The teicoplanin pharmacokinetics (PK) of 30 febrile and severely neutropenic patients (polymorphonuclear count, <500/mm 3 ) with hematologic malignancies were compared with those determined for five healthy volunteers (HV). Neutropenic patients were given piperacillin combined with amikacin, and teicoplanin was added to the regimen the day fever developed in patients suspected of having a staphylococcal infection or 48 h later. Teicoplanin was given intravenously at a dosage of 6 mg/kg of body weight at 0, 12, and 24 h and once a day thereafter. Five to eleven blood samples per patient were collected. Teicoplanin concentrations were measured by liquid chromatography. A bicompartmental model was fitted to the data by a nonlinear mixedeffect-model approach. Multiple-linear regression analysis was applied in an attempt to correlate PK parameters to nine covariates. The mean trough concentrations of teicoplanin 48 h after the onset of treatment and 24 h after the last injection (last trough) ؎ standard deviations were 8.8 ؎ 4.1 and 17.5 ؎ 13.5 mg/liter, respectively. A significant increase was noted in the mean rate of elimination clearance of teicoplanin in neutropenic patients compared with that of HV (0.86 versus 0.73 liter/h, P ؍ 0.002), as was the case with rates of distribution clearance (5.89 versus 4.94 liter/h, P ؍ 0.002); the mean half-life of distribution was significantly shorter in patients than in HV (0.43 versus 0.61 h, P ؍ 0.002). In contrast, the volumes of the central compartment (ca. 5.8 liters for both groups), the volumes of distribution at steady state (HV, 37.6 liters; patients, 55.9 liters), and the elimination half-lives (HV, 39.6 h; patients, 52.7 h) were not significantly different between HV and neutropenic patients. Interindividual variabilities of rates of clearance (coefficient of variation [CV], 43%) and elimination half-lives (CV, 56%) were mainly explained by the variabilities among rates of creatinine clearance. Interindividual variabilities of the volumes of the central compartment (CV, 33%) and the volumes of distribution at steady state (CV ؍ 51%) were correlated to interindividual variabilities among numbers of leukocytes and the ages of patients, respectively. On the basis of the population PK model of teicoplanin, simulations were made to optimize the dosing schedule. A supplemental 6 mg/kg dose of teicoplanin at 36 h resulted in a trough concentration at 48 h of 16.0 ؎ 4.5 mg/liter, with only 7% of patients having a trough concentration of less than 10 mg/liter, compared with 46% of patients on the usual schedule. * Corresponding author. Phone: 33 1 48 95 56 61. Fax: 33 1 48 95 56 59. 1242 on July 4, 2020 by guest http://aac.asm.org/ Downloaded from