580 Background: While colorectal cancer (CRC) has classically been categorized on the basis of oncogenic mutations such as KRAS and BRAF, proteomic analyses directly elucidate the functional state of the cancer cell’s protein signaling, as recently described in a pan-cancer cohort and with mass-spectroscopy in a small CRC cohort. We performed an antibody-based proteomic analysis (reverse-phase protein array; RPPA) of a large cohort at MD Anderson (MDACC) and The Cancer Genome Atlas (TCGA) to determine patterns of protein expression in CRC. Methods: 725 archived CRC tumor samples (263 MDACC discovery set, and 462 TCGA validation set) underwent protein extraction and RPPA at MDACC to determine levels of 127 proteins. With unsupervised hierarchical clustering, samples dichotomized with distinct patterns of protein expression. The proteins with highest discriminatory utility were identified by LIMMA in the discovery set and confirmed in the validation set. Clinical variables and DNA sequencing results were available for correlation. Results: Among the top 30 discriminant proteins for the dichotomized groups in each dataset, 18 were common to both and tended to correlate with each other. One group was notable for high EMT (high fibronectin and collagen VI, low E-cadherin), while the other group was notable for high Akt/TSC/MTOR (high AKT, MTOR, Tuberin), and high RTK pathway components (high BRAF, HER2, HER3). This latter group also was notable for elevated beta-catenin and low CHK1, implicating differential activation of Wnt and cell cycle pathways, and intriguingly had elevated phospho-AMPK and phospho-NFkB. In the MDACC cohort, this latter group was more likely to have mucinous histology (p=0.009 by Fisher’s exact test) and lack lymphovascular invasion (p=0.026). When both TCGA and MDACC cohorts were examined, there was no significant difference in microsatellite instability, PIK3CA, KRAS, or BRAF mutations between the two proteomic groups. Conclusions: CRCs appear to be classifiable into distinct subsets by proteomic features. These findings reflect distinct differences in cellular signaling that are independent of common oncogenic driver mutations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.