Arzneimittelforschung 2012; 62(04): 194-201
DOI: 10.1055/s-0031-1299772
Original Article
© Georg Thieme Verlag KG Stuttgart · New York

The Observed Correlation between in vivo Clinical Pharmacokinetic Parameters and in vitro Potency of VEGFR-2 Inhibitors

Can this be used as a Prospective Guide for the Development of Novel Compounds?
B. Benjamin
1   Vanthys Pharmaceutical Development [P] Ltd., Phoenix Pinnacle, Ulsoor Road, Bangalore, India
,
M. Sahu
1   Vanthys Pharmaceutical Development [P] Ltd., Phoenix Pinnacle, Ulsoor Road, Bangalore, India
,
U. Bhatnagar
1   Vanthys Pharmaceutical Development [P] Ltd., Phoenix Pinnacle, Ulsoor Road, Bangalore, India
,
D. Abhyankar
1   Vanthys Pharmaceutical Development [P] Ltd., Phoenix Pinnacle, Ulsoor Road, Bangalore, India
,
N. R. Srinivas
1   Vanthys Pharmaceutical Development [P] Ltd., Phoenix Pinnacle, Ulsoor Road, Bangalore, India
› Author Affiliations
Further Information

Publication History

received 27 November 2011

accepted 22 December 2011

Publication Date:
30 January 2012 (online)

Abstract

Literature data on the clinical pharmacokinetics of various VEGFR-2 inhibitors along with in vitro potency data were correlated and a linear relationship was established in spite of limited data set. In this work, a model set comprised of axitinib, recentin, sunitinib, pazopanib, and sorafenib were used. The in vitro potencies of the model set compounds were correlated with the published unbound plasma concentrations (Cmax, Cavg, Ctrough). The established linear regression (r2>0.90) equation was used to predict Cmax, Cavg, Ctrough of the ‘prediction set’ (motesanib, telatinib, CP547632, vatalanib, vandetanib) using in vitro potency and unbound protein free fraction. Cavg and Ctrough of prediction set were closely matched (0.2–1.8 fold of reported), demonstrating the usefulness of such predictions for tracking the target related modulation and/or efficacy signals within the clinically optimized population average. In case of Cmax where correlation was least anticipated, the predicted values were within 0.1–1.1 fold of those reported. Such predictions of appropriate parameters would provide rough estimates of whether or not therapeutically relevant dose(s) have been administered when clinical investigations of novel agents of this class are being performed. Therefore, it may aid in increasing clinical doses to a desired level if safety of the compound does not compromise such dose increases. In conclusion, the proposed model may prospectively guide the dosing strategies and would greatly aid the development of novel compounds in this class.