Antimicrobial Agents: August 2018

Flucloxacillin was no better than cefazolin for MSSA bacteremia or endocarditis in a cohort of over 7,000 Australian patients. This study included every unique episode of methicillin-sensitive Staphylococcus aureus bacteremia identified at 27 Australian microbiology labs over a six-year period. The primary outcome of interest, 30-day mortality, was similar whether patients received flucloxacillin or cefazolin (11.2% vs 10.7%; p>0.05) and in propensity-adjusted analysis, mortality was nonsignificantly lower in the cefazolin group (aOR 0.86). The flucloxacillin recipients were more likely to be pediatric patients, less likely to be on hemodialysis, slightly less likely to have a device-related infection (33% vs 37%), and trivially more likely to have endocarditis (8% vs 6%); otherwise the groups were similar. These results agree with previous observational studies showing similar or slightly inferior outcomes when using antistaphylococcal penicillins versus cefazolin for MSSA bacteremia (a nice meta-analysis of studies on the topic was published in JAC this month: see 30085140).

This study adds two things to the neverending ASP versus cefazolin debate. One is a tremendous sample size; if cefazolin was inferior to antistaphylococcal penicillins for staphylococcal bacteremia, we really ought to have seen a statistically significant difference here. The second is data on treatment outcomes in MSSA endocarditis, which usually comes up as the prototypic “high inoculum infection” where ASPs ought to be superior on account of the cefazolin inoculum effect. In a subgroup analysis of the 571 patients who had endocarditis, there was no difference in 30-day mortality between fluclox and cefaolin, with a nonsignificant trend favoring cefazolin (10.6% died in the cefazolin group vs 15.6% in the flucloxacillin group; OR 0.64, which reduced to aOR 0.49 in multivariate analysis). Yes, this is retrospective data and could be influenced by selection bias. It would be wonderful to have an RTC to put this question to bed, and I'm told the Australians are working on it. 29499317

In the meantime, if you’re still worried about the CIE, a paper in AAC this month shows that adding clavulanic acid to cefazolin restores cefazolin’s bactericidal activity against MSSA strains that exhibit the CIE (disclaimer: this is another study from my colleagues over at UT-Houston). So, one approach could be to give amoxicillin-clavulanate along with cefazolin for a couple of days until the bacterial burden drops. Alternately, I’ve seen some experts advocating for using a staphylococcal penicillin until the patient’s blood cultures are sterile, then switching to cefazolin. 30150459

How are the clinical outcomes with ceftazidime-avibactam for OXA-48 type CREs? Pretty good! You’ll recall that there are several different families of carbapenemases, and that this is a clinically relevant distinction because patients in/from different parts of the world carry different carbapemenases producing different antibiotic susceptibility patterns. Briefly, carbapenemases fall into the beta-lactamase Ambler classes A, B, and D. Class A includes the classic “KPC” family, which makes up most of the CREs in the US and is susceptible to ceftazidime-avibactam. Class B includes the metallo-beta-lactamases like VIM, IMP, and NDM, which are resistant avibactam but do not confer resistance to monobactams; hence the popularity of the new sneaky combination of ceftazidime-avibactam and aztreonam, which overcomes the metallo-beta-lactamase with aztreonam and uses avibactam to protect aztreonam from other beta-lactamases. Class D includes the OXA family, which ceftazidime-avibactam should cover per the package insert. So, how did it do in practice?

This prospective observational study included 57 patients with OXA-type CRE infections, of which 28% were intraabdominal, 26% respiratory, and 25% urinary; 54% of the cohort met criteria for sepsis. The majority of patients (81%) received ceftazidime-avibactam as monotherapy; the median duration of therapy was 13 days. Infection-attributable 30-day mortality was 14% and not associated with monotherapy, the 90-day recurrence rate was 10%, and there was no evidence of resistance to the drug. Given that CRE infections are routinely associated with mortality rates in the 20-60% range (see this recent meta-analysis and specifically table 2), 14% 30-day mortality ain’t bad. 30099490

