Quantifying Adipose Tissue Compartments

Staging and pre-treatment unenhanced CT images were retrospectively analyzed by a single trained operator (EB), supervised by a single radiologist (GB), both blinded to clinical data and outcomes, using the OSIRIX-Lite software V5.0 (Pixmeo, Sarl, Switzerland). A single slice at the level of the third lumbar vertebrae (L3), with both transverse processes visible, was selected. Abdominal fat compartments were selected after applying Hounsfield Unit (HU) thresholds ranging from −190 to −30 [24 (link)]. The total adipose tissue area (TAT), subcutaneous adipose tissue area (SAT), and visceral adipose tissue area (VAT) at the level of L3 were obtained through autosegmentation, and manual contour correction when necessary [25 (link)]; intra-organ fat (e.g., intrarenal adipose tissue) was manually excluded. The mean radiodensities were collected from the same regions of interest used for adipose tissue compartment areas (total adipose tissue density, TATd; subcutaneous adipose tissue density, SATd; visceral adipose tissue density, VATd) (Supplementary Figure S1). The VAT/SAT ratio was calculated as an indicator of adipose tissue distribution in different body fat compartments.