Prognostic significance of positive lymph node ratio in resected esophageal cancer.

100 Background: Nodal involvement in esophageal cancer is associated with poor survival. We aim to determine whether the ratio of metastatic to examined lymph nodes (the lymph node ratio [LNR]) is a better predictor of survival as compared to the number of positive lymph nodes in resected esophageal cancer. Methods: 1,149 patients with resected esophageal cancer from 1995 to 2009 were identified from the VA Central Cancer Registry (VACCR) database. The patients were further characterized to 3 lymph node quartiles based on LNR and their median survivals were calculated using the Kaplan-Meier method. Results: Out of 1149 patients 26.4% patients (303) had squamous cell carcinoma and 73.6% (846) were of adenocarcinoma histology. Median age of diagnosis is 63 years. 353 (31%) are stage 1, 384 (33%) are stage 2, and 412 (36%) are stage 3. Majority of them 71% arise in lower third of esophagus followed by 13% in middle third, 4% in upper third and 12 % had unknown site of origin. The group was subdivided into 3 quartiles with 62.7% in LNR1 (0.0-0.1), 25.6% in LNR2 (0.1-0.5) and 11.7% in LNR3 (0.5-1.0). 13.7% had less than 2 nodes removed, 29.3% had 3-6 nodes and 57% had >7 nodes examined. 28% of them had tumor invading sub mucosa, 23.5% had tumor invading muscularis mucosa, 43.2% had involvement of adventitia and 5.3% had penetrating tumor at the time of diagnosis The 5 year survivals based on number of lymph nodes examined, number of positive lymph nodes and positive lymph node ratio are listed in the table. The median overall survival for resected esophageal cancer based on LNR quartiles was 37 vs 14 vs 11.5 months (p<0.0001). Conclusions: Number of positive lymph nodes and positive lymph node ratio correlated with survival outcomes but number of lymph nodes retrieved did not predict any survival differences. However LNR was a better predictor of survival when compared to number of positive nodes. Further validation of this observation needs to done in large multicenter studies. [Table: see text] No significant financial relationships to disclose.