Intraoperative pathologic examination of the sentinel lymph node (SLN) drai
ning a primary breast carcinoma allows an SLN-positive patient to undergo c
omplete axillary lymphadenectomy as part of the same surgical procedure. Ho
wever the optimal technique for rapid SLN assessment has not been determine
d We reviewed our results with imprint cytology, (IC) and frozen section (F
S) examination of SLNs from 278 patients. Compared with H&E-stained paraffi
n sections, IC and FS had an overall accuracy of 93.2%. The false-reassuran
ce rate (false-negative results/all negative results) was 8.4%. It correctl
y identified 98% of macrometastases but only 28% of micrometastases. There
were no false-positive results. Compared with paraffin-section cytokeratin
immunohistochemistry results, the IC-FS false-reassurance rate increased to
25.8%. The false-reassurance rate decreased with smaller primary tumor siz
e (TI vs T2/3) and ductal type, smaller diameter of the SLN (less than or e
qual to 2.0 cm), and greater pathologist experience. IC combined with 2-lev
el FS reliably identifies SLN macrometastases but commonly fails to detect
SLN micrometastases. If SLN micrometastasis is used to determine the need f
or further Lymphadenectomy, more sensitive intraoperative methods will be n
eeded to avoid a second operation.