LIDに報告された、CR-GNRに対する、Colistin vs Colistin + MEPM のRCT
Colistin–carbapenem combinations are synergistic in vitro against carbapenem-resistant Gram-negative bacteria. We aimed to test whether combination therapy improves clinical outcomes for adults with infections caused by carbapenem-resistant or carbapenemase-producing Gram-negative bacteria.
A randomised controlled superiority trial was done in six hospitals in Israel, Greece, and Italy. We included adults with bacteraemia, ventilator-associated pneumonia, hospital-acquired pneumonia, or urosepsis caused by carbapenem-non-susceptible Gram-negative bacteria. Patients were randomly assigned (1:1) centrally, by computer-generated permuted blocks stratified by centre, to intravenous colistin (9-million unit loading dose, followed by 4·5 million units twice per day) or colistin with meropenem (2-g prolonged infusion three times per day). The trial was open-label, with blinded outcome assessment. Treatment success was defined as survival, haemodynamic stability, improved or stable Sequential Organ Failure Assessment score, stable or improved ratio of partial pressure of arterial oxygen to fraction of expired oxygen for patients with pneumonia, and microbiological cure for patients with bacteraemia. The primary outcome was clinical failure, defined as not meeting all success criteria by intention-to-treat analysis, at 14 days after randomisation. This trial is registered at ClinicalTrials.gov, number NCT01732250, and is closed to accrual.
Between Oct 1, 2013, and Dec 31, 2016, we randomly assigned 406 patients to the two treatment groups. Most patients had pneumonia or bacteraemia (355/406, 87%), and most infections were caused by Acinetobacter baumannii (312/406, 77%). No significant difference between colistin monotherapy (156/198, 79%) and combination therapy (152/208, 73%) was observed for clinical failure at 14 days after randomisation (risk difference −5·7%, 95% CI −13·9 to 2·4; risk ratio [RR] 0·93, 95% CI 0·83–1·03). Results were similar among patients with A baumannii infections (RR 0·97, 95% CI 0·87–1·09). Combination therapy increased the incidence of diarrhoea (56 [27%] vs 32 [16%] patients) and decreased the incidence of mild renal failure (37 [30%] of 124 vs 25 [20%] of 125 patients at risk of or with kidney injury).
Combination therapy was not superior to monotherapy. The addition of meropenem to colistin did not improve clinical failure in severe A baumannii infections. The trial was unpowered to specifically address other bacteria.
EU AIDA grant Health-F3-2011-278348.
P : adults with bacteraemia, ventilator-associated pneumonia, hospital-acquired pneumonia, or urosepsis caused by carbapenem-non-susceptible Gram-negative bacteria
E : intravenous colistin (9-million unit loading dose, followed by 4·5 million units twice per day)
C : colistin with meropenem (2-g prolonged infusion three times per day)
O : No significant difference between colistin monotherapy (156/198, 79%) and combination therapy (152/208, 73%) was observed for clinical failure at 14 days after randomisation (risk difference −5·7%, 95% CI −13·9 to 2·4; risk ratio [RR] 0·93, 95% CI 0·83–1·03).
Combination therapy increased the incidence of diarrhoea (56 [27%] vs 32 [16%] patients) and decreased the incidence of mild renal failure (37 [30%] of 124 vs 25 [20%] of 125 patients at risk of or with kidney injury)
Six hospitals, three in Israel, two in Greece, and one in Italy.
RCT. ITT. open-label.
Our trial faced several limitations. We did not measure drug concentrations in real time to direct treatment, but administered colistin in doses that are expected to result in plasma concentrations that are e ective against colistin-susceptible bacteria in most patients.5 We did not use central microbiology laboratory MIC determinations to direct treatment; these measurements and analyses are ongoing and will be reported separately. Of the isolates in which MIC was determined locally, very few had MICs of 8 mg/L or less (9/290); thus, we cannot establish the e cacy of the combination therapy in the lower MIC range. Our primary composite outcome was not previously validated, and we observed very high failure rates (302/406, 74% overall; 252/312, 81% in A baumannii infections). In our trial, clinical failure re ected death or a non-improving SOFA score among survivors; thus, we consider our primary outcome clinically relevant, and the high rates re ective of disease severity. We did not identify di erences in the emergence of colistin resistance; however, resistance outcomes are di cult to assess in randomised trial design, requiring long-term follow-up after the implementation of a treatment policy. Our nding of a potential protective e ect of combination therapy on renal function was unexpected, and should be further investigated. Without biological rationale, the e ect on renal function should not constitute a reason to use combination therapy.
Evidence to improve the treatment of infections caused by carbapenem-resistant Gram-negative bacteria
WHO recognises carbapenem-resistant Acinetobacter baumannii, Pseudomonas aeruginosa, and Entero- bacteriaceae as pathogens of crucial importance for the development of novel antibiotics.1 These bacteria manifest resistance to our most trusted therapeutics and, as a result of their increasing prevalence and transmissibility, are becoming an overwhelming health-care burden.2 Over the past decade, polymyxins (an old class of cationic antimicrobial peptide antibiotics) became the drugs of last resort to treat carbapenem- resistant Gram-negative bacteria. Although advances have been made in understanding the pharmacology and toxicity of polymyxin B and polymyxin E (ie, colistin) many questions remain; among them, does the use of polymyxins in combination with other antibiotics lead to better clinical outcomes?3
In The Lancet Infectious Diseases, Mical Paul and colleagues4 report on the AIDA trial: a randomised controlled trial, at six sites in Israel, Greece, and Italy, comparing colistin and meropenem with colistin alone for the treatment of serious infections caused by carbapenem-resistant Gram-negative bacteria. In patients with mostly ventilator-associated pneumonia or hospital-acquired pneumonia and bloodstream infection caused by carbapenem-resistant A baumannii, the combination of colistin and meropenem did not result in fewer clinical failures than colistin alone. These results are in concordance with two randomised controlled trials in patients infected with carbapenem- resistant A baumannii treated with colistin in combination with rifampicin versus colistin alone,5,6 and with a smaller trial that randomised patients to receive either colistin plus fosfomycin or colistin alone.7 Given the evidence, the authors state that additional antibiotics should not be used in combination with colistin for the treatment of carbapenem-resistant A baumannii, to avoid potential drug-related adverse events, selection of resistant bacteria, and Clostridium di cile infection.
