Background: Extensive drug resistance of Acinetobacter baumannii is a serious problem in the clinical setting. It is therefore important to find active antibiotic combinations that could be effective in the treatment of infections caused by this problematic 'superbug'. In this study, we analyzed the in vitro activities of three colistin-based combinations and a minocycline-based combination against clinically isolated extensive drug resistant Acinetobacter baumannii (XDR-AB) strains. Methods: Fourteen XDR-AB clinical isolates were collected. The clonotypes were determined by polymerase chain reaction-based fingerprinting. Susceptibility testing was carried out according to the standards of the Clinical and Laboratory Standards Institute. Activities of drug combinations were investigated against four selected strains and analyzed by mean survival time over 12 hours (MST_{12 h}) in a time-kill study. Results: The time-kill studies indicated that the minimum inhibitory concentration (MIC) of colistin (0.5 or 0.25 μg/mL) completely killed all strains at 2 to 4 hours, but 0.5×MIC colistin showed no bactericidal activity. Meropenem (8 μg/mL), minocycline (1 μg/mL) or rifampicin (0.06 μg/mL) did not show bactericidal activity. However, combinations of colistin at 0.5×MIC (0.25 or 0.125 μg/mL) with each of the above were synergistic and shown bactericidal activities against all test isolates. A combination of meropenem (16 μg/mL) with minocycline (0.5×MIC, 4 or 2 μg/mL) was synergitic to all test isolates, but neither showed bactericidal activity alone. The MST_{12 h} values of drug combinations (either colistin- or minocycline-based combinations) were significantly shorter than those of the single drugs (p < 0.01). Conclusions: This study indicates that combinations of colistin/meropenem, colistin/rifampicin, colistin/minocycline and minocycline/meropenem are synergistic in vitro against XDR-AB strains.

Meropenem is used to manage postneurosurgical meningitis, but its population pharmacokinetics (PPK) in plasma and cerebrospinal fluid (CSF) in this patient group are not well-known. Our aims were to (i) characterize meropenem PPK in plasma and CSF and (ii) recommend favorable dosing regimens in postneurosurgical meningitis patients. Eighty-two patients were enrolled to receive meropenem infusions of 2 g every 8 h (q8h), 1 g q8h, or 1 g q6h for at least 3 days. Serial blood and CSF samples were collected, and concentrations were determined and analyzed via population modeling. Probabilities of target attainment (PTA) were predicted via Monte Carlo simulations, using the target of unbound meropenem concentrations above the MICs for at least 40% of dosing intervals in plasma and at least of 50% or 100% of dosing intervals in CSF. A two-compartment model plus another CSF compartment best described the data. The central, intercentral/peripheral, and intercentral/CSF compartment clearances were 22.2 liters/h, 1.79 liters/h, and 0.01 liter/h, respectively. Distribution volumes of the central and peripheral compartments were 17.9 liters and 3.84 liters, respectively. The CSF compartment volume was fixed at 0.13 liter, with its clearance calculated by the observed drainage amount. The multiplier for the transfer from the central to the CSF compartment was 0.172. Simulation results show that the PTAs increase as infusion is prolonged and as the daily CSF drainage volume decreases. A 4-hour infusion of 2 g q8h with CSF drainage of less than 150 ml/day, which provides a PTA of >90% for MICs of ≤8 mg/liter in blood and of ≤0.5 mg/liter or 0.25 mg/liter in CSF, is recommended. (This study has been registered at ClinicalTrials.gov under identifier NCT02506686.)