Moran Elishmereni (1), Yuri Kheifetz (1), Michael Amantea (2), Reza Khosravan (2), Zvia Agur (1)
(1) Optimata Ltd., Israel, (2) Department of Clinical Pharmacology, Pfizer, Inc., California, USA
Objectives: Sunitinib malate, a receptor tyrosine kinase inhibitor with multiple antitumor and anti-angiogenic properties, is an approved therapeutic agent for renal cell carcinoma (RCC), imatinib-resistant gastrointestinal stromal tumors (GIST), and pancreatic neuroendocrine tumors (NET). Despite its promising potency, sunitinib has not shown sufficient response rates in patients with non-small cell lung cancer (NSCLC). We investigated the efficacy in NSCLC, and explored alternative sunitinib regimens for this indication, by developing and simulating a semi-mechanistic mixed-effects model for sunitinib in NSCLC.
Methods: A system of ordinary differential equations describing the pharmacodynamic (PD) effects of sunitinib on tumor progression, combined with a former pharmacokinetic model for the drug [1], was designed and implemented on a Monolix platform. Data from clinical trials of sunitinib in advanced NSCLC patients post-chemotherapy, consisting of patient-specific tumor size measurements and longitudinal plasma profiles of the angiogenesis-related biomarkers, vascular endothelial growth factor (VEGF), soluble receptors for VEGF (sVEGFR2 and sVEGFR3) and KIT (sKIT), were used for fine-tuning PD model parameters. Diverse assumptions were tested, e.g. relative importance of sunitinib versus its active metabolite, delayed PD effects of the drug, development of drug resistance, etc. The model also examined the predictive significance of the soluble biomarkers in the interplay with the tumor response and the potential efficacy of altered sunitinib protocols on the patient response to treatment.
Results: Analysis of our model suggests that sunitinib has an immediate inhibitive effect on tumor growth (more than its metabolite). This effect is limited by the gradual development of resistance to the drug, a process which is reversed upon termination of therapy. The model was well fit with the tumor growth dynamics observed in the patients (R2=0.99), and also retrieved the dynamics of the soluble receptors for VEGF and KIT. In contrast to the effects of sunitinib in GIST, the soluble biomarker parameters were not found to be inter-correlated.
Conclusions: Our results suggest that angiogenesis-related soluble biomarkers may play a less important role in lung cancer, and potentially explaining the lower efficacy of sunitinib in NSCLC.
References:
[1] Houk, B.E., et al., A population pharmacokinetic meta-analysis of sunitinib malate (SU11248) and its primary metabolite (SU12662) in healthy volunteers and oncology patients. Clin Cancer Res, 2009. 15(7): p. 2497-506.
Reference: PAGE 23 () Abstr 3006 [www.page-meeting.org/?abstract=3006]
Poster: Drug/Disease modeling - Oncology