Abstract Abstract #1018 Background: Sentinel lymph node (SLN) biopsy has become the standard of care for the evaluation of axillary metastasis in breast cancer. The accurate determination of metastasis in SLNs at the time of surgery is required in order to properly select those patients who should undergo axillary node dissection (AND).
 Methods: We performed a retrospective review of 402 breast cancer patients who underwent SLN biopsy with evaluation of 931 nodes by touch prep from June 2000 through January 2007. Results of touch prep diagnosis at the time of surgery were compared to final pathologic evaluation by permanent H&E. All false negative (FN) results were re-reviewed to determine possible reasons for errors in diagnosis.
 Results: At least one SLN was successfully identified in 379 patients. Of 68 patients with a true positive SLN, all underwent AND and in 46% the SLN was the only node in which metastases were identified. A total of 32 patients (8.44%) had at least one FN result. In 931 nodes evaluated, there were 38 (4.1%) FN results and no false positives. There were no significant differences between FN results in patients with invasive ductal cancer (26 of 284 patients, 4.3% nodes) compared to invasive lobular cancer (5 of 46 patients, 5.6% nodes). The remaining patients had DCIS or other primary diagnosis with two FN SLNs. Of the 38 SLNs with a FN result, 30 (79%) had only micromets on permanent H&E. Of the remaining 8 SLN with a FN touch prep, 3 of 703 SLNs (0.4%) occurred with invasive ductal cancer and 4 of 108 SLNs (3.7%) occurred with invasive lobular cancer, demonstrating a significant difference, p=0.01. In the 32 patients with a FN SLN result, 19 patients underwent completion AND and in 12 patients (63%), the SLNs were the only nodes with cancer. Examining the 19 patients with FN SLN who underwent completion AND, 14 patients had invasive ductal cancer with only 4 (29%) having cancer found in non-SLN compared to 5 patients with invasive lobular cancer with 3 (60%) having cancer in non-SLN. Touch prep evaluation of SLN in breast cancer had an overall sensitivity of 72%, specificity of 100%, positive predictive value of 100% and negative predictive value of 95.4%. The cost for touch prep is $32 less per node compared to frozen section or an average cost reduction of $74 per patient.
 Conclusion: The accuracy of touch prep for the evaluation of SLN in breast cancer compares favorably to the reported results for frozen section with a lower cost. In patients with a FN SLN result, the SLNs are likely to be the only nodes involved with metastases, especially if the SLN is found to have only micromets. However, patients with invasive lobular cancer were more likely to have a FN finding in the setting of marcomets and were more likely to have metastases in non-SLNs. Patients with invasive lobular cancer who have a negative SLN by touch prep should have the SLN evaluated by frozen section to confirm the diagnosis. The data also suggest that patients with invasive lobular cancer and a FN SLN by touch prep may benefit from AND. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 1018.