Abstract Background: Accurate intraoperative diagnosis of sentinel lymph node (SLN) metastases enables the selection of patients for axillary lymph node dissection during the same operation and reduces the need for re-operation. Touch imprint cytology (TIC) serves as a main intraoperative assessment of SLNs in our institute for over five years. The purpose of the present study is to evaluate the clinical value of TIC as an intraoperative assessment for the diagnosis of SLN. Methods: Patients treated for early-stage breast cancer between Feb-2005 and May-2010 enrolled in the study. TIC was routinely performed intraoperatively, the result of which was correlated with definitive histological assessments of serial section with Hematoxylin-Eosin staining. Subsequent immunohistochemistry staining with CK-19 and MUC-1 were performed for research purposes. The Memorial Sloan-Kettering Cancer Center (MSKCC) nomogram was applied in this retrospective study to estimate probability of SLN involvement of each case. Results: A total of 1,077 patients with early-stage breast cancer enrolled in the study, and 3,048 SLNs were successfully harvested during the surgeries. Among those, 265 (24.6%) patients proved to have at least one SLN that was positive for carcinoma. Altogether, 397 (13.0%) involved nodes were removed from patients in the aforementioned patient pool. Based on the final pathology report, the sensitivity, specificity and overall accuracy of TIC was 83.4%, 99.0% and 95.2%, respectively on a per patient basis, and 78.3%, 99.4% and 96.7%, respectively on a per node basis. The sensitivity for macrometastasis and micrometastasis are 88.6% and 39.3%, respectively on a per patient basis, while 87.4% and 31. 3%, respectively on a per node basis. Of the patients included in this study, 98.7% had a positive SLN within their first three harvested SLNs. All the patients who were at a <10% chance of SLNs metastases according to the MSKCC nomogram were proved to be node negative by the final pathology. Conclusion: TIC is feasible and is able to detect macrometastasis in SLNs with an acceptable accuracy for clinical use in early-stage breast cancer patients while its ability to detect micrometastasis is limited. Limiting intraoperative TIC to the first three harvested SLNs in the diagnosis of SLN metastasis may make this diagnostic procedure significantly cheaper and easier for pathologists to perform. The MSKCC nomogram could be applied as a screening tool. Intraoperative assessment for SLNs, such as TIC, could be spared for patients with extremely low MSKCC scores (<10%). Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P1-01-07.