Abstract 4605: Exploring resistance mechanisms in cancer using patient-derived organoids reveals replication stress and EMT as potential biomarkers

Abstract Mutational processes such as chromosomal instability and replication stress are important molecular features of head and neck squamous cell carcinoma (HNSCC) and non-small cell lung cancer (NSCLC), contributing to poor patient survival. Their changing influences during cancer development creates challenges in identifying response biomarkers. Patient-derived organoid 3D cancer models (PDOs) have been shown to match their primary source material and to reflect tumour biology more faithfully than cell lines. In our companion abstract, we detail how we have built the Digital Biological Twin lab to study PDOs for drug testing and biomarker discovery. In this work, we used PDOs to investigate resistance to PARP inhibitors (PARPi), which are mechanistically linked to DNA repair and replication. 10 PDOs (6 HNSCC, 4 NSCLC) generated from untreated tumor resections were bulk RNA-sequenced and treated with Niraparib. IC50, cell viability measured by flow cytometry and protein expression levels were used to classify 3 PDOs as sensitive to treatment. Expression analysis, stratifying by PARPi sensitivity, resulted in 287 differentially expressed genes (DEGs). Enrichment analyses found involvement of the endothelial-to-mesenchymal transition (EMT) and small molecule transport pathways. Knowing that EMT is linked to multi-drug resistance, we treated the PDOs with Cisplatin and found a significant correlation to PARPi IC50 values. By examining the biology of the DEGs, we categorized them using MSigDB into five gene sets, each with 1 to 14 genes. We functionally validated the gene sets by assessing PARPi susceptibility in 45 cell lines pre and post a genome-wide CRISPR knockout, finding four significantly enriched in PARPi gene dependencies. For two replication stress-derived gene sets, we demonstrated significant concordance between PARPi sensitivity status of PDOs and cell lines from DepMap, and with progression free interval of HNSCC and NSCLC patients from TCGA. Leveraging Cox analysis on these patients, one replication stress gene set stratified the patients by showing a significant decrease in relapse probability. In summary, our results indicate that PARPi resistance is associated with the deregulation of genes related to small molecule transport, multi-drug resistance and EMT. The consistent findings around replication stress across disease models strongly suggests a conserved and central role in resistance mechanisms. We also identify a replication stress gene set that significantly correlates with resistance in cancer models and relapse in patients. Further work is needed to translate our findings into a predictive model for drugs linked to replication stress. Here we showed that PDOs can be successfully used for furthering our understanding of resistance mechanisms while improving molecular stratification of patients with direct clinical implications. Citation Format: Ruben M. Drews, Finnian Firth, Paul R. Barber, Anna Pasto, Tony T. Ng. Exploring resistance mechanisms in cancer using patient-derived organoids reveals replication stress and EMT as potential biomarkers [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 4605.