A large randomized phase III assessment of the role of irinotecan added to fluorouracil/leucovorin as adjuvant treatment for colon cancer. The results presented herein explore the relationship between CNA, mRNA and outcome, and contribute to a comprehensive molecular overview of stage-II/III colon cancer, which is paramount for refining patient classification and effective treatment. The most prominent novel amplicons identified in this study include 12 p13.33 and multiple loci on 20 q. The 12 p13.33 amplicon encodes the intriguing candidate WNK1, a member of the WNK family of serine/threonine kinases which affect MAPK signaling and a variety of cancer hallmarks including cell cycle progression, evasion of apoptosis, invasion and metastasis. The complex pattern of gains and amplification on chromosome 20 q suggest multiple oncogenic drivers on this chromosome arm, consistent with observations in breast tumors and other cancer types. The 20 q13.12 amplicon, which was observed in multiple tumors and is the most significant GISTIC peak on 20 q, encodes 11 genes, none of which have been unequivocally described as oncogenic drivers in colon cancer. Nonetheless, the reported functions of some of these genes suggest that further investigation is warranted. For example, the transcription factor HNF4A controls epithelial cell polarity and promotes gut neoplasia in mice. WISP2 regulates the activity of the transforming growth factor aˆ signaling pathway and expression of genes associated with the epithelial-to-mesenchymal transition. The peak at 20 q13.31 encodes BMP7, a member of the TGFaˆ superfamily of proteins whose overexpression in colorectal cancer significantly correlates with markers of pathological aggressiveness such as liver BKM120 metastasis and is an independent prognostic factor of overall survival. Functional characterization of these and other candidate oncogenes in colon cancer cell culture, patient-derived xenografts, or genetically engineered mouse models will help elucidate potential functional implications. Pathway analysis presented previously provides not only a better understanding of the possible biological context of candidate CNA drivers but also help to infer other genes on the altered pathway for which therapeutic options may be available. On the other hand, survival analysis shows improved overall survival for the sample segment with chr20 q13.33 amplification. This association contrasts with findings of another group who reported amplification of 20 q13 is indicating worse overall survival in sporadic colorectal cancers. The exact basis for this discrepancy with our findings for is not clear, although the analyses of Aust et al. were on a substantially smaller cohort. Our analyses of associations between CNA and outcome in this set of stage II/III colon cancers revealed three loci that were significantly associated with overall survival or recurrence free survival. Deletion of the distal tip of chromosome 10 p was associated with poor OS and RFS, while an interstitial deletion of chromosome 19 p was associated with poor OS, and gain of 20 q was associated with significantly better OS in stage III tumors. While 10 p deletions, 19 p deletions, and 20 q gains have been previously reported in stage II/III colon cancers, none of these loci have been previously linked to outcome in these tumors. Conversely, we did not observe significant associations of outcome to previously reported CNAs such as deletion of 16 p13.2 in stage II/III colon cancer, or deletion of 5 q34 and gain of 13 q22.1 in stage II tumors. One potential explanation for these apparent discrepancies may relate to the limited power of the respective studies.