In vitro effect of physiological concentrations of human albumin on the antibacterial activity of tigecycline

Abstract

To determine C(max) tigecycline activity in the presence/absence of physiological concentrations of human albumin with free fraction concentrations as controls. Killing curves (final inoculum: 1.0-5.0 x 10(7) cfu/mL) were performed with 0.88 mg/L final concentrations (serum C(max) after a 100 mg 1 h infusion) in Mueller-Hinton broth supplemented with Ca(2+) and Mg(2+) (MH) and in MH with 4 g/dL human albumin. Controls were curves in MH with concentrations similar to the free fraction (fC(max) = 0.17 mg/L) calculated using protein binding. Activity was measured as log(10) initial inoculum reduction (log(10) initial inoculum-log(10) at 12 h/24 h). Target strains (tigecycline MIC/MBC; mg/L) were: methicillin-resistant Staphylococcus aureus heteroresistant to vancomycin (0.12/0.25); Enterococcus faecium (0.12/0.25); Escherichia coli producing extended-spectrum beta-lactamase (0.12/0.25); and Acinetobacter baumannii (0.25/1). At 24 h the fC(max) produced mean decreases of < or =0.1 cfu/mL for all strains, in contrast to the bactericidal activity (mean >3 log(10) reduction) provided by C(max) concentrations in the presence or absence of albumin for E. coli and E. faecium, and an activity nearly bactericidal for S. aureus (mean approximately 2.8 log(10) reduction). In the case of the A. baumannii isolate the C(max) in the presence or absence of albumin produced a mean reduction of 2.56 log(10) cfu/mL at 12 h (time of one dosing interval), with a bacteriostatic profile when considering 24 h colony counts (similar counts at 0 and 24 h). Correcting the total concentration for the reported literature binding values is unreliable since tigecycline antibacterial activity was greater than that suggested by the free fraction of the drug.