Leading and Following: Busting a Move for Breakpoint Interpretations
I was recently introduced to leadership lessons from a shirtless dancing guy. If you’ve not seen this video, stop reading and watch it! I’ll wait here while you do. No, seriously – you have to watch the video before you read the rest of this blog – it’s only 3 minutes of your life.
All done? First off - how awesome was that? (Although, shirtless dancing guy seriously needs some sunscreen. I’m concerned.)
Secondly- HOW AM I JUST FINDING OUT ABOUT THIS NOW?! I feel like I need to go back and re-write every blog on AST I’ve done so far! For example:
Colistin testing – for lone nuts or leaders?!
ESBL testers- spandex-clad weirdos - or, luminary geniuses?
Reporting MICs without a breakpoint- embarrassing exhibitionists, or limber groovers?
Putting something on the internet is akin to writing it in stone (maybe worse since stones aren’t particularly portable), and rewriting all my old posts is not possible. Nonetheless, I’m going to revisit a recent topic I covered on the Bugs & Drugs Blog – the issue of the “intermediate” category. More specifically, the use of “susceptible-dose dependent” (aka S-DD), which is particularly topical given the deliberations at the CLSI AST Subcommittee meeting in June on the fate of intermediate.
The first application of S-DD as an interpretive category for susceptibility testing was done by the outliers of the susceptibility testing world – the antifungal group (no offense Shawn and Audrey!). This decision is nicely explained in this paper by Art Barry. Because not many people were doing fungal susceptibility testing in the late 1990s, and only a very select group of physicians were even looking at susceptibility reports for Candida spp. (how things have changed!), this decision sort of went under the radar.
The second use of S-DD (the first follower, if you like) was suggested by Dr. John Rex, as a possible resolution to the CLSI AST Subcommittee’s struggles on how to best revise the cefepime breakpoints for the Enterobacteriaceae, given the myriad of cefepime dosing regimens that are FDA approved and/or used clinically.
As a bit of background, breakpoints are defined based on the dose of drug administered, as this impacts the exposure of the pathogen to the drug in the host. This exposure is what ultimately determines if a treatment will be successful or not (ok, some other things come into play, but don’t tell the PK/PD mafia that). The activity of cefepime, like most beta-lactams, is best predicted by the percent of time the cefepime concentration in the host is above the pathogen’s MIC. This % T>MIC changes as one adjusts the dose or the dosing regimen. Bigger doses, more frequent doses, and extended infusions can all increase the % T>MIC. The actual % T> MIC needed to obtain a bacteriostatic (prevents growth) or bactericidal (kills the bugs) response varies by the organism group (one reason why all bugs don’t have the same breakpoint for a given drug). When several doses and/or regimens are FDA-approved, selecting the “best” breakpoint can be a challenge, as one needs to pick a dose to use in the PK/PD calculations.
The lowest approved dose of cefepime (1 g q 12) is predicted to treat infections caused by Enterobacteriaceae with MICs <=2 µg/mL. However, this breakpoint significantly limits the number of isolates that would be considered susceptible to cefepime, potentially pushing people to use a carbapenem (a drug class we want to preserve!). If the highest dose of 2g q8 h was applied to the model (a dose, incidentally, recommended by many infectious diseases pharmacists), the breakpoint predicted to define treatment success is <=8 µg/mL. This higher breakpoint would allow more isolates to be considered susceptible, and fewer people resorting to use of a carbapenem.
Unfortunately, the 1g q 12 dose is used to treat many serious infections caused by Enterobacteriaceae and the 2 g q 8h dose all too often only for its label indication – neutropenic fever. The probability of target attainment (50% T>MIC in this case) is very low (<1%) for an isolate with an MIC of 8 µg/mL when using 1 g q12 cefepime. In contrast, with a dose of 2 g q 8, it’s closer to 100% probability of target attainment. Historically, the CLSI would have addressed this issue through the use of “intermediate”—i.e., isolates with MICs of 4-8 µg/mL would be called intermediate.
However, for most physicians, “I” = “R”. As such, “S-DD” was applied for the first time in the bacteriology world, for isolates of Enterobacteriaceae with MICs of 4 – 8 µg/mL. The presence of S-DD on the laboratory report was thought to serve as an alert to providers that a higher dose was needed to treat such isolates. And, perhaps, since it was something “new,” people would actually pay attention, or at least phone a friend (hopefully an infectious diseases trained one) if they didn’t understand.
However – the going has not been easy for this first follower of the S-DD category:
- Laboratories have struggled with programming “S-DD” as an interpretive category to the LIS, and either continued with “I” or called it something new (like “D”).
- Some (including me) were concerned that S-DD implies there is no test variability (since there is no intermediate category) and that the reported MIC is “true.”
