Intraoperative cytologic examination of breast sentinel lymph nodes: test utility and patient impact

The sentinel lymph node (SLN) procedure is a method for ascertaining the axillary lymph node status in patients with breast cancer. Intraoperative examination of the SLN may be important, because a positive result directs surgery to a complete axillary lymph node dissection. Intraoperative cytologic examination (IOCE) is a method of intraoperative evaluation, although little data are available regarding the sensitivity of the method with respect to tumor size and the size of the SLN metastasis. All SLN cases for the years 1997-2002 at Magee-Womens Hospital were tabulated for primary breast carcinoma size, IOCE result, final histologic result, and size of the SLN metastasis. All SLNs had IOCE with touch imprints. Scrape SLN preparations and frozen sections were strongly discouraged. There were 748 SLN cases comprising 1576 SLNs that had IOCE, and there were 247 true positive SLN cases comprising 522 SLNs. Of the 247 true positive SLN cases, 111 had a positive IOCE (111/247; 45% sensitivity overall) and there were 136 false negatives. Of the 247 cases, 164 were SLN micrometastases < or =2.0 mm in size, and 44 (27%) of these were detected by IOCE, while the remaining 120 cases were false negative. Of the 83 SLN macrometastases (>2.0 mm), 66 (80%) were detected by IOCE, with 17 false negatives. In this series, 15 cases (2%) were given the IOCE diagnosis of atypical/defer, and all of these permanent sections were histologically positive. There were five IOCE-positive cases that were histologically negative. Of the 164 SLNs with micrometastases < or =2.0 mm, 17.6% (29/164) were < or =0.5 mm (6/29 [21.4%] were IOCE positive), 5.5% (9/164) were 0.51-1.0 mm (3/9 [33%] were IOCE positive), and 3.6% (6/164) were 1.1-2.0 mm (2/6 [33%] were IOCE positive). There were 83 SLNs with macrometastases larger than 2.0 mm, and 66/83 (80%) were detected by IOCE. In this group, 22% (18/83) were 2.1-5.0 mm (8/18 [44.4%] were IOCE positive) and 57.8% (48/83) were larger than 5.0 mm (41/48 [85%] were IOCE positive). The mean primary breast tumor size was 15.4 mm, with a mean SLN tumor size of 1.4 mm. There was a significant correlation with tumor size and the presence of SLN metastasis, and a significant correlation with tumor size and size of the SLN metastasis. There was a significant t correlation of primary tumor size and positive IOCE, with the group of negative IOCE cases having a mean tumor size of 14 mm and the positive IOCE group having a mean tumor size of 22 mm. The overall sensitivity of the method was 45%, specificity 99%, positive predictive value 0.99, and negative predictive value 0.80. Sensitivity of the IOCE procedure based on SLN tumor size is as follows: < or =0.5 mm, 21.4%; 0.51-1.0 mm, 33%; 1.1-2.0 mm, 33%; 2.1-5.0 mm, 44.4%; and >5.0 mm, 85%. Primary tumor size correlates with a positive SLN status and size of the SLN metastasis. Most false-negative IOCEs are due to micrometastases. Positive IOCE cases had a significantly larger SLN metastasis size (mean 8.0 mm) than the false-negative IOCE group (mean 1.4 mm). The IOCE of SLNs has a high negative predictive value, but this is a poor test for the detection of micrometastases, as this group accounts for the majority of false-negative IOCEs of breast SLNs.