Pharmacokinetics of bevacizumab in haemodialysis

Sir, Bevacizumab, a recombinant humanized monoclonal antibody against VEGF (rhuMAb VEGF, Avastin , Genentech, South San Francisco, CA), was approved as a treatment for metastatic renal cell carcinoma (RCC) [1]. However, no bevacizumab pharmacokinetic data are available for patients with renal failure. We report a pharmacokinetic study of bevacizumab in a patient with renal insufficiency requiring haemodialysis. Bevacizumab was instituted at a dose of 5mg/kg every 2 weeks for a 23-year-old patient with mRCC. Bevacizumab serum concentrations were determined after 6 months of treatment, in a pharmacokinetic study over 2 weeks. Blood samples were collected just before and 5, 15, 30min, and 1, 6, 12 and 24 h after the end of infusion. Additional blood samples were collected over the dosing interval, before, during and after dialysis sessions. Paired arterial and venous blood samples were performed simultaneously 2 h after the start of haemodialysis. Haemodialysis was performed for 4 h using a F60 polyacrylonitrile dialyser (surface area 1.6m) every 2 days with a double-needle access to a radial arteriovenous fistula with a constant dialysate flow rate of 500ml/min and a blood flow rate of 250–300ml/min. Bevacizumab serum and dialysate concentrations were measured using an ELISA technique. Bevacizumab pharmacokinetics were analysed by both a non-parametric and a compartmental approach using WinNonLin software (Pharsight Corporation). Pharmacokinetic parameters obtained for our patient were compared with those of subjects with normal renal function [2] receiving 10mg/kg every 14 days (Table 1). Bevacizumab concentrations were 90.5 and 125.1mg/ml; 154.0 and 195.8mg/ml; and 145.0 and 173.0mg/ml before and after the three haemodialysis sessions, respectively. Two hours after the start of haemodialysis, arterial and venous concentrations were 109.7 and 122.3mg/ml; 146.4 and 202.2 mg/ml, and 145.0 and 175.0mg/ml, respectively. Values of E and CLHD of bevacizumab were 0% and 0ml/min, respectively. FHD was 0%, i.e. well below the 25% limit value above which haemodialysis clearance should be considered clinically significant. The bevacizumab pharmacokinetic parameters of our haemodialysed patient was therefore similar to the reference values reported in patients with normal renal function and, because he was treated by a dose of 5mg/kg/14days instead of 10mg/kg twice monthly, his bevacizumab area under the curve (AUC) was twice lower than published values (45205 vs 97488 mg h/ml, respectively). Despite a two times lower AUC, the patient had bevacizumab concentrations above the 50% of maximum bevacizumab-induced inhibition (IC50, 104.1 mg/ml) during the first 10 days following the infusion. Since his pharmacokinetic parameters were equivalent to those of patients with normal renal function, we think that the dose of 5mg/kg would be adapted to the haemodialysed patient. Furthermore, bevacizumab seems not to be dialysable and administration may, thus, be performed anytime before or after the session on haemodialysis days.