Effect of quantitative and semi-quantitative elastography methods for management of borderline lesions on ultrasonography

Abstract

Background: Elastography (strain or shear-wave) is a method that estimates tissue stiffness. Introduction: Our aim in this study to evaluate the quantitative and semi-quantitative ultrasound elastography methods for diagnosis of category BI-RADS 4a and BI-RADS 3 lesions, which are borderline for biopsy and follow-up. Material and Methods: 175 consecutive women with 193 ultrasound-visible breast lesions were classified on : Conventional B-mode Ultrasonography (CUS) according to BI-RADS scoring system. Quantitative and semiquantitative values from ultrasound elastography in the form of strain Elastography Ratio (SER), shear Wave Elastography (SWE) and Shear Wave Elastography Ratio (SWER) were obtained. The lesions categorized as BI-RADS 4a and BI-RADS 3 on ultrasound were subsequently re-categorized according to elastography values. Results: Except for the 13 BI-RADS 2 lesions, the remaining 180 lesions were biopsied. Pathology showed 83 lesions to be benign and 97 to be malignant. The sensitivity and specificity of the CUS was 96.9% and 75.0%, respectively with an accuracy of 86.0%. Cut-off points calculated based on ROC curves were 56.8 kPa for SWE, 3.53 for SWER and 3.81 for SER. When we downgraded BI-RADS 4a lesions based on elastography results, the specificity (CUS+SER 96.9%, CUS+SWE 91.7%, and CUS+SWER 90.6%) and the accuracy (CUS+SER 95.3%, CUS+SWE 92.7%, and CUS+SWER 92.2%) are better than CUS. When we upgraded BI-RADS 3 lesions based on elastography results, the sensitivity of combined sets of SWE (99, 0%) and SWER (100,0%) are better than CUS. Conclusion: The rate of false negative biopsies can be decreases with combined use of elastography and ultrasonography.