On the Edge of Endoscopic Therapy: Risk of Lymph Node Involvement in Early Adenocarcinoma of the Esophagus and Cardia

Introduction: Endoscopic resection (ER) of high-grade dysplasia or mucosal adenocarcinoma in the esophagus or cardia is a valid alternative to surgery, given the minimal risk for lymph node involvement (N+). However, cancer infiltrating the muscularis mucosa with a poor differentiation grade or with vaso-invasion (M3+), and submucosal cancers (SM) are considered only relative indications for ER. Aim: To evaluate the risk of N+ in patients with M3+ or SM disease diagnosed in ER specimens of adenocarcinoma of the esophagus or cardia. Methods: Patients undergoing diagnostic ER of adenocarcinoma of the esophagus or cardia between Jan '00-Mar '08 in 2 Dutch centers, were included if the ER-specimen showed M3+ (i.e. cancer infiltrating the muscularis mucosa with either poor differentiation or vaso-invasion) or SM disease. SM infiltration depth was classified as SM1 (<0.5 mm or upper third) or SM2/3 (>0.5 mm or beyond upper third). Exclusion criteria were chemo- or radiotherapy and irradical deep resection margins. The presence of N+ disease was evaluated in surgical resection specimens in the case of an esophagectomy, or during endoscopic follow-up in the case of endoscopic therapy. Results: Diagnostic ER showed M3+ or SM invasion in 49 patients of which 5 were excluded because of chemo- or radiotherapy. 15/44 patients had vertical resection margins positive for cancer. 29 patients were included (24 M, median 68 yrs, 27 esophageal adenocarcinoma). For the purpose of this study, all ER specimens were revised by an expert GI pathologist: M3+ (n=5), SM1 (n=12), SM2/3 (n=12). All M3+ tumors were poorly differentiated and 3 showed vaso-invasion. SM1 tumors had a median infiltration depth of 0.2 mm (IQR 0.1-0.4), 3/12 were poorly differentiated and 1/12 showed vaso-invasion. SM2 tumors had a median infiltration depth of 0.8 mm (IQR 0.8-1.0), 5/12 were poorly differentiated and 1/12 showed vaso-invasion. After the diagnostic ER 8 patients underwent surgery and 21 patients were managed endoscopically. No N+ was found in a total of 154 lymph nodes in the esophagectomy specimens after diagnostic ER. None of the endoscopically treated patients were diagnosed with N+ during a median follow-up of 12 months (IQR 7-28). There was 1 tumor unrelated death. Three patients developed a local recurrence. One patient was managed endoscopically, one was referred for surgery and one was managed expectative due to co-morbidity. Conclusion: In this study, no N+ was found for M3+ and SM disease. Further evaluation is necessary to assess if endoscopic management may be a valid alternative to surgery in patients with M3+ and SM disease diagnosed with ER.