Abstract 5587: Personalized risk prediction tool for esophageal adenocarcinoma in Caucasian population

Abstract

Abstract Backgrounds: Esophageal adenocarcinoma (EAC) is a deadly disease with the greatest increasing incidence rate among all cancers in the past few decades in the U.S. We aimed to develop a powerful risk prediction tool for EAC to assist in the prevention and early detection of this disease. Methods: A large case control study consisting of 924 EAC patients with 925 frequency-matched healthy controls was used to identify potential risk factors for EAC by univariate logistic regression. Stepwise logistic regression was used to select significant predictors in the final multivariate model. Discriminatory accuracy was measured by generating the receiver operator curve (ROC) and calculating the area under the curve (AUC). We also calculated the 5-year and 10-year absolute risk of developing EAC for hypothetical individuals with varying risk factor combinations. A simplified risk score (RS) was developed for easier risk communications and disseminations. Results: BMI, physical activity, smoking, family history, dietary intake, exposure to exhaust, Barrett's esophagus, hernia, esophageal ulcer, and GERD were included in the final model. The discriminative ability is excellent AUC of 0.884. Individuals with the highest RS have a 99% chance to be EAC cases compared to 1.4% for those with the lowest RS. Conclusions: We developed a comprehensive risk prediction model for EAC with excellent model performance. This model has the potential to be utilized in the clinic and public health settings to provide evidence-based risk mitigation strategy and identify high-risk population for targeted screening or prevention trials and for early detection of EAC. Note: This abstract was not presented at the meeting. Citation Format: Xia Pu, Jaffer A. Ajani, Jian Gu, Xiangjun Gu, Yuanqing Ye, Xifeng Wu. Personalized risk prediction tool for esophageal adenocarcinoma in Caucasian population. [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 5587. doi:10.1158/1538-7445.AM2015-5587