Serum fibrosis index-based risk score predicts hepatocellular carcinoma in untreated patients with chronic hepatitis B
Lilian Yan Liang 1 2 3 , Hye Won Lee 4 , Vincent Wai-Sun Wong 1 2 3 , Terry Cheuk-Fung Yip 1 2 3 , Yee-Kit Tse 1 2 3 , Vicki Wing-Ki Hui 1 2 3 , Grace Chung-Yan Lui 2 3 , Henry Lik-Yuen Chan 1 2 3 , Grace Lai-Hung Wong 1 2 3
Affiliations
Affiliations
1
Institute of Digestive Disease, The Chinese University of Hong Kong, Hong Kong SAR.
2
Department of Medicine and Therapeutics, The Chinese University of Hong Kong, Hong Kong SAR.
3
Medical Data Analytic Centre (MDAC), The Chinese University of Hong Kong, Hong Kong SAR.
4
Department of Internal Medicine, Yonsei University College of Medicine, Seoul, South Korea.
PMID: 33631920 DOI: 10.3350/cmh.2020.0333
Abstract
Background/aim: Serum fibrosis scores comprised of common laboratory tests have high utility to assess severity of liver fibrosis. We aimed to derive and validate a hepatocellular carcinoma (HCC) risk score based on serum fibrosis scores to predict HCC in treatment-naïve chronic hepatitis B (CHB) patients.
Methods: 15,187 treatment-naïve adult CHB patients were identified to form the training cohort in this retrospective study. Individual fibrosis score was included to construct a new HCC prediction score. The score was externally validated in an independent treatment-naïve Korean CHB cohort.
Results: 180/15,187 patients (1.2%) in training cohort and 47/4,286 patients (1.1%) in validation cohort developed HCC during a mean follow-up of 52 and 50 months, respectively. The newly developed HCC risk score, Liang score, is composed of gender, age, HBV DNA, FIB-4 index, and ranges from 0 to 22. Area under the time-dependent receiver operating characteristic curve of Liang score was 0.79 (95% CI: 0.70-0.89). A cutoff value of 9 provided an extremely high negative predictive value of 99.9% and high sensitivity of 90.0% at 5 years in the validation cohort; which were higher than those of CU-HCC score, GAG-HCC score and REACH-B score. Patients with Liang score ≤9 had HCC incidence <0.2% per year in both training and validation cohorts, in whom HCC surveillance might be exempted.
Conclusion: A novel HCC risk score, Liang score, based on FIB-4 index, is applicable and accurate to identify treatment-naïve CHB patients with very low risk of HCC to be exempted from HCC surveillance.