The absolute risk of carcinogenic progression given Barrett's esophagus (BE) is low and a majority of esophageal adenocarcinoma (EA) cases are diagnosed at first endoscopy without having had a prior diagnosis of BE. Therefore the research field has largely utilized case-control study designs with the implicit assumption that each case-group assessed (reflux, BE, EA) is representative of that stage in a linear progression towards cancer. To test the assumption that EA cases are a homogeneous group and therefore research findings may contribute to evidence-based prevention programs for diagnosed BE populations, we conducted a case-case study comparing the clinical presentation, medical history and survival of EA patients with and without a prior diagnosis of BE in the SEER-Medicare database. Eligible EA cases were diagnosed at ages > = 68 years during 1994-2009. Exposures were captured for a 2 year period starting 3 years prior to diagnosis. There were 5,271 EA patients in this study, 87% (n = 4,609) of which lacked a previous diagnosis of BE (“EA-no prior BE”). In univariate case-case comparisons, patient characteristics (age, sex, race, etc.) were mostly similar with the exceptions that EA-prior BE cases had lower comorbidity scores and a higher average number of physician visits. Conversely, tumor and treatment characteristics were starkly different; EA-prior BE cases were typically diagnosed with lower stage, lower grade, smaller tumors, and were less likely to have metastatic disease. EA-prior BE cases were also more likely to receive surgery and less likely to receive chemotherapy and radiotherapy. Multivariable case-case comparisons of exposures evidenced positive associations of GERD, ever-cigarette smoking, hypertension, dyslipidemia, weight-loss, peptic ulcer, pernicious anemia/celiac disease, and irritable bowel disease each in relation to EA-prior BE compared with EA-no prior BE. Obesity, metabolic syndrome, impaired fasting glucose, diabetes and depression did not differ between the case groups. EA-prior BE cases had better overall mean survival (2.5 vs. 1.4 years), particularly for localized disease (3.3 vs 2.3 years). Such a survival advantage persisted after adjustment for age at diagnosis, sex, race, education and registry (hazard ratio (HR) = 0.53, 95% confidence interval (CI): 0.48, 0.58). Further adjustment for stage, grade, tumor size, metastatic disease, surgery, chemotherapy, radiotherapy, & Charlson comorbidity score attenuated but did not diminish the association (HR = 0.70, 95%CI: 0.64, 0.78). In sum, our study provides evidence that EA cases occurring among patients previously diagnosed with BE are dramatically different from those that were not, differing in clinical presentation, medical history, and survival. These findings underscore the importance of a prudent approach when attempting to reconstruct the natural history of this malignancy with utility for cancer prevention.

Citation Format: Michael B. Cook, Jennifer Drahos, Shannon Wood, Lindsey Enewold, Ruth Parsons, Neal D. Freedman, Philip R. Taylor, Winnie Ricker, Christian C. Abnet. Pathogenesis and progression of esophageal adenocarcinoma by prior diagnosis of Barrett's esophagus. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 837. doi:10.1158/1538-7445.AM2015-837