Assessment of 142 stellate lesions with imaging features suggestive of radial scar discovered during population-based screening for breast cancer.

Because some lesions diagnosed as radial scars (RS) on core biopsy have been found to be malignant on excision, core biopsy has not had an established role in the assessment of RS. In our breast cancer-screening program, we have avoided core biopsy if RS is suspected on imaging. Recently, two reports have expanded the experience with core biopsy of RS, prompting this review of our assessment protocols for lesions suspected as being RS. Between January 1996 and January 2003, stellate lesions with imaging features of RS in which core biopsy was omitted because of a presumptive radiologic diagnosis of RS are included. Demographic, radiologic, and cytologic data were correlated with the histologic findings in the excised specimen. On imaging, 9% (142) of all stellate lesions were suspected to be RS. Only 66.2% (94) were confirmed as RS on histology; 38 cases (28.6%) were carcinomas (36 invasive, 2 in situ) and 7% showed benign fibrocystic changes; 87.1% of the carcinomas required further surgery for positive margins. Axillary staging was also needed for the invasive cancers. Among the histologically proven RS, 28 of 94 (29.8%) showed areas of atypical ductal hyperplasia, lobular neoplasia, ductal carcinoma in situ, or invasive carcinoma. These proliferations were typically focal and unpredictable and were usually completely excised by the initial diagnostic biopsy. Core biopsy would be valuable in the assessment of lesions with imaging features suggestive of RS since 28.6% of such lesions are indeed carcinomas that mimic RS. Identification of these cancers would permit one stage breast and axillary surgery to be planned. The policy of mammographic surveillance for lesions with nonmalignant core biopsies remains controversial because of the paucity of data. Ongoing evaluation is needed as more experience is reported.