The correlation between the spread of metastases by level in the axillary nodes and disease-free survival in breast cancer. A multifactorial analysis.

Axillary lymph nodes were separated from 492 radical or modified radical mastectomies for primary breast cancer and examined according to their anatomical level corresponding to their position along the theoretical pathway of lymph drainage from the breast. The patterns of metastasis and the relationship between metastatized levels and disease-free survival were investigated to see whether complete axillary dissection is necessary for the staging and the planning of adjuvant therapy in breast cancer. Progressive involvement from level I (proximal) to level III (distal) was found in 206 specimens (80.8% of tumors with axillary metastases), while discontinuous or "skip" metastases were present in 49 (19.2%), including 38 (14.9%) with positive nodes at level II or III but not at level I. "Skip" metastasis was more frequent when fewer than four nodes were positive, and not related to either the size of the primary tumor or its location. The effect of age, menopausal status, tumor size, node status, number of positive nodes, anatomic level of axillary node involvement, estrogen and progesterone receptors, and adjuvant therapies on disease-free survival was evaluated using a multivariate proportional hazard model and life table analysis. This showed that disease-free survival was strongly related to the number of positive nodes (P less than 0.001), tumor size (P = 0.001) and level of node involvement (P = 0.01) as independent prognostic factors. Moreover, the subset of patients with four or more positive nodes and involvements of level III had a higher risk of recurrence (25% recurrence-free patients 5 years after mastectomy). The high frequency of "skip" metastases and the prognostic value of both the level of involvement and the number of metastatic nodes suggest that a complete axillary dissection is needed in the surgical management of breast cancer to obtain all the data useful in the planning of adjuvant therapy.