2022-RA-448-ESGO Venous Thromboembolic Disease in Ovarian Cancer: Incidence, Impact on Overall Survival and Development of a Predictive Score

Introduction/Background Venous thromboembolism disease (VTE) is a major cause of morbidity and mortality in patients managed for ovarian cancer. The first objective of this study is to assess the incidence of thromboembolic events and the impact of VTE occurrence in ovarian cancer patients on overall survival (OS). The secondary objective is to identify predictive factors for VTE to establish a predictive nomogram at the time of ovarian cancer diagnosis. Methodology A retrospective study from a prospective cohort of patients managed for ovarian cancer in the gynecologic oncologic surgery department of the Georges Pompidou European Hospital between January 2003 and December 2020 was performed. A survival analysis by Kaplan Meyer and Cox model and a multivariate logistic regression analysis were used. A nomogram to predict the risk of VTE at the time of ovarian cancer diagnosis was created. Results Among the 615 patients included, the incidence of VTE was 17.7%. Of 109 VTEs identified, 77 (70.9%) were observed at the time of ovarian cancer diagnosis and 49.5% of patients were asymptomatic. Patients with VTE had a significantly shorter OS compared to patients without thromboembolic events (HR = 1.62, 95% CI 1.06 – 2.49, p = 0.03). Peritoneal carcinosis index, body mass index, moderate to severe renal failure, weight loss, ASA score and histological type were associated with the occurrence of VTE at the diagnosis of ovarian cancer. VTE predictive nomogram created in this population had good internal agreement (AUC = 0.81, CI95% 0.73 – 0.89). Conclusion The incidence of VTE is high during ovarian cancer management, with many asymptomatic events. The impact of VTE is unfavorable on OS. The use of a nomogram could allow an earlier screening and thus improve the management and prognosis of patients. These results raise the question of systematic screening and its modalities in this population.