K. Han,1 T. Peyret,2 A. Quartino,1 N. H. Gosselin,2 M. Mouksassi,2 S. Girish,1 D. E. Allison,1 J. Jin1; 1Genentech, South San Francisco, CA, 2Pharsight, Montreal, QC, Canada

BACKGROUND: Bevacizumab (BEV) is approved for various cancers. BEV pharmacokinetics (PK) was previously analyzed based on limited data. This analysis aimed to comprehensively characterize BEV PK and explore the influence of patient variables on BEV PK based on a large database.
METHODS: Rich and sparse BEV serum concentration (Cs) data were collected from phase I, II and III studies. BEV was given intravenously as a single agent or in combination with chemotherapy for both single- and multiple-dose schedules. Population PK analysis was performed using NONMEM. Precision of parameter estimation was evaluated using bootstrapping.
RESULTS: In total, 9,742 BEV Cs from 1,792 patients with colorectal, breast, non-small cell lung and prostate cancer were analyzed. BEV doses ranged from 1 to 20 mg/kg at a dosing frequency ranging from every 1 to 3 weeks. A two-compartment model best described the data. The population estimates (SE%) of BEV clearance (CL), central (V1) and peripheral volume of distribution for a 70 kg patient are 9.01 mL/h (1.2%), 2.88 L (0.9%) and 2.57 L (3.1%), respectively. The half-life is 19.6 days. CL and V1 increase with weight, and are higher in males than females by 14% and 18%, respectively. CL decreases with albumin and increases with alkaline phosphatase. BEV exposure is higher in patients receiving interferon treatment than other treatments. Visual predictive check and goodness-of-fit plots confirmed adequate fit and superiority of the final model.
CONCLUSION: A robust BEV population PK model was developed and can be used to simulate BEV exposure to optimize BEV dosing strategies. BEV PK and factors correlated with BEV PK are similar with previous analyses. BEV PK showed dose linearity within the wide dose range. Short-term PK up to 30 days after dose and long-term PK up to 6 months after dose are comparable.