Comparison of Complete and Representative Frozen Section Sampling of Breast Cancer Sentinel Lymph Node
Weisheng Xu, Karen Kostroff, Tawfiqul Bhuiya. Hofstra North Shore-LIJ School of Medicine, Lake Success, NY; Hofstra North Shore-LIJ School of Medicine, New Hyde Park, NY
Background: Intraoperative examination of sentinel lymph nodes (SLN) is critical in surgical management of breast cancer. Positive SLN usually leads to axillary lymph node dissection. Various methodologies are employed in handling SLN during frozen section. In this study, we compared the diagnostic outcomes between freezing the entire lymph node and freezing a representative section of the node.
Design: 958 breast cancer cases with 2211 SLN were collected from two tertiary hospitals from 01/01/2008 to 12/31/2010, including 325 cases with 956 SLN (2.9/case) from hospital 1 (H1), and 633 cases with 1255 SLN (2.0/case) from hospital 2 (H2). In H1, a lymph node was sliced at 2 mm and entirely frozen; while in H2, half of a small node or one slice of a large node cut at 2 mm intervals was frozen. Both hospitals performed H&E and AE1/3 immunohistochemistry examinations on 3 levels of permanent sections for frozen section negative cases as final SLN analysis. The results were statistically analyzed with Chi square test.
Results: For H1, 110 of 956 (11.5%) SLN were positive in final analysis, including 91 SLN with macrometastasis (macromet), 9 with micrometastasis (micromet) and 10 with isolated tumor cells (ITC); 87 of the 110 positive SLN (79.1%) were positive on frozen (80 macromet, 6 micromet, 1 ITC). For H2, 186 of 1255 (14.8%) SLN were positive in final analysis (125 macromet, 26 micromet, 35 ITC); 116 of the 186 positive SLN (62.4%) were positive on frozen (114 macromet, 1 micromet, 1 ITC). The sensitivity, specificity and negative predictive value of SLN frozen evaluation in H1 were (%): for macromet 94.5/100/99.4, micromet 66.7/100/99.7, and ITC 10.0/100/99.1; and in H2, for macromet 94.4/100/99.4, micromet 3.8/100/98.0, and ITC 2.9/100/97.3. Chi square statistical analysis showed a significant difference in detection of micromet in SLN between the two protocols (p<0.001), but no significant difference for the detection of macromet (p=0.43) and ITC (p=0.33).
Conclusions: Frozen examination of the entire SLN detects significantly more SLN with micromet than representative sampling. Both methods are equally effective in detecting macromet, and inadequate in ITC detection. As the paradigm changes for the indications for an axillary dissection, the significance of a micromet found at frozen section will probably change. In the meantime, saving a patient a second surgery by complete evaluation of SLN at frozen section remains a desirable practice.
Monday, March 19, 2012 9:30 AM
Poster Session I Stowell-Orbison/Surgical Pathology/Autopsy Awards Poster Session # 23, Monday Morning