COMMENT
REVISITing treatment of metallo-β-lactamases
Emily L Heil, Erin K McCreary
Lancet Infect Dis 2024 Online first October 07, 2024
https://doi.org/10.1016/S1473-3099(24)00561-9
Metallo-β-lactamase enzymes, named because of the presence of one or two zinc ions at the catalytic site, are responsible for β-lactam antibiotic hydrolysis and are increasing in prevalence worldwide.1 Despite the increase in antibiotic development over the past decade, a β-lactamase inhibitor that binds to metallo-β-lactamases is not commercially available. Treatment options for metallo-β-lactamases are restricted to non-β-lactam antibiotics, cefiderocol, or the combination of ceftazidime–avibactam and aztreonam.2Aztreonam is stable against metallo-β-lactamase-mediated hydrolysis, and avibactam inhibits most other clinically relevant β-lactamase enzymes, making this an enticing combination.
In The Lancet Infectious Diseases, Yehuda Carmeli and colleagues present the results of the REVISIT study,3 a prospective, multi-national, open-label, randomised trial in which patients who were admitted to the hospital with complicated intra-abdominal infection (cIAI) or hospital-acquired or ventilator-associated pneumonia (HAP–VAP), suspected or confirmed to be caused by Gram-negative bacteria, were randomised in a 2:1 ratio to aztreonam–avibactam or meropenem with or without colistin. Formal hypothesis testing was not planned for the study and there was no power calculation completed a priori. 422 patients were enrolled and randomly allocated, with 271 having at least one Gram-negative pathogen identified at baseline. Only 80 isolates were tested for carbapenemases, of which 19 were positive: nine (11%) for serine carbapenemases and ten (12%) for metallo-β-lactamases (including one patient with the intrinsically metallo-β-lactamase-positive Stenotrophomonas maltophilia). Patients were excluded if they had received more than 24 h of any systemic antibiotic within 48 h before randomisation, unless the patient had a previous systemic antibiotic treatment failure based on worsening signs and symptoms of infection or lack of improvement despite a minimum of 48 h of treatment (which occurred in 75 patients [27%] of 282 in the aztreonam–avibactam group and 30 [21%] of 140 in the meropenem group).
Most patients were not critically ill, with approximately 60% of patients in both groups having an Acute Physiology and Chronic Health Evaluation II score of 10 or under and a primary diagnosis of cIAI. Overall, 26 (19%) patients in the meropenem group received concomitant colistin; only four (1%) patients in the entire cohort received a concomitant aminoglycoside, all in the aztreonam–avibactam group. In the intention-to-treat analysis cohort, the clinical cure rate at the test-of-cure visit (day 28) was 68·4% in the aztreonam–avibactam group and 65·7% in the meropenem group (treatment difference 2·7% [95% CI –11·4 to 17·8]). While a primary endpoint of clinical cure might be scrutinised in an open-label study, clinical response was assessed by a masked independent adjudication committee, resulting in an unbiased adjudication of the primary objective. Consistent with other clinical trials, patients with HAP–VAP had lower clinical cure rates and a higher proportion of previous treatment failure than patients with cIAI. 28-day all-cause mortality rates were low (five [3%] of 177 for aztreonam–avibactam vs six [6 %] of 94 for meropenem) in the microbiological intention-to-treat cohort (ie, those with at least one Gram-negative pathogen identified at baseline).
In the microbiologically evaluable analysis set with metallo-β-lactamase-positive pathogens, clinical cure rates were 50·0% (two of four) and 0% (zero of one) in the aztreonam–avibactam and meropenem groups, respectively. One patient with clinical failure in the aztreonam–avibactam group had S maltophilia (metallo-β-lactamase subtype L1) with an aztreonam–avibactam minimum inhibitory concentration (MIC) of 2 mg/L and a non-metallo-β-lactamase-producing Pseudomonas aeruginosa co-infection. Another patient with clinical failure in the aztreonam–avibactam group had Verona integron-encoded metallo-β-lactamase-producing Klebsiella pneumoniae with an aztreonam–avibactam MIC of 2 mg/L. All patient isolates were within the proposed susceptibility breakpoint for aztreonam–avibactam (ie, ≤4/4 μg/mL for Enterobacterales and ≤8/4 μg/mL for P aeruginosa and S maltophilia).4 Notably, patients with monomicrobial P aeruginosa infection were excluded from the study; patients with polymicrobial infections including P aeruginosa were excluded from the clinically evaluable group. The co-formulation of aztreonam and avibactam was given at a pharmacokinetically and pharmacodynamically optimised ratio of 3:1 (1500 mg aztreonam plus 500 mg avibactam every 6 h for patients with creatinine clearance >50 mL/min) to optimise the pharmacokinetics and pharmacodynamics of both drugs, and therefore it is unlikely that failure to achieve appropriate drug exposure contributed to clinical failure.5
Adverse drug events in the study were common, including relatively high rates of Clostridioides difficile infection (8% in the aztreonam–avibactam group), hypersensitivity events (13% in the aztreonam–avibactam group, although none were reported to be anaphylaxis or angioedema), and mild-to-moderate liver-related adverse events (18%). Encouragingly, hepatic aminotransferase elevations were similar between groups, potentially due to the lower dose of aztreonam used compared to previous reports.6
Can the co-formulation of aztreonam–avibactam replace the use of ceftazidime–avibactam plus aztreonam in practice? A co-formulation of aztreonam–avibactam provides an optimised ratio of drugs and circumvents the administration challenges of simultaneous ceftazidime–avibactam plus aztreonam.7 However, aztreonam–avibactam requires two loading doses, administered sequentially over 30 min and then 3 h, with the maintenance regimen starting 3 h after the end of the second loading dose. Additionally, there might be some benefit with the addition of ceftazidime, as demonstrated in hollow-fibre infection models comparing ceftazidime–avibactam plus aztreonam with aztreonam–avibactam against regrowth of New Delhi metallo-β-lactamase 1-producing Enterobacteriaceae.8 Ultimately, the small number of patients with metallo-β-lactamases in the study restricts the ability to draw conclusions about the utility of aztreonam–avibactam in eradicating these difficult-to-treat pathogens. However, years of experience with the combination of ceftazidime–avibactam plus aztreonam suggest this combination reduces 30-day mortality compared with colistin-based regimens in patients with carbapenem-resistant Enterobacterales infections.9 Importantly, the data evaluating ceftazidime–avibactam plus aztreonam are all observational in nature, and dosing is not standardised across studies.
Developing a β-lactamase inhibitor for metallo-β-lactamases is challenging for numerous reasons, including structural diversity and toxicity concerns related to their similarities to human enzymes.10 In the interim, aztreonam–avibactam might be a single drug option for culture-confirmed metallo-β-lactamase-producing monomicrobial infections.
