In vivo fitness of carbapenem-resistant Acinetobacter baumannii strains in murine infection is associated with treatment failure in human infections

Objectives Mortality among patients with carbapenem-resistant Acinetobacter baumannii (CRAB) infections varies between studies. We examined whether in vivo fitness of CRAB strains is associated with clinical outcomes in patients with CRAB infections. Methods Isolates were collected from patients enrolled in the AIDA trial with hospital-acquired pneumonia, bloodstream infections and/or urinary tract infections caused by CRAB. The primary outcome was 14-day clinical failure, defined as failure to meet all criteria: alive; haemodynamically stable; improved or stable Sequential Organ Failure Assessment (SOFA) score; improved or stable oxygenation; and microbiological cure of bacteraemia. The secondary outcome was 14-day mortality. We tested in vivo growth using a neutropenic murine thigh infection model. Fitness was defined based on the CFU count 24 hours after injection of an inoculum of 105 CFU. We used mixed-effects logistic regression to test the association between fitness and the two outcomes. Results The sample included 266 patients; 215 (80.8%) experienced clinical failure. CRAB fitness ranged from 5.23 to 10.08 log CFU/g. The odds of clinical failure increased by 62% for every 1-log CFU/g increase in fitness (OR 1.62, 95% CI 1.04–2.52). After adjusting for age, Charlson score, SOFA score and acquisition in the intensive care unit, fitness remained significant (adjusted OR 1.63, 95% CI 1.03–2.59). CRAB fitness had a similar effect on 14-day mortailty, although the association was not statistically significant (OR 1.56, 95% CI 0.95–2.57). It became significant after adjusting for age, Charlson score, SOFA score and recent surgery (adjusted OR 1.88, 95% CI 1.09–3.25). Conclusions In vivo CRAB fitness was associated with clinical failure in patients with CRAB infection.