Preoperative sonographic classification of axillary lymph nodes in patients with breast cancer: node-to-node correlation with surgical histology and sentinel node biopsy results.

OBJECTIVE

The purpose of this study was to prospectively evaluate the role of axillary lymph node classification by sonography in breast cancer patients by node-to-node correlation with surgical histology and sentinel node biopsy results.

SUBJECTS AND METHODS

Between June 2006 and December 2006, preoperative axillary sonography was performed in 191 consecutive breast cancer patients (median age, 46 years; age range, 24-79 years) who had been scheduled to undergo breast cancer surgery with sentinel node biopsy. The axillary lymph node that had the thickest cortex or that was closest to the primary tumor was prospectively classified and then removed through sonographically guided needle localization. Correspondence about and histologic results for the needle-localized nodes and the radioactive sentinel nodes were analyzed. The rate of malignancy, according to the sonographic classification, and the area under a receiver operating characteristic curve were analyzed.

RESULTS

Of the 191 needle-localized nodes, 41 (21%) had metastases and 150 (79%) did not have metastases. When a cutoff point of a cortical thickness of 2.5 mm was used, sonographic classification showed 85% (35/41) sensitivity, 78% (117/150) specificity, and an area under the curve of 0.861 (95% CI, 0.796-0.926). Of the 54 patients with metastases at sentinel node biopsy or axillary lymph node dissection, 13 (24%) had false-negative results of sonographically guided needle localization. Unsuccessful lymphatic mapping because of absent radiotracer uptake during sentinel node biopsy was found in 4% (7/191), whereas all needle-localized nodes with a cortical thickness of more than 2.5 mm were confirmed as metastases.

CONCLUSION

Sonographic classification of axillary lymph nodes is effective for predicting the presence of metastases to avoid sentinel node biopsy or to reduce unsuccessful lymphatic mapping during sentinel node biopsy.