Evaluation of ethnic influence in the application of a hepatocellular carcinoma predictive model for chronic hepatitis C.

ÌÇÐÄ´«Ã½

Abstract

Currently, there is no well-established algorithm predicting HCC development in untreated HCV patients. We aimed to validate an algorithm (REVEAL-HCV: age, AST, ALT, HCV RNA, HCV genotype, and cirrhosis) developed in Taiwanese patients.We analyzed 1,381 (50.1% White, 14.7% Hispanic, 13.8% Asian of diverse origin, and 7.8% African-American) adult treatment-naïve HCV patients (no viral co-infection, no HCC within 6 months) at 4 U.S. and one Hong Kong centers (11/1994-10/2017).Compared to the non-Asian cohort, the Asian cohort had higher percentage of patients in the low-risk group (46.1% vs. 26.1%) and lower percentage in the high-risk group (12.0% vs. 20.3%, p<0.01). Overall, 5-year HCC incidence were 1.75%, 4.71%, and 24.4% for low, medium and high-risk patients, respectively (p<0.0001). For the overall cohort, AUROC for HCC prediction were 0.83 (95% CI: 0.72-0.93), 0.82 (95% CI: 0.75-0.88), and 0.84 (95% CI: 0.77-0.89) for 1-year, 3-year and 5-year HCC risk, respectively. There was slightly lower AUROC for Asian compared to the non-Asian cohort at 3 years (0.75 vs. 0.83) and 5 years (0.78 vs. 0.84), though this was not statistically significant. In multivariable analysis, we found male sex, presence of metabolic syndrome as well as the risk score categories to be independently associated with HCC but not ethnicity.The REVEAL-HCV risk score has good validity for both Asian and non-Asian populations. Further studies should consider additional factors such as sex, metabolic syndrome and treatment status. This article is protected by copyright. All rights reserved.

View details for

View details for