2022-RA-1605-ESGO Precision-predicting risk of endometrial cancer in asymptomatic women. (20th October 2022)

• Record Type: Journal Article
• Authors: Kitson, Sarah Joanne; Crosbie, Emma; Gareth Evans, D; Lophatananon, Artitaya; Muir, Kenneth; Ashcroft, Darren; Kontopantelis, Evangelos; Martin, Glen
• Is Part Of: International journal of gynecological cancer. Volume 32(2022)Supplement 2
• Page Start: A388
• Page End: A389
• Publication Date: 2022-10-20
• Subjects: Generative organs, Female -- Cancer -- Periodicals; 616.99465
• DOI: 10.1136/ijgc-2022-ESGO.833
• Languages: English
• ISSNs: 1048-891X

Abstract

Abstract : Introduction/Background: Global endometrial cancer (EC) cases continue to increase, placing a significant health and financial burden on individuals and healthcare services. Effective primary disease prevention strategies are urgently required but remain under-researched. Identifying high-risk women for intervention would ensure therapies are targeted at those most likely to benefit. This study aimed to develop a well calibrated EC risk prediction model based on routinely collected data and to validate it in an independent cohort. Methodology: Data from the UK Biobank, comprising 222, 031 females ages 45–60 years and 902 incident EC cases, were used to build a flexible parametric survival model using EC risk factors identified through a systematic review of the literature. Model fit was improved with variable transformation and stepwise backward selection. Missing data were dealt with using multiple imputation and bootstrapping (100-fold) was applied for internal validation. Model calibration was assessed using flexible calibration plots and discrimination through calculation of the C-statistic. The model is being externally validated in the Clinical Practice Research Datalink, using data from 3, 094, 371 women, of whom 20, 882 have developed EC. Results: Age, body mass index, waist circumference, age at menarche, age at last birth, late menopause (≥55 years), current hormone replacement therapy or tamoxifen use, prolonged oral contraceptive pill use (≥5 years), type 2 diabetes, smoking and family history of bowel cancer were incorporated into the model. Based on these variables, the model had an adjusted C-static of 0.75 and was well calibrated, with a calibration slope of 0.97 after internal validation. Conclusion: Our model, using easily measurable anthropometric, lifestyle and reproductive variables alongside personal and family medical history, accurately identifies women at high-risk of EC. External validation will determine whether it can be used to determine eligibility for primary EC prevention trials and reduce the size and costs associated with such studies.