Prospective Evaluation of Factors Influencing Success Rates of Sentinel Node Biopsy in 814 Breast Cancer Patients

Publisher Website
Google Scholar
Abstract
Objective: This prospective multicenter study was performed to assess the reliability of sentinel lymph node (SLN) biopsy in breast cancer and to analyze factors potentially influencing success rates. Methods: In 21 departments, SLN biopsy and consecutive axillary lymph node dissection were performed in 814 breast cancer patients. The 80 surgeons involved were free in the choice of lymphography technique. The detection rate and the sensitivity, as well as the impact of lymphography technique, patient selection, technical procedure and learning curves, were evaluated. Results: The blue dye technique was used in 137 patients, radiocolloid in 169 patients, and combined blue dye/radiocolloid in 508 patients. The identification rate for the sentinel node was 83.9% for the entire group and showed a significant dependence on the lymphography technique (blue dye, 71.6%; radiocolloid, 78.8%; combined blue dye and radiocolloid, 89.6%). The overall sensitivity in detecting lymph node metastases was 91.3%. Immunostaining for cytoceratine revealed micrometastases in 19 (5.1%) of 374 patients in whom H/E staining was negative. The combined subdermal/peritumoral injection of the colloid showed a significantly higher identification rate than subdermal or peritumoral injection alone (96.8%, 84.6%, 78.6%; p < 0.001). There was also a significant higher detection rate in cases of SLN biopsy performed prior to lumpectomy, compared to SLN biopsy following lumpectomy (94.7% versus 82.8%; p < 0.001). Furthermore, there was a close correlation between the number of performed examinations and the detection rate. Conclusion: SLN mapping predicts the axillary lymph node status accurately. Learning curves and several technical features influence the detection rate significantly. However, the false negative rate was independent of experience and injection technique.

This publication has 27 references indexed in Scilit: