We performed several prospective clinical studies to evaluate the role of sonoelastography for the characterization of nonpalpable breast lesions.
First, 842 consecutive women who were scheduled to undergo US-guided core biopsy due to 864 breast lesions(79 cancers and 785 benign lesions) detected by supplemental screening US were examined with a commercialized sonoelastography. For the BI-RADS category 4a lesions, 28.5% (216 of 757) had a normal strain and 99.1% (214 of 216) of lesions with normal strain found out to be benign.
Second, 300 women with 339 nonpalpable lesions were prospectively examined with sonoelastograpy and color Doppler US (CDUS) prior to biopsy. Of the 301 lesions with BI-RADS category 4a, 76 (25%) lesions showed normal strain on elastography and no vascularity on CDUS and all of them were found to be benign. Breast lesions diagnosed with DCIS by US-guided core biopsy with invasive components at surgery show less strain than pure DCIS at elastography regardless of the lesion size and B-mode US findings.
Third, we evaluated the accuracy of neural network analysis of elastographic features at sonoelastography for the classification of 113 benign and 68 malignant breast tumors. Az values of the three elastographic features- mean (0.87), median (0.86) and mode (0.83)-were significantly higher than the Az values for the six B-mode features (0.54-0.69).
In conclusion, addition of elastography to breast US has potential to reduce benign biopsy for BIRADS category 3 or 4a lesions that are detected by supplemental screening US.