Sentinel nodes in breast cancer: relevance of axillary level II nodes and optimal number of nodes that need to be removed.

BACKGROUND: In some patients, the radiocolloid used to perform sentinel lymph node biopsy (SLNB) for breast cancer appears in a number of lymph nodes and in different levels of the axilla. Most positive sentinel lymph node specimens (SLNSs) removed during SLNB are identified in level I of the axilla and within the first 4 SLNSs. Our objective was to verify the staging accuracy of harvesting only the first 4 SLNSs and to determine the relevance of SLNSs that reside in level II of the axilla. METHODS: A prospective database documenting the method of identification, radioisotope count, order of retrieval, and axillary level of SLNSs from 893 SLNBs was analyzed. RESULTS: A median of 2 SLNSs (range 1-9) were removed per patient. More than 4 SLNSs were found in 8.0%. All SLNSs harboring the largest nodal metastases were identified within the first 4 harvested. Twenty-one percent (184 of 870) of patients had level II SLNSs; 4.9% (9 of 184) were positive. When SLNSs were positive in both levels I and II, the nodal metastases were always of greater or equal size in the level I nodes. Only one patient (0.5%) had a positive level II SLNS macrometastasis (> 2 mm, pN1), with a negative level I SLNS, but it was the hottest node and was removed first. CONCLUSIONS: Removal of more than the first 4 hottest SLNSs does not improve staging accuracy. Level II nodes can be ignored if a hotter level I SLNS is first identified.