Detecting endometrial cancer from urine

Laura Costas

Bellvitge Biomedical Research Institute (IDIBELL), l’Hospitalet de Llobregat, Spain

Endometrial cancer is the fifth most common tumour in women in Spain and its incidence is increasing. Although it is usually diagnosed during the postmenopausal period, cases also occur in younger women.

Currently, the diagnosis of this cancer requires tests such as pelvic ultrasonography, which has low specificity, or endometrial biopsy, which is quite invasive. Moreover, in a large number of women it is either inconclusive or even impossible to perform. In such cases, other types of examinations must be performed under general anaesthesia. This makes diagnosis difficult and places a high economic burden on healthcare systems.

In previous studies, the team leading this project has demonstrated that it is possible to obtain genomic tumour information from urine samples. This information allows tumours to be classified molecularly with great precision into four subtypes, using a classification equivalent to that obtained from tissue biopsies. In this project, genomic analysis and computer vision techniques will be used to develop a tool for non-invasive detection of endometrial cancer. The ultimate goal is to enable early diagnosis of the disease, which will open the door to personalised treatments and the possibility of controlling and monitoring its evolution.

Leveraging Genomics and AI for Non-Invasive Detection of Endometrial Cancer, a Urine Sample Solution

Stage 1