Abstract 2322: Rapid discovery of novel antiangiogenic drugs using a phenotypic profiling platform with high predictive value for preclinical models of cancer

Abstract

Abstract Deregulation of angiogenesis plays a major role in a number of human diseases, most notably cancer. While angiogenesis inhibitors are promising anticancer drug candidates, existing FDA approved therapeutics have shown limited efficacy in the clinic. The majority of angiogenesis inhibitors under clinical development have been designed with the help of high-throughput screening techniques targeting single molecular moieties. These methods have proven effective but it has been long recognized that they lack correlation with activity in preclinical models of cancer which results in high levels of attrition during early stages of drug discovery and phase II clinical trials. We propose that multi-parameter analysis discovery methods are required to capture and translate the tumor angiogenesis complexity into the drug discovery process. Here we introduce a novel high-content cell-specific fluorescence platform for discovery of antiangiogenic agents with highly predictive value for preclinical end points. We have validated this technology by screening the 1970 small molecules of the NCI Developmental Therapeutics Program Diversity Set. This platform features a primary screening based on high content growth and tube formation assays using VEGFR2-nonexpressing, fluorescent-reporter endothelial cells. Over 30 morphological parameters were quantitatively assessed in the tube formation assays in an automated fashion with the newly developed image analysis software AngioApplication. 2.3% (46) of all the small molecules in the library showed growth inhibitory activity, 3.5% (70) significantly blocked tube formation and 0.5% (11) demonstrated both activities. None of the lead compounds were found to interfere with tubulin polymerization or inhibited receptor tyrosine kinase activity. Seven lead compounds were evaluated in a leiomyosarcoma xenograft model of angiogenesis. All showed anti-tumor activity, and two of the compounds (CID 5458317 and CID 429599) blocked tumor growth with comparable efficacy as bevacizumab. An antiangiogenic score computed based on tube formation assessment and gene expression profiling was highly predictive of tumor growth inhibition in vivo. The antiangiogenic mechanism of action of these compounds was validated using tumor onplants on chicken chorioallantoic assays which consistently tracked observations in xenograft models. Intravital microscopy assessment of xenograft tumors treated with CID 5458317 showed massive leakiness and vascular regression of tumor vasculature but not in adjacent normal vasculature. In conclusion, we have developed a drug discovery phenotypic screening platform which shows high predictive value for preclinical models of cancer. Using this platform, several small molecules with novel antiangiogenic mechanisms of action have been identified. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2322. doi:1538-7445.AM2012-2322