During the past month, the Environmental Working Group (‘EWG’) released a report (pdf) about 2011 Retail Meat Annual Report of the National Antimicrobial Resistance Monitoring System (‘NARMS’; pdf). In short, EWG mistated the problem, demonstrated some astonishing ignorance (or cynicism), and, most importantly, has missed the boat on how NARMS could be used, with appropriate political pressure, to actually demonstrate a ‘farm-to-fork’ transfer of antibiotic resistance.
Let’s start with some of the things EWP got wrong (boldface mine):
We do not believe that EWG fully considered important factors that put these results in context, including:
- whether the bacterium is a foodborne pathogen. The report highlights resistance to Enterococcus, but this is not considered a foodborne pathogen. Instead, we include it because its behavior is helpful in understanding how resistance occurs.
- which drug(s) the bacterium is naturally resistant to. For example, most Enterococcus faecalis is naturally resistant to the antibiotic class of lincosamides. Because we know and expect to see this resistance, we are not as concerned with resistance in this species the way we would be with resistance in true pathogens like Salmonella and Campylobacter.
- why NARMS includes certain drugs in its testing design. We include some antibiotics for epidemiology purposes– to track the spread of certain bacteria or certain genes. But resistance to these antibiotics doesn’t reflect a danger to public health.
- whether the antibiotics that are commonly used to treat patients are still effective. NARMS data indicates that first-line treatments for all four bacteria that we track (Salmonella, Enterococcus, Escherichia coli and Campylobacter) are still effective.
- what the 2011 data indicate relative to similar data reported for prior years.
Additionally, we believe that it is inaccurate and alarmist to define bacteria resistant to one, or even a few, antimicrobials as “superbugs” if these same bacteria are still treatable by other commonly used antibiotics. This is especially misleading when speaking of bacteria that do not cause foodborne disease and have natural resistances, such as Enterococcus.
The last part about the misuse of the term superbugs really, ahem, bugged me. Every isolate they looked at, every single one, could be treated. In addition, among E. coli and Salmonella, there were no ESBL or carbapenem-resistant isolates (these are usually resistant to all but one or all antibiotics, respectively).
There is some good and bad news from the report. The good news:
- In the critically important class of antimicrobials, the 2011 data showed no fluoroquinolone resistance in Salmonella from any source. This is the drug of choice for treating adults with Salmonella.
- Trimethoprim-sulfonamide is another drug used to treat Salmonella infections and resistance remains low (0% to 3.7%).
- Fluoroquinolone resistance in Campylobacter has stopped increasing and remained essentially unchanged since the FDA withdrew the use of this drug class in poultry in 2005.
- Macrolide antibiotic resistance in retail chicken isolates remains low, with 2011 results at 0.5% of Campylobacter jejuni and 4.3% of Campylobacter coli. The macrolide antibiotic erythromycin is the drug of choice for treating Campylobacter infections.
These are the findings that bother me (boldface mine):
- Multidrug resistance is rare in Campylobacter. Only nine out of 634 Campylobacter isolates from poultry were resistant to 3 or more antimicrobial classes in 2011. However, gentamicin resistance in Campylobacter coli markedly increased from 0.7% in 2007 (when it first appeared in the NARMS retail meat report) to 18.1% in 2011. Gentamicin has been suggested as a possible second-line therapy for Campylobacter infections, although it is not commonly used.
- Resistance to third-generation cephalosporins, which are used to treat salmonellosis, has increased in Salmonella from chicken (10 to 33.5%) and turkey (8.1 to 22.4%) meats when comparing 2002 and 2011 percentages. FDA noted this development in previous years and has already taken action by prohibiting certain extra-label uses of cephalosporins in cows, pigs, chickens and turkeys, and is continuing to closely monitor resistance to these drugs.
While there is increased betalactam resistance (betalactams are derivatives of penicillin), we are not observing any ESBL or carbapenem resistance (newer derivatives of penicillin). What is disturbing is that there seems to be some sort of selective pressure leading to an increase of resistance to betalactams. If an ESBL or carbapenem resistant isolate were to get to the farm from the clinic (note the direction), this selection pressure could very well lead to an increase of these resistant strains (or the genes that encode these resistances) on the farm. In turn, this reservoir could then ‘reseed’ the hospital and cause resistant infections–something we’ve already seen with S. aureus ST398.
This is the nightmare scenario regarding drug resistance. Instead, EWP focuses on what’s on the meat as ‘making you sick’, which is either incendiary fear-mongering to achieve a desirable policy outcome or a fundamental misunderstanding of the scope of the problem. If resistance to foodborne pathogens were spiraling out of control, that would be showing up in the hospital data. That’s not what NARMS is about.
Much of what NARMS is trying to do is track the spread of resistance from the farm to the clinic–and, from personal experience, I can tell you that industry is terrified of NARMS accomplishing this. For years, they have spent a lot of time lobbying to prevent NARMS from being able to focus resources on this problem. Based on some data I’ve recently seen, if NARMS could get this aspect of its work adequately funded so they could perform the required genetic analyses, the links between agricultural and clinical resistance would be pretty devastating. That’s where groups like EWP should be putting their efforts–because the industry lobbyists never quit and will always pounce on inaccurate science.
On a related front, industry lobbyists have successfully shuttered the program that tracked actual foodborne pathogens. Didn’t hear from EWP then.
Reducing the use of agricultural antibiotics is critical (and something I’ve been a part of in the past). But to be effective (at policy, not fundraising), there needs to be a strategy, one that isn’t based on inaccurate hyperbole, but on good science.