A. baumannii Bloodstream Infection: Mortality and Resistance

Key Takeaways
- At a Beijing hospital, patients with A. baumannii bloodstream infection frequently died during hospitalization.
- Older age, carbapenem resistance, and greater illness severity were associated with death; an initial association with ventilation weakened after further adjustment.
- IC2/Pasteur ST2 predominated, with blaOXA-23 associated with carbapenem resistance in that lineage.
- Cefiderocol and sulbactam–durlobactam showed in vitro activity, although non-susceptible isolates occurred with both agents.
In a Beijing study of A. baumannii bloodstream infection, investigators retrospectively examined the first culture-confirmed episode in each of 131 patients at China–Japan Friendship Hospital during 2016–2025 and paired each patient with a bloodstream isolate. Infection onset was the collection of the first positive blood culture; in-hospital death was the event of interest and live discharge a competing event in Fine–Gray models. Carbapenem-resistant A. baumannii (CRAB) and carbapenem-susceptible infections were comparison groups, not treatment arms. Whole-genome sequencing (WGS) with Pasteur and Oxford typing assessed relatedness. Susceptibility was tested in triplicate by broth microdilution, using iron-depleted broth for cefiderocol and Clinical and Laboratory Standards Institute M100 36th-edition breakpoints where applicable. Public bloodstream genomes, including U.S. isolates, supplied geographic genomic context, but the Beijing outcome and susceptibility estimates cannot be assumed to apply to U.S. patients.
In-hospital deaths occurred in 100/131 patients (76.3%). In the primary Fine–Gray model, CRAB was associated with a higher cumulative incidence of in-hospital death (subdistribution hazard ratio 3.60, 95% confidence interval 1.26–10.35). Older age and higher Sequential Organ Failure Assessment (SOFA) scores at BSI onset remained positively associated with death after expanded adjustment. Mechanical ventilation at onset was associated with death in the primary model, but that association weakened after additional adjustment.
International clone 2 (IC2), specifically Pasteur sequence type 2 (ST2), accounted for 116/131 bloodstream isolates (88.5%). Among ST2 isolates, 113/116 (97.4%) both carried blaOXA-23 and had the CRAB phenotype. The ST2 population comprised several sublineages rather than a single uniform hospital clone. In vitro, 129/131 isolates (98.5%) were susceptible to cefiderocol, versus 106/131 (80.9%) to sulbactam–durlobactam; non-susceptible isolates occurred with both. Sulbactam–durlobactam had lower minimum inhibitory concentrations (MICs) than sulbactam alone, with a further MIC decrease in combination with meropenem. Conventional agents showed limited activity. Patient treatments were neither randomized nor compared for efficacy.
The investigators noted that the retrospective, limited single-center cohort left room for residual confounding and did not establish causality or outcomes at other hospitals. Additional adjustment weakened the ventilation association, while sparse later follow-up complicated its interpretation. Treatment timing, dosage, duration, and initial antimicrobial activity did not permit comparisons of regimen effectiveness. Closely related isolate pairs alone could not establish hospital transmission without epidemiologic links. Laboratory susceptibility and MIC shifts likewise could not demonstrate patient benefit.
The authors concluded that high in-hospital mortality accompanied a locally IC2/ST2-dominated, carbapenem-resistant A. baumannii population. Cefiderocol and sulbactam–durlobactam retained laboratory activity against many bloodstream isolates despite detected non-susceptibility.
Clinician Questions
How was carbapenem-resistant A. baumannii defined in the Beijing bloodstream infection cohort?
Among A. baumannii bloodstream isolates from China–Japan Friendship Hospital in Beijing, carbapenem resistance meant meeting Clinical and Laboratory Standards Institute resistance criteria for either imipenem or meropenem; a resistance gene alone did not define the phenotype.
Does PBP3 T526S predict sulbactam–durlobactam response in A. baumannii bloodstream infection?
All observed A. baumannii bloodstream isolates carrying PBP3 T526S were non-susceptible to sulbactam–durlobactam in laboratory testing. Carriers were few and belonged to the dominant ST2 lineage, and genetic relatedness was not accounted for in the ST2-restricted analysis. Neither a causal mechanism nor reliable prediction of patient treatment response was established.
This content is intended for healthcare professionals for educational purposes. The article and its image were developed with AI assistance and reviewed by members of the ReachMD team prior to publication. Please refer to the original source linked for the complete and most up-to-date information.