Prediction of sentinel lymph node-only disease in women with invasive breast cancer.

Background: The purpose of this study was to create a model that predicts which breast cancer patients will have sentinel lymph node (SLN)-only metastasis.

Methods: SLN-positive breast cancer patients (N = 1,253) were analyzed. Multivariate analysis was performed to identify factors predicting SLN-only disease; a prediction rule was created.

Results: Median tumor size was 2 cm. The median number of SLNs removed was 2; the median number of positive SLNs was 1. Multivariate analysis found tumor size, number of positive SLN, and proportion of SLN positive were significant predictors of SLN-only disease (P < .001). A prediction rule with 1 point being given for >1 positive SLN, 1 point for >50% of SLN positive, and up to 4 points for tumor size (T1a = 1, T1b or T1c = 2, T2 = 3, and T3 = 4) was established. Ninety-five percent of patients with 1 point had SLN-only disease (P < .0001).

Conclusion: An integer-based model may predict SLN-only disease and may be useful in determining whether completion axillary lymph node dissection is required.