Concordance and Diagnostic Accuracy of Ultrasonography and Mammography Findings with Pathology Results in Breast Cancer

Abstract

Background: Ultrasonography and mammography are two radiologic approaches for screening breast cancer; however, the pathology report is required for the ultimate diagnosis of malignancy. This study aimed to assess the concordance of ultrasonography (US) and mammography with the pathology in breast cancer.
Methods: A cross-sectional study was conducted to assess the breast US and the mammography findings based on the BI-RADS model in comparison with the definitive pathology reports in a single medical center. The sensitivity, the specificity, positive (PPV) and negative predictive value (NPV) and also the concordance between the US and the mammography data were analyzed.
Results: In this study, 126 patients were included. The sensitivity, specificity, PPV, and NPV for the US were 69.8, 71.9, 75.6 and 81.3 and for mammography were 91.9, 76.6, 80.8 and, 94.6 percent, respectively. The ROC-curve for either the US or the mammography showed that the BI-RADS 4 was accompanied with the highest sensitivity and specificity for the screening of the malignant breast lesions regarding the final diagnosis. Although an overall higher correlation between mammography report and presence of a malignant lesion was observed, the total relative concordance between the results of US and mammography as screening tools proved to be statistically significant (P<0.01).
Conclusion: Both the US and the mammography were sensitive and specific screening tools, particularly for the malignant breast lesions. Furthermore, when evidence of the BI-RADS?4 in either the mammography or the US was present, utilization of the other test could be ignored before biopsy.