HISTOPATHOLOGIC VALIDATION OF THE SENTINEL LYMPH-NODE HYPOTHESIS FOR BREAST-CARCINOMA

Background and Objective The sentinel node hypothesis assumes that a p rimary tumor drains to a specific lymph node in the regional lymphatic basin. To determine whether the sentinel node is indeed the node most likely to harbor an axillary metastasis from breast carcinoma, the au thors used cytokeratin immunohistochemical staining (IHC) to examine b oth sentinel and nonsentinel lymph nodes. Methods From February 1994 t hrough October 1995, patients with breast cancer were staged with sent inel lymphadenectomy followed by completion level I and II axillary di ssection. if the sentinel node was free of metastasis by hematoxylin a nd eosin staining (H&E), then sentinel and nonsentinel nodes were exam ined with IHC. Results The 103 patients had a median age of 55 years a nd a median tumor size of 1.8 cm (58.3% T1, 39.8% T2, and 1.9% T3), A mean of 2 sentinel (range, 1-8) and 18.9 nonsentinel (range, 7-37) nod es were excised per patient. The H&E identified 33 patients (32%) with a sentinel lymph node metastasis and 70 patients (68%) with tumor-fre e sentinel nodes. Applying IHC to the 157 tumor-free sentinel nodes in these 70 patients showed an additional 10 tumor-involved nodes, each in a different patient. Thus, 10 (14.3%) of 70 patients who were tumor -free by H&E actually were sentinel node-positive, and the IHC lymph n ode conversion rate from sentinel node-negative to sentinel node-posit ive was 6.4% (10/157). Overall, sentinel node metastases were detected in 43 (41.8%) of 103 patients. In the 60 patients whose sentinel node s were metastasis-free by H&E and IHC, 1087 nonsentinel nodes were exa mined at 2 levels by IHC and only 1 additional tumor-positive lymph no de was identified. Therefore, one H&E sentinel node-negative patient ( 1.7%) was actually node-positive (p < 0.0001), and the nonsentinel IHC lymph node conversion rate was 0.09% (1/1087; p < 0.0001). Conclusion s If the sentinel node is tumor-free by both H&E and IHC, then the pro bability of nonsentinel node involvement is <0.1%. The true false-nega tive rate of this technique using multiple sections and IHC to examine all nonsentinel nodes for metastasis is 0.97% (1/103) in the authors' hands. The sentinel lymph node is indeed the most likely axillary nod e to harbor metastatic breast carcinoma.