Can axillary dissection be avoided in patients with sentinel lymph node metastasis?

Background: Who should undergo a completion dissection following identification of a +sentinel lymph node (SLN) is controversial. Methods: The records of 1,133 patients who underwent SLN mapping were reviewed. The association between patient, tumor, and treatment characteristics and the presence of +SLNs and +nonSLNs was analyzed using two-way tables of frequency counts and Pearson χ² test. Possible predictors of +SLNs and +nonSLNs were analyzed using simple and multiple logistic regression, Results: One thousand one hundred forty-eight SLN procedures were performed. 367 procedures (32%) yielded +SLNs. For patients with a +SLN, on multiple logistic regression analysis LVSI, increasing numbers of +SLNs, decreasing numbers of negative SLNs, and increasing size of the largest SLN metastasis were statistically significantly associated with increased likelihood of nonSLN involvement. No subgroup was identified that did not have a significant rate of nonSLN involvement on completion axillary dissection, except those who had a large number of negative SLNs (≥3) and small size of the largest SLN metastasis (<10 mm). Conclusions: A definitive answer to the question of who needs a completion axillary dissection awaits the results of ongoing trials. In the interim, our data does not support eliminating dissection for any subgroup of patients with +SLNs.