RESUMEN
UNLABELLED: The purpose of the study was to compare the accuracy of breast MRI and ultrasonography in assessing the tumor focality and tumor size of newly diagnosed non-high risk breast cancer patients. METHODS: The tumor focality status and the maximal tumor diameter by MRI and ultrasonography were retrospectively compared with the corresponding histopathological findings as reference. Test characteristics concerning the tumor focality status were calculated. Bland-Altman plots were generated to evaluate the agreement of the tumor size measurements by imaging and histopathology. The t-test for dependent samples and the Fisher exact test were used to test differences between groups for statistical significance. The Pearson correlation coefficient r was calculated to measure the degree of association between the tumor diameter by imaging and histopathology. RESULTS: Sixty-four patient diagnosed between 2011 and 2013 were analyzed. MRI showed a good sensitivity of 83% for detecting multifocal disease (ultrasonography, 75%). The positive predictive value was 67% and the ratio of true-positive to false-positive findings 2.0. MRI showed better limits of agreement (-21 to 26 mm versus -29 to 26 mm) and a better correlation (r=0.77 versus r=0.66) with the histopathological tumor diameter compared to ultrasonography. The mean differences between the tumor diameter by MRI and histopathology and ultrasonography and histopathology were not significantly different (p=0.09). The T classification (T1a, T1b, T1c, T2, T3) was correctly estimated by MRI in 43 patients (67.2%) and by ultrasonography in 39 patients (60.9%) (p=0.58). CONCLUSION: In our patient cohort only a modest diagnostic advantage of MRI compared to ultrasonography could be detected.
