Abstract C057: Tumor molecular features associated with estimated genetic ancestry of patients with pancreatic ductal adenocarcinoma

Abstract

Abstract Pancreatic cancer is the third leading cause of cancer-related death in the US, and the incidence rates of this disease have continued to rise over the past decade. Epidemiological analyses have revealed that different racial groups are affected by this cancer disproportionately, with the Black or African American population being the most affected. Currently, it is unclear to what extent biological pathways contribute to race-associated disparities in pancreatic cancer development. Research efforts to address this topic require analyses of patient-derived clinical data and face many challenges, such as low enrollment rates of minority patients and incomplete records of self-reported race informationIn this project, we utilized DNA and RNA sequencing data generated by the Know Your Tumor Program from the Pancreatic Cancer Action Network (PanCAN) to: 1) classify each patient into their most similar genetically assessed population (GAP) based on continent-level ancestry using targeted region sequencing of tumor genomes and 2) identify gene expression differences between GAPs. Our sample size consisted of 409 individuals with confirmed pancreatic cancer and targeted tumor DNA sequencing, 226 of whom also had tumor RNA-Seq data. Reference GAPs were defined as the five 1000 Genome Project (1KG) super-populations, and each PanCAN subject was assigned to one of these reference GAPs based on their top three genome-wide principal components and k-nearest neighbors clustering, where k=5 1KG reference subjects. Gene expression differences between pairs of GAPs was performed using a negative binomial model implemented in DESeq2, and differential expression of splice variants was performed using suppa2. Both gene expression analyses accounted for pathologic stage of disease and adjusted for multiple comparisons using the false discovery rate (FDR). For the 409 PanCAN participants, the GAP assignments to the five 1KG super-populations were as follows: Africans/African Americans (AFR, n=19), Admixed Hispanic/Latino Americans (AMR, n=22), East Asians (EAS, n=17), South Asians (SAS, n=7), and Europeans/European Americans (EUR, n=344). Among the 184 participants (45%) that provided self-reported race information, GAP assignment was in high agreement (99%)with their self-reported race. Due to sample size, gene expression analyses were restricted to the AFR, EAS, and EUR GAPs. At an FDR<0.05, 605, 128, and 695 differentially expressed genes were identified for the AFR/EUR, EAS/EUR, and AFR/EAS pairwise comparisons, respectively. In particular, differentially expressed genes between AFR and EUR GAPs were significantly enriched for multiple immune related pathways that regulate B cell functions. For this GAP comparison, RNA splicing patterns of multiple genes, such as CDK4, were also differentially expressed. The RNA transcript variants differentially expressed between these groups can be candidates for future studies to understand their roles in driving biological changes that contribute to racial disparities in pancreatic cancer. Citation Format: Mohammed H M Bharmal, Ian Loveless, Howard Crawford, Nina Steele, Albert Levin, Ling Huang. Tumor molecular features associated with estimated genetic ancestry of patients with pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research; 2024 Sep 15-18; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl_2):Abstract nr C057.