Achim Fritsch (1), Friederike Kanefendt (1), Andreas Lindauer (1), Martina Kinzig (2), Fritz Sörgel (2), , Ulrich Jaehde (1)
(1) Institute of Pharmacy, Clinical Pharmacy, University of Bonn, Germany, (2) IBMP - Institute of Biomedical and Pharmaceutical Research, Nürnberg-Heroldsberg, Germany
Objectives: Oral multi-tyrosinekinase inhibitors are widely used and established treatment options in numerous tumour entities. Yet, response rates are often low and valuable surrogate endpoints which indicate therapy response are lacking. However, there is increasing evidence that several circulating proteins may serve as predictive biomarkers in pharmacokinetic/pharmacodynamic (PK/PD) models.
The aim of this investigation was the development of PK/PD models to describe the anti-angiogenic response to sunitinib and to identify potential predictors for efficacy in patients with metastatic colorectal cancer and renal cell cancer.
Methods: 21 patients with CRC and liver metastases receiving a daily dose of 37.5 mg sunitinib on a 4 weeks on/2 weeks off treatment schedule in addition to FOLFIRI participated in this prospective, open label, single arm, multicentric study. Concentrations of sunitinib, its active metabolite SU12662, measured using LC-MS/MS. VEGF-A, VEGFR-2 and 3 were quantified using commercially available ELISA kits and validated immunoassays, respectively. PK/PD models were developed using NONMEM (v 7.1.2). Potential predictors for time to progression (TTP) were analyzed using Cox regression and a model-based approach.
As part of the non-interventional EuroTARGET project, which aims at identifying and characterizing host and tumour-related biomarkers in metastatic renal cell cancer (mRCC), these models will be further extended.
Results: Biomarker concentration-time courses could be well described by an indirect response model. Minimum concentrations relative to baseline were estimated as 0.63 and 0.59 for sVEGFR 2 and sVEGFR 3, respectively. Concentrations of both biomarkers were highly correlated, however did not predict TTP. Instead, higher exposure to the unbound active drug (sum of sunitinib and SU12662) was identified as predictor for TTP (HR: 0.49 (95% CI: 0.27-0.88), p=0.013). Within the EuroTARGET project, recruitment of mRCC patients receiving sunitinib is still ongoing.
Conclusions: Concentration-time profiles of drug, metabolite, and biomarkers could be well described by PK/PD models. The extent of biomarker response was comparable with healthy volunteers but did not predict tumour response in patients with mCRC. In contrast, TTP was correlated with active drug pharmacokinetics. The developed PK/PD models will be extended using drug and biomarker concentrations from mRCC patients in order to analyze differences between the two tumor entities
Reference: PAGE 24 () Abstr 3610 [www.page-meeting.org/?abstract=3610]
Poster: Drug/Disease modeling - Oncology