Differential diagnosis of benign and malign breast masses and histopathologic correlation elastography diagnostic evaluation of contribution

ABSTRACT Objectives: Our purpose was to research the contribution of 5-point scoring system and strain index by sonoelastography in differentiating between benign and malignant breast masses and to compare both of the methods' diagnostic performance. Material and Methods: The study was prospectively carried out with the research ethics committee approval in Dicle University Faculty of Medicine Department of Radiology between June, 2013 and October, 2013. Eighty women patients with breast masses were included in this study. These patients' age average was between 42 ± 13 (range; 14-81). Sonographic and sonoelastographic examinations for all the patients were done before the biopsy. The lesions were scanned with the real-time elastography-Siemens Acuson S2000 and 14 MHz Linear probe. On B-mod sonography, the lesions were evaluated according to the Breast Imaging Reporting and Data System and their BI-RADS scores were identified. After B-mod sonographic evaluation, elastographic technique was used with the same probe and elastographic images were obtained. Afterwards, elasticity scores were determined with a 5-point scoring method known as "Tsukuba". The mass and its surrounding adipose tissue elasticity ranges were measured quantatively through static elastogram image by means of ROI (Region of interest); afterwards, the device was automatically calculated. The diagnostic efficiency of B-mod sonography, sonoelastragrophic 5-point scoring and strain index methods was compared with histopathological findings. Results: According to the histopathological evaluation; the 59 (%73,75) patients' lesions in 80 ones in this study were benign, and 21 (%26,25) were malign. Most of the benign lesions included fibroadenoma, and the malign lesions included invasive ductal carcinoma. The sensitivity of the B-mod sonography was %100; its specificity was %69,49; positive predictive values were %53,85; negative predictive values were %100, and its diagnostic accuracy was % 77. According to the results of the ROC analysis, 5- point scoring method under the ROC curve was 0,900; The strain index was 0,961. The sensitivity of 5-point scoring method was %85,7; its specificity was %86,4; its accuracy was %86,2; its positive predictive values were %69,2; its negative predictive values were %94,4. Strain index average were detected as 10,45 ± 7,04 (range; 3,4-25,1) in the malign lesions and 2,88±2,56 (range; 0,5-19,18) in the benign. The strain index averages of the malign lesions were detertemined as statiscally and significantly higher than the benign lesions (p<0,005). The best cut off point for the strain index was 4, 29. The method's sensitivity was detected as %95,2; its specificity as %93,2; its accuracy as %93,7; its positive predictive values as %83,3 and its negative predictive values as %98,2. The sensitivity of B-mod US had higher sensitivity than the one of both 5-point scoring and strain index methods. The B-mod US specificity was lower than both 5-point scoring and strain index. B-mod US had the highest sensitivity and strain index method had the highest specificity. The diagnostic performance of strain index performance had higher values than 5-point scoring method's. Conclusion: We detected in our study that the diagnostic performance of strain index had higher values than the ones of 5-point scoring method. We think that the diagnostic accuracy of sonoelastography in differentiating between benign and malign breast masses increases by means of adding strain index measurement to 5-point scoring method. Therefore; Sonoelastographic examination, which is used with the combination of the two methods (strain index and 5-point scoring methods) that are carried out after B-mod sonographic scanning can contribute to the planning of doing breast masses biopsy and the prevention of the potentially unnecessary attempts Key words: Elastography, breast masses, scoring, biopsy