PURPOSE: While positive resection margin (RM) in women undergoing breast-conserving surgery (BCS) represents a clear indication for re-resection, there is no unequivocal recommendation regarding the extent of the clear RM. The aim of this study was to define the optimal extent of the RM and the risk factors for close or positive RM. METHODS: Patients scheduled for BCS had diagnosis confirmed before BCS (lumpectomy and quadrantectomy) by core biopsy. Sentinel lymph node biopsy followed BCS, and in case of positive findings axillary lymph node dissection followed. According to RM patients were categorized into 4 groups: 1) Patients with positive RM; 2) Clear RM < 2 mm; 3) Clear RM of 2-5 mm; and 4) RM > 5 mm. In the first 3 groups where re-resection was indicated, the presence of tumor cells in the re-resection specimen was determined. All patients were followed for local recurrence. RESULTS: 330 patients undergoing BCS were studied. Median follow up was 39.6 months (range 12-70). Lumpectomy was performed in 111 cases and quadrantectomy in 219. In 19 cases the final procedure was mastectomy due to the impossibility to achieve negative RM. In 78 cases re-resection followed the primary procedure due to close or positive RM. Clear RM was < 2 mm in 12 cases (15%), 2-5 mm in 56 (72%) and positive margin in 10 (13%). Positive re-resection specimen was detected in 31 cases (39.7%) (in 10 cases with positive RM after primary procedure, in 3 with negative margin < 2 mm and in 18 with 2-5 mm margin). The re-resection rate according to the location of the primary tumor was 77% (n=60) in the upper outer quadrant, 8% (n=6) in the lower outer quadrant, 6% (n=5) in the upper inner quadrant, 4% (n=3) in the lower inner quadrant, and 5% (n=4) in centrally located tumors. Multicentric/ multifocal tumor was diagnosed in 16 cases from which re-resection was indicated in 12 cases (75%). The number of re-resection according to tumor size was as follows: Tis 8 cases (30.7%), T1a none, T1b 14 (20.2%), T1c 34 (22.5%), T2 22 (28%). Re-resection was performed in 8 cases (31%) of ductal carcinoma in situ (DCIS), in 53 (22%) of ductal carcinoma, in 10 (37%) of lobular carcinoma, and in 7 (15%) of other histology. Five cases with local relapse were detected during follow up. CONCLUSION: The generally recommended clear RM of 1-5 mm is not sufficient because of the high number of positive specimens in the case of clear RM of 2-5 mm. The risk factors for close or positive RM are multicentric tumors and upper outer location of the primary tumor. Longer follow up will be needed to analyze local relapse rate according to RM status.

Department of Surgery Atlas Hospital Tomas Bata University in Zlin Zlin Czech Republic