- Some felt doses should be maximized for all infections, and that a 1 g q 12 dose was never appropriate.
- Published data that the cefepime S-DD breakpoint was not appropriate for all Enterobacteriaceae led to confusion on treating S-DD isolates.
- Many just didn’t get it, and so S-DD wound up reading like R, as evidenced by the number of people just not using cefepime if the S-DD category was reported.
In fact, CLSI recently debated whether we should drop the S-DD category altogether, given the lukewarm reception in the bacterial AST community with cefepime. In the end, we decided to forge onward with the S-DD concept – due in large part to the second and third followers, which were approved by the CLSI AST Subcommittee for publication in the M100S 29th edition in January.
That’s right – there will be two more breakpoints that include the S-DD category!
- Daptomycin for Enterococcus spp. (S-DD at 2-4 ug/mL)
- Ceftaroline for Staphylococcus aureus (S-DD coincidentally also defined by MICs of 2-4 ug/mL)
Do we have an official S-DD movement?? Will M100 soon be wriggling with S-DD interpretive categories? I’ll confess – I’m still standing on the edge of this S-DD group, occasionally nodding my head to the tune, but also glancing around discreetly to make sure no one is watching my stunted S-DD grooves too closely.
Let’s start with the daptomycin S-DD breakpoint. First off, THANK GOODNESS (and maybe David Nicolau and Joe Kuti, and their fellows) that we finally have a revised daptomycin breakpoint for the enterococci!!!
Why does S-DD make sense for Enterococcus spp? It is pretty clear that the FDA-approved daptomycin doses are not effective for the enterococci. Even the “high” label dose of 6 mg/kg really only covers Enterococcus faecalis isolates. Of course, it’s not E. faecalis we want to treat with daptomycin, but rather vancomycin-resistant E. faecium. We know there is a good amount of data that demonstrate elevated doses – i.e., 10-12 mg/kg are associated with much better outcomes for E. faecium as compared to the label dose. The animal data that were originally used to establish the PD to inform Enterococcus breakpoint are somewhat suspect, given mice are so unaffected by E. faecium that the poor things just barely survive in the mouse, let alone need much daptomycin to kill them. This likely over-estimated the activity of daptomycin vs. E. faecium in the early days. When a more sophisticated clinical cut-off was used, the data supported a <=1 µg/mL susceptible breakpoint with a 6 mg/kg dose, and 2-4 µg/mL was predicted to be susceptible with a dose of 10-12 mg/kg. The rub here? Less than 10% of E. faecium have MICs of 1 µg/mL or lower, effectively making all E. faecium S-DD or R to daptomycin, pushing providers to use that higher, more appropriate dose. The MICs of E. faecalis are lower, allowing the 6 mg/kg to be used for this species. Isn’t it nice how that worked out?! In reality, I think it makes the most sense to think of E. faecium as ONLY being S-DD to daptomycin, and to use higher doses, which appear to be tolerated, although off-label.
What about ceftaroline? Suffice to say, outside the US, a higher ceftaroline dose (600 mg q 8h, infused over 2 h) is approved for skin and skin structure infections, and these are predicted to treat isolates with an MIC of 2 – 4 µg/mL. In South America and Asia, several isolates of S. aureus have MICs of 2 µg/mL (these are rare in the U.S.), and this change to the breakpoint will allow these to be treated with ceftaroline, at the higher dose.
Sounds great, right? So why am I not yet busting out some awesome S-DD moves? (Like the floss? Which, as an aside is MUCH harder than it looks—despite what my 10-year old son says). Here is my major reservation: susceptibility testing is challenging. An “MIC” is never an absolute value, and I worry that assigning a dose to an MIC is somewhat misleading. The intent is to not use dose X for MIC Y, but rather if the MIC is in the S-DD range, the maximal dose should be used. Both daptomycin and ceftaroline have significant testing issues. To demonstrate this, for fun, Kyle Spafford and I threw the MIC variability into a Monte Carlo simulation to look at probability of target attainment, with this additional variable. We did this by using data compiled by Joe Kuti for the recent Enterococcus daptomycin breakpoint revision. MIC variability data was based on a collection of 40 E. faecium isolates tested in replicate for daptomycin MICs by broth microdilution using 3 brands of cation-adjusted Mueller-Hinton broth, across three testing laboratories. The spread was pretty incredible, with some isolates having 6, 7 or even 8 MICs.
The bottom line: things look much better when the higher dose is used. So, perhaps rather than trying to finesse an imperfect test too drastically, we should focus on maximizing the dose, in conjunction with evaluation of the clinical scenario.
But then again, maybe I am the one standing alone and looking ridiculous, missing out on the awesome S-DD dance party.
The above represent the views of the author and does not necessarily reflect the opinion of the American Society for Microbiology.