Characterization of colorectal cancer and identification of prognostic determinants

Hyde, Angela J. (2012) Characterization of colorectal cancer and identification of prognostic determinants. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF (Migrated (PDF/A Conversion) from original format: (application/pdf)) - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (116MB) [English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. (Original Version)

Abstract

Newfoundland and Labrador has the highest rate of colorectal cancer (CRC) in Canada, and its rate of familial CRC is among the highest worldwide. The goal of this PhD programme was to further characterize CRC in this province, to gain a better understanding of the heterogeneity of this disease, and to identify clinically relevant differences in the tumours that could be exploited to personalize and improve clinical care. -- The molecular genetic, protein expression, and histological heterogeneities of colorectal cancer were evaluated in three large projects. First, the molecular features related to mismatch repair proficiency were assessed and correlated with family history. In a population-based cohort of 296 patients younger than 70 years at diagnosis, 60% met high or intermediate risk of hereditary CRC according to family history criteria, while only 13% of the associated tumours had defects in the mismatch repair system compatible with Lynch Syndrome. Although Newfoundland has a very high proportion of familial colorectal cancer, Lynch Syndrome does not account for the majority of these cancers; there are likely other, novel hereditary factors contributing to the burden of disease. -- Next, the heterogeneity in protein expression was assessed, using tissue microarray technology and immunohistochemistry, in two studies. To determine if the expression of selected proteins can predict patient outcome, expression scores for 12 proteins were assessed together in unsupervised hierarchical clustering analysis using data from 280 tumours from population-based patients along with another 111 from unselected CRC patients diagnosed younger than age 75 years. This analysis identified three tumour subgroups, one ofwhich had statistically significant reduced disease-specific survival after adjustment for known prognosticfactors (adjusted hazard ratio 1.82). The conclusions were that expression of specific proteins studied may have prognostic utility, and that the statistical methods used were valuable for identifying prognostically relevant subgroups of CRC. -- In the same cohort, lamin A/C protein expression in tumour nuclei was assessed to determine whether it predicted prognosis. Stage III patients whose tumours expressed lamin A/C in >25% of cells were four times more likely to die of their disease when compared to stage III patients with lower levels of expression (adjusted hazard ratio 4.14). -- Finally, to determine whether features of tumour histology were good predictors of microsatellite instability, hemotoxylin and eosin-stained tumour sections were studied from a population-based group of 710 incident CRCs. A number of histological features were associated with microsatellite instability, and an algorithm (PREDICT) was developed to predict this status using multivariate modelling. In a validation cohort of 280 population-based incident CRCs, PREDICT had a sensitivity of 96.9% and a specificity of 76.6% for identifying microsatellite instable tumours. Microsatellite instability can be used to assist in the identification of Lynch Syndrome, and in both the prediction of patient outcome and response to standard adjuvant chemotherapy. This algorithm has been adopted into clinical practice in parts of Ontario, and as part of a Community Genetics initiative in Newfoundland and Labrador.

Actions (login required)

View Item