Vikram Kotwal1, Chimezie Mbachi2, Yuchen Wang3, Bashar Attar3, Tejinder Randhawa2, Estefania Flores2, Julian Robles4, Craig Rosenstengle4, Melchor Demetria3, Oluwatoyin Adeyemi5, Gregory Huhn5, Arvind R Murali6. 1. Division of Gastroenterology and Hepatology, Cook County Health and Hospitals System, 1950 West Polk Street, 6th Floor, Chicago, IL, USA. drvikramkotwal@gmail.com. 2. Department of Medicine, Cook County Health and Hospitals System, Chicago, IL, USA. 3. Division of Gastroenterology and Hepatology, Cook County Health and Hospitals System, 1950 West Polk Street, 6th Floor, Chicago, IL, USA. 4. Department of Medicine, University of Iowa Hospitals and Clinics, Iowa, USA. 5. Division of Infectious Disease, Cook County Health and Hospitals System, Chicago, IL, USA. 6. Division of Gastroenterology and Hepatology, University of Iowa Hospitals and Clinics, Iowa, USA.
Abstract
BACKGROUND AND AIMS: Several criteria have been described to noninvasively predict the presence of high-risk esophageal varices in patients with compensated advanced chronic liver disease (cACLD). However, a recent study showed that treatment with β blockers could increase decompensation-free survival in patients with clinically significant portal hypertension, thereby making it important to predict the presence of any esophageal varices. We aimed to develop a simple scoring system to predict any esophageal varices. METHODS: We retrospectively reviewed patients who had vibration-controlled transient elastography (VCTE) at Cook County Hospital, Chicago, USA. Patients with cACLD and liver stiffness measurement (LSM) ≥ 10 kPa with esophagogastroduodenoscopy performed within one year of VCTE were analyzed. We generated a novel score to predict esophageal varices, using the beta coefficient of predictive variables. The score was validated in an external cohort at the University of Iowa Hospital, USA. RESULTS: There were 372 patients in the development cohort and 200 patients in the validation cohort. LSM, platelet count, and albumin were identified as predictors of esophageal varices and were included for generating the Cook County score as "platelet count * - 0.0155872 + VCTE score * 0.0387052 + albumin * - 0.8549209." The area under receiver operating curve for our score was 0.86 for any varices and 0.85 for high risk varices and avoided more endoscopies than the expanded Baveno VI criteria while maintaining a very low miss rate (negative predictive value > 99%). CONCLUSION: We propose a new, highly accurate, and easy-to-use scoring system to predict the presence of not only high-risk but any esophageal varices in patients with cACLD.
BACKGROUND AND AIMS: Several criteria have been described to noninvasively predict the presence of high-risk esophageal varices in patients with compensated advanced chronic liver disease (cACLD). However, a recent study showed that treatment with β blockers could increase decompensation-free survival in patients with clinically significant portal hypertension, thereby making it important to predict the presence of any esophageal varices. We aimed to develop a simple scoring system to predict any esophageal varices. METHODS: We retrospectively reviewed patients who had vibration-controlled transient elastography (VCTE) at Cook County Hospital, Chicago, USA. Patients with cACLD and liver stiffness measurement (LSM) ≥ 10 kPa with esophagogastroduodenoscopy performed within one year of VCTE were analyzed. We generated a novel score to predict esophageal varices, using the beta coefficient of predictive variables. The score was validated in an external cohort at the University of Iowa Hospital, USA. RESULTS: There were 372 patients in the development cohort and 200 patients in the validation cohort. LSM, platelet count, and albumin were identified as predictors of esophageal varices and were included for generating the Cook County score as "platelet count * - 0.0155872 + VCTE score * 0.0387052 + albumin * - 0.8549209." The area under receiver operating curve for our score was 0.86 for any varices and 0.85 for high risk varices and avoided more endoscopies than the expanded Baveno VI criteria while maintaining a very low miss rate (negative predictive value > 99%). CONCLUSION: We propose a new, highly accurate, and easy-to-use scoring system to predict the presence of not only high-risk but any esophageal varices in patients with cACLD.