Evaluation of a breast cancer nomogram for predicting the likelihood of additional nodal metastases in patients with a positive sentinel node biopsy.

Aim: Completion axillary lymph node dissection (CALND) performed as a standard procedure after a positive sentinel node biopsy (SLNB) in breast cancer patients results, in almost 40-70% of cases, in no additional positive nodes. A nomogram has been developed at Memorial Sloan Kettering Cancer Center (MSKCC) to predict the likelihood of nonsentinel node metastases (NSLNM) after a positive SLNB. Aim of study was to assess the accuracy of MSKCC nomogram in our community breast cancer population.
Material of Study: From a retrospective database of 276 breast cancer patients we evaluated 62 consecutive cases who underwent CALND after a positive SLNB. Patient and tumor characteristics were collected and the nomogram was used to calculate the probability of NSLNM. The accuracy of MSKCC nomogram was tested by the Receiver Operating Characteristic (ROC) curve. The Area Under the Curve (AUC), sensitivity and specificity were calculated for a 10% cut-off value.
Results: Presence of macrometastases (p=0.03) and its extranodal extension (p=0.013) in sentinel node were associated with NSLNM, while other tumor and patient characteristics were not. The accuracy of MSKCC nomogram as measured by AUC was 0.67. The nomogram showed 95% sensitivity and 14% specificity. We revised the nomogram by incorporating the presence of extranodal extension and we obtained a new test with improved specificity (84%).
Discussion: The modified predictive model is a useful tool in predicting the likelihood of NSLNM in our cohort of patients and may help decision regarding the need of completion axillary lymph node dissection.