Histologic evaluation of sentinel and non-sentinel axillary lymph nodes in breast cancer by multilevel sectioning and predictors of non-sentinel metastasis.

Sentinel lymph node (SLN) provides accurate nodal staging for breast cancer. This technique has been introduced in Siriraj Hospital since 1998. The goal of this study is to assess its accuracy in predicting the state of the axilla, and compare the results of standard examination and multilevel sectioning. A retrospective analysis of 195 breast cancer patients who underwent both SLN biopsy (using dye alone as the lymphatic mapping) and axillary node dissection during 1998-2002 were analyzed. All slides including SLNs and the non-SLNs (NSLNs) were reviewed and multilevel study was performed on all SLNs and NSLNs [four levels of hematoxylin-eosin (HE) at 200 µm interval and keratin stains on the first and fourth levels]. Of 195 patients, 30% of cases were SLN-positive (32 NSLN-positive and 27 NSLN-negative). Additional study could detect positive axillary nodes 10.8% (4 SLN-positive and 5 NSLN-positive) more than standard HE stain. The false negative rate increased from 20.3% to 24.1%. The concordance between SLN and NSLN statuses was 89.7%. The sensitivity was 75.9%. By multivariate analysis, the significant predictors for axillary node metastasis were tumor size of more than 2.2 cm, histologic type of invasive ductal carcinoma (IDC), not otherwise specified (NOS) and lymphovascular invasion (LVI). By univariable analysis, the significant predictors of NSLN metastasis after positive-SLN were outer location of the tumor, LVI and perinodal extension. In conclusion, use of multilevel and immunohistochemistry increased detection of positive-SLNs. Caution should be kept in accepting SLN biopsy using peritumoral dye technique alone as the procedure for staging due to a high false-negative rate. The concordance rate of 89.7% confirmed the reliability of SLN. Outer location of tumor, LVI and perinodal extension is significant predictors of positive-NSLN after positive-SLN.