TEICOPLANIN PHARMACOKINETICS AND DOSAGE RECOMMENDATIONS IN CHRONIC-HEMODIALYSIS PATIENTS AND IN PATIENTS UNDERGOING CONTINUOUS VENOVENOUS HEMODIALYSIS

Multiple-dose pharmacokinetics of teicoplanin, a glycopeptide antibiot ic against gram-positive infections, were studied in 9 chronic hemodia lysis (HD) patients and in 7 patients with an acute renal failure (ARF ) treated by continuous veno-venous hemodialysis (CVVHD). After a load ing dose of 800 mg i.v. the 400 mg maintenance doses were administered according to a target trough concentration of 5-15 mg/l. Using the Ba yesian estimation method implemented in the computer program Abbottbas e Pharmacokinetic System (PKS), we defined an open three-compartment k inetic model for teicoplanin and calculated the individual pharmacokin etics. The mean terminal elimination half-life was 176 +/- 41.3 h in t he HD group and 99 +/- 22.3 h in the CVVHD group (p <0.005). The total body clearance (CL) was 4 +/- 1.2 ml/min and 9.2 +/- 1.7 ml/min in th e HD and CVVHD patients respectively (p <0.001). The mean reduction of the serum levels during a HD session was 9.1% in the patients dialyse d with a F8 filter and 20.2% with a high-flux F60 filter (p <0.001). T he resulting extraction rate was 10 +/- 3.6% (F8) which is similar to the unbound fraction. The elimination of teicoplanin during CVVHD ther apy strongly depended on the ultrafiltration rate (UFR) (r = 0.923, p <0.05). An UFR of 15.6 l/24 h resulted in a removal of 32%/24 h of a 4 00 mg dose and an UFR of 6.2 l/24 h in 9.5%/24 h respectively. Dosage simulations were performed and dosage recommendations can be given: ch ronic HD patients should receive a loading dose of 800 mg on day 1 and 400 mg on the days 5 and 10 followed by a maintenance therapy of 400 mg once weekly. In CVVHD patients a higher dosage has to be administer ed with a loading dose of 800 mg on day 1 and 400 mg on the days 2 and 3 followed by a maintenance dose of 400 mg every second to third day. The correlation of the 549 measured teicoplanin serum levels to the l evels predicted by the PKS program was statistically significant (p <0 .01). No differences were found comparing the results obtained by the Bayesian method to the results of a nonlinear least squares regression analysis. Therefore, routine therapeutic drug monitoring (TDM) for te icoplanin can be performed using this kinetic model.