BI-RADS 3, 4, and 5 lesions: value of US in management--follow-up and outcome.

PURPOSE To evaluate the use, final outcome, and positive biopsy rate of American College of Radiology ultrasonographic (US) Breast Imaging Reporting and Data System (BI-RADS) categories 3, 4, and 5 recommended for breast masses. MATERIALS AND METHODS At US, consecutive masses, palpable and nonpalpable, categorized as BI-RADS 3, 4, and 5 between January 1, 2003, and December 31, 2004, were retrospectively reviewed with institutional review board approval. Medical records provided imaging and histologic information. RESULTS After patients lost to follow-up were excluded, the study population was 767 patients with 926 masses (476 palpable, 450 nonpalpable). In BI-RADS 3 masses (n = 356), imaging follow-up of 252 masses documented stability for 6-24 months. Aspiration of 24 masses revealed cysts. Biopsy in 80 masses revealed three malignancies, all of which were diagnosed within 6 months of the index examination, were smaller than 1 cm, and were node negative (negative predictive value = 99.2%). In BI-RADS 4 masses (n = 524), aspiration results indicated 35 cysts; biopsy in 455 revealed 85 malignancies (positive predictive value [PPV] = 16.2%). Imaging follow-up only in 34 revealed no cancers 2 and more years later. Among BI-RADS 5 masses (n = 46), 43 were malignant and three benign (PPV = 93.4%). CONCLUSION Inconsistent use of BI-RADS category 3 occurred in 14.0% of cases when biopsy was recommended. Although biopsy was performed in almost equal numbers of palpable and nonpalpable masses, only 11% of palpable BI-RADS 3 and 4 masses were malignant, as compared with 22% of nonpalpable masses. Strict adherence to lexicon characteristics of probably benign lesions should improve specificity.