Background: The identification of lymph node metastases in colorectal resection specimens is necessary for accurate tumor staging. However, routine lymph node dissection by the pathologist yields only a subset of nodes removed surgically and may not include those nodes most directly in the path of lymphatic drainage from the tumor. Intraoperative mapping of such sentinel lymph nodes (SLNs) has been reported in cases of melanoma and breast cancer. We applied a similar method to cases of colorectal carcinoma, with emphasis on the pathology of the SLNs.
Methods: Eighty-three consecutive patients with colorectal carcinoma were evaluated after intraoperative injection of 1 to 2 mL of 1% isosulfan blue dye (Lymphazurin) into the peritumoral subserosa. Blue-stained lymph nodes were suture-tagged by the surgeon within minutes of the injection for identification by the pathologist, and a standard resection was performed. Designated SLNs were sectioned at 10 levels through the block; a cytokeratin immunostain (AE1) was also obtained. To evaluate the possibility that increased detection of metastases in the SLN might be solely due to increased histologic sampling, all initially negative non-SLNs in the first 25 cases were sectioned also at 10 levels.
Results: Sentinel lymph nodes were identified intraoperatively in 82 (99%) of 83 patients and accounted for 152 (11.9%) of 1275 lymph nodes recovered, with an average of 1.9 SLNs per patient. A total of 99 positive lymph nodes (38 positive SLNs and 61 positive non-SLNs) were identified in 34 node-positive patients. The SLNs were the only site of metastasis in 17 patients (50%), while 14 patients (41%) had both positive SLNs and non-SLNs. Three patients (9%) had positive non-SLNs with negative SLNs, representing skip metastases. In patients with positive SLNs, 91 (19%) of 474 total lymph nodes and 53 (12%) of 436 non-SLNs were positive for metastasis. In patients with negative SLNs, 8 (1%) of 801 total lymph nodes and 8 (1.2%) of 687 non-SLNs were positive for metastasis. Multilevel sections of 330 initially negative non-SLNs in the first 25 patients yielded only 2 additional positive nodes (0. 6%). All patients with positive SLNs were correctly staged by a combination of 4 representative levels through the SLN(s) together with a single cytokeratin immunostain.
Conclusions: Intraoperative mapping of SLNs in colorectal carcinoma identifies lymph nodes likely to contain metastases. Focused pathologic evaluation of the 1 to 4 SLNs so identified can improve the accuracy of pathologic staging.