Marion Bouillon-Pichault (1), Claire Brillac (1), Christine Veyrat-Follet (1), Céline Amara (1), Nathalie Fagniez (1) and Céline Nicolazzi (1)
(1) Sanofi, Paris, France
Objectives: In oncology drug development, tumor growth inhibition (TGI) in xenograft mouse models are typically used to assess efficacy in preclinical setting but there is a long held debate over their clinical predictive capabilities. Therefore, a significant challenge is the selection of optimal dose and dosing regimen for the clinic. The aim of this study was to develop a translational oriented population PK/PD model of TGI for a new antibody drug conjugate (ADC), SARx, in SCID mice bearing patient-derived primary colon tumor, and then to apply a translational strategy to help selecting appropriate dosing schedule for first in human study.
Methods: First, a population PK model for SARx has been developed using data from 113 xenograft mice from a pharmacology study where SARx was administered as single i.v. bolus administration at doses of 0.75, 2.5 and 5 mg/kg. Then, a Simeoni tumor growth model was applied to tumor volume data coming from the previously described study (470 tumor volume measurements following single injection of SARx) and another efficacy pharmacology study in 42 mice evaluating single and repeated administrations from 5 to 20 mg/kg with different schedules, every 2 weeks, weekly and bi-weekly (673 tumor volume measurements). Of 1143 observations, 28.6% of tumor volumes measured were below the limit of palpation. In parallel, TK monkey data (single and multiple doses) were modelled and were used to predict SARx PK in man based on allometric scaling. The population PK and PK/PD analyses were performed using the Stochastic Approximation Expectation Maximization algorithm for nonlinear mixed-effects models implemented in MONOLIX.
Results: The TGI model was found to predict well both single and repeated dose tumor volume data. An increase of the drug potency (K2) was observed with the number of administrations when the dose is split. The typical estimated threshold concentration (CT) after repeated doses was 4.6 µg/mL. This increase in K2 is largely supported by the use of repeated dosing regimen demonstrating the interest in future clinical trial of the investigations of split administrations. The predicted PK was used to propose the minimal dose in human allowing maintaining plasma concentrations above the established CT.
Conclusions: This translational strategy may be valuable tool to help designing future clinical studies of our ADC compounds and to select the most appropriate dosing regimen for tumor eradication.
Reference: PAGE 23 (2014) Abstr 3014 [www.page-meeting.org/?abstract=3014]
Poster: Drug/Disease modeling - Oncology