Imprint cytology of axillary lymph nodes as an intraoperative diagnostic tool

Aims: Frozen section histology of lymph nodes intraoperatively is associated with a false negative rate of twenty-five per cent. Imprint cytology is reported to have greater accuracy than frozen section. Accurate intraoperative examination of lymph nodes could help to prevent some patients from having a second axillary operation in breast cancer. This study assesses the sensitivity of imprint cytology to lymph node metastases using two different staining techniques. Methods: Imprint cytology of 238 freshly excised axillary lymph nodes in 53 patients with stage T1-2, N0, M0 breast cancer was performed. Imprints were stained using toluidine blue dye and anti-pancytokeratin immunoglobulin to compare the two staining methods. A consultant histopathologist blinded to the routine histology results examined each set of imprints. A non-pathologist also examined each set of imprints to determine whether a technician could be used to screen slides in order to decrease the workload of the pathologist. Results: Sensitivity was 82% with toluidine blue and 36.4% with anti-pancytokeratin immunoglobulin. Positive predictive values were 100% and 57% respectively for toluidine blue and anti-pancytokeratin. The false-negative rate was 18% for toluidine blue and 63.6% for anti-pancytokeratin antibody when examined by a consultant histopathologist. Conclusions: Imprint cytology using toluidine blue is as sensitive as frozen section in the detection of axillary lymph node metastases. A positive result may be acted upon with acceptable safety. Immunohistochemistry using antibody to pancytokeratin did not improve imprint sensitivity in this study. Imprint cytology could help to prevent a second operation in 80% of breast cancer patients with positive lymph nodes.