Skip nodes: Importance in EUS nodal staging of gastroesophageal cancer

Introduction: There is currently no standardized method for evaluating nodal involvement in gastroesophageal cancer by EUS. If gastroesophageal cancer invades regional lymph nodes in a contiguous fashion, then the evaluation of regional lymph nodes by EUS can start in the area of the tumor and proceed until uninvolved nodes are identified. Alternatively, if gastroesophageal cancer is associated with skip nodes, a complete mapping of all regional nodes at risk is required for accurate staging. It is known that esophageal squamous cell carcinoma and distal gastric adenocarcinoma can be associated with skip nodes (20-25%). However, the prevalence of this phenomenon in the more common types of gastroesophageal cancers (esophageal adenocarcinoma and proximal gastric adenocarcinoma) is not known. Hypothesis: Skip nodes are found in esophageal adenocarcinoma and proximal gastric adenocarcinoma. Methods: A retrospective review of the surgical pathology from resected esophageal adenocarcinoma and proximal gastric adenocarcinoma from 7/94 to 7/96 was performed at our institution. The location of the primary tumor and the pattern of nodal involvement were recorded. The presence of skip nodes was confirmed only if: 1) the specimen contained an uninvolved (skip) lymph node group nested between two involved lymph node groups, 2) one of the involved lymph node groups was proximal to the skip lymph node, 3) one of the involved lymph node groups was distal to the skip lymph node, and 4) all three of the lymph node groups were located along the expected direction of lymphatic spread of the cancer from its origin toward the central lymph nodes. Results: 31 patients with esophageal or proximal gastric adenocarcinoma resection were identified. 11 patients were excluded from the evaluation (9 had no nodal involvement, 2 had inadequate surgical specimen). Of the 20 patients with an evaluable surgical specimen, 5 had esophageal adenocarcinoma and 15 had adenocarcinoma of the gastric cardia. The overall prevalence of skip nodes was 3/20 (15%) which included 1/5 patients with esophageal adenc-carcinoma (20%), and 2/15 patients with proximal gastric adenocarcinoma (13%) Conclusions: The prevalence of skip nodes in esophageal and proximal gastric adenocarcinoma is significant. Accurate EUS staging of lymph node status in this group of patients should routinely involve the mapping of all "at risk" regional lymph nodes.