Does the AIDA trial close the door on combination therapy for the treatment of all carbapenem-resistant A baumannii or other Gram-negative bacteria? No. The molecular mechanism of carbapenem resistance in
A baumannii was not included in this analysis. Whether certain genotypes or phenotypes di er in their response to combination therapy remains a crucial question in the pharmacotherapy of Acinetobacter spp. Additionally, the pathophysiology of infection by A baumannii is di erent than other Gram-negative bacteria. The shedding of lipopolysaccharide and bacterial density modulate virulence and might be a ected by the type and timing of antibiotic therapy. Even the identi cation of A baumannii is challenging, because the taxonomy of the genus is quite complex.8 In the AIDA trial, were all infections caused by A baumannii, and were all patients infected by the same species of Acinetobacter? We believe these factors do not come into play with P aeruginosa or Enterobacteriaceae.
We are awaiting results of a randomised controlled trial,9 conducted in the USA, Israel, Taiwan, and Thailand, examining colistin plus meropenem versus colistin alone to treat carbapenem-resistant Gram-negative bacteria (NCT01597973, ClinicalTrials.gov). In the AIDA trial, the number of patients infected with carbapenem- resistant Enterobacteriacea or P aeruginosa was too small to draw conclusions about the bene t of combination therapy. Some outcomes (28-day mortality and clinical failure) were more favourable in the subgroup of patients with infections caused by Enterobacteriaceae treated with colistin–meropenem combination therapy than in patients treated with colistin alone; however, these di erences were not statistically signi cant. An analysis incorporating propensity score matching of a di erent retrospective cohort of patients with carbapenem-resistant Enterobacteriaceae bloodstream infection found that combination therapy was advantageous in patients who had a high pretreatment probability of death on the basis of a previously validated mortality score.10
The advent of new antibiotics might preclude the question of how to improve colistin-based therapy: ceftazidime–avibactam o ers superior outcomes than colistin-based regimens, as shown in an observational study of patients infected with carbapenem-resistant Enterobacteriaceae, speci cally Klebsiella pneumoniae carbapenemase (KPC)-producing K pneumoniae.11
Meropenem–vaborbactam is also active against carbapenem-resistant Enterobacteriaceae if the re- sistance is mediated by KPC-2 and KPC-3, but not by the oxacillinase OXA-48 or metallo-β-lactamases (eg, New Dehli metallo-β-lactamases and Verona integron-encoded metallo-β-lactamases).12 Similarly, ceftolozane–tazobactam can be used to treat carbapenem-resistant P aeruginosa in the absence of acquired carbapenemases.13 The avoidance of colistin- based regimens can be advantageous because it reduces the risk of kidney injury, which was measured at 30% in the AIDA trial.
The high patient mortality rate (44% at 28 days) in the AIDA trial is sobering—considering that infection with bacteria susceptibile to colistin was a criterion for inclusion and that colistin dosing was carefully controlled—but is not surprising. Infection with carbapenem-resistant A baumannii is associated with a risk of mortality that is twice that of infection with their carbapenem-susceptible counterparts.14 Patients in the AIDA trial had relatively low Charlson and SOFA indexes, indicating that underlying comorbidities and severity of illness are unlikely explanations for such high mortality rates. It appears that colistin, either as a monotherapy or combined with a carbapenem, is not that e ective. Polymyxin B might have pharmacological characteristics that render it superior to colistin, and it might be advantageous to administer polymyxins directly into the site of infection (eg, inhaled colistin for pneumonia).2
The AIDA trial underscores the urgent need for new antibiotics with activity against carbapenem-resistant A baumannii, typically producing OXA-23, OXA-24/40, and OXA-58 carbapenemases. Antibiotics that retain stability or inhibitors against these enzymes have proven particularly elusive. Ce derocol, a novel siderophore cephalosporin, has promising in vitro activity against carbapenem-resistant A baumannii, as well as metallo-β-lactamase-producing organisms, but regulatory approval is still needed and its e cacy in the clinical arena is yet to be shown.14
In the AIDA trial, we are concerned that the time to administration of e ective antibiotic therapy exceeded 2 days in 50% of patients. This gure highlights that it is imperative to adopt and re ne methods to rapidly detect carbapenem resistance in Gram-negative bacteria. In combination with epidemiological studies establishing the prevalence of resistance and clinical studies of outcomes preferred by patients and clinicians, these diagnostic platforms can assist decision making regarding appropriate antibiotic therapy.15,16 Similarly, the identi cation of genotypes and phenotypes linked to resistance can lead to the development of rational antibiotic regimens de ned by molecular targets, and inform the design of clinical trials. Thus, antimicrobial chemotherapy can advance towards the paradigm of precision medicine.17 Investments to usher this transformation are urgently needed to o set the staggering global e ect of antimicrobial resistance—projected to cost US$100 trillion and result in 10 million deaths per year by 205018—and to produce the high-quality evidence necessary to guide therapeutic decisions in our patients with carbapenem-resistant Gram-negative bacteria.