Background: Rates of endometrial cancer incidence and mortality are rising as a result of increasing rate of obesity and an aging population. Diagnosed early, endometrial cancer has an excellent prognosis but those with advanced disease have a very poor outlook. Epidemiological risk factors fail to accurately stratify women and there is no standardised screening programme for endometrial cancer. A growing body of evidence suggests that genetic factors contribute to the risk. The aim of this project was to investigate the potential benefit of genomic tools for predicting the risk of endometrial cancer to enable targeted and personalised prevention. Methods: i) A systematic review of the literature was performed to identify a panel of robust single nucleotide polymorphisms (SNPs) contributing to endometrial cancer risk. ii) A genome-wide association study (GWAS) was conducted in a case-control study from the North West of England to identify and validate SNPs associated with endometrial cancer. iii) A polygenic risk score (PRS) was developed and refined in this cohort followed by validation in additional datasets. iv) The relationship between endometrial cancer risk and pathogenic BRCA variants was investigated by comparing the prevalence of cases identified in a BRCA carrier database to the general population. DNA sequencing was undertaken to examine the presence of any somatic pathogenic mutations. Results: 24 SNPs most likely to influence endometrial cancer risk were identified through the systematic review. Six risk regions of suggestive significance were identified in the Manchester GWAS. 72 externally curated SNPs were investigated where a significant association was confirmed for ten SNPs. A refined PRS consisting of 40 SNPs was developed in our independent cohort resulting in 62% discriminatory ability. The PRS was applied to two other datasets with low to moderate success. No evidence for an increased risk of endometrial cancer in BRCA pathogenic carriers was observed. No pathogenic BRCA somatic mutations were identified in serous subtype tumours. Conclusions: In this project, we report strong evidence in favour of SNPs influencing endometrial cancer risk and have independently validated the most robust SNPs. A PRS based on the most predictive SNPs was moderately capable of risk stratification at a level similar to published multivariable risk prediction models. If successfully validated, the PRS may be useful in improving the accuracy of risk prediction models in risk-based care. Risk-reduction measures and incorporation into risk prediction models are not warranted for carriers of pathogenic BRCA mutations.
- polygenic risk score
- endometrial cancer
- genomics
Utilising Genomics for Risk Prediction and Targeted Prevention of Endometrial Cancer
Bafligil, C. (Author). 31 Dec 2021
Student thesis: Phd