Systematic review: Accuracy of the enhanced liver fibrosis test for diagnosing advanced liver fibrosis and cirrhosis

Enhanced liver fibrosis (ELF) score predicts hepatic decompensation and mortality

Background & Aims

In community pathways for detection of liver disease the most common reason for referral is fibrosis assessment. We investigated the impact of adding the Enhanced Liver Fibrosis (ELF) score as a second-line test (subsequent to an indeterminate or high Fibrosis-4 index [FIB-4] and/or non-alcoholic fatty liver disease fibrosis score) to guide referral and prognostication in our multi-aetiology pathway.

Methods

Patients with ELF results from the intelligent Liver Function Testing (iLFT) pathway were recruited. Case note review was undertaken to compare ELF with endpoints of cirrhosis, hepatic decompensation, and mortality (liver-related and all-cause death).

Results

In total, 1,327 individuals were included with a median follow-up of 859 days and median ELF score of 10.2. Overall sensitivity for cirrhosis at the 9.8 threshold was 94% (100% for metabolic-associated steatotic liver disease, 89% for alcohol-related liver disease). Determination of the ELF score as a second-line test reduced the referral rate by 34%. ELF scores predicted hepatic outcomes; each unit change was associated with increased decompensation (adjusted Hazard Ratio [aHR] 2.215, 95% CI: 1.934-2.537) and liver-related mortality (aHR 2.024, 95% CI: 1.674-2.446). ELF outperformed FIB-4 for risk of liver-related mortality, particularly in the short-term (area under the curve [AUC] 94.3% vs. 82.8% at six months). Where FIB-4 was indeterminate, ELF had higher AUC for all outcomes within at least 2 years. ELF ≥13 was associated with particularly high rates of decompensation (26% within 90 days) and all-cause mortality (38% at 1 year).

Conclusions

The addition of ELF reduced the number of individuals referred for fibrosis assessment following iLFT pathway testing and provided useful prognostic information. Individuals with ELF scores ≥13 were considered at high-risk of negative outcomes warranting urgent clinical assessment.

Impact and implications

Primary care pathways for suspected liver disease are increasingly common and often lead to increased specialist hepatology referrals for fibrosis assessment. This study, using clinical follow-up for liver-related outcomes, provides further evidence supporting ELF testing to safely reduce referrals in a two-step approach when combined with other simple fibrosis markers. Additionally, ELF scores predict liver-related morbidity and mortality, with ELF scores ≥13 indicating particularly high-risk patients. This study may help inform the implementation of diagnostic pathways for early detection of liver disease and highlights the need for urgent review of individuals with very high ELF scores.

Intelligent Liver Function Testing (iLFT): An Intelligent Laboratory Approach to Identifying Chronic Liver Disease

The intelligent Liver Function Testing (iLFT) pathway is a novel, algorithm-based system which provides automated laboratory investigations and clinical feedback on abnormal liver function test (LFT) results from primary care. iLFT was introduced to NHS Tayside, Scotland, in August 2018 in response to vast numbers of abnormal LFTs, many of which were not appropriately investigated, coupled with rising mortality from chronic liver disease. Here, we outline the development and implementation of the iLFT pathway, considering the implications for the diagnostic laboratories, primary care services and specialist hepatology clinics. Additionally, we describe the utility, outcomes and evolution of iLFT, which was used over 11,000 times in its first three years alone. Finally, we will consider the future of iLFT and propose areas where similar 'intelligent' approaches could be used to add value to laboratory investigations.

Circulating fatty acids and risk of hepatocellular carcinoma and chronic liver disease mortality in the UK Biobank

Nuclear magnetic resonance (NMR)-based plasma fatty acids are objective biomarkers of many diseases. Herein, we aim to explore the associations of NMR-based plasma fatty acids with the risk of hepatocellular carcinoma (HCC) and chronic liver disease (CLD) mortality in 252,398 UK Biobank participants. Here we show plasma levels of n-3 poly-unsaturated fatty acids (PUFA) and n-6 PUFA are negatively associated with the risk of incident HCC [HRQ4vsQ1: 0.48 (95% CI: 0.33-0.69) and 0.48 (95% CI: 0.28-0.81), respectively] and CLD mortality [HRQ4vsQ1: 0.21 (95% CI: 0.13-0.33) and 0.15 (95% CI: 0.08-0.30), respectively], whereas plasma levels of saturated fatty acids are positively associated with these outcomes [HRQ4vsQ1: 3.55 (95% CI: 2.25-5.61) for HCC and 6.34 (95% CI: 3.68-10.92) for CLD mortality]. Furthermore, fibrosis stage significantly modifies the associations between PUFA and CLD mortality. This study contributes to the limited prospective evidence on the associations between plasma-specific fatty acids and end-stage liver outcomes.

