Purpose: Sentinel lymph node mapping accurately predicts nodal status in > 90 percent of melanoma and breast and colorectal cancers. However, because of anatomic differences, sentinel lymph node mapping of rectal cancers has been considered inaccurate and difficult relative to colon. A prospective study was undertaken to identify differences in sentinel lymph node mapping between patients with colon cancer and those with rectal cancer.
Methods: At operation 1 to 3 ml of 1 percent isosulfan blue dye was injected subserosally around colon cancers. The first to fourth blue-staining nodes seen within ten minutes of injection were marked as sentinel lymph nodes. For cancer of the mid-rectum to low rectum, the dye was injected submucosally via rigid scope and spinal needle. The mesorectum was dissected ex vivo to identify blue nodes nearest the tumor as sentinel lymph nodes. Multilevel microsections of sentinel lymph nodes were stained with hematoxylin and eosin and immunostained for cytokeratin, and standard examination of the entire specimen was performed.
Results: There were 407 consecutive patients (336 with colon and 71 rectum). The sentinel lymph nodes were identified in 99.1 percent of colon and 91.5 percent of rectal patients (P < 0.0001). Skip metastases were found in 3.6 percent of colon vs. 2.8 percent of rectal patients (P = 0.16). Occult micrometastases were found in 13.4 percent of colon vs. 7.0 percent of rectal patients (P = 0.24). Except for success rates, no other parameters were statistically different between colon and rectum. Lower success in sentinel lymph node identification in rectal cancer may have been related to neoadjuvant chemoradiation received in all six of the patients with sentinel lymph node mapping failures.
Conclusion: Despite higher success rates in sentinel lymph node identification for colon patients, sentinel lymph node mapping was highly successful (91.5 percent) in rectal patients. Nodal upstaging, skip metastases, and occult metastases were similar.