OGC O05 - Utilising patient derived organoids to predict the response to treatment in gastric cancer

Abstract Background Gastric adenocarcinoma remains a significant global health burden. For advanced disease, treatment remains multi-modal with neoadjuvant chemotherapy followed by resectional surgery. Although the FLOT4-AIO trial has established the regimen of 5-fluorauracil, Leucovorin, Oxaliplatin and Docetaxel (FLOT) as the gold standard, some patients demonstrate a poor or no response to treatment, resulting in adverse oncological outcomes. Currently there are no accurate biomarkers to assess or predict the treatment response to neoadjuvant therapy. This study aims to explore the potential of patient-derived organoids as a predictive model to assess the treatment response to neoadjuvant chemotherapy in gastric adenocarcinoma. Method Patient derived organoids are a novel, 3 dimensional in-vitro model which recapitulates the different cells that constitute the tissue of origin. From June 2023 to June 2024, 25 patients with a confirmed diagnosis of gastric adenocarcinoma, were recruited for tissue acquisition for organoid generation. Endoscopic biopsies of cancer tissue taken during OGD and staging laparoscopy were used to generate organoids from each patient. Organoids were subsequently expanded to a sufficient volume to allow cytotoxicity assays in response to different doses of the FLOT regimen. The organoid cytotoxicity response was compared to the patients radiological and pathological response following treatment. Results Organoids were successfully generated from 76% of recruited patients (n=19). Of these, 4 patients and were excluded due to being unfit for neoadjuvant chemotherapy, whilst 2 were excluded due to the confirmation of metastatic disease, precluding neoadjuvant treatment. Finally a single patient was excluded due to mortality during neoadjuvant treatment. Of the remaining organoids, 5 were subjected to FLOT treatment within 2 weeks of tissue acquisition. Organoids demonstrated unique dose response profiles following with statistically significant differences in the IC50 values (p =0.048). The organoid responses corresponded to the clinical responses demonstrated by the patient they were derived from. Conclusion In conclusion, this study underscores the potential of patient-derived cancer organoids as a predictive model for assessing the response to neoadjuvant chemotherapy in gastric adenocarcinoma. The generation of organoids from a significant proportion of patients and the subsequent drug response testing within 2 weeks of tissue acquisition suggests this novel method has the potential to become a valuable tool for predicting patient-specific responses and guiding personalised treatment plans. As such, further research is warranted to further assess the validity of this model, refine the existing techniques and explore its potential integration into clinical practice.