239 Background: Tumor heterogeneity and clonal evolution are key drivers of cancer progression and therapeutic resistance. In metastatic prostate cancer, the aggressive variant prostate cancer (AVPC) phenotype, often defined by concurrent alterations in TP53, RB1, and PTEN, is associated with poor clinical outcomes and rapid progression under standard treatments. However, the evolutionary trajectories that sustain resistance and promote relapse in AVPC remain incompletely characterized. To elucidate the clonal and genomic mechanisms underlying therapeutic failure, we performed a whole-genome analysis of AVPC tumors before and after chemotherapy. Methods: We analyzed 40 AVPC tumors from frozen metastatic biopsy cores collected at baseline in men enrolled in the MDA 2017-0133 (CCOLA) trial. High-depth whole-genome sequencing (~80–120×) was performed on all baseline samples. A longitudinal subset of nine patients also underwent matched post-chemotherapy biopsies after six cycles of carboplatin–cabazitaxel, yielding nine post-treatment tumors; per protocol, post-chemotherapy biopsies were obtained only from patients who progressed on treatment. Somatic mutations, copy number alterations and tumor purity estimates were integrated to reconstruct clonal/subclonal architecture and quantify genomic instability. Results: Concurrent inactivation of ≥2 of TP53, RB1, and PTEN, were frequent and associated with adverse clinical outcomes. Despite chemotherapy, overall clonal diversity persisted, indicating sustained subclonal complexity under treatment pressure. In longitudinal analyses, post-treatment founder clones commonly descended from pre-existing minor subclones that accrued additional mutations during therapy. Progression-specific TP53 mutations were identified in 14/25 (56%) evaluable patients and were also observed in 3/15 (20%) patients with prolonged progression-free survival, underscoring the mechanistic heterogeneity of resistance. The cohort exhibited pronounced genomic instability with extensive subclonal copy number alterations, exceeding prior reports and revealing the underestimated complexity of AVPC. Conclusions: Therapy-resistant subclones emerge from pre-existing tumor lineages that evolve under selective pressure, contributing to progression and resistance in AVPC. The persistence of clonal diversity despite chemotherapy shows that many tumors retain evolutionary capacity, which limits the reliability of single time-point genomic models for prognosis or treatment selection. These findings support earlier, serial detection of emergent resistant lineages and the use of genomic profiling to guide timely, individualized treatment strategies. Clinical trial information: NCT03263650 .
Tran et al. (Sun,) studied this question.