Abstract PR011: The carcinoma of prostate sequencing of tumor and clinical endpoints (CAPSTONE) project: A clinico-genomic resource to enable patient-centric genomic research and improve the actionability of genetic testing in metastatic prostate cancer

Abstract

Abstract Introduction Key barriers restrict the utility (or use) of genetic information in making treatment decisions in metastatic PCa. First, the lack of platforms that link genomic and clinical outcome data limits our ability to uncover clinico-genomic associations. Second, predictions based on univariate biomarkers (e.g. individual mutations), are insufficient given the heterogeneity of metastatic genomes. Third, lack of predictive biomarkers for standard therapies limits the use of testing to prioritize patients for trials. We initiated CAPSTONE, a clinico-genomic resource and research study, to overcome these barriers. Methods CAPSTONE comprises three complementary discovery n=325, validation n=429, and real-world n=1,800 clinico-genomic cohorts, reflecting data characteristics in the research and real-world contexts. The discovery cohort includes PCa patients with RNA-seq and matched tumor germline exome sequencing (WES) from UM, the validation cohort has similar WES/RNA-seq data from non-UM mCRPC patients from the SU2C study. The real-world dataset is from PROMISE, a multi-institutional effort among 25+ PCa centers across the US utilizing commercial tests. Genetic and clinical data are harmonized through a uniform computational workflow and interoperable clinical data model. Results In this early report, we present results from the analysis of the discovery cohort with a median follow of 106 months (IQR, 90-121) and 91% PCa adenocarcinoma histology (n=292/325). 36 samples were from patients with localized disease, 101 from metastatic hormone sensitive PCa (mHSPC) and 191 from metastatic castrate resistant PCa (mCRPC). We called driver mutations, measured chromosomal instability (CIN), and obtained RNA-seq based readouts of pathway activity and immune infiltration. We were able to confirm all reported clinico-genomic associations in metastatic PCa, and identify multiple new ones. AR amplifications were found in 10% of mHSPC and were associated with significantly shorter survival on first-line therapy (p=0.0061). TP53 mutations were associated with worse OS from the time of biopsy on multivariable analysis (HR: 2.2; 1.7-2.9, p<0.001). OS from biopsy was worse for dual TP53/RB1 mutants when compared to TP53 or RB1 mutants alone, independent of the disease state at time of biopsy (HR, 4.3; 95% CI: 2.7-7.0, p<0.001). Surprisingly, both TP53 loss (p<0.001) and AR (p<0.001) amplification was associated with increased CIN (wGII). CIN increased in progression from localized to mHSPC to mCRPC (p=3e-5) and was independently associated with worse OS when controlled for covariates (HR: 3.28; 95% CI: 1.39-7.80, p=0.007). wGII-high tumors show upregulation of EMT-associated genes (q=6.8e-14), downregulation of AR-target genes (q=2.7e-6), and show lower CD8+ T cell levels (R=-0.24,p=7.7e-4). Conclusions We demonstrate the utility of CAPSTONE in identifying single-gene and complex genetic biomarkers associated with patient outcomes. Most notably, we identify CIN as one of the strongest independent genetic predictors of poor patient outcomes in PCa. Citation Format: Marcin Cieslik, Ryan Rebernick, Liat Hammer, Matt McFarlane, Thomas Westbrook, Yi-Mi Wu, Dan Robinson, Dan Spratt, Ajjai Alva, William Jackson, Zachery Reichert, Arul Chinnaiyan, Joshi Alumkal, Robert Dess. The carcinoma of prostate sequencing of tumor and clinical endpoints (CAPSTONE) project: A clinico-genomic resource to enable patient-centric genomic research and improve the actionability of genetic testing in metastatic prostate cancer [abstract]. In: Proceedings of the AACR Special Conference: Advances in Prostate Cancer Research; 2023 Mar 15-18; Denver, Colorado. Philadelphia (PA): AACR; Cancer Res 2023;83(11 Suppl):Abstract nr PR011.