RESUMO
<p><b>OBJECTIVE</b>To compare the effectiveness and accuracy of the use of vacuum-assisted biopsy (VAB) versus wire localization (WL) in the diagnosis of non-palpable breast lesions (NPBL).</p><p><b>METHODS</b>Ninety-seven consecutive women with NPBL were randomized into VAB group and WL group. All specimens were identified by mammography. The patients were requested to score the cosmetic appearance of their breast after operation, and a numerical rating scale was used to measure pain on the first postoperative day. Underestimation rates for atypical ductal hyperplasia (ADH) and ductal carcinoma in situ (DCIS) were recorded if open surgical biopsy revealed DCIS and invasive cancer, respectively. Clear margins were also recorded in the two groups.</p><p><b>RESULTS</b>VAB and WL located all the NPBL successfully. In the VAB group, the specimen volume was smaller than that of the WL group (2.3 cm(3) vs. 18.4 cm(3), P = 0.03). Underestimation rates of ADH and DCIS in the VAB group were 16.7% and 11.1%, respectively. The diagnostic accordance rate of VAB was 97.9%, the false negative rate was 2.1%, and there was no false positive case. The means of the numerical rating pain scale were different in both groups (1.7 for VAB vs. 2.5 for WL, P = 0.02). When cosmetic results were taken into account, 40 VAB patients had excellent outcomes and 8 good outcomes, compared with 25 excellent and 24 good for the WL group. There were better cosmetic outcomes with the VAB procedure (P < 0.05).</p><p><b>CONCLUSION</b>VAB is highly reliable and may avoid diagnostic open surgery in the majority of patients with benign lesions. However, because of the underestimation of histologic diagnosis and tumor margin involvement, VAB can not be used to completely substitute wire localization.</p>