Description

For women in high-income countries, ovarian/fallopian tube and endometrial cancers are within the top four cancers in terms of incidence, death and healthcare expenditure. The deaths associated with these cancers are largely caused by Stage III/IV disease, for which cure rates have not changed in three decades, despite escalating costs of treatment. Attempts at early detection have been ineffective in reducing mortality, because the high-grade subtypes, which account for the majority of deaths, metastasize while the primary cancer is still small, has not caused symptoms, and is undetectable by imaging or blood tumour markers.

In recent years, the recognition that somatic mutations are early steps in carcinogenesis has led to a shift from tests such as imaging and non-specific blood tumour markers to technology that detects cancer-associated mutations in cervical, uterine, or blood samples. Several DNA-tagging technologies have been shown to be capable of identifying small amount of cancer DNA among thousands of normal cells, the proverbial needle in a haystack.

This investigation aims to develop and validate a high-sensitivity capture using a panel of genes involved in ovarian and endometrial carcinogenesis, low-pass whole genome sequencing, coupled with a machine-learning derived classifier for discriminating cancer from benign gynecologic disease prevalent in peri/post-menopausal women.