Unique genomic and neoepitope landscapes across tumors: a study across time, tissues, and space within a single Lynch Syndrome patient

To investigate the longitudinal mutational patterns arising in Lynch Syndrome associated tumors we interrogated the genomes of five different cancers that arose over a period of 10 years in a patient who underwent resection in the absence of chemotherapy and radiation for each cancer. These included a papillary transitional cell carcinoma (PTCC) in the renal pelvis, a duodenal carcinoma, two separate CRCs that arose 3 years apart, and multiple regions of a triple negative breast cancer (TNBC). In each case we flow sorted tumor fractions from archived formalin fixed paraffin embedded (FFPE) tissue and profiled the tumor genomes with whole genome copy number variant (CNV) arrays and whole exome sequencing. These data were then used to identify the pathogenic variant underlying the diagnosis of LS, and to compare and contrast the CNV, mutational and neoepitope patterns across these divergent tumors that arose over a 10 year period. These results provide a unique analysis of distinct MSI+ tumors arising in a single LS patient. We applied DNA content based flow sorting to isolate the nuclei from biopsies of five distinct tumors that arose in a patient with Lynch Syndrome over a 10 year period. We coupled this strategy with oligonucleotide array CGH (aCGH) and whole exome sequencing (WES), thereby obtaining high definition genomic profiles of from each tumor. The aCGH data was assessed with a series of QC metrics then analyzed using an aberration detection algorithm (ADM2). Reads from the WES BAM files were stripped using XYalign version 1.1.5 then mapped to the 1000 genomes version of GRCh38 using bwa-mem version 0.7.17. We applied EpitopeHunter to predict neoepitopes.