Sentinel lymph node dissection (SLND) for small, early-stage breast cancer is well accepted. However, the role of SLND for large primary breast cancer is controversial. We investigated the feasibility and clinical applicability of SLND in patients with large (> or = 5 cm) breast cancers and clinically negative axillae.

A prospectively entered database was used to identify all patients who underwent surgical management of histopathologically confirmed primary breast carcinomas > or = 5 cm in diameter between September 1991 and August 2000. Patients who had clinically negative axillae and underwent SLND followed by completion axillary lymph node dissection (ALND) were selected for the study. The positivity rate, accuracy rate, and false-negative rate of SLND were determined.

Of the 41 patients selected for the study, 24 had infiltrating ductal carcinoma and 17 had infiltrating lobular carcinoma. Mean tumor size was 7.12 cm (range, 5-23 cm). At least one sentinel lymph node (SLN) was identified in all cases. Thirty patients had tumor-positive SLNs. Axillary metastasis was also identified in one patient who did not have a positive SLN. Thus, SLN status accurately predicted regional nodal status in 98% (40 of 41) of cases. The false-negative rate of SLND was 3% (1 of 31). None of the three patients with SLN micrometastasis, defined as a tumor focus < or = 2 mm, had tumor deposits in nonsentinel axillary lymph nodes. Only SLN macrometastasis (> 2-mm tumor deposit) and primary tumor size > or = 7 cm predicted nonsentinel axillary metastasis with significance on multivariate analysis (P = .008 and P = .046, respectively).

SLND is feasible and accurate in patients with large breast cancers and clinically negative axillae. Axillary lymph node dissection can be avoided in nearly one third of patients by focused examination of the SLN.