Quantitative signal intensity ratios to distinguish between subfascial lipoma and atypical lipomatous tumor/well-differentiated liposarcoma using short-tau inversion recovery (STIR) MRI.

To establish simple quantitative variables at short-tau inversion recovery (STIR) magnetic resonance imaging (MRI) to identify lipomas with high specificity in patients with indeterminate subfascial lipomatous tumors. The MRI examinations of 26 patients (14 men, 12 women; mean age 63±12.5 [SD] years; range: 40-84years) with histopathologically proven subfascial atypical lipomatous tumors/well-differentiated liposarcomas (ALT/WDLs) and those of 68 patients (32 men, 36 women; mean age, 56±13.5 [SD] years; range: 21-83years) with lipomas were retrospectively reviewed. Ratios derived from region of interest based signal intensity (SI) measurements of tumors and adjacent fat on STIR images were calculated and maximum tumor diameters were noted. Diagnostic parameter capabilities were assessed using ROC curve analysis. Interreader agreement was evaluated by calculation of intraclass correlation coefficients (ICC). Using a cut-off value of 1.18, STIR-SI ratios allowed discriminating between lipoma and ALT/WDL (AUC=0.88; P<0.001) yielding 93% specificity (95% CI: 77-99%) and 74% sensitivity (95% CI: 61-84%) for the diagnosis of lipoma. Interreader agreement was excellent (ICC=0.93). A significant difference in maximum tumor diameter was found between ALT/WDLs (mean: 18.1±6.0 [SD] cm; range: 5.6-33.1cm) and lipomas (mean: 9.7±5.0 [SD] cm; range: 2.9-29.1cm) (P<0.001). Using a cut-off of 11cm, maximum tumor diameter allowed discriminating between lipoma and ALT/WDLs with 92% specificity (95% CI: 75-99%) and 69% sensitivity (95% CI: 57-80%). The combination of a STIR-SI ratio<1.4 and maximum tumor diameter<11cm yielded 100% specificity (95% CI: 87-100%) and 65% sensitivity (95% CI: 54-77%) for the diagnosis of lipoma. The combination of STIR-SI ratio and maximum diameter allows discriminating between lipoma and ALT/WDL in initially indeterminate lipomatous tumors.

Copyright © 2020 Société française de radiologie. Published by Elsevier Masson SAS. All rights reserved.