부산대학교 도서관

Observational studies suggest incidental use of aspirin or statins after colorectal cancer diagnosis may improve survival but the epidemiological data to support the repurposing of these commonly prescribed cardiovascular medications lacks consistency. The role of biomarkers to predict the anti-cancer effect of these medications in colorectal cancer has also been incompletely investigated. For these reasons the aim of this thesis was to examine the association between aspirin and statin use and colorectal cancer prognosis using methods that incorporated systematic review, pharmacoepidemiological and molecular pathological epidemiology techniques. Firstly, neither low-dose aspirin nor statin use after diagnosis were associated with improved colorectal cancer-specific survival in a Scottish colorectal cancer cohort containing 8,391 patients with stage l-lll disease. Similarly, in an updated systematic review and meta-analysis there was.consistent evidence that statin use before but not after diagnosis was associated with improved cancer-specific survival. On this basis there is currently insufficient evidence to recommend clinical trials assessing adjuvant statin therapy in the treatment of surgically resected colorectal cancer. Within the Northern Ireland Colon Cancer Tissue and Data Resource (n=680 patients with stage II and III colon cancer) low-dose aspirin use was associated with better survival and the association was more pronounced in tumours with higher levels of PTGS2 expression. In contrast, no significant interaction was observed among patients with PIK3CA mutant tumours. This raises further questions for trials assessing adjuvant aspirin solely in PIK3CA mutant tumours. It also suggests that planned molecular subgroup analyses in trials that do not restrict recruitment on the basis of tumour molecular profile may be more appropriate. Finally, statin use was not associated with improved survival this cohort. However the differences in association between statin use and colon cancer survival by tumour KRAS mutation status, tumour HMGCR expression and allelic variation in rs12654264 are worthy of further investigation.