Abstract IA18: Probing tumor heterogeneity with mouse models

Abstract

Abstract Although the majority of pancreatic cancers harbor mutations in a small set of cancer genes, individual tumors vary widely with respect to histology, gene expression patterns, and response to therapy. This inter-tumoral heterogeneity is likely due to mutations or epigenetic events affecting less commonly-mutated cancer genes, resulting in tumors that can be broadly grouped into three or four categories based upon transcriptional profiling. To determine the extent to which the KPC model captures the inter-tumoral heterogeneity of pancreatic cancer, we performed RNA-sequencing on isolated tumor cells from 11 murine pancreatic tumors. Using unsupervised clustering, tumors fell into two broad categories: cluster 1 tumors, which were poorly-differentiated and well-vascularized, and cluster 2 tumors, which were moderately- to well-differentiated and poorly-vascularized. Differences in gene expression could account for some of these differences; for example, cluster 1 tumor cells exhibited low levels of Shh expression, consistent with the paucity of stroma. Moreover, mechanisms of EMT differed between the two classes of tumors. Gene set variation analysis with whole transcriptome data from human PDA revealed a correlation between cluster 1 tumors and the squamous or quasi-mesenchymal subtype. Likewise, cluster 2 tumors were correlated with the classical, ADEX, and pancreatic progenitor subtypes of human PDA. These data suggest that the KPC mouse model, despite its foundation on Kras and p53 mutations alone, captures the heterogeneity of human PDA and may therefore be useful for understanding heterogeneous responses to therapy. Citation Format: Ben Z. Stanger.{Authors}. Probing tumor heterogeneity with mouse models. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2016 May 12-15; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2016;76(24 Suppl):Abstract nr IA18.