Derivation and validation of Transform equations to convert historical Enhanced liver fibrosis (ELF) scores to modern equivalents

BACKGROUND AND AIMS

The Siemens Healthineers ELF™ Test was designed in 2004 with 2 algorithms to allow choice in histological alignment. Consequently, historical and modern algorithms are not fully harmonized, complicating comparisons involving early datasets. We derived transform equations to equate all ELF score versions, allowing historical data to be used in systematic reviews and meta-analyses.

METHODS

Historical ELF equations were graphed pairwise versus their modern equivalent to assess correlation and derive four transforms. Transforms were validated using multiple datasets and evaluated for median absolute bias, number of samples reflecting clinically significant bias, number of discordant samples, bias at established cutoffs, and regression slope and y-intercept.

RESULTS

Three transforms were validated equating Scheuer-aligned and/or age-included historical ELF equations (Immuno 1) to later equations aligned to Ishak and omitting age. A fourth transform corrected ADVIA Centaur® / Atellica® IM ELF scores miscalculated using the Scheuer Immuno 1 equation. Transformed data were well within allowable ELF bias limits.

CONCLUSIONS

All transforms enabled accurate comparison of ELF scores generated by all historical algorithms to the current ADVIA Centaur / Atellica IM Analyzer ELF score. The transforms presented in this report should be used in systematic reviews and meta-analyses to facilitate comparisons to historical data.

Investigation of Serum Markers of Hepatic Fibrosis in Equids

Liver disease is common in equine practice, and treatment and prognosis are dependent on histopathologic examination of biopsies. Liver biopsy is invasive and expensive which restricts its use. Serum markers are used to predict hepatic fibrosis in humans. This study aimed to investigate the enhanced liver fibrosis (ELF) test, based on serum Hyaluronic Acid (HA), procollagen III N-terminal peptide (PIIINP), and tissue inhibitor of metalloproteinase 1 (TIMP-1) to detect hepatic fibrosis in equids. Four groups were included; two with increased serum concentrations of liver-derived enzymes and a liver biopsy (group H; 10 horses and ponies and group D; 10 donkeys) and two without any evidence of liver disease (group HC; 10 horses and ponies and group DC; 10 donkeys). All samples were analyzed for concentrations of HA, PIINP, and TIMP-1. Given the failure to detect TIMP-1 in most subjects, a novel eELF (equid ELF) score was calculated, based on HA and PIIINP. HA and PIIINP concentrations and the eELF score, were compared with determined hepatic fibrosis. HA, PIIINP, and eELF were significantly greater in horses and ponies with a histopathologic fibrosis score ≥ 2 compared with those < 2. A similar observation was found with donkeys for HA and eELF. A significant correlation was found between fibrosis score and HA, PIIINP, and eELF for horses and ponies, and between fibrosis score and HA and eELF in donkeys. Serum HA and the eELF score might be useful serum markers to predict and monitor hepatic fibrosis in horses, ponies, and donkeys.

Combined use of the ELF test and CLivD score improves prediction of liver-related outcomes in the general population

BACKGROUND AND AIMS

Effective and feasible population screening strategies are needed for the early detection of individuals at high risk of future severe liver-related outcomes. We evaluated the predictive performance of the combination of liver fibrosis assessment, phenotype profile, and genetic risk.

METHODS

Data from 5795 adults attending the Finnish Health 2000 Survey were linked with healthcare registers for liver-related outcomes (hospitalization, hepatocellular cancer, and death). Fibrosis was assessed using the enhanced liver fibrosis (ELF) test, phenotype profile by the chronic liver disease (CLivD) risk score, and genetic risk by a validated Polygenic Risk Score (PRS-5). Predictive performance was assessed by competing-risk analyses.

RESULTS

During a median 13-year follow-up, 64 liver-related outcome events were recorded. ELF, CLivD score, and PRS-5 were independently associated with liver-related outcomes. The absolute 10-year risk of liver-related outcomes at an ELF value of 11.3 ranged from 0.3% to 33% depending on the CLivD score. The CLivD score added 51% of new predictive information to the ELF test and improved areas under the curve (AUCs) from 0.91, 0.81, and 0.71 for ELF alone to 0.95, 0.85, and 0.80, respectively, for ELF combined with the CLivD score at 1, 5, and 10 years. The greatest improvement was for 10-year predictions (delta-AUC 0.097, p < .0001). Adding PRS-5 did not significantly increase predictive performance. Findings were consistent in individuals with obesity, diabetes, or alcohol risk use, and regardless of whether gamma-glutamyltransferase was used in the CLivD score.

