Comparison of two models for predicting non-sentinel lymph node metastases in sentinel lymph node-positive breast cancer patients.

Several models for the prediction of non-sentinel lymph node (SLN) metastasis in SLN-positive breast cancer patients have been proposed. In this study, the authors evaluate the Memorial Sloan-Kettering Cancer Center (MSKCC) nomogram and an axilla scoring system from Tenon Hospital to predict the probability of having non-(SLN) involvement and to define their actual usefulness when applied to the subgroup of patients with micrometastasis in SLN. The study population consisted of 103 patients: 74 with macrometastasis and 29 with micrometastasis in the SLN. The receiver operating characteristic (ROC) curves were drawn and the area under the curves (AUCs) was calculated to assess the discriminative power of the nomograms. Both the ROCs and relative AUCs were calculated for all the patients in the study and for the two subgroups. The AUC for the entire study population was 0.712 in the MSKCC nomogram and 0.759 in the Tenon score. In 74 patients with macrometastasis in SLN, the AUC was 0.760 in MSKCC nomogram and 0.707 in Tenon score. Micrometastasis in SLN was found in t29 patients: AUC was 0.577 in the MSKCC nomogram and 0.738 in the Tenon score. Several institutions have tested the MSKCC nomogram, with AUC ranging from 0.58 to 0.84. It was not validated by four studies, which did not recommend its use in patients with micrometastasis. In our results, the validation of the Tenon score confirmed its relevance even in patients with micrometastasis in SLN.