Jiankang Zhao, Yanyan Fan, Bin Cao
Mechanisms other than acquired carbapenemases-consistent with chromosomal determinants-accounted for most non-susceptibility in this predominantly northern Chinese respiratory tract CRPA collection. Carbapenemase types determine distinct drug susceptibility patterns. Routine carbapenemase detection fails to identify most CZA-resistant strains. FDC retained the broadest in vitro activity and is a reliable therapeutic option pending clinical outcome data, and ST270-associated resistance deserves continuous surveillance.
BACKGROUND: β-lactam/β-lactamase inhibitor combinations (BL/BLIs) and siderophore cephalosporin serve as critical polymyxin alternatives for carbapenem-resistant Pseudomonas aeruginosa (CRPA). However, systematic multicenter clinical data from China and the resistance mechanisms underlying non-susceptibility to these agents remain poorly defined.
METHODS: We collected 442 CRPA isolates from 10 hospitals across eight Chinese provinces (2017-2025), determined minimum inhibitory concentrations (MICs) for cefiderocol (FDC), ceftazidime-avibactam (CZA), imipenem-relebactam (IMR), and ceftolozane-tazobactam (CT) through broth microdilution, and performed whole-genome sequencing. Resistance genes, virulence factors, and sequence types (STs) were identified. A chromosomal co-selection score (eight non-mobile resistance genes) was constructed. Multivariable regression and receiver operating characteristic (ROC) analysis identified predictors of non-susceptibility.
RESULTS: Carbapenemase genes were detected in only 33 isolates (7.5%): metallo β lactamase (MBL) genes in 28, blaKPC-2 in 2, and blaGES 5/14 in 3; the ST463 clone was absent. MBL producers were resistant to CZA, IMR, and CT (0-4% susceptible) but 100% susceptible to FDC. KPC 2 and GES producers were CZA-susceptible (100%) but IMR- and CT-resistant. Among 127 CZA non-susceptible isolates, 99 (78%) lacked carbapenemase genes; ST270 (from a single center) accounted for 64.6% of these. The chromosomal score independently predicted MIC (ρ: CT 0.64, FDC 0.59, IMR 0.50, CZA 0.25-largely attributable to ST270; all p < 0.001). ExoU positive isolates had lower MICs (FDC median 0.12 vs. 0.25 mg/L, p < 0.001). FDC remained active against 97.3% of all isolates and 100% of carbapenemase-positive ones. Genotype-based prediction yielded area under the ROC curve (AUC) values of 0.86 (CT) and 0.80 (CZA).
CONCLUSIONS: Mechanisms other than acquired carbapenemases-consistent with chromosomal determinants-accounted for most non-susceptibility in this predominantly northern Chinese respiratory tract CRPA collection. Carbapenemase types determine distinct drug susceptibility patterns. Routine carbapenemase detection fails to identify most CZA-resistant strains. FDC retained the broadest in vitro activity and is a reliable therapeutic option pending clinical outcome data, and ST270-associated resistance deserves continuous surveillance.