To: Miljenko Zuanic who wrote (19 ) 4/6/2001 12:10:34 AM From: Miljenko Zuanic Respond to of 1840 A Biomathematical Model of Neoplastic Cell Growth and Prediction in Silico of Effective Doses of ABX-EGF in Cancer Patients Lorin Karsten Roskos, Michelle Lohner, Gisela Schwab, Xiao-Dong Yang, Abgenix, Inc., Fremont, CA. A biomathematical model of A431 tumor growth in a mouse xenograft model, suppression of A431 growth by ABX-EGF, a fully-human monoclonal antibody against the EGF receptor, and pharmacokinetics (PK) of ABX-EGF in mice and monkeys were used to predict dosing regimens that maintain effective levels of ABX-EGF in patients. The novel model of neoplastic cell mitosis, aging, and apoptosis was linked to a PK model of ABX-EGF in mice to calculate an IC50 for suppression of cell proliferation. From the IC50 and other model parameter estimates, two additional parameters were calculated: (1) IC90, the serum level of ABX-EGF producing 90% (near maximum) suppression of mitosis and (2) Cerad, the serum level that, if maintained, would produce complete tumor eradication. Dose-ranging PK data from monkeys were modeled. Clearance of ABX-EGF in monkeys was by parallel linear clearance and Michaelis-Menten (km=4.98 mcg/mL) processes. Allometric scaling of clearance from monkeys to humans was conducted, and simulations of weekly dosing in humans were conducted for the range of doses proposed for a Phase 1 clinical study (0.01 to 6 mg/kg/week). Preliminary PK data from the Phase 1 study at doses up to 0.3 mg/kg/week (current dose cohort) were nearly superimposable on the simulations, indicating that the PK in monkeys is a good predictor of PK in humans. The simulations predict that trough levels of ABX-EGF will exceed the mouse-model estimate of IC90 at the 1 mg/kg/week dose, and the Cerad will be exceeded at a dose of 3 mg/kg/week. Further simulations indicate that biweekly dosing regimens will be feasible. Additional model analysis suggests a strong rationale for combination studies of ABX-EGF with cytotoxic chemotherapeutic agents, since combination with therapeutics that increase neoplastic cell death rate may lower the threshold for Cerad.