MET overexpression as a hallmark of the epithelial-mesenchymal transition (EMT) phenotype in colorectal cancer.

Abstract

334 Background: Dysregulation of the proto-oncogene MET (mesenchymal-epithelial transition factor gene) has been implicated in tumorigenesis and correlates with worse survival and chemo/radio-resistance in colorectal cancer (CRC). EMT has been identified as a dominant molecular characteristic of a subset of CRC tumors and represents a key feature in the developing colorectal taxonomy. The purpose of this study was to compare protein expression of MET with protein/gene expression of EMT markers and other clinicopathological characteristics, and to evaluate its impact on overall survival (OS). Methods: We performed an exploratory analysis of 590 CRC samples using data from The Cancer Genome Atlas. Fisher-exact test and Pearson’s method was used to determine the relationship between MET protein expression, clinicopathological characteristics and EMT marker protein expression by reverse-phase protein array (RPPA) and EMT-associated gene expression by RNA-sequencing. Regression tree method was applied to find the best cutoff point for MET using patients with available survival data. Overall survival (OS) was estimated non-parametrically using Kaplan-Meier curve and log-rank test was used to evaluate hazard ratio. Results: MET expression by RPPA did not correlate with traditional clinicopathologic characteristics. MET was overexpressed in 17% of CRC tumors and was significantly associated with OS (HR 2.92; 95% CI: 1.45 - 5.92). Correlation analysis of MET levels with gene expression of EMT markers AXL, CDH1, FGFR1, SNAIL, TWIST1/2, VIM, SLUG, ZEB1/2, FN1 demonstrated that the highest quartile of MET protein expression was associated with a 1.5 fold increase in ZEB1 (p = 0.002), a 1.4 fold increase in AXL (p = 0.005) and ZEB2 (p = 0.008), and a 1.3 fold increase in VIM (p = 0.02). MET expression also correlated strongly with protein expressions of SNAIL (transcription factor for EMT) (r = 0.96) and ERCC1 (r = 0.83) (a marker for oxaliplatin chemo-resistance). Conclusions: Increased MET protein expression is seen in 17% of CRC tumors and strongly correlates with a molecular EMT phenotype and poor survival in patients with CRC. MET protein expression may be a surrogate biomarker for this unique subset of CRC.