Evaluating the dıagnostıc accuracy of birads assessment of mammography and ultrasound according to ultrasound- guıded core needle biopsy: 9 years-experience in a small town

Abstract

Aim: To examine the diagnostic accuracy of the BI-RADS (Breast Imaging Reporting and Data System) assessment of ultrasonography and/or mammography according to US-guided core needle biopsy.
Materials and Methods: 463 patients who had ultrasonography and/or mammography imaging and subsequent US-guided core needle biopsy between June 2011 and June 2020 in the Radiology Departments of two different centers in our town were studied retrospectively. BI-RADS assessment of ultrasonography and/or mammography were compared with histopathological diagnoses. When both examination existed, the higher score was determined as the final BIRADS category. The diagnostic efficiency of ultrasonography and/or mammography for determining malignancy were calculated.
Results: Of the 463 lesions 222 (47.9%) were malignant and 241(52.1%) were benign. Based on ultrasonographic examination, malignancy was observed in 5 of the 86 (5.8%), 89 of 178 (50%), 4 of 63 (6.3%), 26 of 26 (100%) and 86 of 86 (100%), cases reported as BI-RADS 3, 4, 4A, 4C and 5. Based on mammography results, all of 50 masses (100%) classified as BI-RADS 5 and 38 of 47 masses (80.8%) classified as BI-RADS 4 were malignant. The frequency of malignancy in BI-RADS 3 masses was 5.8% based on ultrasonographic examination and 0.7% with both modalities. Sensitivity, specificity, positive predictive value, negative predictive value and diagnostic accuracy for ultrasonography were 92.2, 61.0, 69.1, 89.2, and 76.2%; for mammography were 96.7, 41.1, 89.8, 70.0, and 88.0%; for combined radiological evaluation were 99.5, 53.5, 66.3, 99.2 and 75.5% respectively.
Conclusion: When mammography and ultrasonography were evaluated together, the rate of malignancy in BIRADS 3 lesions decreased from 5.8% (only ultrasonography) to 0.7%. In addition, sensitivity and negative predictive value increased statistically significantly (p<0.001). This finding showed the importance of the co-evaluation of mammography and ultrasonography.