2774. Predictors of Treatment Outcomes in Bloodstream Infections Due to OXA-48-producing Klebsiella pneumoniae

BACKGROUND: Klebsiella pneumoniae causes various infections and the rising prevalence of carbapenem-resistant K. pneumoniae (CRKP) constitutes a major public health threat. Carbapenemase production is the most common mechanism of carbapenem resistance in CRKP, and OXA-48-like-producing strains accounted for about one-fifth of carbapenemase-producing CRKP worldwide. Despite clinical evidence supporting the use of ceftazidime-avibactam (CZA) in treating carbapenem-resistant Enterobacterales (CRE) infections, studies regarding its effectiveness for OXA-48-producing CRKP is limited. OXA-48-producing CRKP may test susceptible to meropenem in vitro, but whether meropenem is effective when phenotypic-genotypic discordance exists is unknown. METHODS: In this retrospective cohort study, we evaluated patients with OXA-48-producing CRKP bacteremia at Taipei Veterans General Hospital from April 2017 through December 2022. Patients were categorized into five groups according to the treatment received: CZA, meropenem, colistin, colistin-meropenem combination, and other treatments. Clinical outcomes were compared according to the antibiotics received, and logistic regression was performed to evaluate risk factors associated with treatment failure. RESULTS: Forty-five patients were included, with 87% (39/45) having received at least one active antibiotic against the OXA-48-producing CRKP isolated from blood. Fourteen- and 30-day mortality rates were 31% and 38%, respectively. The clinical failure rate was 27% (4/15) for patients treated with CZA, 40% (4/10) with meropenem, 100% (8/8) with colistin, 75% (3/4) with meropenem-colistin-combination, and 38% (3/8) with other treatments (P = 0.010). The clinical failure rate for meropenem-treated patients dropped to 25% (2/8) if only CRKP strains with meropenem minimal inhibitory concentration (MIC) ≤ 4 mg/L were analyzed. In multivariate analysis, clinical failure was associated with Apache II Score (P = 0.022) and colistin use (P = 0.002). Rates of clinical failure across treatment regimens [Figure: see text] Rates of clinical failure at different meropenem MIC values in patients receiving meropenem treatment [Figure: see text] MIC, minimum inhibitory concentration Multivariable analysis of risk factors associated with clinical failure at 30 days [Figure: see text] CI, confidence interval CONCLUSION: CZA is an important option in the treatment of OXA-48-producing CRKP bacteremia, and meropenem may remain a reasonable alternative if meropenem MIC was ≤ 4 mg/L. Use of colistin is an independent risk factor for treatment failure and should be avoided. DISCLOSURES: Po-Han Huang, MD, MSc, Pfizer: Honoraria