Whole exome correlates in metastatic hormone sensitive prostate cancer (mHSPC): Results from E3805 CHAARTED.

5088 Background: The landscape and clinical impact of genomic variants in mHSPC are incompletely understood. We have shown that deleterious germline BRCA2 alterations (alts) associate with shorter time to castration resistant prostate cancer (TTCRPC) on androgen deprivation therapy (ADT), but not on ADT plus docetaxel (ADT+D) (ASCO 2023). Somatic variation from tissue of patients (pts) with mHSPC is not well characterized and may influence therapeutic outcomes. Methods: We performed whole exome sequencing of HSPC specimens obtained at initial diagnosis and germline DNA from whole blood from pts in the CHAARTED trial (NCT00309985) of ADT vs ADT+D. Somatic single nucleotide variants (SNV) and allele-specific copy number alts (CNAs) were identified, including estimation of tumor mutational burden (TMB) and copy number burden (CNB). Prognostic effects were evaluated in both arms. Loss of selected tumor suppressor genes was defined by monoallelic or biallelic loss. TTCRPC and overall survival (OS) were estimated by Kaplan-Meier method. Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were estimated by Cox models. Results: After quality control, 68 tumor-normal cases were analyzed. The majority of pts had synchronous (51.5%) and high volume (58.8%) disease. Most frequent somatic SNVs were TP53 (33.8%), PTEN (7.4%), PIK3CA (5.9%) and SPOP (5.9%). AR amplification was uncommon (4.4%), however frequent MYC gain (60.3%) was seen as well as monoallelic deletion of NKX3-1, TP53, PTEN and BRCA2. Whole genome doubling occurred in 14.7%. Median TMB and CNB were 4.7 mut/Mb and 9.1%, respectively. CNB was elevated in synchronous (p=0.01) and high volume (p=0.046) disease. CNB greater than median was associated with shorter TTCRPC in the overall cohort (HR 2.06, 95%CI 1.13-3.75, p=0.015) and ADT+D arm (HR 2.98, 95%CI 1.17-7.58, p=0.016). The effect in the overall cohort was reduced after adjustment for volume and presentation (HR 1.72, p=0.091). A compounding effect of PTEN and TP53 alts was seen in the ADT arm where median TTCRPC for wild-type (WT), 1-gene hit, 2-gene hit was 19.6 mos, 8.5 mos, 6.3 mos, respectively. The median OS of WT vs 1-hit/2-hit loss was 47.1 mos vs 26.8 mos (HR 2.05, 95% CI 0.90-4.68, p=0.087). Similar differences in TTCRP and OS by PTEN/ TP53 status were not observed with ADT+D. Conclusions: The genomic landscape of mHSPC is characterized by frequent alts in putative drivers known to be enriched in mCRPC (Table). We observed a low rate of genomic instability and AR amplification/mutation similar to non-metastatic HSPC. Concordant with the STAMPEDE trial, greater CNB confers a higher risk of progression. Combined tumor suppressor gene alts associate with prognosis in mHSPC, which may differ by therapy intensification. Table: see text