Evaluating shear wave elastography for differentiating lipomas from low to intermediate grade liposarcomas: is it reproducible and reliable?

Abstract

To evaluate if imaging findings and shear wave elastography (SWE) can differentiate between lipoma and low-to-intermediate grade liposarcoma and to assess for SWE reproducibility across different manufacturer machines. Consecutive patients referred for lipomatous soft tissue mass biopsy prospectively underwent B-mode ultrasound (US), SWE on two machines (LOGIQ-E9 and SSI-Aixplorer), and MRI. Three musculoskeletal radiologists independently scored images and gave an assessment of malignancy on a four-point scale. Inter-reader agreement was evaluated with kappa statistic (US/US + MRI) and ICC (SSI-Aixplorer velocity/stiffness; cross-machine comparisons). Diagnostic utility of shear wave velocity (SWV) was evaluated using area under receiver operating characteristic (AUROC) curve against pathology. Among 269 patients (mean age 58.8 ± 15.7 years, 34.6% (93/269) female), 22% (59/269) of lesions were malignant. Inter-reader agreement on SSI-Aixplorer was good for velocity and stiffness with ICC 0.87 (95% CI 0.75-0.94) and 0.93 (95% CI 0.87-0.97) respectively. Agreement between machines was moderate for velocity and stiffness with ICC 0.62 (95% CI 0.53-0.69) and 0.66 (95% CI 0.58-0.73) respectively. SWV poorly predicted malignancy (AUROC 0.57, 95% CI 0.49-0.65), and failed to differentiate benign from malignant lipomatous tumours at previously defined 2.02 m/s threshold, misclassifying 9.5% (4/42) as benign. SWV failed to further stratify radiologist grouped benign/probably benign lesions on US (AUROC 0.59, 95% CI 0.49-0.68). Using MRI to further stratify lesions classified as benign/probably benign on US with SWV ≤ 2.02 m/s had a sensitivity of 61.8% and specificity of 80.2%. SWE did not enhance diagnostic accuracy in differentiating lipoma from low-to-intermediate grade liposarcoma and lacked reproducibility across different machines.