Gastric cancer (GC) has a poor overall survival at 5 years. More recently, the incidence levels in patients under 50 years is also rapidly rising. A significant proportion of patients have little to no response to neoadjuvant treatment with curative intent. There is an urgent need to improve treatment options and overcome treatment resistance. In this translational project, we will use a personalised approach to understand resistance to tailor more effective treatments using novel machine learning algorithms and patient-derived organoid models for drug testing. This is an international collaboration bringing expertise together from Trinity St. James’s Cancer Institute, Technical University Dresden, Germany, and with Vivan Therapeutics, UK. We hypothesise that Artificial Intelligence (AI) machine learning tools will identify chemotherapy-refractory patients at diagnosis. In collaboration with the Technical University Dresden, combining radiology, histology, and other clinical data, this multimodal AL model will integrate multiple data types to generate a novel algorithm for treatment resistance. The development of patterns of treatment resistance will also be identified through performing genomic data assessments. In collaboration with Vivan therapeutics, this data will be examined against their library of avatars and treatment recommendations will be given to test in our generated gastric cancer patient organoids. Molecular subtypes of gastric cancer will also be assessed using immunohistochemistry for mismatch repair proteins / Claudin 18.2 / FGFR2b / PD-L1 CPS score & HER2 to inform treatment testing. Using this data, new combination treatments will be tested using patient-derived organoid models. These novel models will guide personalised treatment options tailored to each gastric cancer patient. The impact of this work will allow the introduction of more effective systemic anti-cancer treatments which will have a positive impact on management and improve outcomes for gastric cancer patients with dismal survival rates.
