Sentinel lymph-node biopsy-based prediction of further breast cancer metastases in the axilla

Background: Histopathological factors may help identifying a subgroup of breast cancer patients with metastases confined to the sentinel lymph nodes (SLNs). Methods: A retrospective analysis was carried out on 111 tumours successfully mapped with Patent blue, 69 of which had SLN metastases. Results: Multivariate analysis revealed that SLN metastases situated in the sinuses and a small tumour size are the two most important predictors of involvement of only one SLN. The metastasis size and a small tumour size were found important in the model discriminating between tumours with metastases to SLNs only and those with non-SLN involvement. Classification of tumours with only one SLN metastasis and those with a multinodal involvement resulted in a smaller error rate, falsely classified as lesser nodal involvement. Patients with tumours <1.8 cm and metastatic to the sinuses of a single SLN had a low probability of non-SLN metastasis, and might be candidates for axillary sparing after a positive SLN biopsy. Conclusions: Further investigations are required to assess the validity of such predictive models for the identification of patients with no metastases beyond the SLN. Axillary treatment must remain the rule until predictive models of non-SLN involvement are fully validated.