Introduction/Background Endometrial carcinoma (EC) is the most common gynaecological malignancy in the developed world. Currently, no valid non-invasive diagnostic or prognostic methods exist, making diagnosis and treatment rely on histopathological and surgical findings. The clinical study ’Biomarkers for Diagnosis and Prognosis of Endometrial Carcinoma’ (BioEndoCar; NCT03553589) addresses this issue.

Methodology A prospective observational case-control study was conducted at six medical centres across Europe. Plasma samples from women with diagnosed EC and controls were examined using non-targeted/targeted metabolomic and semi-quantitative immune-based proteomic approaches. The blood metabolomics (>850 metabolites) and proteomics (>900 proteins) data together with clinical and epidemiological data, were analysed using advanced artificial intelligence (AI) and machine learning (ML) methods to develop new diagnostic/prognostic models for early EC diagnosis and identifying patients with low/high risk for cancer progression and recurrence.

Results BioEndoCar has recruited over 440 patients, with strict standard operating procedures for sample collection, processing, and storage. The diagnostic/prognostic models based on all data developed using AI/ML methods showed promising characteristics with a repeated k-fold cross-validation AUC > 0.8. The developed models will undergo further validation using both statistical (AI/ML) approaches to confirm which subset of proteomic and metabolomic data could serve as diagnostic and prognostic biomarkers in endometrial cancer.

Conclusion The BioEndoCar study has completed the initial phase of identifying and validating diagnostic/prognostic models for early EC diagnosis and identifying patients with low/high risk for cancer progression and recurrence using artificial intelligence and machine learning methods. If validated, the models including a subset of proteomic and metabolomic data could serve as a foundation for developing valuable non-invasive tools for the diagnosis and prognosis of EC.

Disclosures The BioEndoCar consortium was a EU-H2020 funded Transcan2 ERA-Net project (2018–2021), with involvement of the national funding agencies: Ministry of Education, Science and Sports Slovenia; Dutch Cancer Society, The Netherlands; Federal Ministry for Education and Research, Germany; Estonian Research Council, Estonia and National Centre for Research and Development, Poland.