Abstract
Most patients with metastatic prostate cancer eventually develop resistance to primary androgen deprivation therapy. To identify predictive biomarker for Abiraterone acetate/prednisone resistance, we screened alternative splice variants between responders and non-responders from the PROMOTE clinical study and pinned down the most significant variant, CENPK-delta8. Through preclinical patient-derived mouse xenograft (PDX) and 3D organoids obtained from responders and non-responders, as well as in vitro models, aberrant CENPK-delta8 expression was determined to link to drug resistance via enhanced migration and proliferation. The FLNA and FLOT1 were observed to specifically bind to CENK-delta8 rather than wild-type CENPK, underscoring the role of CENPK-delta8 in cytoskeleton organization and cell migration. Our study, leveraging data from the PROMOTE study, TCGA, and TCGA SpliceReq databases, highlights the important function of alternative splice variants in drug response and their potential to be prognostic biomarkers for improving individual therapeutic outcomes in precision medicine.