researchers have been documenting urinary tract infections (UTIs) caused by E coli strains that also appear to be linked to exposure to food-producing animals.
These extraintestinal infections can occur when people consume contaminated meat or vegetables and the bacteria moves from the gut—where it may be harmless—into the urinary tract, causing an infection. If the infection progresses into the kidneys, it can then spill over into the bloodstream. And if the E coli strain is antibiotic-resistant, it increases the risk of treatment failure and mortality.
The question is how often these infections involve animal-related E coli strains, and what particular strains of disease-causing E coli have been passing between animals and humans. With E coli now the most common bacterial bloodstream pathogen in England, and roughly 10% of those infections caused by ESBL-producing E coli (which are resistant to penicillins and cephalosporins), these have become critical public health questions.
To answer these questions, a team led by researchers from Public Health England and the University of East Anglia set out to create a national map of ESBL E coli reservoirs and compare the bacteria from these reservoirs with ESBL E coli isolated from clinical bloodstream infections. In 2013 and 2014, they collected 293 ESBL E coli bloodstream isolates from National Health Service labs in five UK regions (London, East Anglia, Northwest England, Scotland, and Wales), along with human feces, sewage, raw meat, fruits and vegetables, dairy farm slurry, and veterinary diagnostic samples from the same regions.
They then sequenced the recovered E coli isolates from the samples and compared the results with the sequenced bloodstream and veterinary diagnostic isolates.
Overall, the team found ESBL E coli in 11% of the 20,243 fecal samples and 65% of 159 chicken samples, but rarely in beef and pork (2% and 3% of 159 and 79 beef and pork samples, respectively) and not at all in fruits and vegetables. Twenty-eight percent of 97 dairy farm slurry samples were positive for ESBL E coli. And although the researchers could not accurately calculate the ESBL E coli prevalence in sewage, a collection of 65 ESBL E coli isolates from sewage was assembled.
Comparison of the sequenced isolates revealed that ST131—an epidemic, multidrug-resistant E coli lineage that is one of the predominant antibiotic-resistant uropathogens circulating in hospitals—accounted for nearly two thirds (64%) of the 293 ESBL E coli bloodstream isolates. It was also the most prevalent strain in feces (36% of 360 isolates) and sewage (22% of 65 isolates). Two other strains—ST38 and ST648—were also widespread in the human-related sources.
The animal-related ESBL E coli isolates, on the other hand, were dominated by different sequence types, mainly ST602, ST23, ST117, and ST10. The top-ranked sequence types in the human-related isolates were rare in the meat, slurry, and veterinary diagnostic samples, and the predominant animal-related strains were rare in humans. While there was overlap with one sequence type—ST10—further analysis found significant differences in the human- and animal-related ST10 isolates.
"In short, what the results show is that there are human-adapted strains of ESBL-E. coli, principally ST131, which dwell in the gut and which occasionally—usually via UTIs—go on to cause serious infections. And that there are animal strains of ESBL-E. coli," lead study author David Livermore, PhD, of the University of East Anglia, said in a university press release. "But—and critically—there's little crossover between strains from humans, chickens and cattle."
Livermore and his colleagues say the findings indicate that human-adapted ESBL E coli are likely being transmitted through the human-to-human oral-fecal route, rather than through the food chain. They say this could account for the higher prevalence of ESBL E coli in fecal samples from London, where the pathogen was found in 17% of fecal samples (compared with 8% to 9%) in other regions. The London samples all came from a hospital that predominantly serves low-income, crowded neighborhoods.
That means, they add, that one of the measures to stop transmission should be an emphasis on better hand hygiene after going to the bathroom, especially in nursing homes, since the most severe E coli infections occur in the elderly.
"In order to limit serious, antibiotic resistant E. coli bloodstream infections, we must focus on thorough hand washing and good infection control, as well as the effective management of urinary tract infections," said study co-author Neil Woodford, PhD, of Public Health England.
Animal-related E coli still a threat
But the study doesn't rule out the possibility that some human ESBL E coli infections are coming from food or animal-related sources, or that future multidrug-resistant E coli lineages could emerge from food animals and become adept at infecting people. The authors cite previous studies in which nearly identical strains of E coli have been isolated from retail chicken meat and human infections. They also encourage efforts to counter the spread of ESBL E coli in animals.
In a related commentary, Amee Manges, MPH, PhD, of the University of British Columbia addresses what the study might have missed. She notes, for example, that contemporaneous sampling of human, food, and environmental isolates doesn't account for past links between human- and animal related E coli strains.
In other words, human-adapted strains of ESBL E coli could have emerged, at some point in the past, from animal sources. An example of this comes from a study published last year in mBio, which found that a sublineage of ST131 (ST131-H22) that has been circulating in poultry since the 1940s and is prevalent in chicken and turkey meat accounted for a small percentage of human UTIs.
Manges also argues that the current study, while sampling a large and diverse collection of reservoirs for bloodstream infection-causing E coli, represents only a fraction of the total diversity of circulating E coli bacteria and ignores non-ESBL E coli found in animals, humans, and the environment that are highly genetically similar. She concludes that surveillance systems that integrate human and animal health systems and food inspection agencies are needed to detect newly emerging multidrug-resistant E coli lineages that could threaten public health.
"A single introduction of a highly successful resistant clone might suffice to trigger the next (non-ST131) pandemic," Manges writes.
Oct 22 Lancet Infect Dis study
Oct 22 Lancet Infect Dis commentary
Oct 22 University of East Anglia press release
Aug 28, 2018, CIDRAP News story, "Analysis ties resistant E coli from poultry meat to human UTIs"