Background and objectives: To determine the sentinel node detection rate and the accuracy with which the sentinel node histology reflects that of the axilla in a series of patients with palpable invasive breast cancer.
Methods: Forty-four patients with clinically node-negative palpable invasive T1 or T2 breast tumors underwent sentinel node biopsy using isosulfan blue dye, followed immediately by either local excision of the primary lesion with standard axillary lymph node dissection or modified radical mastectomy. All surgeries were performed at Northwest Hospital, Seattle, Washington, between January 1996 and October 1997.
Results: The sentinel node was successfully identified in 73% of the patients (32/44). The frequency of sentinel node detection was greater for tumors in the outer quadrants than the inner quadrants (z-test, P < 0.001). Of the 32 patients in whom a sentinel node was identified, 10 (31%) had histologically positive sentinel nodes: 5 (16%) by frozen section, 2 additional patients (6%) after permanent hematoxalin-eosin (H&E) stained sections, and the remaining 3 (9%) after immunohistochemical stains for cytokeratins when the FS and permanent H&E-stained sections were benign. Twenty patients had benign axilla. The sentinel node was falsely negative in 2 patients, yielding an accuracy of 93.8%, sensitivity of 83.3%, and negative predictive value of 91%.
Conclusions: Lymphatic mapping is technically feasible for patients with small (T1 or T2) palpable invasive breast tumors. The sentinel node can be reliably identified in the majority of these patients, and its histology reflects that of the axilla with a high degree of accuracy. Immunohistochemical stains and permanent H&E-stained sections of the sentinel node increased the test's ability to correctly identify axillary metastases. Improving this sensitivity remains a primary goal, however, if benign sentinel node histology is to be used as a criterion to preclude axillary dissection.
Copyright 1999 Wiley-Liss, Inc.