Predicting early appropriate therapy for patients infected by carbapenem-resistant Gram-negative pathogens in intensive care units in Italy.

Abstract

Background: Antibiotic resistance among Gram-negative bacteria in intensive care units (ICUs) is linked with high morbidity and mortality in patients. In this study, we estimated the therapeutic coverage of various antibiotics, focusing on cefiderocol and comparators, administered empirically against an infection of unknown origin in the ICU.

Methods: In the ARTEMIS surveillance study, susceptibilities of 624 Italian Gram-negative isolates to amikacin, aztreonam-avibactam, cefiderocol, ceftazidime-avibactam, ceftolozane-tazobactam, colistin, imipenem-relebactam, meropenem, and meropenem-vaborbactam were tested by broth microdilution, and results were interpreted by European Committee on Antimicrobial Susceptibility Testing breakpoints. The susceptibility rates from the ARTEMIS study were extrapolated to Gram-negative isolates obtained from 5,774 patients in Italian ICUs in 2021. The sum of the predicted susceptibilities of individual pathogens represented the overall likelihood of in vitro activity of each antibiotic as early targeted therapy for ICU patients.

Results: A total of 624 Italian Gram-negative isolates included 206 Pseudomonas aeruginosa, 138 Acinetobacter baumannii, 187 Klebsiella pneumoniae, and 93 Escherichia coli. Against A. baumannii, K. pneumoniae, P. aeruginosa, and E. coli, the overall susceptibility rates for cefiderocol were 87.7%, 96.8%, 99%, and 100%, respectively; and for comparator agents, 8.7-96.4%, 25.7-100%, 73.3-100%, and 89.2-100%, respectively. Among the subset of meropenem-resistant isolates, susceptibility rates of A. baumannii, K. pneumoniae, and P. aeruginosa to cefiderocol were 86.4%, 96.2% and 100%, respectively. Corresponding susceptibility rates to comparator agents were 0-96.8%, 0-100%, and 6.4-100%, respectively. There were no meropenem-resistant isolates of E. coli. The extrapolation of data to isolates from Italian ICUs showed that the highest likelihood of therapeutic coverage, both overall and among meropenem-resistant isolates, was reported for colistin (96.8% and 72.2%, respectively) and cefiderocol (95.7% and 71.4%, respectively). All other antibiotics were associated with a likelihood below 73% overall and between 0% and 41.4% for meropenem-resistant isolates.

Conclusions: Based on confirmed susceptibility rates and reported ICU prevalence of multiple Gram-negative species, cefiderocol showed a higher predicted therapeutic coverage and utility in ICUs compared with comparator beta-lactam-beta-lactamase inhibitor antibiotics. Cefiderocol may be a promising early treatment option for patients at high risk of carbapenem-resistant Gram-negative bacterial infections in the ICU.