Identification of predictive response and resistance factors to targeted therapy in gastric cancer using a systems medicine approach

Background: Progress in treatment of gastric cancer (OC) has been limited due to molecular and clinical heterogeneity. Novel drugs targeting the receptor tyrosine kinases (RTKs) HER2 and EGFR have shown mixed success in clinical trials. While the HER2 antibody trastuzumab has been approved for GC treatment, the EGFR antibodies cetuximab and panitumumab failed to improve patient outcomes. The SYS-Stomach consortium, supported by the Federal Ministry of Education and Research (BMBF), aims to investigate differences in the mode-of-action of both treatments and to unravel primary and secondary resistance mechanisms. Material and Methods: We apply systematic molecular multi-omics and ceO phenotypic measurements to GC cell lines. From these we derive mechanistic and statistical models of the signalling networks coupled to cellular phenotypes and agent-based cellular behaviour models. The models will be validated against cell culture and clinical sample derived molecular and morphological tumour characteristics based on MALDI imaging mass spectrometry, a powerful toot to investigate the distribution of molecules in tumour sample sections. Validated models will be used to predict potential response and resistance factors of EGFR- and HER2- directed treatment. These response predictors will be validated in tumour samples from GC patient cohorts treated with cetuximab or trastuzumab. Results: We established a link between motility-focused phenotypic properties of GC cell lines with molecular characteristics in response to cetuximab and developed a GC specific semantic network connecting EGFR signalling to the regulation of cellular motility. This semantic model has been used to inform the development of a mechanistic mathematical model for the EGFR pathway which describes the measured kinetic and dose response data obtained for cetuximab responder and non-responder cell lines. This mechanistic model is linked to the phenotypic measurements motility and invasiveness using a regression model. An agent-based model for gastric tumour growth in 3D was developed. It adapts the motility of each tumour cell according to its micro-environment, through employing the above-mentioned mechanistic model for all agents (cells). The clinical observational VARIANZ study which aims to assess resistance mechanisms in HER2+ tumour samples from patients receiving trastuzumab, has been established, HER2 testing in a central pathology institute has been put in place. Recruitment is ongoin