CONCLUSION

A combination of ELF and CLivD score predicts liver-related outcomes significantly better than the ELF test alone.

Gas6 in chronic liver disease-a novel blood-based biomarker for liver fibrosis

The expression of the receptor tyrosine kinase Axl and its cleavage product soluble Axl (sAxl) is increased in liver fibrosis, cirrhosis, and hepatocellular carcinoma (HCC). In this multicenter study, we evaluated the diagnostic value of Gas6, the high-affinity ligand of Axl, in patients with chronic liver disease. Levels of sAxl and Gas6, and their albumin (alb) ratios were analyzed in serum samples of patients with biopsy-proven liver fibrosis, end-stage liver disease, HCC, and healthy controls, and were compared to Fibrosis-4 (FIB-4), enhanced liver fibrosis (ELF™) test, Child-Pugh score (CPS), model of end-stage liver disease (MELD) score, hepatic venous pressure gradient, and α-fetoprotein, respectively. A total of 1111 patients (median age 57.8 y, 67.3% male) was analyzed. Gas6/alb showed high diagnostic accuracy for the detection of significant (≥F2: AUC 0.805) to advanced fibrosis (≥F3: AUC 0.818), and was superior to Fib-4 for the detection of cirrhosis (F4: AUC 0.897 vs. 0.878). In addition, Gas6/alb was highly predictive of liver disease severity (Odds ratios for CPS B/C, MELD ≥ 15, and clinically significant portal hypertension (CSPH) were 16.534, 10.258, and 12.115), and was associated with transplant-free survival (Hazard ratio 1.031). Although Gas6 and Gas6/alb showed high diagnostic accuracy for the detection of HCC in comparison to chronic liver disease patients without cirrhosis (AUC 0.852, 0.868), they failed to discriminate between HCC in cirrhosis versus cirrhosis only. In conclusion, Gas6/alb shows a high accuracy to detect significant to advanced fibrosis and cirrhosis, and predicts severity of liver disease including CSPH.

Emerging concepts in the detection of liver fibrosis in non-alcoholic fatty liver disease

INTRODUCTION

The non-invasive identification of liver fibrosis related to Non-Alcoholic Fatty Liver Disease is crucial for risk-stratification of patients. Currently, the reference standard to stage hepatic fibrosis relies on liver biopsy, but multiple approaches are developed to allow for non-invasive diagnosis and risk stratification. Non-invasive tests, including blood-based scores and vibration-controlled transient elastography, have been widely validated and represent a good surrogate for risk stratification according to recent European and American guidelines.

AREAS COVERED

Novel approaches are based on 'liquid' biomarkers of liver fibrogenesis, including collagen-derived markers (PRO-C3 or PRO-C6), or 'multi-omics' technologies (e.g. proteomic-based molecules or miRNA testing), bearing the advantage of tailoring the intrahepatic disease activity. Alternative approaches are based on 'dry' biomarkers, including magnetic resonance-based tools (including proton density fat fraction, magnetic resonance elastography, or corrected T1), which reach similar accuracy of liver histology and will potentially help identify the best candidates for pharmacological treatment of fibrosing non-alcoholic steatohepatitis.

EXPERT OPINION

In the near future, the sequential use of non-invasive tests, as well as the complimentary use of liquid and dry biomarkers according to the clinical need (diagnosis, risk stratification, and prognosis, or treatment response) will guide and improve the management of this liver disease.

Noninvasive Tests Used in Risk Stratification of Patients with Nonalcoholic Fatty Liver Disease

As the prevalence of obesity and type 2 diabetes increases around the world, the prevalence of nonalcoholic fatty liver disease (NAFLD) has grown proportionately. Although most patients with NAFLD do not experience progressive liver disease, about 15% to 20% of those with nonalcoholic steatohepatitis can and do progress. Because liver biopsy's role in NAFLD has become increasingly limited, efforts have been undertaken to develop non-invasive tests (NITs) to help identify patients at high risk of progression. The following article discusses the NITs that are available to determine the presence of NAFLD as well as high-risk NAFLD.

