Stereotactic 11-gauge vacuum-assisted breast biopsy: Influence of number of specimens on diagnostic accuracy

PURPOSE: To determine whether number of specimens obtained at stereotactic 11-gauge vacuum-assisted breast biopsy with the patient prone influences diagnostic accuracy and to determine whether this number varies depending on mammographic appearance of lesions as masses or microcalcifications. MATERIALS AND METHODS: Biopsy was prospectively performed in 100 patients (median age, 55 years; range, 31-81 years) with 100 lesions that were mammo-graphically evident as masses (n = 50) and microcalcifications (n = 50) with standardized protocol to acquire 20 specimens per lesion in three 360degrees probe rotations at one skin entry site. Specimens were histologically evaluated sequentially, and findings were compared with results of surgical excision or of mammographic follow-up for at least 24 months. Differences in diagnostic yield after each probe rotation and differences in diagnostic yield between masses and microcalcifications were determined with chi(2) test. RESULTS: Up to 12 specimens harvested within two 360degrees probe rotations were necessary to yield correct diagnosis in 96% of patients with masses and 92% of patients with microcalcifications. Diagnostic yield was not improved with more than 12 specimens for masses or microcalcifications. In two (4%) of 47 patients with lesions that were eventually diagnosed as cancer, results at stereotactic biopsy indicated they were benign. Underestimation of diagnosis of lesions as atypical ductal hyperplasia and ductal carcinoma in situ occurred in two (50%) of four and two (17%) of 12 lesions, respectively. With 20 specimens harvested during three probe rotations, there was no statistically significant difference in diagnostic yield between patients with masses and those with microcalcifications (P =.68). CONCLUSION: At 11 -gauge vacuum-assisted biopsy, highest diagnostic yield was achieved with 12 specimens per lesion, independent of mammographic appearance of the lesion. Even with standardized retrieval of 20 specimens per lesion, underestimation of disease still occurs. (C) RSNA, 2004.