Print, CLasham, AMehta, Sunali2013-09-252013http://hdl.handle.net/2292/20770Background Colorectal cancer (CRC) affects over one million patients each year worldwide. CRC patients with localised disease are treated with surgery, sometimes followed by adjuvant chemotherapy in an attempt to prevent future relapse. However, not all CRC patients benefit from this chemotherapy and it is difficult to identify which patients benefit based on currently available clinical and pathological information. Objective To gain a better understanding of CRC biology and use this to develop an improved prognostic tool to assist with clinical decision makings for patients diagnosed with CRC. Method In order to better predict CRC progression, we are combining RNA abundance signatures from tumours with clinical data from patients, and laboratory‐defined metagenes using microarray data from HT‐29 CRC cells treated with: (1) siRNAs to knockdown the levels of specific transcripts, (2) conditions that modify proliferation and differentiation, and (3) chemotherapeutic drugs. Results Combined clinicopathological and molecular pathway signatures could cluster CRC patients into groups significantly associated with survival outcomes. Putative classifiers that used these pathway signatures were able to predict CRC patient disease free survival (DFS) as effectively as previously‐developed molecular classifiers or classifiers that use clinicopathological information alone. However, no classifiers assessed could predict DFS effectively for stage II CRC patients. Conclusion Molecular pathway information inferred from tumour genomics, especially when combined with clinicopathological information, appears to be informative about individual tumours and predictive of clinical outcome for some patients. As cancer genomics expands, molecular pathway‐based signatures provide a promising method to understand the biology of individual tumours and thereby personalise patient care. We hope that after rigorous validation, these multi‐modal signatures can better predict CRC outcome in the clinic.Items in ResearchSpace are protected by copyright, with all rights reserved, unless otherwise indicated. Previously published items are made available in accordance with the copyright policy of the publisher.https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htmThesisCopyright: The Authorhttp://purl.org/eprint/accessRights/OpenAccessQ112200851