LAUSR

Objectives: Visceral adiposity has been established as a predictor of outcomes in various cancers; however, its role in predicting outcomes in advanced stage endometrial cancer patients is unknown. We aimed to determine the association of radiographic measurements of visceral fat with clinical outcomes in patients with endometrial cancer. Methods: A retrospective review of stage III-IV endometrial cancer patients who underwent surgery after preoperative computed tomography (CT) scans between 2005 and 2014 was performed. Follow-up was continued until 10/2020. Visceral fat area (VFA) and subcutaneous fat area (SFA) were assessed in CT scans by a board certified radiologist. Total fat area (TFA) is the sum of VFA and SFA. Exploratory analysis was performed to establish the optimal cut-off values for VFA to SFA ratios to identify patients with poor prognostic body composition. Survival rates were analyzed using Kaplan-Meier analysis and log-rank tests. Results: Ninety-two patients were included. Forty-two (51%) patients had a low VFA/SFA ratio ( 0.45). There were no significant differences in demographics or clinical characteristics including grade and stage and adjuvant therapy between the groups. The mean age was 62.7 years and the mean BMI was 33 kg/m2. The majority of patients had endometrioid histology (57, 68.7%) and grade 1 (25, 30.1%) and grade 3 (19, 22.9%) disease. Twenty four (28.9%) patients had mixed history and two (2.4%) had high grade serous histology. The mean VFA, SFA, and TFA were 176.3 cm2, 379.3 cm2, and 555.3 cm2 respectively. Compared to patients with low VFA/SFA ratios, patients with high VFA/SFA ratios had a shorter recurrence-free survival (median 29.6 vs 32.3 months, P=0.01) and shorter overall survival (median 56 vs 93.7 months, P=0.03). Conclusions: Visceral fat measurements are predictive of outcomes in patients with endometrial cancer. Specifically, VFA:SFA ratios are predictive of overall survival in this cohort. Future studies should be pursued to validate this tool prospectively in order to determine the independent prognostic significance.