There’s a new paper out in the CDC’s journal Emerging Infectious Diseases that makes a provocative claim: There is enough similarity between drug-resistance genes in  E. coli carried by chickens and  E. coli infecting humans that the chickens may be the source of it.

If it is correct — and it seems plausible and is backed by past research — the claim provides another piece of evidence that antibiotic use in agriculture has a direct effect on human health.

Here are the details:

The paper is a collaboration by researchers from several hospitals in the Netherlands, plus the Netherlands’ National Institute of Public Health and Environmental Protection, the University of Birmingham and a section of the UK’s National Health Service. They isolated E. coli from patients in four Dutch hospitals over 2.5 months in 2009, and compared those with E. coli strains isolated from randomly chosen supermarket meat that was bought in the hospitals’ local areas during the same time period. They compared both those sets of isolates against a third set, of E. coli from blood cultures taken from patients in the hospitals during the same months.

In each set of samples, they were looking at the E. coli to see whether they harbored genes for the type of resistance known as ESBL, for “extended-spectrum beta-lactamase,” an enzyme that denatures a category of drugs used for serious infections that occur mostly in hospitals. When the extended-spectrum beta-lactams no longer work, only a few last-resort drugs are left. (Back in the 1980s, the most common genes for ESBL were bla_TEM or bla_SHV, but in the past 10 or so years there has been a rapid global increase in the occurrence of a different ESBL gene, bla_CTX-M.)

Here’s what they found:

• Out of 876 patients tested by rectal swab — because E. coli is a gut bacterium, carried in and spread by feces — 45 (5 percent) harbored ESBL genes.
• Out of 31 blood cultures in the hospitals’ labs, 23 (74 percent) contained ESBL genes.
• Out of 262 meat samples, 79 (30 percent) harbored an ESBL gene. Broken down by type of meat, there was ESBL in 80 percent of the chicken samples, 5 percent of the beef, 2 percent of the pork, and 9 percent of ground or otherwise mixed meat.

When they broke down the organisms by type, they looked like this. Note the amount in each pie chart that is given over to the ESBL genes bla_CTX-M, and the significant correspondence of bla_CTX-M-1 in red.

When they put the genes through a second level of genetic analysis, multi-locus sequence typing, 57 percent of the rectal specimens and 57 percent of the blood cultures were closely related to the strains in the chicken meat.

There’s an important backdrop to this research. The Netherlands has one of the lowest rates of human antibiotic resistance in the world, thanks to especially stringent infection control and drug-conservation policies. Paradoxically, it has the highest rate of antibiotic use in agriculture in Europe. As a result, when something starts to move into humans, it is easier to distinguish, because there is no “background noise” of high rates of hospital and community drug resistance such as there are in the US. And because there are no competing resistance factors from other sources, it is easier to identify and explain.

Thus, the researchers can comfortably say:

We conclude that the high rate of ESBL contamination of retail chicken meat in the Netherlands, which involves many of the same ESBL genes present in colonized and infected humans, is a plausible source of the recent increase of ESBL genes in the Netherlands. The similarity of E. coli strains and predominant drug resistance genes in meat and humans provides circumstantial evidence for an animal reservoir for a substantial part of ESBL genes found in humans.

If something about this research sounds familiar, it’s because a similar study was published a few months ago, also from the Netherlands, with a partially overlapping analysis: chicken meat and blood-culture records, but no swabs from simultaneous patients. That study too found a high degree of correlation between ESBL-containing organisms in humans and in chickens.

These findings won’t come as a surprise to anyone who accepts — as most good science and a number of public health authorities do — that antibiotic overuse in large-scale farming creates drug-resistant organisms that affect human health. The question, for those who don’t accept such a link, is: How much evidence is enough?

(Footnote: In addition to being published in EID, this study was also presented by Dr. Jan Kluytmans, the senior author, during the World HAI Forum taking place this week in France. I’ll have more on the HAI Forum in a future post.)

Cite: Overdevest I et al. Extended-spectrum ?-lactamase genes of Escherichia coli in chicken meat and humans, the Netherlands. Emerg Infect Dis. 2011 Jul. http://www.cdc.gov/EID/content/17/7/1216.htm

See Also:

Flickr/StevenW./CC