Mammographic Findings Predicting an Extensive Intraductal Component in Early Stage Invasive Breast Cancer: Analysis on Microcalcification

Kim, Jeong Ah; Kim, Mi Hye; Lee, Mi Kyung; Oh, Ki Keun; Kim, Eun Kyung

Abstract

Purpose: To analyze the mammographic findings of extensive intraductal compenent (EIC)-positive early invasive breast carcinoma and to determine the mammographic features which predict an EIC positivity in aninvasive carcinoma. MATERIALS AND METHODS: The mammographic and pathologic findings in 71 patients aged 34 -79 (mean 50) years in whom stage I or II invasive breast carcinoma had been diagnosed were etrospectively analysed. The m a mmographic findings were assigned to one of three groups: mass, mass with microcalcification, or microcalci-fication only. The shape and distribution of a calcification were classified according to the BI-RADS TMl exicon, and its extent was classified as either more or less than 3cm. To detect the presence or absence of EIC and the type of ductal carcinoma in situ (DCIS), the findings were reexamined by means of slide mappings. Results: Twenty-eight of 71 patients (39%) showed EIC positivity. The mammographic findings of EIC-posi-tive invasive cancer (n=28) were mass with microcalcification (n=14), microcalcification only (n=7) and mass only (n=7). The mammographic finding which predicted EIC positivity was mass with microcalcification (P-PV: 0.67, NPV: 0.33, p=0.02). A mammographic of mass only (n=39) showed a significantly high negative predictive value for EIC positivity. (PPV 0.18, NPV 0.82, p<0.01). A comparison of cases with or without calci-fication showed that those with microcalcifications (n=32) showed a significantly high PPV of 0.66 (NPV: 0.34, p<0.01 ) while those without calcification (n=39) showed a significantly high NPV of 0.82 (PPV:0.18, p<0.01 ). There were no significant differences in positive predictive values for EIC between the shape, distribution and extent of calcifications. Conclusion: Whenever microcalcification with or without mass is seen on mammograns obtained during early breast cancer, we can predict EIC-positivity, regardless of shape or distribution according to the BI-RADS TM lexicon.