What is the most effective antimicrobial regimen for carbapenem resistant Acinetobacter baumannii (CRAB) pneumonia? The authors performed a retrospective cohort study including three Taiwanese hospitals over a 5-year period. They looked at ICU patients with pneumonia from whom CRAB had been cultured who were treated with at least one antibiotic to which the isolate was susceptible for at least 48 hours. Among 238 cases, the antibiotic regimens used were tigecycline monotherapy (35%), tigecycline plus colistin (18%), colistin alone (14%), colistin with some other antibiotic (14%), tigecycline with some other antibiotic (13%), and sulbactam (5%). In multivariate analysis, tigecycline-based therapy was associated with higher ICU mortality (aOR 2.4) and more treatment failure (aOR 2.51). Colistin monotherapy was associated with lower ICU mortality than tigecycline monotherapy (aOR 0.3), and overall colistin-based therapy was associated with lower rates of treatment failure (aOR 0.48).

It’s important to note that the patients in this study received the standard dose of tigecycline (100mg loading dose and then 50mg every 12 hours). Some experts recommend double-dosing tigecycline (i.e. 200mg load and then 100mg every 12 hours) for severe infections, and perhaps outcomes would have been better with this regimen. Also, while colistin-based regimens may have had the “best” therapeutic outcomes, outcomes were poor across the board; 35% of patients died in the ICU, 64% had treatment failure, and 20% of those successfully treated had recurrent CRAB pneumonias. So, while this study makes a case for using colistin over tigecycline for CRAB pneumonia, it’s really a cry for help. 29108947

Beta-lactams prevent Enterococcus faecium from exerting an inoculum effect against daptomycin. Disclaimer: this work came out of the UT Health McGovern medical school, across the street from Baylor, so I know and have attended lectures from some of the co-authors. Recent literature suggests that higher doses of daptomycin (e.g. 10 -12 mg/kg daily) are associated with better outcomes than standard dosing for enterococcal and staphylococcal bacteremias. One reason might be an inoculum effect – that is, higher bacterial concentrations exhibiting higher MICs to a drug.

The authors demonstrated that a daptomycin E. faecium strain with mutations common among clinical isolates indeed exhibited an inoculum effect in a simulated endocardial vegetation PK/PD model. At an inoculum of 10^9 cfu/g, daptomycin concentrations equivalent to 6-8 mg/kg daily led to selection for daptomycin resistance and bacterial regrowth, whereas bactericidal activity was retained for a 10^7 cfu/g inoculum; in contrast, daptomycin at concentrations equivalent to 10mg/kg dosing had bactericidal activity against both the high and low-concentration inoculums. This inoculum effect proved therapeutically relevant in a rat infection model. Interestingly, adding any one of several beta-lactams to the PK/PD model (ampicillin, ceftaroline, or ertapenem – all previously shown to have the best synergy with daptomycin for enterococcus) abolished the inoculum effect and prevented emergence of resistance with the lower doses of daptomycin.

I mention this study for three reasons. One, I thought it was fascinating that beta-lactams still had clinically relevant activity against Enterococcus faecium even though the organism had wildly high MICs to these drugs. Two, it gives us a mechanistic explanation for why higher doses of daptomycin might be associated with better outcomes in enterococcal bacteremia. Three, it suggests that combining daptomycin with ampicillin, ceftaroline, or ertapenem might be useful strategy for persistent E. faecium bacteremias, which are awful to deal with, particularly in elderly and multicomorbid patients deemed too sick for the surgical source control procedures they need to clear their infections. 29760141

Oral fosfomycin is poorly tolerated at daily and every other day intervals. A recent trial showed that a single 3g dose of fosfomycin was less effective than five days of nitrofurantoin for uncomplicated UTI. Naturally, someone has looked at whether patients might tolerate more frequent fosfomycin dosing, which, given that more than a third of the drug is excreted in urine within the first 24 hours, could significantly improve urine drug exposure. The authors randomized 18 healthy subjects to receive 3g of fosfomycin every other day or daily for a week, after which the groups crossed over for another week. Serum drug levels at day 1 and day 5 were almost identical between the groups, but for me, the bottom line is the adverse event profile: the percentage of diarrhea-free days was not great for either regimen (61% for daily versus 77% for every other day). While it’s an apples-to-oranges comparison, the total incidence of GI symptoms quoted in a review of adverse events with traditional dosing of fosfomycin was only 6.5% (source here; see table 6). So, increasing the dosing frequency of fosfomycin makes it a drug with a lot more GI side effects. 29891606