Abstract 2189: Molecular similarity of Barrett’s-associated and gastro-esophageal junction adenocarcinomas

Abstract

Abstract The incidence of Barrett’s-derived esophageal adenocarcinomas (EAC) and gastroesophageal junction adenocarcinomas (GEJAC) are rapidly increasing in western countries. Our recent efforts identified cell surface peptides that target Barrett’s(BE)-associated EAC for early detection by fluorescent endoscopy however these markers only identify a subset of tumors due to molecular heterogeneity. We also observe that >30% of esophagectomy patients present with GEJAC with no evidence of Barrett's metaplasia. Although the etiology of both cancers appears similar, a detailed molecular comparison of EAC and GEJAC is lacking. Our goal was to compare molecular profiles for these two cancer types to develop a panel of either universal, or cancer-type specific cell surface markers to facilitate early detection. We utilized two pre-existing datasets that performed whole-exome sequencing and copy number variation (CNV) analysis in a combined cohort of GEJAC and EACs. When assessing the rate of non-silent mutations, gene specific mutations, and CNV changes, no significance differences were observed between GEJAC and EAC. For a comparative characterization of the expression profiles of EAC and GEJAC we utilized Affymetrix mRNA analysis of 123 GEJAC and EAC. Normal gastric and esophageal squamous tissues were used as controls. Principal component analysis of the tumors showed little separation between the two tumor classes (PC#1 and #3, p-value = .03 for t-test of EAC vs. GEJAC). When performing ANOVA analysis of GEJAC and EAC we observed 305 genes with p-value1.3. Unsupervised hierarchical clustering of the 123 tumors revealed two main clusters (Cluster 1 n = 58, Cluster 2 n = 64) with no enrichment of tumor type in either cluster (Clus. 1 vs. Clus. 2, GEJAC and EAC p-value = 0.067). These observations suggest that there few differences at the transcription level between GEJAC vs. EAC. Multi-platform (mutation, CNV and mRNA expression) molecular characterization of GEJAC and EAC tumors revealed no significant differences between tumor types suggesting that shared molecular processes are likely involved in the development of these tumors. We then used the 123 GEJAC/EAC expression data, along with our previous cohort of 46 samples representing BE progression to EAC, in order to identify potential cell surface coding genes specifically overexpressed within the combined cancer cohort. We identified >20 cell surface coding genes that were over-expressed (FC>2, p-value<0.01) in both GEJAC and EAC, when compared to normal tissue and BE mucosa. Three of the candidates were chosen for qPCR validation and also protein expression using commercial antibodies and an esophageal tumor tissue microarray. Developing a comprehensive panel of cell surface markers able to discriminate the majority of EAC and GEJAC will be an essential step towards integration of image-based technology into existing early detection screening programs. Supported NCI U54 CA163059 Citation Format: Daysha Ferrer-Torres, Derek Nancarrow, Rork Kuick, Ernest Nadal, Thomas D. Wang, Andrew Chang, Jules Lin, Rishindra M. Reddy, David G Beer. Molecular similarity of Barrett’s-associated and gastro-esophageal junction adenocarcinomas. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 2189. doi:10.1158/1538-7445.AM2015-2189