Head and neck cancer long term survival has only experienced marginal gains over the past 3 decades. Further, there is inadequate understanding of the biology of carcinogenesis and recurrence, as well as its relation to the microenvironment. Deeper understanding of these areas would provide improved molecular target identification. In pursuit of this goal, a small clinical trial collected lesion and adjacent normal-appearing mucosa for subsequent RNA-seq analysis. Patients were grouped by post-surgical pathology as either precancer (hyperplasia - severe dysplasia) or cancer (carcinoma in situ - early stage invasive cancer). Following identification of differentially expressed (DE) genes, DE genesets were submitted for Ingenuity Pathway Analysis (IPA). Hierarchical clustering illustrates distinct separation between lesion and perilesional normal mucosa of the top 100 DE genes. Among the top 25 dysregulated pathways, 50% were associated with creation of fibrotic tumor microenvironment (TME), 20% were related to changes in immune populations inhabiting the TME and 10% devoted to metabolism changes. Subgroup analysis, (precancer vs. cancer), revealed dysregulation of metabolism (~50%) predominating in precancer. Metabolism remained an important dysregulation at 30% of the top 25 pathways in cancer. Protein network analysis, (Metascape on-line tool), confirmed IPA results, illustrating an extensive, previously undescribed, interconnectedness of fibrosis with shifts in fatty acid metabolism from oxidative to gluconeogenesis, providing a foundation for choosing targets amenable to cancer prevention. Several notable pathways are likely contributed to by inflammatory and other cells in the milieu, not precancer cells themselves. So, we dove deeper, using EpIC (Epitope Immunogenicity Characterization) algorithm to assess relative percentages of non-tumor cells based on 20-count gene signatures. Gene expression favored a profile of significantly increasing cancer-associated fibroblasts, decreasing CD-8 killer T cells, and increasing vascular endothelial cells during progression, with macrophage content slightly increasing in cancer specimens. These findings suggest interaction between immunoinflammatory milieu and precancerous cells promoting malignancy. Several high yield target pathways are related to published mechanisms of action for drugs of high interest to our cancer prevention program (pioglitazone/metformin). Further, we confirmed protein network analysis in an additional oral carcinoma dataset from Conway et. al, (Oncotarget 2015; doi: 10.18632/oncotarget.5529). Abstract 2590
Smart Citations