Angiogenic gene expression in primary neuroendocrine tumors and their metastases.

Abstract

200 Background: Neuroendocrine tumors (NETs) are neoplasms arising from the cells of the nervous, endocrine and hormonal systems. They most commonly originate in the small bowel (SB) and frequently metastasize to the lymph nodes (LNs) and/or liver (LV). Current treatment of metastatic NETs involves a variety of approaches including antiangiogenic therapies. Our group demonstrated that there are significant histologic and functional differences between the primary NETs and their nodal or organ metastases. We hypothesized that sampling of multiple tumor sites within the same individual will reveal differential expression profiles of angiogenesis-related genes. Methods: Tissue-matched normal and tumor tissue samples were obtained from patients with well differentiated NETs who underwent simultaneous removal of their primary tumor, nodal, and organ metastasis. High quality RNA was extracted from each tumor site using TRIzol and RNeasy Mini kit. Gene expression of 28 well-documented angiogenesis-related genes was assessed using Custom Quantitative RT-PCR array. These gene expression trends were validated by Illumina microarray and TaqMan analysis. Immunohistochemistry (IHC) staining was performed using Avidin-Biotin-Peroxidase complex, with the markers: SSTR2 and FGFR3. Results: Normal SB, LN, and LV gene expression of 28 genes was compared to that of the tumor sample at each tumor site [4-fold change, p ≤ 0.01]. A consistent up-regulation of SSTR2 and SSTR1 was seen in 18/24 (75%) samples. Up-regulation of FGFR3 and SSTR5 was observed in 13/24 (50%) of tumors. TGFA and IGF were consistently down-regulated in 12/24 (50%) and 10/24 (42%) of tumor samples, respectively. Six genes expression trends were validated by TaqMan analysis and Illumina microarray analysis. IHC staining revealed higher SSTR2 and FGFR3 protein expression in all three tumor sites compared to the control. Conclusions: Expression of angiogenesis-related genes varies between the primary tumor (SB) and metastatic sites (LV, LN) within the same individual. We found a positive correlation between SSTR2 and FGFR3 gene and protein expression levels in all three tumor sites.