Lizard-Tailed Squirrels: Do Taxonomical Challenges Influence Antimicrobial Breakpoints?
This summer, my family spent a couple of days in Paris. We spent one evening picnicking and relaxing at the foot of the Eiffel Tower, waiting for the sun to set. At one point, my son, who was rummaging through the bushes as nine-year-old boys do, called over to us about a squirrel he found. "It's super cool, mom--it has a lizard tail," he exclaimed, convinced he had discovered a new squirrel species that would lead to his fame and fortune.
Let's take a step back and evaluate how clinical breakpoints are established. Traditionally, the Enterobacteriaceae breakpoints are lumped into one big group, under the premise that all members of the family behave in relatively the same way. The data for the breakpoints are driven by the usual suspects: E. coli, K. pneumoniae and E. cloacae, for both in vivo human outcome data and animal pharmacodynamics experiments. Similarly, in vitro MICs are evaluated, with the bulk of isolates evaluated being these three heavy hitters. Generally speaking, while variability of around 10% exists even among different isolates of the same species, no real data have emerged to indicate that a problem exists with this approach. Indeed, PK/PD experts, like Dr. David Nicolau, who serves on the CLSI AST Breakpoint Working Group, feel pretty comfortable extrapolating these data to other members of the family: "However, one cannot go so far as to extrapolate PK/PD data from E. coli and Klebsiella to organisms like Pseudomonas aeruginosa or Acinetobacter baumannii." This is one problem associated with setting breakpoints--very few such PD studies have been done with species like A. baumannii. Fortunately, some newer antimicrobials that target multidrug resistant (MDR) gram-negative pathogens, like cefdericol, are starting to be evaluated with these species.Alright, I know you're wondering where I'm going with this. I promise--it's related to AST. Really, to do with naming conventions, and the conundrums they present when it comes to the performance and interpretation of AST. There has been a recent explosion in eager taxonomical splitters. Cutibacterium acnes anyone? (They may be cute, but they're still Props to me!) We laboratorians are in part to blame for this, with our eager adoption of MALDI-TOF mass spectrometers in the laboratory, giving us the ability to more accurately identify isolates to the species level, whether we should or not. Another recent taxonomical challenge is the re-classification of bacteria--for example, the addition of Plesiomonas to family Enterobacteriaceae (see right). Or, how about new members of this family? I've never even heard of Izhakiella, let alone Pseudocitrobacter spp. While the first is unlikely to be isolated from clinical specimens (unless you happen to be in the practice of culturing sick mirid bugs), the latter is has been found as a colonizer of patients in Pakistan, harboring the NDM-1 gene to boot. If we were to recover one of these species, is it appropriate to interpret AST data using the Enterobacteriaceae breakpoint?
If all members of the Enterobacteriaceae behave the same way, why then are the fluoroquinolone breakpoints for Salmonella different from the rest of the Enterobacteriaceae? Breakpoints, incidentally, that the CDC suggested should also apply to Shigella spp. This reflects the fact that breakpoints are a continually evolving entity, with changes being driven by clinical data, emerging resistance mechanisms and newer analytical methods. The Salmonella breakpoint, for example, was established in response to reports of fluoroquinolone treatment failures for typhoid fever in cases with "susceptible" MICs. So does that mean that Salmonella spp. respond differently than the rest of the Enterobacteriaceae to fluoroquinolones? The answer is likely, no. Newer PK/PD analyses indicate the fluoroquinolone breakpoints for the whole of the Enterobacteriaceae need to be adjusted downward, a proposal under review with CLSI and already adopted by EUCAST and USCAST.
Let's complicate things a little more, and look at how AST devices are cleared by FDA. The FDA looks at the species that are listed in the indications for use for the given antimicrobial agent, and only allows device clearance for those species. So, for example, you might have an Enterobacteriaceae breakpoint, but if a Serratia marcescens infection was not seen in the clinical trial for the drug, there would be no S. marcescens indication for the drug, and you could not get an AST device (Vitek2, Microscan, etc) approved for that drug and S. marcescens. This poses a challenge, as performance issues for the system and these given species may not be communicated to the laboratory. To be fair, these are listed as limitations in the package insert for the AST device (or programmed in the system to not be reported). Such restrictions are not at all helpful, as clinical infections do occur due to these species, and knowledge of MIC data for these is more helpful than no data.
Where does that leave us with new species assigned to the Enterobacteriaceae or those recently renamed? You'll note there is no footnote anywhere in the CLSI M100S 27th Edition Table 2A-1 that specifies which Enterobacteriaceae are included in the breakpoints--so, maybe we can just apply the breakpoint and call it a day. As my colleague (and all-around antibiotic badass) Karen Bush told me when I asked her about this, no data exist to suggest this approach is wrong. After all, a lizard-tailed squirrel is really just a rat by a fancier name.I digress--my point is that if we can't get tests for our more familiar members of the Enterobacteriaceae, then the chances of getting one for a weirdobacterium member (as Dr. Lars Westblade fondly calls these bugs) is even more remote.
This is part of a monthly series from Romney Humphries covering specialized concerns in antibiotic susceptibility testing. The above represents the opinions of Dr. Romney Humphries, and is not necessarily representative of ASM.