Patterns of care and treatment outcomes of patients with stage I esophageal cancer: A National Cancer Database analysis.
4035 Background: The aim of this study was to examine current patterns of care and associated outcomes for patients with stage I esophageal cancer (EC) treated in the United States. Methods: The National Cancer Data Base (NCDB) was queried for patients diagnosed with clinical stage T1-2N0 EC from 2004-2012. Patients were categorized into four treatment groups: observation without definitive therapy (Obs), chemoradiotherapy (CRT), local excision (LE), and esophagectomy (Eso). Patient, tumor, and treatment parameters were compared between groups. Kaplan-Meier 5-year overall survival (OS) estimates, postoperative 30- and 90-day mortality comparisons, and multivariate Cox proportional hazards modeling are reported. Results: A total of 5,460 patients met the criteria. Of these, 21% were observed, 14% underwent CRT, 23% LE, and 42% Eso. Median age and follow up were 67 years and 28 months, respectively. Eso was the primary treatment for patients of age ≤ 80 while 48% of patients age > 80 were observed. Age, race, comorbidity score, tumor location within the esophagus, type of medical insurance, median income, type of facility (academic vs. non-academic), and distance from treating facility were significant factors for predicting receipt of local therapy over observation. Postoperative 30-day mortality between the LE and Eso groups was 0.5% and 2.9%, respectively ( P< .001), which increased to 1.4% and 5.5% at 90 days ( P< .001). Five-year OS was 21% for Obs, 26% CRT, 64% LE, and 63% Eso ( P < .001). Multivariate analyses demonstrated improved OS with any form of local definitive therapy: CRT ( HR: 0.54, 95% CI [0.48 - 0.61], P< .001), LE ( HR: 0.24, [0.20 - 0.27], P< .001), Eso (HR: 0.31, [0.28 - 0.35], P< .001). Age, comorbidity score, facility type, distance, median income quartile, and insurance status were also independently associated with OS. Conclusions: Management of stage I EC is influenced by several demographic and socioeconomic factors. Clinical observation yields suboptimal outcomes compared to any local therapy, and a surgical approach should be considered over CRT whenever feasible.