Intraoperative evaluation of axillary sentinel lymph nodes using touch imprint cytology and immunohistochemistry. Part II. Results.

AIM In order to operate selectively on positive axillae during the initial operative session for early breast cancer, an accurate and rapid intraoperative method to examine an axillary node sample (ANS) or a sentinel node biopsy (SNB) is required. The aim of this study was to determine the feasibility and accuracy of Immunohistochemistry (IHC)-stained touch imprints in detecting metastatic axillary nodes intraoperatively. MATERIAL AND METHODS Four hundred and thirty-two axillary nodes from 52 patients (23 axillary node clearance (ANC), 15 ANS and 14 SNB) were bisected, imprinted and stained with anti-cytokeratin 19 IHC. Results were compared with those of routine haematoxylin and eosin (H&E)-stained sections. RESULTS IHC imprints detected 32 positive nodes from 12 patients. H&E sections detected 31 positive nodes from 11 patients. IHC imprints missed metastases in three nodes and missed the diagnosis of positive axillae in two patients. H&E missed metastases in four nodes and missed the diagnosis of positive axillae in 3 patients. On a node-basis, sensitivities were 91.4 and 88.5%, negative predictive values (NPV) were 99.2 and 99.0% and overall accuracies were 99.3 and 99.1% for IHC imprints and H&E sections, respectively. On a patient-basis, sensitivities were 85.7 and 78.5%, NPVs were 95.2 and 93.1% and overall accuracies were 96.1 and 94.2% for IHC imprints and H&E sections respectively. There were no false positives. Interpretation of the results by a non-histopathologist was concordant with that of a histopathologist. Results might be obtained within 30-45 min depending on the number of examined nodes. CONCLUSION Intraoperative IHC staining of touch imprints of axillary sentinel nodes is feasible and is a reliable method for evaluating axillary nodes. Slides can be reliably interpreted by a trained non-histopathologist.