Analysis of Frozen Section Results of Sentinel Lymph Node in 896 Patients with Breast Cancer
A Kim, WJ Sung, YK Bae, SH Kang, SJ Lee. Yeungnam University College of Medicine, Daegu, Korea
Background: Intraoperative frozen section (FS) of sentinel lymph nodes (SLNs) is widely used to determine whether axillary dissection should be followed in breast cancer patients. However there have been no standard guidelines for the number of sections, cutting interval and the use of immnunohistochemistry. We evaluated the usefulness and limitations of FS protocol which has been used in our institution for intraoperative SLN examination.
Design: We analyzed FS results of 896 patients who underwent intraoperative SLN biopsy between January 2005 and December 2007. Our institutional protocol is as follows. SLNs larger than 5mm were sliced in 2mm interval and all slices from one SLN were embedded in one block. Then two serial sections were cut for hematoxylin and eosin (HE) stain without trimming away the embedded tissue. The rest of the frozen tissue was formalin-fixed and paraffin-embedded for permanent section (PS). Two serial PSs from each block were taken, one for HE and the other for cytokeratin (AE1/AE3) immunostain. Cytokeratin stain was done in the cases with negative FS results. Metastatic SLNs were graded according to the AJCC cancer staging manual (6th edition) as isolated tumor cells (ITCs), micrometastasis (>0.2 to 2mm) and macrometastasis (>2mm). The results of FS and PS were compared with regard to the pathologic diagnosis and the standard was based on the results of PS.
Results: Among 896 patients, 810 patients (90.4%) had invasive carcinoma and 86 (9.6%) patients had ductal carcinoma in situ. The average number of SLNs was 3 per patients. Total 240 (26.8%) patients were found to have metastatic SLN(s) in the final pathology report. In 34 patients, the FS diagnoses were negative but the PS diagnoses were positive (false negative rate, 3.8%). Another 10 (1.1%) cases which were negative on FS had ITCs on the cytokeratin immunostains. The rate of micrometastasis in the false negative cases was 85.3%. False negative results were caused by interpretation error (29.4%) or technical problems (no tumor cells visible on FS, but tumor cells visible on PS) (70.6%).
Conclusions: The false negative rate of our protocol for FS of SLN was very low. The failure of FS was largely caused by the failure to detect micrometastasis. Therefore, FS is a reliable method for intraoperative SLN examination if a very stringent protocol is used.
Tuesday, March 10, 2009 9:30 AM
Poster Session III # 65, Tuesday Morning