Primary sarcoma of prostate: A genomic landscape study.

Abstract

181 Background: Primary sarcoma of prostate is exceedingly rare and not well-studied. We sought to describe the genomic landscape of this rare entity and identify potential therapy targets. Methods: From 19,057 cases of prostate cancers, only 11 (<.01%) cases of primary sarcomas were identified and underwent comprehensive genomic profiling (CGP) using FDA-approved hybrid capture-based system to assess all classes of genomic alterations (GA). Genomic-based ancestry, genomic signature, gLOH, MSI and TMB status were determined by CGP. Germline status was predicted using a Somatic-Germline-Zygosity algorithm. PD-L1 expression was determined by IHC (Dako 22C3 TPC scoring). Results: The routine histology and IHC stains from all 11 cases were reviewed centrally. There were 9 stromal sarcomas, and 1 each leiomyosarcoma and rhabdomyosarcoma. All patients (median age 57) were clinically advanced stage at the time of CGP. The primary tumor site was used in 6 patients and metastatic sites (3 bone, 1 ureter and 1 lung metastasis) in 5 patients. The mean number of GA per case was 2.3 (1 to 4 GA/sample). The most frequent GA were in TP53 (36.4%), RB1 (27.3%), ATRX (18.2%). Potentially targetable GA were rare and included MTOR pathway inhibitors for GA in TSC2 and PTEN (1 case each) and PIK3CA inhibitors (GA in PIK3CA in 1 case). There were 3 cases with non “targetable” gene rearrangements including a STAT6- NAB2 fusion in a stromal sarcoma (possible solitary fibrous tumor of prostate), a BCOR- MAML3 fusion and a TMPRSS2- ERG fusion in a stromal sarcoma with an adjacent focus of Gleason 6 prostate adenocarcinoma. 1 case featured a predicted germline mutation of the FLCN gene. All cases were microsatellite stable. TMB ranged from 0 to 9.8 mutations/Mb with median of 3.4 mutations/Mb with no cases at ≥10 mutations/Mb. gLOH scores were low, ranging from 0% to 10.9% (median 2.6%). Genomic ancestry was European in 7 patients and Admixed American in 4 others. There were no specific genomic signatures identified. PD-L1 was tested in 4 cases (all negative). Conclusions: Prostate sarcoma is an exceedingly rare primary cancer of the prostate with limited opportunities for targeted therapy or immunotherapy strategies. These tumors do not appear to be driven by “targetable” gene fusions and individual “targetable” mutations are uncommon. [Table: see text]