Background: Sentinel lymphadenectomy is highly accurate for identifying axillary metastasis from a primary breast carcinoma. Nonsentinel axillary lymph nodes (NSNs) are unlikely to contain tumor cells if the axillary sentinel node (SN) is tumor free. We previously showed that the size of the primary tumor and the size of its SN metastasis predict the risk of NSN tumor involvement detected by hematoxylin and eosin staining. This study used immunohistochemical staining (IHC) to determine the likelihood of NSN axillary metastasis in the presence of SN metastasis.
Methods: Between 1991 and 1997, axillary lymphadenectomy was performed in 156 women (157 axillary basins) who had primary breast carcinoma with SN metastasis. By hematoxylin and eosin staining, we identified NSN metastasis in 55 axillae (35%). IHC was then used to re-examine all NSNs (1827 lymph nodes) from the remaining 102 axillae. The incidence of IHC-detected NSN involvement was analyzed with respect to clinical and tumor characteristics.
Results: By using IHC, we identified NSN metastasis in 15 (14.7%) of the 102 axillae. By multivariate analysis, the size of the SN metastasis (P = .0001) and the size of the primary tumor (P = .038) were the only independent variables predicting NSN metastasis determined by using either hematoxylin and eosin staining or IHC. Only the number of SN metastases (1 vs. >1) was a significant (P = .04) predictor of IHC-detected NSN metastasis.
Conclusions: Use of IHC increases the likelihood of detection of NSN metastasis, and the risk of IHC-detected metastasis increases with the size of the SN metastasis and the size of the primary tumor. If SN involvement is micrometastatic (< or =2 mm) or detected by using IHC, tumor cells are unlikely to be found in other axillary lymph nodes in patients with a small primary tumor. The clinical significance of micrometastatic disease in lymph nodes is controversial, and a prospective randomized study is necessary to resolve this important issue.