The OxIMP network met on Thursday for the first of this term's seminars. Andrew Singer, Senior Scientist at the Centre for Ecology and Hydrology started things off with a talk about antimicrobial resistance (AMR) in the wider environment, and what it means for environmental policy.
AMR is now widely recognised as a pressing concern for public health agencies worldwide. In the U.K. the government has prioritised policy responses that focus on surveillance, reducing infection, the prudent use of antibiotics, and education. However, current policy responses have largely been limited to "upstream" issues around the use of antibiotics, and have paid little attention to the "downstream" issues of AMR in the wider environment.
Andrew told us that 30-90% of the antibiotics we take are excreted into the sewage system. They tend to pass through the sewage system unaltered and enter the UK's rivers. Andrew's work (based on analyses of sediment in rivers) reveals that proximity to a sewage works explains up to 50% of the prevalence of AMR genes in a river - and this is despite the variable and weather-dependent effects of agricultural run-off.
EU regulation is based on the notion that the threshold for intervention is when the concentration of antibiotic is high enough to kill more than 5% of the bacteria present. The 5% threshold is an arbitrary risk assessment measure, is based on the assumption that bacteria in the environment will respond exactly as those in a laboratory (with no supporting evidence that this is true), and tells us nothing about what that concentration of antibiotics does in terms of selection pressure for resistance genes in bacteria. Recent research suggests that the "minimum selective concentration" (the amount of antibiotic required to provide a selective advantage to a resistant microbe relative to a nonresistant microbe) could be ten times lower than the 5% mortality threshold. On that measure, our rivers are in trouble. For example, looking at the Thames catchment, three quarters of the rivers are "at risk" of quinolone (a type of antibiotic) resistance. If you wanted to reduce that level to say 10% of rivers having resistance pressures, current calculations using Andrew and his colleagues' model suggests that antibiotic use in humans would need to be reduced by as much as 80-88% - an unrealistic number in contemporary public health.
The scale of AMR selection pressures raises significant questions about how policy can address these environmental concerns - particularly when you also consider the selection pressures introduced to the environment by biocides in many different forms. Biocides act in a similar way to select for AMR in microbes, because resistance genes often confer 'co-resistance' and 'cross-resistance' (i.e. the group of genes, or single gene, that is selected for by resistance pressures confers resistance to both biocides and antibiotics). And in fact, the use of metals as biocides might by far more important in co-selecting for resistance in the environment than antibiotic use in humans and animals.
The missing link in this story is transmission of AMR bacteria from the environment to humans. This is much harder to show, and is a serious data gap at present- we just don't know how much AMR in the environment links to resistant bacterial infections in humans. There are some very credible reasons to suggest this link exists, and some clues towards transmission pathways, but Andrew suggested that at present the uncertainty itself is perhaps reason enough to invoke the precautionary principle.
After the talk, Andrew and members of the OxIMP network reconvened to discuss some of the issues raised by his fascinating talk, including reducing antibiotic and biocide usage, quantifying environmental transmission, rethinking sewage systems and pollution policy, and the prospects for long term sustainable health systems.
The next seminar in the series is on Wednesday 15th February with Hannah Landecker - more details are available here.