Model for Identifying Individuals at Risk for Esophageal Adenocarcinoma

Andrew T Kunzmann, Aaron P Thrift, Chris R. Cardwell, Jesper Lagergren, Shaohua Xie, Brian T Johnston, Lesley A Anderson, John Busby, Una C. McMenamin, Andrew D Spence, Helen G Coleman

Research output: Contribution to journalArticlepeer-review

41 Citations (Scopus)
617 Downloads (Pure)


Background & Aims: The prognosis for most patients with esophageal adenocarcinoma (EAC) is poor because they present with advanced disease. Models developed to identify patients at risk for EAC and increase early detection have been developed based on data from case–control studies. We analyzed data from a prospective study to identify factors available to clinicians that identify individuals with a high absolute risk of EAC. Methods: We collected data from 355,034 individuals (all older than 50 years) without a prior history of cancer enrolled in the United Kingdom Biobank prospective cohort study from 2006 through 2010; clinical data were collected through September 2014. We identified demographic, lifestyle, and medical factors, measured at baseline, that associated with development of EAC within 5 years using logistic regression analysis. We used these data to create a model to identify individuals at risk for EAC. Model performance was assessed using area under the receiver operating characteristics curve (AUROC), sensitivity, and specificity analyses. Results: Within up to 5 years of follow up, 220 individuals developed EAC. Age, sex, smoking, body mass index, and history of esophageal conditions or treatments identified individuals who developed EAC (AUROC, 0.80; 95% CI, 0.77–0.82). We used these factors to develop a scoring system and identified a point cut off that 104,723 individuals (29.5%), including 170 of the 220 cases with EAC, were above. The scoring system identified individuals who developed EAC with 77.4% sensitivity and 70.5% specificity. The 5-year risk of EAC was 0.16% for individuals with scores above the threshold and 0.02% for individuals with scores below the threshold. Conclusion: We combined data on several well-established risk factors that are available to clinicians to develop a system to identify individuals with a higher absolute risk of EAC within 5 years. Studies are needed to evaluate the utility of these factors in a multi-stage, triaged, screening program.
Original languageEnglish
JournalClin. Gastroenterol. Hepatol.
Early online date17 Mar 2018
Publication statusEarly online date - 17 Mar 2018


Dive into the research topics of 'Model for Identifying Individuals at Risk for Esophageal Adenocarcinoma'. Together they form a unique fingerprint.

Cite this