September 2019: Antimicrobial agents

Oral lefamulin is non-inferior to moxifloxacin for the production of very expensive diarrhea – uh, I mean, treatment of community-acquired pneumonia (CAP).  In August 2019 the FDA made lefamulin the first pleuromutilin antibiotic approved for systemic use (repatamulin, a topical antibiotic, has been around since 2007).  Pleuromutilins act by binding to the 50S subunit of the bacterial ribosome, much like clindamycin, linezolid, and macrolides.  Practically speaking, you can think of lefamulin as lovechild of azithromycin and linezolid – it has great gram-positive activity including MRSA and VRE, but also covers atypical organisms and prolongs the QT interval like a macrolide.  Lefamulin doesn’t need renal dosing but should be dose-reduced (IV form) or avoided (PO form) in advanced liver disease.

Anyway, the LEAP-2 trial randomized 738 patients with CAP (defined by radiography, >3 clinical symptoms of pneumonia, and >2 vital sign abnormalities) to receive either 5 days of oral lefamulin to 7 days of oral moxifloxacin.  The main outcome measure was “early clinical response,” defined as improvement in two or more symptoms of pneumonia within 4 days of receiving the first antibiotic dose with no worsening symptoms or need for additional antibiotics; investigator-assessed clinical cure at 5 to 10 days after completing therapy was a secondary endpoint.  The study used a non-inferiority design with the standard 10% noninferioritty margin.

A total 685 patients completed the trial; about a third were >65yo, half were women, half had baseline renal impairment, and half had CAP with a PORT risk class of III or greater.  In the mITT population, the rates of early clinical response were 91% in both groups (risk difference 0.1%, one-sided 97.5% CI -4.4% to ∞), and rates of cure at 5-10 days after therapy were 88% with lefamulin vs 89% vs moxifloxacin (risk difference -1.6% with one-sided 97.5% CI -6.3% to ∞).  Analysis by pathogen revealed that clinical response rates were lower with lefamulin among penicillin-susceptible pneumococcus (77% vs 95%); however, within the subset the patients receiving lefamulin had more PORT class >3 pnuemonias (69% vs 37%), so in the absence of a compelling biologic hypothesis for why lefamulin would fail here I don’t know that it means much. 

The kicker, though, was drug tolerability: 33% of lefamulin recipients had an adverse event vs 25% of the moxifloxacin recipients, largely explained by lefamulin’s higher incidence of diarrhea (12% vs 1%), nausea (5% vs 2%), and vomiting (3% vs 1%).  Add in the difference in cost - using the prices quoted in UpToDate, you’re looking a $1650 for a course of lefamulin vs $272 for an equivalent course of moxifloxacin (or, you know, $30 for amoxicillin plus doxycycline) – and I think it’s hard to see lefamulin as anything more than “the mother-in-law option” when it comes to routine treatment of CAP.  31560372

(High dose) ceftolozane-tazobactam works as well as meropenem for nosocomial pneumonia.   I think this is an odd study, because where I trained we only broke out ceftolo/tazo when we had a highly suspected or proven infection due to MDR P.aeruginosa, which I think is the proper stewardship of this drug.  And in that case, the question is not whether to use ceftolo/tazo rather than a carbapenem (to which MDR Pseudomonas seems invariably resistant- thanks, OprD), but whether to use ceftolo/tazo rather than last-line drugs like aminoglycosides and colistin – and that answer seems an obvious yes.  But I suppose it’s still nice to see that ceftolo/tazo works just as well any other antipseudomonal beta-lactam for HAP/VAP, particularly since the MERINO trial suggested the inferiority of another tazobactam-based drug versus meropenem for ESBLs.

In the ASPECT-NP trial, authors randomized 726 adults with VAP (a third with an APACHEII score >20; over 90% in the ICU) to receive either 3g ceftolo/tazo or 1g meropenem every eight hours for 8-14 days.  The primary outcome was 28-day mortality, and clinical response at 7-14 days after the end of therapy was recorded as a secondary endpoint; the study used a noninferiority design.  The authors found no difference in the rate of 28-day mortality with ceftolo/tazo versus meropenem (24% vs 25; weighted risk difference 1.1% with 95% CI -5.1% to 7.4%), meeting the noninferiority threshold.  Rates of clinical response were also identical (53% vs 54% with weight risk difference 1.1%; 95% CI -6.2% to 8.3%). 

The accompanying editorial (which, disclosure, is co-authored by one of my UNMC colleagues) points out that the rate of treatment-related serious adverse events was higher in the ceftolo/tazo arm (2% vs 1% with p >0.05 - keep in mind this study wasn’t powered to detect differences in safety outcomes).  Moreover, the non-inferiority of ceftolo/tazo was shown in the context of likely underdosed meropenem (2g IV q8h would be preferable for a critically ill patient with good baseline renal function) and a higher rate of proven treatment-resistant infections in the meropenem arm (13% vs 3% for ceftolo/tazo).  So on a fair playing field, maybe ceftolo/tazo could be inferior to meropenem for the treatment of VAP.  But, I tell you what: I’d still take it over amikacin plus colistin.  31563344

SPEAKING OF, Clinical Infectious Diseases had a retrospective multicenter cohort study out last month reporting the outcomes of MDR Pseudomonas aeruginosa infection treated with either ceftolo/tazo or polymyxin or aminoglycoside-based regimens.  A total 200 patients were included in the study; 69% were in the ICU and 42% were in septic shock when starting treatment.  More patients in the Polymixin or AG arm received combination antibiotics (72% vs 15% with p<0.001), and after adjusting for between-group differences, receipt of ceftolo/tazo instead of an alternative regimen was associated with clinical cure (aOR 2.63 with 95% CI 1.3-5.3) and a more than ten-fold lower rate of AKI (aOR 0.08 with 95% CI 0.03-0.22).  Mortality was also lower in the ceftolo/tazo arm (aOR 0.62 with 95% CI 0.3-1.2); the authors report this as no difference on account of the 95% CI crossing one, but IMO is this a great example of statistical significance and clinical significance not being the same thing.  We have plenty of reason to suspect that AG and polymixin-based treatment would be inferior to an antipseudomonal beta-lactam for severe MDR P. aeruginosa infections, so I think the burden of proof ought to be on anyone who wants to keep using their awful toxic salvage regimens. 31545346

Are bacteriophages the future of therapy for orthopedic hardware infections?  I don’t usually include case reports these days, but thought this one was worth sharing.  As I’ve begun to learn on the UNMC bone and joint ID service, prosthetic joint and other orthopedic hardware infections are an absolute disaster to manage when they’re due to highly antibiotic-resistant pathogens – or really, anything that isn’t streptococcus or a rifampin-susceptible staph.  But maybe that’s about to change?  In last month’s AAC, a group out of Germany reports their experience using adjunctive bacteriophage therapy in a patient with chronic relapsing knee arthroplasty infection and associated osteomyelitis due to MDR Pseudomonas aeruginosa.  The patient had two strains of PSA, one of which was resistant to all antibiotics other than colistin, and was managed with two-stage exchange using a gentamicin-impregnated antibiotics spacer (worth noting because perhaps local gent concentrations were enough to overcome the aminoglycoside resistance).  Anyway, the team applied phage during the explantation surgery and then instilled additional phage via surgical drain every eight hours for five days.  This was combined with triple IV therapy with colistin, meropenem, and ceftazidime, and yielded in microbiologic cure of the Pseudomonas infection.  The authors point out that at least some phages have been shown to work synergistically with antibiotics to erode biofilm – and in this regard, they may be particularly useful for treating orthopedic hardware infections.  31527029