Exploring the Cancer Genome Atlas (TCGA) for the molecular profile of young onset colorectal cancers.

Abstract

3548 Background: Colorectal cancer (CRC) incidence and mortality has been declining, in part due to increased implementation of screening, but the incidence among patients under 50 (young onset, YO) is increasing at a rate of 2% per year. The cause of this increasing incidence remains poorly understood, but differences in mutation profiles can help understand pathogenesis, prognosis, and identify targets for therapy. Methods: Genomic and clinical data for 488 TCGA CRC patients was used to evaluate differences in genetic alterations between YO and patients over 50. Chi-squared tests were used to determine differences in somatic mutation frequency in critical pathways implicated in CRC: DNA MMR, P53, WNT, RAS-MAPK, PI3K/AKT/mTOR, and TGF-B pathways. For 85 of the patients, proteomic data via RPPA was also available and analyzed. Results: The average age of included patients was 66 (SD 12.8). 76 (12.2%) were under 50 at time of diagnosis. When comparing YO with those over 50, there were no differences in microsatellite instability, histologic type (adeno or mucinous), location (colon or rectal), tumor size, or metastasis. YO patients were more likely to have nodal involvement (p = 0.007) and higher histological grade (p = 0.022). YO patients were more likely to have mutations in the MMR pathway (43% vs 23%, p = 0.002) and the PI3K/AKT/mTOR pathway (70% vs 54%, p = 0.024). Specifically, YO were more likely to have mutations in MSH2 (7% vs 1%, p = 0.001), MSH6 (24% vs 7%, p = 0.000); ATM (46% vs 30%, p = 0.015); FZD10 (7% vs 2%, p = 0.007); ERBB2 (15% vs 7%, p = 0.027); PIK3R1 (20% vs 9%, p = 0.014), PTEN (61% vs 35%, p = 0.000), and TGFBR2 (13% vs 4%, p = 0.004). When looking at proteomic data, YO were more likely to have decreased expression of MSH2 (p = 0.003) and MSH6 (p = 0.005). Conclusions: Patients with YO CRC are more likely to have somatic mutations in genes involved in the MMR pathway and the PI3K/AKT/mTOR pathway. Specifically, in MSH2, MSH6, ATM, FZD10, ERBB2, PIK3R1, PTEN and TGFBR2. When including proteomic data, significant differences were only seen in expression of MSH2/6. Some of these genes (e.g. ERBB2/HER2) are targets for existing therapies, and others are being actively investigated as potential therapeutic targets. Establishing differences in tumor genetic profiles is a first step towards understanding the increase in YO CRC, and simultaneously identifies targets for therapy. However, because of post-transcriptional changes (e.g. RNAi, methylation), genetic profiling alone cannot always reliably establish differences in protein expression, and thus therapy targets.