Accuracy of percutaneous core biopsy of isolated breast microcalcifications identified by mammography. Experience with a vacuum-assisted large-core biopsy device

PURPOSE: To evaluate the diagnostic accuracy of 11-G vacuum-assisted stereotactic core biopsy (VAB) of isolated clusters of microcalcifications identified by mammography. MATERIALS AND METHODS: Retrospective analysis of 364 consecutive procedures from February 1999 to June 2002. Final outcome was histological diagnosis at surgery or mammographic follow-up. Linkage with local cancer registry was available. Diagnostic accuracy and upgrading of atypical ductal hyperplasia (ADH) to carcinoma or of ductal carcinoma in situ (DCIS) to invasive carcinoma (INV) was evaluated. The positive predictive value (PPV) of radiological judgement (score 1 to 5) and of the cluster volume (only for unifocal lesions) for ADH or more severe lesions was also considered. RESULTS: A total of 364 consecutive VAB procedures were evaluated (average age 54.9, range 33-81). VAB report was negative, ADH, DCIS or INV in 192, 22, 126 or 24 cases, respectively. Of 188 cases with ADH or more severe reports at VAB or surgical biopsy 16 had an originally negative VAB report, yielding a sensitivity of 91.4%. Upgrading at surgical biopsy for cases with follow-up was 29.4% (5/17) for ADH (DCIS=2, INV=3) and 17.3% (20/115) for DCIS. Of 221 cases with known outcome and mammograms available for review PPV was 37.1%, 65.9%, 90.9%, and 89.4% for radiological suspicion degrees 2,3,4 and 5, respectively (chi squared for trend = 32.44, p<10(-6)) and was 70.0%, 72.4% and 89.4% for cluster volumes of 0-60, 61-500 and >500 mmc, respectively (chi squared for trend= 2.36, p=0.12) among 195 unifocal clusters. No microcalcifications were found at core radiography in 20 cases (VAB negative=18, ADH=1, INV=1) with DCIS or INV occurring in 4 or 1 case, respectively, at further surgery. CONCLUSIONS: Core biopsy avoids unnecessary surgery in many subjects with suspicious microcalcifications, although it implies a non negligible risk of false negative report. Surgical biopsy in VAB negative cases could be indicated according to other variables (e.g. the degree of radiological suspicion). VAB has relevant limits in grading breast lesions, as ADH or DCIS are associated to a considerable risk of upgrading at surgical biopsy. VAB reports other than INV need to be confirmed at surgery before an individual treatment strategy may be defined.