Integrative transcriptomics defines the culture-imposed identity of ovarian cancer organoids

Background: Patient-derived organoids have transformed preclinical oncology by providing three-dimensional tumor models that preserve genomic fidelity and heterogeneous drug responses. However, whether these models faithfully recapitulate the transcriptomic identity of their primary tumor counterparts—particularly with respect to molecular subtypes and pathway architecture—remains an unresolved question of fundamental translational relevance. Methods: We conducted an integrative transcriptomic comparison of 17 high-grade serous ovarian cancer (HGSOC) organoids and 271 high-purity The Cancer Genome Atlas (TCGA) -OV primary tumor samples. Following variance-stabilizing transformation and ComBat batch correction, we applied ConsensusOV molecular subtype classification, unsupervised hierarchical clustering, differential gene expression analysis, gene set variation analysis, and gene set enrichment analysis to characterize patterns of concordance and divergence. Results: Unsupervised clustering demonstrated that all 17 organoids clustered exclusively within cluster C1. Cluster C1 was characterized by a more transcriptionally mixed profile, higher stromal interaction scores, and enrichment of epithelial-to-mesenchymal transition and hypoxia pathways, reflecting adaptation to the collagen-rich extracellular matrix environment of organoid culture. In contrast, cluster C2—comprising only TCGA samples—exhibited transcriptomic features consistent with near-pure tumor cell populations dominated by proliferative programs. Within C1, organoids showed marked upregulation of HALLMARKMITOTICSPINDLE and HALLMARKG2MCHECKPOINT relative to C1 primary tumors. Conclusion: These findings indicate that HGSOC organoids acquire an extracellular matrix-driven, stromal-interactive transcriptional phenotype that clusters with heterogeneous primary tumors, while simultaneously exhibiting a culture-induced proliferative program. This work provides a quantitative framework for defining the appropriate scope and limitations of organoid models in translational ovarian cancer research.