Background: Sentinel lymph node (SLN) dissection has been investigated after neoadjuvant chemotherapy and has shown mixed results. Our objective was to evaluate SLN dissection in node-positive patients and to determine whether postchemotherapy ultrasound could select patients for this technique.
Methods: Between 1994 and 2010, 150 patients with biopsy proven axillary metastasis underwent SLN dissection after chemotherapy and 121 underwent axillary lymph node dissection (ALND). Clinicopathologic characteristics were analyzed before and after chemotherapy. Statistical analyses included Fisher's exact test for nodal response and multivariate logistic regression for factors associated with false-negative events.
Results: Median age was 52 years. Median tumor size at presentation was 2 cm. The SLN was identified in 93 % (139/150). In 111 patients in whom a SLN was identified and ALND performed, 15 patients had a false-negative SLN (20.8 %). In the 52 patients with normalized nodes on ultrasound, the false-negative rate decreased to 16.1 %. Multivariate analysis revealed smaller initial tumor size and fewer SLNs removed (<2) were associated with a false-negative SLN. There were 63 (42 %) patients with a pathologic complete response (pCR) in the nodes. Of those with normalized nodes on ultrasound, 38 (51 %) of 75 had a pCR. Only 25 (33 %) of 75 with persistent suspicious/malignant-appearing nodes had a pCR (p = 0.047).
Conclusions: Approximately 42 % of patients have a pCR in the nodes after chemotherapy. Normalized morphology on ultrasound correlates with a higher pCR rate. SLN dissection in these patients is associated with a false-negative rate of 20.8 %. Removing fewer than two SLNs is associated with a higher false-negative rate.