Angiosarcoma heterogeneity and potential therapeutic vulnerability to immune checkpoint blockade: insights from genomic sequencing

Fig. 1. Distribution of mutations identified in the cohort of 48 angiosarcoma samples (The Angiosarcoma Project—https://ascproject.org/). a Number of mutations per sample in the Angiosarcoma Project collection of 48 tumors (mean = 260 mutations/sample); the mean in the subgroup of face and scalp angiosarcoma is 925 mutations/sample (N = 11 samples). In comparison, the mean number of mutations per sample in The Cancer Genome Atlas (TCGA) pan-cancer cohort is 212 mutations/sample (N = 10,106 samples; p value = 0.019), in the sarcoma cohort is 74 mutations/sample (N = 235 samples; p value = 0.008), in the melanoma cohort is 742 mutations/sample (N = 441 samples; p value = 0.496), and in the Angiosarcoma Project non-face/non-scalp subgroup is 63 mutations/sample (N = 37 samples; p value = 0.007). b–d Classification of mutations observed in 48 angiosarcoma samples by physicochemical properties (nucleotide, mutation types, and amino acid changes). A total of 12,499 variants were analyzed. Hydrophobicity increase reflects greater immunogenicity, as described by [13–15]

Fig. 2. Mutational signature analysis of primary angiosarcoma (N = 48 samples, 12,499 single nucleotide variants analyzed). The figure shows that face and scalp angiosarcomas often have high or intermediate mutation burden and UV light signature whereas other angiosarcomas have mostly lower mutation burden and aging signatures. Each etiology was studied individually. The signature probability was calculated using the entire set of variants in each sample. The probability (see green spectrum) is given as a number between 0 and 1: 0 corresponds to the lowest probability that the causative event occurred in the tumor and 1 corresponds to the highest probability that the causative event occurred in the tumor. Aging (as demonstrated by spontaneous demethylation of CpG islands), ultra-violet (UV) light exposure, and defective mismatch repair (MMR) are the top-3 etiologies proposed. Abbreviations: APOBEC, apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like; Mb, megabase; MMR, mismatch repair; MSI, microsatellite instability; POLE, polymerase epsilon; POLH, polymerase eta; UV, ultra-violet