Genome and transcriptome profiles of matched ovarian primary and metastasis sites in three patients

Abstract

Metastasis to the peritoneum and lymph node is associated with poor survival in patients with epithelial ovarian cancer (EOC). To better understand EOC metastasis, we carried out exome sequencing, RNA sequencing and copy number variation (CNV) studies on ovary, peritoneum and lymph node tumors from three patients. To analyze single nucleotide polymorphisms (SNPs) from exome and RNA sequencing data, we use the ratio of reference allele reads to the total number of reads as a measure of a site's allele distribution. This approach is more appropriate for analyzing heterogeneous cancer samples than standard SNP calling methods. Analysis of significant SNP, CNV and gene expression shifts across tumor sites and patients reveals a large degree of heterogeneity between patients. The small number of shifting alleles and differentially expressed genes conserved across patients could be interesting genes for further study. In addition, using both the exome and RNA sequencing data, we identify specific alleles that are expressed at significantly higher and lower levels than expected from the genomic allele ratios. These differentially expressed alleles are very interesting candidates for further study as they may indicate tumor selection for particular alleles. Overall, our results indicate a remarkably heterogeneous landscape of ovarian cancer primary tumors and metastasis and suggest that a more patient-specific research and treatment approach is advisable to achieve better patient outcomes.