Single and multi-hit PIK3CA short variant (SV) genomic alterations (GA) in clinically advanced prostate cancer (CAPC): A genomic landscape study.

258 Background: Tumors harboring 2 or more PIK3CA Short variants (SV) (“Multi-hit”) have been described in breast cancer as linked to enhanced clinical outcome from anti-PIK3CA targeted therapies including alpelisib and investigational agents in clinical trials. The landscape and clinical implications of multi-hit PIK3CA alterations in other tumors, including in CAPC, are underexplored. Methods: 19,978 CAPC samples underwent hybrid capture based comprehensive genomic profiling (CGP) to evaluate all classes of GA and determine tumor mutational burden (TMB), microsatellite instability (MSI), genomic ancestry, signature and loss of heterozygosity (gLOH). Tumor cell PD-L1 expression was determined by IHC (Dako 22C3). Results: 18,741 (93.8%) CAPC were PIK3CA wild type (WT), 1,155 (5.8%) featured a single PIK3CA SV and 82 (0.4%) featured multi-hit PIK3CA SVs. The median ages of CAPC patients with single hit (69.1 yrs) or multi-hit (68.7 yrs) PIK3CA SV were older than the PIK3CA WT (67.6 yrs) CAPC (p<.0001 for both). At 14.0%, African ancestry was more frequent in PIK3CA WT CAPC than in single-hit (10.4%; p=.001) and multi-hit (10.2%; not significant). Single-hit (81.4%; p<.0001) and multi-hit (85.2% p=.05) PIK3CA SV CAPC featured significantly more MMR trinucleotide genomic signatures than PIK3CA WT (64.8%). Single-hit (2.7%; p=.02) and multi-hit (5.8% p=.05) PIK3CA SV CAPC featured significantly more POLE trinucleotide genomic signatures than PIK3CA WT (0.4%). MSI high status was significantly more common in both PIK3CA single-hit (12.4% vs 2.5%; p<.0001) and multi-hit (35.4% vs 2.5%; p<.0001). Mean TMB was also significantly higher in single-hit PIK3CA (11.1 vs 3.5 mut/Mb; p<.0001) and multi-hit (42.9 vs 3.5 mut/Mb; p<.0001). Noteworthy differences in GA of potential importance for CAPC pts included significantly higher frequencies of GA in BRCA2 in multi-hit vs WT (18.3% vs 8.5%; p=.019), ATM in multi-hit vs WT (13.4% vs 5.6%; p=.02) and PTEN in single-hit vs WT (40.2% vs 30.1%; p<.0001) and lower frequencies of GA in CDK12 (3.6% vs 5.6%; p=.009) and SPOP (7.4% vs 9.8%; p=.012) in single-hit vs WT. There were no differences in gLOH or PD-L1 expression among the 3 groups. Conclusions: Although uncommon, the identification of multi-hit PIK3CA GA in CAPC highlights a potentially unique subtype of this disease that may be associated with enhanced responsiveness to anti-PIK3CA targeted therapy strategies.