Abstract 2725: Kinase activity profiling combined with genotyping as a tool for predictive biomarker discovery for the treatment of gastroesophageal adenocarcinoma (GEC)

Abstract

Abstract Background: Molecularly targeted therapy holds promise for the treatment of GEC, but interpatient molecular heterogeneity has proven to be a hurdle towards their successful implementation. Here we describe a new approach towards patient selection where genotyping is combined with kinase activity profiling. Kinases are the main targets for the new precision drugs being developed currently. The possibility of ex vivo testing of these new kinase inhibiting drug is a useful feature of the PamChip kinase assay applied, which may be predictive for treatment. Methods: Nine explanted tumors were genotyped (cMET, kRAS, PI3K, HER2, FGFR2, EGFR) and evaluated for kinase activities by analyzing cell lysates using dynamic peptide microarrays (PamChip assay). This profiling was used to determine i) if there are discriminatory profiles that identify tumor cells with various known genomic drivers; ii) if specific inhibitors abrogated these activity profiles selectively; iii) if patient samples (ascites fluid, cultured patient cells and/or human tumor xenografts, and frozen samples directly from patients) were amenable to this analysis with reproducible results. Cells or tissue cryosections were lysed in buffer with phosphatase and protease inhibitors. Protein tyrosine kinase (PTK) activity profiles were generated on PamChip® peptide microarrays comprising peptidic substrates derived from known human phosphorylation sites. The direct enzymatic effect of the kinase inhibitors crizotinib, affatinib and a dual cMET/RON kinase inhibitor was tested by directly spiking these compounds (at 0.3, 3.0 and 30 uM) into the lysate just before analysis and comparing the results to DMSO control. Peptide phosphorylation was monitored using fluorescently labeled antibodies and quantified with Bionavigator software. Results: Kinase activity profiles were highly reproducible (r2=0.997) when repeated on GEC cell lysates. Baseline activity profiles differed between molecular categories within GEC, including HER2 amplification (amp+), MET amp+, PI3K mutated, FGFR2 amp+ and KRAS amp+/mutated subsets. Specific inhibitors towards the cMET or EGFR driver events decreased substrate phosphorylation concentration dependently. Lysates from cell lysates with known MET amp+ were selectively sensitive to crizotinib and much less to affatinib. Conclusions: Kinase activity profiling with the multiplex PamChip kinase assay can categorize cell lines and tumors into different molecular subsets. Treatment of lysates with specific kinase inhibitors and combining this functional proteomics information with genomics data, is a new strategy we propose here to predict responses to pharmacotherapy. Citation Format: Daniel V.T. Catenacci, Peng Xu, Les Henderson, Dirk Pijnenburg, Adrienne van den Berg, Rob Ruijtenbeek. Kinase activity profiling combined with genotyping as a tool for predictive biomarker discovery for the treatment of gastroesophageal adenocarcinoma (GEC). [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2725. doi:10.1158/1538-7445.AM2014-2725