INTRODUCTION: Beta-lactam antibiotics are essential in the treatment of Gram-negative bloodstream infections. The effect of beta-lactam antibiotics depends on the time of unbound antibiotic concentration above the minimal inhibitory concentration (MIC). An antibiotic concentration above MIC during the whole dosing interval (100% ƒT >
MIC) has been suggested as a target for severe infections. The aim of the present study was to compare target attainment using targets derived from known MICs with standard generic targets. METHODS: In this prospective, single-center study, adult patients with Gram-negative bloodstream infection treated with cefotaxime, piperacillin/tazobactam or meropenem were eligible for inclusion. Trough antibiotic concentrations were collected during a single dosing interval and actual MIC values for the antimicrobial agent against the infecting isolate were obtained using E-tests. Epidemiological cut off values, ECOFFs, were applied from European Committee on Antimicrobial Susceptibility Testing, EUCAST, tables for isolates within the wild-type distribution. Antibiotic concentrations were measured using Liquid Chromatography tandem Mass Spectrometry. Free concentrations were estimated based on total concentrations. Two targets based on actual MICs were assessed: free trough concentrations above (1) four times the actual MIC (100% ƒT >
4MIC) or above (2) the ECOFF (100% ƒT >
ECOFF). Proportions of target attainment for the MIC-based targets were compared with attainment using clinical breakpoints or PK/PD breakpoints. Treatment response was defined as clinical resolution at day 7 (No persisting signs or symptoms of infection). RESULTS: We included 98 patients with a median age of 72 years. The most common microbiological finding was Escherichia coli (63%) followed by Klebsiella pneumoniae (12%). Of all patients, 77/98 patients (79%) attained 100% ƒT >
4MIC and 80/98 (82%) attained 100% ƒT >
ECOFF, compared with 57/98 (58%) using 100% ƒT >
EUCAST clinical breakpoints. Clinical resolution at day 7 was significantly associated with target attainment applying the target 100% ƒT >
4MIC (p = 0.013), but this was not the case when 100% ƒT >
ECOFF was applied (p = 0.50). CONCLUSIONS: In our material, higher target attainment rates were seen using targets derived from actual MICs, compared to EUCAST clinical breakpoints. Attaining 100% ƒT >
4MIC was associated with resolution of infection, but the latter finding should be interpreted cautiously.