Excess Mortality Associated With Colistin-Tigecycline Compared With Colistin-Carbapenem Combination Therapy for Extensively Drug-Resistant Acinetobacter baumannii Bacteremia: A Multicenter Prospective Observational Study

Objectives: Since few therapeutic options exist for extensively drug resistant Acinetobacter baumannii, an emerging threat in ICUs worldwide, and comparative prospective studies of colistin-based combination therapies are lacking, our objective was to compare the outcomes of patients with extensively drug-resistant A. baumannii bacteremia, treated with colistin-carbapenem and colistin-tigecycline combinations. Design: Prospective, observational, multicenter study. Setting, Patients, and Interventions: Adults with extensively drug-resistant A. baumannii bacteremia were prospectively followed from 2010 to 2013 at three hospitals in Taiwan. Extensively drug-resistant A. baumannii was defined as A. baumannii (genospecies 2) nonsusceptible to all drug classes except for colistin and tigecycline, and standard combination therapy as use of parenteral colistin-carbapenem or colistin-tigecycline for at least 48 hours after onset of bacteremia. Measurements and Main Results: Primary outcome measure was 14-day mortality. Of the 176 episodes of extensively drug-resistant A. baumannii bacteremia evaluated, 55 patients with a median (interquartile range) age of 62 years (44-79 yr) and Sequential Organ Failure Assessment score of 9 (5-13) points received standard combination therapy: colistin-tigecycline in 29 patients and colistin-carbapenem in 26. Crude 14-day and in-hospital mortality rates for patients receiving colistin-tigecycline versus patients receiving colistin-carbapenem were 35% versus 15% (p = 0.105) and 69% versus 50% (p = 0.152), respectively. Breakthrough extensively drug-resistant A. baumannii bacteremia under steady state concentrations of combination therapy for colistin-tigecycline group was 18% and for colistin-carbapenem group was 0% (p = 0.059). Eleven patients (20.0%) developed nephrotoxicity. After adjusting for age, sex, comorbidity, initial disease severity, loading colistin dose, polymicrobial infection, and primary infection site, excess 14-day mortality was associated with the use of colistin-tigecycline in the subgroup with tigecycline minimum inhibitory concentration greater than 2 mg/L compared with the use of colistin-carbapenem (hazard ratio, 6.93; 95% CI, 1.61-29.78; p = 0.009). Conclusions: Increased 14-day mortality was associated with colistin-tigecycline therapy given tigecycline minimum inhibitory concentration greater than 2 mg/L compared with colistin-carbapenem therapy for extensively drug-resistant A. baumannii bacteremia.