Validation of ex vivo lymphatic mapping in hematoxylin-eosin node-negative carcinoma of the colon and rectum.

Background: Substantial evidence supports that detailed analysis of the regional lymphatics will identify previously unrecognized micrometastatic disease in colorectal cancer. In order to determine whether the sentinel lymph node(s) (SLNs) harvested by ex vivo lymphatic mapping in node-negative colorectal cancer (CRC) are the most likely node(s) to harbor micrometastatic disease, we examined all nodes in CRC specimens in an identical fashion.

Methods: One hundred twenty-four specimens from patients with colorectal cancer were delivered to pathology in the fresh state and underwent ex vivo sentinel lymph node mapping. If negative by routine hematoxylin and eosin (H&E) analysis, the SLNs and non-SLNs were subjected to further analysis by level section H&E and immunohistochemical (IHC) analysis.

Results: A mean of 30 nodes were harvested (range, 5-111). Fifty-one patients (41%) were found to be node-positive by routine H&E analysis. SLNs were identified in all but three specimens. A total of 2177 nodes were analyzed from the 66 H&E node-negative specimens (1883 non-SLNs and 294 SLNs). Overall, metastases were identified in 13 of 278 SLNs and in only 5 of 1829 non-SLNs (P <.001). Only 5 of 66 patients (7.5%) had evidence of metastatic disease in non-SLNs when the SLNs were negative. Thirteen apparently node-negative patients (19.3%) were upstaged by IHC analysis of the SLNs (P =.04).

Conclusions: If the SLN is negative by both H&E and IHC analysis, the probability of finding metastases in a non-SLN is remote. If microstaging is demonstrated to be prognostically relevant, focused examination should be of the SLN(s).