Evaluation of the in vitro efficacy of aztreonam–zidebactam and comparator β-lactam/β-lactamase inhibitor combinations against carbapenem-resistant Klebsiella pneumoniae

Carbapenem-resistant Klebsiella pneumoniae producing metallo-β-lactamases poses a global threat, as it is unresponsive to most β-lactam/β-lactamase inhibitor (BL/BLI) combinations. This study evaluated the in vitro activity of aztreonam–zidebactam (AZT + ZID) and comparator BL/BLI combinations against clinical K. pneumoniae and identified genomic determinants of susceptibility. Minimum inhibitory concentrations (MICs) were tested for 130 K. pneumoniae isolates. Resistance genomic predictors were identified by whole-genome sequencing (WGS) of 80 selected strains, statistically analyzed, and different combinations were compared. Two AZT + ZID formulations were tested (both at equal drug ratios versus fixed 4 µg/ml ZID), both of which resulted in significant reductions in MICs compared with AZT monotherapy. AZT + ZID (4 µg/ml) achieved median fold improvements of 4129 × against NDM producers, 2048 × against OXA-48-like producers, and 4267 × against non-carbapenemase isolates, being 16-fold superior to the equal-ratio formulation. Aztreonam–avibactam (AZT + AVI) outperformed ceftazidime–avibactam, which was inactive against NDM producers, while cefepime–zidebactam showed activity comparable to the AZT + ZID equal-ratio combination. WGS identified three independent resistance determinants: bla NDM , bla CTX-M , and OmpK35 disruption. S137P in penicillin-binding protein 2 (PBP2), exclusive to the ST231 clonal lineage, was the only PBP mutation with independent MIC effects, producing a 2787-fold increase in MIC for AZT + ZID (4 µg/ml) while retaining susceptibility to AZT + AVI, demonstrating formulation-specific resistance amenable to alternative therapy. A single pan-resistant isolate, combining four β-lactamase classes with dual porin loss, failed to respond to the inhibitor-based strategy. In conclusion, AZT + ZID fixed-concentration formulation is significantly more potent than the equal-ratio formulation and warrants clinical validation. Integrating genomic surveillance with phenotypic susceptibility testing enables mechanism-guided treatment selection.