Validation of a Predictive Tool for Assessing the Risk of Non-Sentinel Node Metastases in Breast Cancer Patients with Sentinel Lymph Node MetastasisAuthor: Marco Gipponi
Background: An original predictive tool based on the primary tumor’s and sentinel node’s features able to assess the risk and the extent of nodal metastases into residual axillary lymph nodes after sentinel node biopsy was proposed. Materials and Methods: A training set of 1,836 breast cancer patients was assessed to select those clinico-pathological factors to be included into a multivariate model that was validated in 201 patients. Results: T stage category (P = .018); histological grading (P = .048), and number of metastatic SN (P = .027) were included into the predictive model and tested into the validation set (AUC of the ROC curve was .638; P = .001). Four categories of patients, based on the individual probability to have one or more additional positive nodes, were identified. Conclusion: This predictive tool may aid in the decision-making as for completion axillary dissection in breast cancer patients with sentinel node metastasis.