Clinical outcomes and emergence of resistance of Pseudomonas aeruginosa infections treated with ceftolozane-tazobactam versus ceftazidime-avibactam

Abstract

ABSTRACT Few studies compare outcomes of patients with difficult-to-treat resistance (DTR) Pseudomonas aeruginosa infections treated with ceftolozane-tazobactam versus ceftazidime-avibactam. A multicenter prospective study was conducted of unique patients with DTR P. aeruginosa infections from 2018 to 2023 receiving >72 h of ceftolozane-tazobactam or ceftazidime-avibactam, with confirmation that the P. aeruginosa isolate was susceptible to the agent administered by broth microdilution. Inverse probability weighting (IPW) incorporating propensity scores was utilized to ensure balanced baseline characteristics. Regression performed on the post-IPW group determined 30-day mortality and subsequent emergence of resistance (i.e., ≥4-fold increase in MIC) to the initial treatment (i.e., ceftolozane-tazobactam or ceftazidime-avibactam). Among 186 eligible patients, 102 (55%) received ceftolozane-tazobactam and 84 (45%) received ceftazidime-avibactam. In the post-IPW cohort, balance was achieved across all variables [e.g., demographics, severity of illness, severe immunocompromise, Charlson Comorbidity Index ≥5, continuous renal replacement therapy (CRRT), source of infection, combination therapy]. Thirty-day mortality was similar between the ceftolozane-tazobactam and ceftazidime-avibactam groups [21% vs 17%; adjusted odds ratio (aOR): 1.01 (95% confidence interval, CI: 0.90–1.14)]. Emergence of resistance was higher in the ceftolozane-tazobactam group [38% vs 25%; aOR: 1.89 (95% CI: 0.98–4.88)], but did not achieve statistical significance. Prolonged treatment durations and use of CRRT were associated with increased emergence of resistance (both P = 0.04). Although the survival of patients with DTR P. aeruginosa infections appears similar regardless of whether ceftolozane-tazobactam or ceftazidime-avibactam is prescribed, the emergence of resistance may be more concerning with the former. Plausible mechanistic explanations support these findings. Modifiable risk factors were identified that may mitigate this risk.