Objective: The purpose of this study was to assess and compare contrast-enhanced ultrasound and MRI patterns in the diagnosis of soft-tissue masses.
Materials and methods: Two hundred fifty-five consecutively registered patients with histologically confirmed soft-tissue masses were included in this retrospective study. The diagnostic properties of four predefined contrast enhancement (CE) patterns were assessed, and logistic regression analysis was performed to determine the correlation between diagnosis and CE pattern, lesion size, and patient age and sex. The influence of lesion size on the occurrence of inhomogeneous CE patterns in malignancies was also determined.
Results: Homogeneous CE patterns were highly specific for benignity, and inhomogeneous CE was moderately specific for malignancy in both ultrasound and MRI. A combination of homogeneous and inhomogeneous CE patterns led to 88.3% and 88.7% sensitivity, 66.7% and 59.7% specificity, 73.4% and 68.2% correct classification, 54.6% and 47.8% positive predictive value, 92.6% and 92.7% negative predictive value, 2.65 and 2.20 positive likelihood ratio, and 0.18 and 0.19 negative likelihood ratio for contrast-enhanced ultrasound and contrast-enhanced MRI. Cases with homogeneous CE in either ultrasound or MRI also were predominantly benign. The occurrence of inhomogeneous CE in malignant lesions increased with size.
Conclusion: CE patterns in ultrasound and MRI offer additional information about the differentiation of an unknown soft-tissue mass. The results of this study showed that homogeneous or absent CE was specific for benign differentiation and that heterogeneous CE was linked to malignancy. The routine analysis of CE patterns should increase diagnostic reliability in unclear soft-tissue masses.
Keywords: classification; contrast enhancement; patterns; sarcoma; soft-tissue tumors.