Prognostic Value of Controlled Attenuation Parameter by Transient Elastography

• Published in: The American journal of gastroenterology Vol. 112; no. 12; pp. 1812 - 1823
• Main Authors: Liu, Ken, Wong, Vincent Wai-Sun, Lau, Keith, Liu, Sienna Du, Tse, Yee-Kit, Yip, Terry Cheuk-Fung, Kwok, Raymond, Chan, Alex Yiu-Wa, Chan, Henry Lik-Yuen, Wong, Grace Lai-Hung
• Format: Journal Article
• Language: English
• Published: United States Wolters Kluwer Health Medical Research, Lippincott Williams & Wilkins 01.12.2017
• Subjects: Cancer; Cardiovascular Diseases - etiology; Cohort Studies; Elasticity Imaging Techniques; Fatty Liver - complications; Fatty Liver - diagnostic imaging; Female; Gastroenterology; Hepatitis; Humans; Liver cancer; Liver diseases; Liver Neoplasms - etiology; Male; Middle Aged; Multivariate analysis; Prognosis; Risk Factors; Sensitivity and Specificity
• DOI: 10.1038/ajg.2017.389
• ISSN: 0002-9270; 1572-0241

Liver stiffness measurement (LSM) by transient elastography (TE) has been shown to predict outcomes in patients with liver disease. While controlled attenuation parameter (CAP) measurement can accurately quantify hepatic steatosis, its prognostic value is unknown. We aim to determine if CAP is predictive for liver-related events (LRE), non-hepatocellular carcinoma (HCC) cancers, and cardiovascular events (CVE). Consecutive patients with both a reliable LSM and ≥10 successful CAP measurements by TE from August 2012 to March 2016 were included in the analysis. LRE were defined as HCC or hepatic decompensation. CVE were defined as acute coronary syndrome (ACS), cerebrovascular accident (CVA), or coronary intervention (stenting or bypass). Of the 5,848 patients that were examined, 4,282 (56.7% male, median age 57 years) had adequate follow-up, reliable LSM (median 6.1 kPa), and ≥10 CAP measurements (median 250 dB/m). Indications for TE were: suspected non-alcoholic fatty liver disease (NAFLD) (40.7%), hepatitis B (HBV) (37.0%), hepatitis C (2.9%), and others (19.4%). During 8,540 patient-years of follow-up, there were 45 patients with LRE (34 HCC, 33 decompensations), 73 with newly diagnosed non-HCC cancers, and 65 with CVE (27 ACS, 25 CVA, and 35 coronary interventions). CAP did not predict LRE, non-HCC cancer, or CVE on univariate analysis. On multivariate analysis, LSM, male sex, platelet count, serum albumin, and HBV etiology independently predicted LRE; age was the only independent predictor of non-HCC cancer; while age, fasting blood glucose, total cholesterol, and creatinine predicted for CVE. Subgroup analyses of viral hepatitis and NAFLD patients revealed similar results. Neither the presence nor the severity of hepatic steatosis as measured by CAP predict LRE, cancer, or CVE in the short term.