Target site antimicrobial activity of colistin might be misestimated if tested in conventional growth media

Cation-dependent inhibition of antimicrobial activity was reported for polymyxin antibiotics. Ca(2+) and Mg(2+) concentrations recommended by the Clinical and Laboratory Standards Institute (CLSI) for the supplementation of Müller-Hinton broth (MHB) are markedly lower than interstitial space fluid (ISF) concentrations in vivo. Hence, it was speculated that the antimicrobial activity of colistin might be overestimated if tested using conventional cation-adjusted MHB. The antimicrobial activity of colistin against n = 100 clinical isolates of Pseudomonas aeruginosa, Acinetobacter baumannii, Klebsiella pneumoniae and Escherichia coli (n = 25 each) was evaluated by broth microdilution and, for selected isolates, by time-kill curves, in MHB without cations (MHB(ONLY)), MHB supplemented with 25 mg/L Ca(2+) and 12.5 mg/L Mg(2+) according to CLSI recommendations (MHB(CLSI)), and in MHB adjusted to 50 mg/L Ca(2+) and 20 mg/L Mg(2+) simulating ISF concentrations (MHB(ISF)). The minimum inhibitory concentration (MIC) values of colistin against the vast majority of isolates of both P. aeruginosa and A. baumannii increased significantly with higher cation concentrations. The susceptibility of K. pneumoniae isolates to colistin did not show significant changes between cation-supplemented media, while the MICs of E. coli decreased with ascending cation concentrations. These findings were confirmed in time-kill studies, where colistin killing against P. aeruginosa 1514 and A. baumannii 1485 declined with increasing cation concentrations. Contrarily, the killing of selected concentrations of colistin against K. pneumoniae 15 and E. coli 16 was enhanced in the presence of increasing cation concentrations. The present data suggest that the clinical antimicrobial activity of colistin might be misestimated in vitro if tested in conventional growth media.