In comparison with FIB-4, NFS at baseline or PW12 was a more effective predictor of LRCs in all patients and those with MS. NFS at PW12 was also a predictor of HCC occurrence in patients with MS. NFS at PW12 had a higher time-dependent AUROC value for the prediction of LRCs compared with NFS at baseline in patients with MS, and time-dependent NFS AUROCs for predicting LRCs at baseline and PW12 were similar in all patients. Therefore, NFS at PW12 can be used to predict LRCs and HCC occurrence in CHC patients with an SVR to DAA therapy, especially in those with MS.
APRI and FIB-4 were initially developed to predict significant fibrosis in patients with HCV infection and with HIV and HCV coinfection, respectively. These noninvasive assessments have been used to predict HCC development and LRCs in patients with CHC. Ioannou et al. revealed that patients with FIB-4 ≥ 3.25 before DAA therapy had a higher incidence of HCC than those with FIB-4 < 3.25. López et al. demonstrated that FIB-4 at baseline and 1 year after EOT were independent factors associated with HCC in patients with CHC with an SVR. Chalouni et al. demonstrated that patients coinfected with HIV and HCV with a higher baseline FIB-4 (>4.00) showed a higher incidence of LRCs, liver-related death and overall mortality. In 2007, Angulo et al. developed NFS to identify advanced liver fibrosis in patients with NAFLD, and Ampuero et al. constructed HFS, which has higher accuracy than FIB-4 and NFS, and applied it for identifying advanced liver fibrosis in patients with NAFLD in a multinational multicentre study in 2020. However, a recent study from Mexico reported that HFS has a low positive predictive value (36.7%, 95% CI: 19.9–56.1). Notably, few studies have investigated the predictive role of NFS in patients with CHC. The present study demonstrated that NFS at baseline or at PW12 is a useful predictor of LRCs in patients with CHC with an SVR to DAA therapy.
Noninvasive fibrosis assessments, such as FIB-4, APRI and liver stiffness measurements at baseline, are widely used to predict HCC development and LRCs in patients with viral hepatitis.[10–12] Several studies have indicated that posttreatment noninvasive tests predict HCC development and LRCs more accurately than those performed at baseline;[11,25–27] noninvasive assessments at baseline may be confounded by hepatic necroinflammation.[28,29] We previously demonstrated that age, platelet count and AFP levels after 12 months of treatment (APA-B model) had a higher predictive accuracy of HCC development in patients with CHB receiving entecavir monotherapy than baseline measurements. We further demonstrated that FIB-4 values and AFP levels at 12 months, rather than those at baseline, were independent predictors of HCC development in patients with CHB beyond year 5 of entecavir therapy. In line with these observations, the present study revealed that the time-dependent AUROC value of NFS at PW12 for the prediction of LRCs was higher than NFS at baseline in patients with MS.
Chen et al. demonstrated that BMI and DM were independently associated with an increased risk of HCC in patients with treatment-naïve CHC, suggesting the synergistic roles of these metabolic factors with HCV in HCC development. Huang et al. revealed that patients with CHC having mild liver fibrosis and impaired fasting glucose, impaired glucose tolerance or subclinical DM diagnosed using an oral glucose tolerance test were at higher risk of HCC despite achieving SVR after pegylated interferon-based therapy. Yen et al. demonstrated that in addition to age, fibrosis stage, HCV genotype 1, steatosis and DM, either alone or in combination, were independent factors associated with HCC development in patients with CHC who had achieved SVR after pegylated interferon-based therapy. These findings suggest that the prevalent severity of liver fibrosis and MS-related steatosis is crucial to the future risk of HCC after HCV is eradicated. Age, AST-to-ALT ratio and platelet count are frequently used for predicting the severity of liver fibrosis, and these variables are also components of FIB-4.[19,35] At baseline or at follow-up, albumin was a predictor of HCC occurrence and LRCs in patients with CHC achieving SVR. NFS consists of these variables with proportionally weighted coefficients; thus, they can serve as predictors of LRCs and HCC occurrence in patients with CHC with an SVR to DAA therapy. Early cirrhosis may be difficult to diagnose clinically unless histological confirmation is attempted. To overcome the diagnostic uncertainty, we used NFS as a simple, objective and discriminatory method of fibrosis measurement for assessing LRCs in patients with chronic liver disease. NFS at PW12 AUROC had a higher value than LC (Figure S5). In a subgroup of patients with available APRI, FIB-4, NFS and HFS (n = 613), NFS at PW12 AUROC had the highest values in patients with and without LC (Figure S6A,B). Therefore, NFS at PW12 may be a more discriminatory marker for predicting LRCs compared to cirrhosis.
Host lipids, lipoproteins and apolipoproteins are essential in several steps of the HCV life cycle, including entry, replication and assembly. In turn, HCV infection impairs the very-low-density lipoprotein release pathway and stimulates the synthesis of cholesterol and fatty acids through the activation of steroid-responsive element-binding proteins. Therefore, HCV infection is associated with hepatic lipid accumulation and dyslipidemia.[6,7] Patients with CHC have a higher prevalence of insulin resistance and DM than healthy controls and a twofold higher incidence of hepatic steatosis than patients with CHB. Recent studies have revealed that insulin resistance decreases and glycaemic control improves following HCV eradication with DAA therapy in patients with or without DM.[36,37] Whether MS abates over time following the eradication of HCV and how changes in metabolic factors modulate the long-term risks of HCC and LRCs remain undetermined.
This study has several limitations. First, it was a single-centre retrospective study with only 590 patients. Second, the median follow-up duration was short (25.27 [17.77–43.26] months), with a proportion of patients with normal AST and ALT levels, platelet count, and lost to follow-up after achieving SVR. The follow-up period should be extended in future studies. Third, a significant proportion of patients without NFS data at baseline or PW12 was excluded. These patients were younger; had lower AST, ALT and FIB-4 values; were less likely to have LC; and had a shorter follow-up duration. This may have affected the estimated HR of the predictive factors in the Cox regression model. Finally, the number of patients without SVR was low, which precluded the assessment of the predictive role of NFS for LRCs and HCC development in patients with CHC who failed DAA therapy.
In conclusion, NFS at baseline or PW12 is a more effective predictor of LRCs compared with FIB-4 values in all patients and those with MS. NFS at PW12 had a higher time-dependent AUROC value for the prediction of LRCs compared with NFS at baseline. Therefore, NFS at PW12 may be a useful predictor of LRCs and HCC occurrence in patients with CHC with an SVR to DAA therapy, especially in those with MS.
This study was supported by a grant (No. DMR-111-030) from China Medical University Hospital in Taichung, Taiwan
AFP, α-fetoprotein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; APRI, aspartate aminotransferase-to-platelet ratio index; AUROC, area under the receiver operating characteristic curve; BP, blood pressure; CHC, chronic hepatitis C; CI, confidence interval; DAA, direct-acting antiviral agent; DLD, decompensated liver disease; DM, diabetes mellitus; EOT, end of therapy; FIB-4, fibrosis-4 index; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HFS, Hepamet fibrosis score; HIV, human immunodeficiency virus; HR, hazard ratio; LC, liver cirrhosis; LRC, liver-related complication; MS, metabolic syndrome; NAFLD, nonalcoholic fatty liver disease; NFS, nonalcoholic fatty liver disease fibrosis score; PW12, 3 or 6 months after DAA therapy; SVR, sustained virologic response.
We thank Yi-Chune Guo, Hung-Yu Kuo and Yi-Ting Lin for their assistance with data collection.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author with approval by the local Research Ethics Committee.
J Viral Hepat. 2022;29(9):785-79. © 2022 Blackwell Publishing