Background: Axillary lymph node dissection (ALND) is the current standard of care for breast cancer patients with sentinel lymph node (SN) involvement. However, the SN is the only involved axillary node in a significant proportion of these patients. Here we examined factors predictive of non-SN involvement in patients with a metastatic SN, in order to develop a scoring system for predicting non-SN involvement.
Materials and methods: This study was based on a prospective database of 337 patients who underwent SN biopsy for breast cancer, of whom 81 (24%) were SN-positive; we examined factors predictive of non SN involvement in the 71 of these 81 women who underwent complementary ALND. All clinical and histological criteria were recorded and analysed according to non-SN status, by using Chi-2 analysis, Student's t-test, and multivariate logistic regression.
Results: Univariate analysis showed a significant association between non-SN involvement and histological primary tumor size (p=0.0001), SN macrometastasis (p=0.01), the method used to detect SN metastasis (H&E versus immunohistochemistry) (p=0.03), the number of positive SNs (p=0.049), the proportion of involved SNs among all identified SNs (p=0.0001) and lymphovascular invasion (p=0.006). Histological primary tumor size (p=0.006), SN macrometastasis (p=0.02) and the proportion of involved SNs among all identified SNs (p=0.03) remained significantly associated with non-SN status in multivariate analysis. Based on the multivariate analysis, we developed an axilla scoring system (range 0-7) to predict the likelihood of non-SN metastasis in breast cancer patients with SN involvement.
Conclusion: In patients with invasive breast cancer and a positive SN, histological primary tumor size, the size of SN metastases, and the proportion of involved SNs among all identified SNs were independently predictive of non-SN involvement.