(P 405) PRIMARY CARDIAC ANGIOSARCOMA- GENOMIC AND MOLECULAR LANDSCAPE

Objective: Primary cardiac angiosarcoma (CAS) is the most common histological type of primary cardiac sarcomas. The most common site of tumor is the right atrium of the heart. The median overall survival with surgical resection ranges from 8-9 months.
The genomic and molecular characteristics of primary CAS are poorly understood with only 3 families with cases of CAS that had exome or POT1 sequencing performed in one study on TP53 negative Li-Fraumeni like families and two CAS patients with exome profiling in the Angiosarcoma Count-Me-In project. Understanding the genomic landscape of primary CAS can be valuable in developing novel therapies and improving survival outcomes for patients.

Methods: Patients with a diagnosis of primary CAS seen at MD Anderson Cancer Center and Houston Methodist Research Institute between 1/1/2001 and 12/31/2015 were identified using an electronic database system. The tissue bank database was searched for availability of tumor tissue samples for these patients. Whole exome sequencing (WES) was performed on 20 primary CAS samples obtained from 13 unique patients (17 frozen, 3 formalin-fixed paraffin embedded). SNP array was performed on 15 samples and RNA sequencing was performed on 3 samples. The reads of WES were mapped to the hg19 reference genome using BWA-MEM, and MuTect was used to call somatic mutations. The DNAcopy R package was used to analyze copy number data derived from SNP array. RNA sequencing reads of the samples were mapped to the hg19 reference genome using the STAR aligner. For calculation of gene expression, raw count data of each gene were first obtained using HTSeq and normalized to TPM (Transcripts Per Million). Finally, we used a rank-based method, single-sample gene set enrichment scores (ssGSEA), to compare the gene expression data of two batches of our samples. Immune infiltration scores of samples across different cancer types were calculated from the gene expression data using ESTIMATE. Clinical data was abstracted manually from the electronic medical records.

Results: The cohort was predominantly male (77%) with a median age of diagnosis of 44 years. The right atrium was the most common site of tumor (n= 11 patients) and 8 patients had metastatic disease at the time of diagnosis (Table 1). The median overall survival was 21.8 months.
We identified 1706 nonsynonymous point mutations with an average of 58.6 mutations/ sample and a median tumor mutation burden of 1.45 mut/Mb. The most frequently mutated genes were KDR, POT1, CDKN2A, NOTCH1 and ZFHX4. 5 of the missense mutations in KDR (4 encoding p.Thr771Arg and 1 encoding p.Thr771Lys) are located in the transmembrane domain. Copy number amplification of KDR was detected in 1 patient. The mutations in POT1 (2 encoding p.Ala230Val, 2 encoding p.Ile220Ser and 1 encoding p.Tyr26Cys) occurred in the portion encoding the OB1 and OB2 region of the protein that makes direct contact with DNA. A germline variant in POT1 p.Arg117His was also detected in 1 patient. Co-occurrence of KDR and POT1 mutations were seen in 2 patients. 4 alterations were detected in NOTCH1 including 1 missense mutation (encoding p.Cys461Trp), 1 in frame deletion and copy number amplifications in 2 patients. 2 patients had a loss of heterozygosity in CDKN2A, while 1 patient had a missense mutation (encoding p.Ala21Val). 1 patient had a mutation in PLCG1 (encoding p.Arg707Gln) (Figure 1). We did not detect any focal copy number amplifications or deletions by SNP array analysis. Gains spanned chromosomes 7 and 8, with RECQL and SLC39A4 being the frequently amplified cancer genes on chromosome 8.
Primary CAS was not significantly different from AS of other sites on performing ssGSEA using RNAseq data from 3 samples. The ESTIMATE immune score for these samples were also high.

Conclusion: Our results highlight the presence of KDR mutations in primary CAS that have been reported predominantly in primary breast AS in previous studies. Primary CAS is characterized by POT1 mutations, which have previously been reported only in CAS in Li-Fraumeni like families. We are limited by the small number of samples with RNAseq data to derive inferences about the immune profile of CAS, which is also confounded by chemotherapy administered to the patients prior to surgical resection. Functional assays to understand the effect of mutations in the transmembrane domain of KDR are needed to determine if it is a potential therapeutic target.