A Breast Cancer Nomogram for Prediction of Non-Sentinel Node Metastasis - Validation of Fourteen Existing Models

Abstract

<b>Background:</b> To avoid performing axillary lymph node dissection (ALND) for non-sentinel lymph node(SLN)-negative patients with-SLN positive axilla, nomograms for predicting the status have been developed inmany centers. We created a new nomogram predicting non-SLN metastasis in SLN-positive patients with invasivebreast cancer and evaluated 14 existing breast cancer models in our patient group. Materials and <br/><b>Methods</b>:Two hundred and thirty seven invasive breast cancer patients with SLN metastases who underwent ALND wereincluded in the study. Based on independent predictive factors for non-SLN metastasis identified by logisticregression analysis, we developed a new nomogram. Receiver operating characteristics (ROC) curves for themodels were created and the areas under the curves (AUC) were computed. <br/><b>Results</b>: In a multivariate analysis,tumor size, presence of lymphovascular invasion, extranodal extension of SLN, large size of metastatic SLN,the number of negative SLNs, and multifocality were found to be independent predictive factors for non-SLNmetastasis. The AUC was found to be 0.87, and calibration was good for the present Ondokuz Mayis nomogram.Among the 14 validated models, the MSKCC, Stanford, Turkish, MD Anderson, MOU (Masaryk), Ljubljana,and DEU models yielded excellent AUC values of > 0.80. <br/><b>Conclusions</b>: We present a new model to predict thelikelihood of non-SLN metastasis. Each clinic should determine and use the most suitable nomogram or shouldcreate their own nomograms for the prediction of non- SLN metastasis.