Colorectal cancer (CRC), the second most commonly diagnosed cancer, is a biologically heterogeneous disease. Studies examining the molecular characteristics of colorectal tumors indicate that there are multiple subtypes of this disease that develop through the activation of different neoplastic pathways. Next-generation DNA sequencing has provided further detail into the heterogeneity of colorectal cancer by identifying mutated genes in colorectal tumors, including well-known genes such as APC, SMAD4, and PIK3CA, as well as genes that have been less well described. Having detailed tumor mutation data allows us to better define more homogeneous subtypes by mutated genes and pathways, which may be informative in terms of a better understanding of etiological and survival differences in colorectal cancer.
When we started this work, somatic mutations had not yet been comprehensively studied in sizable datasets with detailed germline genetic risk factors, and environmental/lifestyle risk factors, and survival outcomes data. The idea of sequencing tumor DNA from a large number of cases came from the Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO), which is an international consortium that initially investigated germline genetic risk factors for colorectal cancer and how known environmental/lifestyle risk factors modified these genetic effects. Over the past decade, we in collaboration with the Colon Cancer Family Registry (CCFR) brought together individual-level genetic and epidemiological data for over 130,000 participants from 70 studies across North America, Australia, Asia, and Europe. We expanded the research of our international team by weaving together tumor data with the germline genetic, epidemiological, and patient clinical outcomes data.
Our paper provides a comprehensive analysis of colorectal tumor profile data with respect to driver genes, impacted pathways, known tumor characteristics, and colorectal cancer survival. Using DNA targeted sequencing data, we identified several new genes significantly mutated in CRC. We found that, among hypermutated tumors, an increased mutation burden was associated with improved colorectal cancer-specific survival. Despite our large sample size, we observed that among the 205 tested genes, only TP53 mutations were associated with poorer colorectal cancer-specific survival, particularly in cases carrying TP53 mutations with loss of transcriptional activity. Lastly, we observed differences in the mutational frequency of several genes and pathways by tumor location, stage, and sex.
Our study findings lay the foundation for a better understanding of CRC subtypes and advancing strategies for CRC prevention, diagnosis, and treatment. Taking advantage of our harmonized and detailed phenotype data, we are continuing with in-depth analyses of the associations between germline genetics and environmental/lifestyle risk factors with the mutational profile of the tumors. The ability to do this work was and continues to be possible because of collaborative efforts to address important research questions that no single study would be able to undertake alone.