Sentinel lymph node evaluation is technically feasible and may improve staging in colorectal cancer

Abstract

PURPOSE: The accurate nodal staging of colorectal cancer (CRC) is important to identify those patients who may benefit from adjuvant chemotherapy. Some have suggested that identification of sentinel lymph nodes (SLN) may improve staging in CRC. We sought to determine: the feasibility of identifying SLN in CRC utilizing isosulfan blue dye; the accuracy of the identified SLN in predicting the status of the remainder of the lymph nodes in CRC; and whether a more thorough evaluation of SLN with serial step sectioning and immunohistochemistry would more accurately stage patients with CRC. METHODS: A pilot trial was initiated at Walter Reed Army Medical Center (WRAMC), and 17 patients with masses on colonoscopy and subsequent tissue diagnosis of CRC were enrolled. Patients underwent standard surgical resection of their CRC with wedge of mesentery containing draining lymph nodes. Isosulfan blue dye was injected around the tumor subserosally/submucosally before dividing the mesenteric portion of the resection (n = 7) or ex vivo (n = 10). Sentinel lymph nodes were defined as all nodes staining blue and were dissected from the mesentery in the operating room. The SLN were sent separately for standard bivalving and hematoxylin and eosin staining (H&E) followed by serial step sectioning and immunohistochemistry (IHC) staining for cytokeratin. RESULTS: Seventeen patients (6 men, 11 women) were enrolled. The average preoperative carcinoembryonic antigen (CEA) was 5.9 (range, 1.2 to 18.9), and the average postoperative CEA was 2.8 (range, 0.7 to 9.1). One patient had a T1 tumor, 6 patients had T2 tumors, and 10 patients had T3 tumors on final pathology. Five cancers were well differentiated, 11 were moderately differentiated, and 1 was poorly differentiated. In all 17 cases, SLN were identified. A mean of 5.5 SLN was found per specimen (range, 2 to 11) with no difference noted between injection techniques (in vivo vs ex vivo). An additional 12 nonsentinel lymph nodes (range, 1 to 29) were identified per specimen. Ten patients had negative SLN and non-SLN. Seven patients were found to have positive SLN (3 by H&E, 2 by serial step sectioning, and 2 by IHC only). CONCLUSIONS: The isosulfan blue technique is technically feasible to allow identification of sentinel lymph nodes. In this study, no false-negative SLN occurred. A total of 7 patients had positive SLN; more importantly, 4 patients were upstaged as a result of serial step sectioning and immunohistochemistry staining. We hypothesize that this method may help pathologists find appropriate lymph nodes for more detailed analysis. As a result, patients may be more accurately staged and counseled for adjuvant chemotherapy, which has been shown to improve survival in node-positive CRC. Further studies should be undertaken to test these preliminary findings.