Low performance of the MSKCC nomogram in preoperatively ultrasonically negative axillary lymph node in breast cancer patients

BACKGROUND AND OBJECTIVES: In order to predict the nonsentinel lymph node (NSLN) metastases in sentinel lymph node (SLN) positive patients a nomogram was created at the Memorial Sloan Kettering Cancer Centre (MSKCC). The aim of our study was to validate the MSKCC nomogram in patients grouped by the preoperative ultrasound (US) examination of the axillary lymph nodes. METHODS: The MSKCC nomogram was validated separately in three groups of patients: (US-0) only clinically preoperatively negative axillary lymph nodes (126 patients), (US-1) US negative axillary lymph nodes (109 patients), and (US-2) US suspicious but fine needle aspiration biopsy (FNAB) negative axillary lymph nodes (41 patients). RESULTS: The predicted probability underestimates the actual probability with the mean absolute error equal to 0.116 in the US-0 group (P = 0.003), and overestimates the actual probability (mean absolute error equal to 0.084) in US-1 group (P = 0.033) and US-2 group (mean absolute error is 0.110) (P = 0.275). CONCLUSION: We found that the MSKCC nomogram overestimates the probability of the NSLN metastases in breast cancer patients with (i) preoperatively US negative or (ii) US suspicious, but FNAB negative axillary lymph nodes. We also found that MSKCC nomogram has only limited value in patients with only clinically negative axillary lymph nodes.