Abstract
BACKGROUND: To determine accuracy of clinical staging of T2N0 esophageal cancer from the National Cancer Database (NCDB). METHODS: The NCDB was accessed to identify patients with T2N0M0 esophageal cancer (adenocarcinoma or squamous cell carcinoma) treated between 2004-2013 that underwent esophagectomy. Pathologic staging was compared to clinical stage. Univariate (UVA) and multivariate analysis (MVA) was performed to identify factors related to pathologic upstaging using Cox proportional hazard ratio. RESULTS: We identified 1,840 patients with T2N0 esophageal cancer who underwent esophagectomy as first line therapy. The median age was 67 years. The vast majority of patients were male and had distal adenocarcinomas. Clinical staging in was accurate pathologically in 30.7% of patients. While there was a trend for worse accuracy with increasing year of diagnosis, there rate of pT0-2N0 was stable. Tumor length >3 cm was significantly associated with tumor upstaging, while poor differentiation was significantly associated with nodal upstaging. UVA and MVA identified younger age, tumor length >3 cm, and poor differentiation were significantly associated with overall upstaging. Gender, tumor location, and tumor histology were not prognostic. CONCLUSIONS: Clinical staging for T2N0M0 esophageal cancer continues to remain highly inaccurate, however, rates of pT0-2N0 have steadily remained over 50%. Tumor length >3 cm and poor differentiation are strongly associated with pathologic upstaging.