BI-RADS categorization as a predictor of malignancy

PURPOSE: To determine the positive predictive value (PPV) of the American C ollege of Radiology Breast Imaging Reporting and Data System (BI-RADS) cate gories 0, 2, 3, 4, and 5 by using BI-RADS terminology and by auditing data on needle localizations. MATERIALS AND METHODS: Between April 1991 and December 1996, 1,400 mammogra phically guided needle localizations were performed in 1,109 patients. Info rmation entered into the mammographic database included where the initial m ammography was performed (inside vs outside the institution), BI-RADS categ ory, mammographic finding, and histopathologic findings. A recorded recomme ndation was available for 1,312;localizations in 1,097 patients who compose d the study population. who composed the study population RESULTS: The 1,312 localizations yielded 449 (34%) cancers (139 [31%] were ductal carcinoma in situ [DCIS]; 310 [69%] were invasive cancers) and 863 ( 66%) benign lesions. There were 15 (1%) category 0 lesions; the PPV was 13% (two of 15 lesions). There were 50 (4%) category 2 lesions; the PPV was 0% (0 of40 lesions). There were 141 (1 1%) category 3 lesions; the PPV was 2% (three of 141 lesions). The three cancers in this group were all non-comedo type DCIS. There were 936 (71%) category 4 lesions; the PPV was 30% (279 of 936 lesions). There were 170 (13%) category 5 lesions; the PPV was 97% (16 5 of 1 70 lesions). CONCLUSION: Placing mammographic lesions into BI-RADS categories is useful for predicting the presence of malignancy. Perhaps, most important, a lesio n placed into BI-RADS category 3 is highly predictive of-benignity ty, and short-term interval follow-up as an alternative to biopsy would decrease th e number of biopsies performed in benign lesions.