A modular patient-derived organoid–xenograft platform reveals molecular and clinical trajectories of prostate cancer progression

Abstract Patient-derived organoids (PDOs) are becoming increasingly important in prostate cancer (PCa) translational research. However, direct proof-of-concept studies demonstrating their ability to model disease evolution, identify relevant biomarkers, and accurately predict treatment response in PCa patients, remain scarce. Here, we report the establishment of serially transplantable xenografts series derived from two advanced PCa PDO lines, which can be further re-cultured as organoids. Newly-generated model series maintain key phenotypic, genomic, and functional characteristics of the original patient tumors, and emulate relevant molecular subtypes of advanced PCa. Single-cell RNA sequencing (scRNA-seq) analysis uncovers transcriptomic differences between xenograft and organoid models, as well as signaling pathways which are largely preserved and can be targeted ex vivo. Functional drug profiles correlate with molecular and clinical attributes, as exemplified by response to androgen receptor (AR) pathway inhibitors and glucocorticoid-mediated AR signaling activation. Longitudinal scRNA-seq analysis of perturbed PDOs identifies a rare PROX1 + /ALDH1A1 + cell population, which pre-exist in the treatment-naïve setting and is significantly enriched upon androgen deprivation. Notably, this cell population is similarly enriched in post-treatment samples of the original patient and is associated with aggressive AR-negative PCa molecular subtypes, suggesting a potential link with PCa progression. Our study provides proof-of-concept evidence that organoids can mirror PCa patient-specific drug sensitivity profiles and molecular paths of disease progression, uncovering pertinent biomarkers. Ultimately, our organoid-xenograft model series provide a modular and scalable platform that can readily be used for mechanistic and translational studies.