Abstract C107: Use of spatial transcriptomics to identify molecular features associated with African American heritage in pancreatic cancer

Yuesong Wu,Aoqi Xie,Ian Loveless,Madison George,Kendyll Gartrelle,Julie Clark,Daniel Salas-Escabillas,Rupen Shah,David Kwon,Ralph Francescone,Débora B. Vendramini Costa,Howard Crawford,Brian Theisen,Yuehua Cui,Ling Huang,Nina G. Steele
DOI: https://doi.org/10.1158/1538-7445.panca2023-c107
IF: 11.2
2024-01-16
Cancer Research
Abstract:Abstract Pancreatic cancer is the third leading cause of cancer-related deaths in the US, with a 5-year survival rate of 12%. Black or African American patients (BAA) have a 20% higher incidence of pancreatic cancer, and recent studies suggest that there are biological factors that may influence this disparity, in addition to complex socioeconomic factors. Pancreatic tumors are highly enriched with stroma and exhibit significant inter- and intra-patient morphological and molecular heterogeneity. To improve the care for all patients, we need to better understand the disease mechanisms by comprehensively mapping different spatial locations within individual patients’ tumors and determining whether there are common patterns of pathogenesis across diverse patient populations. In this study, we utilized the 10X Visium spatial transcriptomics platform on eight treatment-naïve primary pancreatic tumors. We tested multiple published bioinformatic packages, including SpatialDE, SPARK-X, Spatial PCA, and Cottrazm, and optimized a multi-sample integrated analysis pipeline that allows us to automatically segment tumor lesions, infer major cell types in regions of interest on tissues, detect spatially variant genes, and assess associations between tissue morphological phenotypes and transcriptomic features. Importantly, our study cohort includes four White patients and four BAA patients. Through the comparison of spatial transcriptomes between White and BAA patients, we identified several genes displaying differential expression in tumor lesions of distinct racial groups, including Trefoil Factor 1 (TFF1) and CYP3A5. These genes present potential candidates for further investigation to elucidate the biological mechanisms underlying racial disparities in pancreatic cancer. Citation Format: Yuesong Wu, Aoqi Xie, Ian Loveless, Madison George, Kendyll Gartrelle, Julie Clark, Daniel Salas-Escabillas, Rupen Shah, David Kwon, Ralph Francescone, Débora B. Vendramini Costa, Howard Crawford, Brian Theisen, Yuehua Cui, Ling Huang, Nina G. Steele. Use of spatial transcriptomics to identify molecular features associated with African American heritage in pancreatic cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr C107.
oncology
What problem does this paper attempt to address?