Objective: To study the pharmacokinetics of vancoymcin in critically ill patients with acute renal failure treated with continuous venovenous haemofiltration (CVVHF). Design: Open-label study. Setting: Hospital pharmacy centre and medical intensive care unit of the University Medical Centre Utrecht. Materials and methods: In a laboratory setting, the sieving coefficient (s) of vancomycin by polyacrilonitrile (PAN) haemofilters of different surface areas was studied. In one patient, the pharmacokinetics of vancomycin were studied following a single dose of vancomycin. Another patient was treated with a vancomycin dosing regimen based on data from the literature, but high trough concentrations made dose reduction necessary after 24 h of withholding therapy. After two doses of 250 mg, serum and ultrafiltrate samples were collected for pharmacokinetic evaluation. Intervention: CVVHF with the following operational characteristics: blood flow 200 ml/min, ultrafiltrate flow 25 ml/min, postdilution, PAN 06 hollow fibre haemofilter. Measurements and results: The average sieving coefficient in vitro was 0.73 +/- 0.06, 0.86 +/- 0.11, and 0.80 +/- 0.06 for the PAN 03, 06, and 10 haemofilters, respectively. Changes in the sieving coefficient by increasing the ultrafiltration rate were not clinically significant. The first patient was given a single dose of vancomycin, 1000 mg by intravenous infusion. The following pharmacokinetic data were obtained: apparent volume of distribution (Vd) 55.81, terminal half-life time (t(1/2) (term)) 15.4 h, total clearance (Cl-tot) 2.5 l/h, CVVHF clearance (CLCVVHF,form 1) 1.4 l/h, and body clearance (Cl-body) 1.1 l/h The average sieving coefficient during the study period was 0.89 +/- 0.03. In the second patient, the pharmacokinetics of vancomycin were studied following dose reduction: Vd 41.7 l, t(1/2 term) 20.3 h, Cl-tot 1.4 l/h, Cl-CVVHF,Cl-form 1 1.4 l/h, and Cl-body < 0.1 l/h. The average sieving coefficient during the study period was 0.88 +/- 0.03. The cumulative amount of vancomycin removed by means of CVVHF during the 12-h study period was 245 mg in patient 1 and 228 mg in patient 2. Conclusion: CVVHF with a PAN 06 haemofilter effectively removed vancomycin in two critically ill patients. The amount of vancomycin removed with CVVHF was about 250 mg per 12 h. A clear difference in body clearance in the two patients was observed. Our dosage recommendation for vancomycin in critically ill patients receiving CVVHF is a loading dose of 15-20 mg/kg followed after 24 h by 250 to 500 mg twice daily with close monitoring of the serum and ultrafiltrate vancomycin concentration.
