Abdominal fat distribution in endometrial cancer: from diagnosis to follow-up

Abstract

Background: The objective of this study is to quantify abdominal obesity markers from computed tomography (CT) scans at primary diagnosis and follow-up in a large endometrial cancer cohort, and to assess temporal change in obesity markers in relation to surgicopathological patient characteristics and outcome.

Methods: Total- (TAV), subcutaneous- (SAV), visceral (VAV) abdominal fat volumes, and visceral-to-total fat percentage (VAV%) were derived from CT scans acquired in an endometrial cancer patient cohort at primary diagnosis (n_primary=293). Temporal (delta, δ) changes in CT obesity markers from primary diagnosis to follow-up were assessed for all patients with a follow-up CT 13 (7, 19) [median (interquartile range)] months after diagnosis (n_follow-up=152/293 patients). The CT obesity markers were assessed in relation to clinicopathological features and progression-free survival (PFS) using Mann-Whitney U-test, and Cox hazard ratios (HRs), respectively.

Results: At primary diagnosis, VAV% was the only marker significantly associated with high-risk histology (median of 33% for endometrioid endometrial carcinoma (EEC) grade 1-2, 36% for EEC grade 3 and 36% for non-endometrioid EC, p = 0.003), myometrial invasion (MI) (median of 34% for MI < 50% vs. 35% for MI ≥ 50%, p = 0.03) and lymphovascular space invasion (LVSI) (median of 34% for no LVSI vs. 36% for LVSI, p = 0.009). High VAV% (≥ 35%) also predicted poor PFS both in univariable analysis (HR = 1.8, p = 0.02), and when stratified for surgicopathological FIGO stage (HR = 3.1, p = 0.03). At follow-up, median TAV, VAV, SAV, and VAV% were significantly lower than at primary diagnosis (p < 0.001 for all). Furthermore, patients with progression had larger reductions in visceral fat compartments (δVAV=-24%, δVAV% =-3%), than patients with no progression (δVAV=-17%, δVAV%=-2%, p ≤ 0.006 for both).

Conclusion: Visceral abdominal obesity (high VAV%) is associated with high-risk histologic features, myometrial invasion, and poor prognosis. Furthermore, high visceral fat loss during/following therapy is associated with disease progression.

Keywords: Adiposity; Computed tomography; Endometrial neoplasms; Intra-abdominal fat; Obesity.

Fig. 1

Left: Abdominal fat compartments are derived from computed tomography (CT) scans covering the level of the upper right diaphragm to the level of vertebra L5/S1 are segmented using the software iNtuition (TeraReon Inc San Mateo, USA) Middle: CT obesity markers derived at primary diagnosis for patient A (63 years, FIGO IIIB, EEC grade 2, BMI 40 kg/m², alive and well at follow-up) and patient B (86 years, FIGO II, NEEC, BMI 25 kg/m², recurrence detected 6 months after primary treatment) Right: Temporal (delta, δ) changes in CT obesity markers from primary treatment to follow-up for patient A and B BMI, body mass index, EEC, endometrioid endometrial carcinoma; NEEC, non-endometrioid endometrial carcinoma; FIGO, international federation of gynecology and obstetrics; SAV, subcutaneous abdominal fat volume; TAV, total abdominal fat volume; VAV, visceral abdominal fat volume; VAV%, visceral-to-total abdominal fat percentage; WC, waist circumference at L3/L4

Fig. 2

Kaplan-Meier curves depicting VAV% (A) and delta [δ]VAV% (C) grouped in quartiles in relation to progression-free survival (PSF). Patients with VAV%≥35% (B) and δVAV%≤-6% (D) have significantly reduced PFS compared to patients with VAV%<35%, and δVAV%>-6%, (p ≤ 0.02 for both)