Abstract 4606: Digital biological twin platform: multiomics in patient derived organoids combined with machine learning

Abstract Patient-derived organoids (PDOs) are in vitro 3D structures that mimic the complexity of tumors, including their molecular and phenotypic features. Recently, numerous studies have demonstrated their potential as preclinical model for drug screening and tailored treatment selection for patients.Non-small cell lung cancer (NSCLC) with its rapid growth and early metastasis onset, represents the most common cause of cancer-related deaths worldwide. Current treatments offer limited long-term survival, necessitating novel approaches. Immunotherapy, specifically anti-PD1 and anti-PD-L1, has transformed NSCLC treatment, but only a small percentage of patients respond. There is an unmet need to develop novel clinically relevant models of NSCLC to improve cancer patient risk stratification as well as to predict future targeted and/or immuno-therapy susceptibilities that may be met with PDOs.We have created the Digital Biological Twin lab, a platform that performs multi-omics analysis on PDOs combining advanced cellular, genomic, and molecular techniques with machine learning models, aiming to improve cancer patient risk stratification and predict future therapy susceptibilities. This platform, part of the Oncology translational research organization at GSK, will provide proof of concept data from patient samples collected in hospital biobank-linked organoid studies. In an ongoing study we are establishing PDO models from 200 NSCLC tissues collected at the Guy’s and St Thomas’ Biobank in London. The PDOs are treated with standard care, then co-cultured with pre-activated immune cells in the presence of IO drugs, and subjected to assays including exosome isolation, protein and DNA content analysis, imaging assays, flow cytometry analysis of tumor markers and immune phenotype, and transcriptomic analysis. This multi-omics approach allows for simultaneous analysis of different biological parameters after treatment to understand tumour cell interaction with immune cells and drug-induced changes. The results will be used to identify potential biomarkers of treatment response and generate hypotheses for drug resistance.Initial results in a pilot study with PDOs from NSCL and head and neck cancers have shown that several factors correlate with and can be combined in a linear elastic-net or robust Bayesian regression model that predicts resistance risk, as measured by PDO cell viability or IC50. These factors include the treatment induced change in CD45+ immune cell infiltrating, changes in tumour PD-L1 and immune PD-1 expression, changes in MYC, TIGIT and CD96 expression. More details of a genomic analysis will be presented in a companion presentation at this conference. Citation Format: Anna Pasto, Paul Barber, Halh Al-Serori, Maria Fankhaenel, Diianeira Maria Tsiridou, Shani Austin-Williams, Finnian Firth, Ruben M. Drews, Amro Ahmed-Ebbiary, Michael Finn, Mint Htun, Ellie N. McDowall, Annie-Rose Henry, Susan Ndagire, Eleni M. Karapanagiotou, Cheryl E. Gillet, Andrea Bille, Tony Ng. Digital biological twin platform: multiomics in patient derived organoids combined with machine learning [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4606.