RESUMO
BACKGROUND: The increasing morbidity and mortality rates associated with Acinetobacter baumannii are due to the emergence of drug resistance and the limited treatment options. We compared characteristics of colistin-resistant Acinetobacter baumannii (CR-AB) clinical isolates recovered from patients with and without prior colistin treatment. We assessed whether prior colistin treatment affects the resistance mechanism of CR-AB isolates, mortality rates, and clinical characteristics. Additionally, a proper method for identifying CR-AB was determined. METHODS: We collected 36 non-duplicate CR-AB clinical isolates resistant to colistin. Antimicrobial susceptibility testing, Sanger sequencing analysis, molecular typing, lipid A structure analysis, and in vitro synergy testing were performed. Eleven colistin-susceptible AB isolates were used as controls. RESULTS: Despite no differences in clinical characteristics between patients with and without prior colistin treatment, resistance-causing genetic mutations were more frequent in isolates from colistin-treated patients. Distinct mutations were overlooked via the Sanger sequencing method, perhaps because of a masking effect by the colistin-susceptible AB subpopulation of CR-AB isolates lacking genetic mutations. However, modified lipid A analysis revealed colistin resistance peaks, despite the population heterogeneity, and peak levels were significantly different between the groups. CONCLUSIONS: Although prior colistin use did not induce clinical or susceptibility differences, we demonstrated that identification of CR-AB by sequencing is insufficient. We propose that population heterogeneity has a masking effect, especially in colistin non-treated patients; therefore, accurate testing methods reflecting physiological alterations of the bacteria, such as phosphoethanolamine-modified lipid A identification by matrix-assisted laser desorption ionization-time of flight, should be employed.