Dr. Tharakeswara Bathala, from MD Anderson Cancer Center, and colleagues conducted a study to evaluate the diagnostic performance of conventional ultrasonography combined with USE for distinguishing benign from malignant soft-tissue masses.
The study included 74 patients with subcutaneous soft-tissue masses who had ultrasound and USE between January 2009 and May 2012. Three radiologists reviewed both sets of images, classifying grayscale ultrasound imaging features as benign, malignant, or indeterminate. The three reviewers also evaluated USE images according to tissue elasticity and classified lesions as soft, intermediate, or hard. Bathala's team used results from either percutaneous biopsy or surgical excision as the reference standard.
Out of 74 lesions, conventional ultrasound found 37 benign, eight indeterminate, and 29 malignant lesions. USE found 35 benign, 14 indeterminate, and 25 malignant lesions. Together, the two modalities classified 40 lesions as benign, four as indeterminate, and 30 as malignant, the researchers found.
The estimated sensitivity for conventional ultrasound was 100%, compared with 95% for ultrasound strain elastography and 100% for ultrasound and USE. Specificity was 71%, 67%, and 77%, respectively.
The data suggest that ultrasound alone is very sensitive for detecting and characterizing subcutaneous soft-tissue lesions, according to Bathala's team. But combining ultrasound and USE did improve specificity: Of the eight patients who had an indeterminate ultrasound, USE characterized four -- three correctly as benign and one incorrectly as malignant.
Adding USE to ultrasound, especially when results are indeterminate on ultrasound alone, offers better diagnostic performance, the researchers concluded.