How to Identify Advanced Nonalcoholic Fatty Liver Disease in the Primary Care Setting

Nonalcoholic fatty liver disease (NAFLD) affects 30 to 40% of the population globally and is increasingly considered the most common liver disease. Patients with type 2 diabetes, obesity, and cardiovascular diseases are at especially increased risk for NAFLD. Although most patients with NAFLD do not have progressive liver disease, some patients progress to cirrhosis, liver cancer, and liver mortality. Given the sheer number of patients with NAFLD, the burden of disease is enormous. Despite this large and increasing burden, identification of NAFLD patients at risk for progressive liver disease in the primary care and diabetology practice settings remains highly suboptimal. In this review, our aim is to summarize a stepwise approach to risk stratify patients with NAFLD which should help practitioners in their management of patients with NAFLD.

Using the ELF test, FIB-4 and NAFLD fibrosis score to screen the population for liver disease

BACKGROUND AND AIMS

There is a need for accurate population screening biomarkers for alcohol-related and non-alcoholic fatty liver disease (ALD, NAFLD). We evaluated the Enhanced Liver Fibrosis (ELF) test compared to FIB-4 and NAFLD fibrosis score (NFS) using transient elastography (TE) as fibrosis screening reference.

METHODS

We prospectively included participants from the general population, and people at risk of ALD or NAFLD. Screening positive participants (TE ≥8 kPa) were offered a liver biopsy. We measured concomitant ELF, FIB-4, and NFS using validated cut-offs: ≥9.8, ≥1.3, ≥-1.45, respectively.

RESULTS

We included 3,378 participants (1,973 general population, 953 at risk of ALD, 452 at risk of NAFLD), with a median age of 57 years (IQR: 51-63). Two hundred-and-forty-two were screening positive (3.4% in general population, 12% in ALD, 14% in NAFLD). Most participants with TE <8 kPa also had ELF <9.8 (88%) despite a poor overall correlation between ELF and TE (Spearman´s rho=0.207). ELF had significantly fewer false positives (11%) than FIB-4 and NFS (35% and 45%), while retaining a low rate of false negatives (<8%). A screening strategy of FIB-4 followed by ELF in indeterminate cases resulted in 8% false positives, 4% false negatives and 88% correctly classified. We performed a liver biopsy in 155/242 (64%) screening positives, of those 54 (35%) had advanced fibrosis (≥F3). ELF diagnosed advanced fibrosis with significantly better diagnostic accuracy than FIB-4 and NFS: AUROC 0.85 (95% CI 0.79-0.92) versus 0.73 (0.64-0.81) and 0.66 (0.57-0.76).

CONCLUSION

The ELF test alone or combined with FIB-4 for liver fibrosis screening in the general population and at-risk groups reduces the number of futile referrals compared to FIB-4 and NFS, without overlooking true cases.

IMPACT AND IMPLEMENTATIONS

We need referral pathways that are efficient at detecting advanced fibrosis from alcohol-related and non-alcoholic fatty liver disease in the population, but without causing futile referrals or excessive use of resources. This study indicates that a sequential test strategy of FIB-4, followed by the ELF test in indeterminate cases leads to few patients referred for confirmatory liver stiffness measurement, while retaining a high rate of detected cases, and at low direct costs. This two-step referral pathway could be used by primary care for mass, targeted, or opportunistic screening for liver fibrosis in the population.

CLINICAL TRIAL NUMBER

Clinicaltrials.gov number NCT03308916.

Meta-analysis: Enhanced liver fibrosis test to identify hepatic fibrosis in chronic liver diseases

BACKGROUND & AIMS

Patients with liver disease can be stratified for risk of liver-related ill health by degree of hepatic fibrosis. The Enhanced liver fibrosis (ELF) test was developed to quantify hepatic fibrosis non-invasively and is widely used. The objective of this review was to identify and synthesise the evidence on the diagnostic accuracy of the ELF test for staging of hepatic fibrosis.

APPROACH & RESULTS

Searches of PubMed and EMBASE were conducted between October 2020 and November 2021 to identify studies reporting the diagnostic accuracy of the ELF test compared to histology in liver disease patients. QUADAS-2 was used to assess risk of bias in each study. Meta-analysis using the multiple thresholds model described by Steinhauser S, Schumacher M, Rücker G. Modelling multiple thresholds in meta-analysis of diagnostic test accuracy studies. BMC Med. Res. Methodol. 2016;16. 10.1186/s12874-016-0196-1 allowed synthesis of 2 × 2 data at different cut-offs. Sixty-three studies were included in this review. These studies included 19,285 patients with or at risk of liver disease from viral hepatitis, Non-Alcoholic Fatty Liver Disease, Alcohol-related Liver Disease and other mixed chronic liver diseases. The prevalence of significant fibrosis, advanced fibrosis and cirrhosis was 47.5%, 39.2% and 4.4%, respectively. Cut-offs with maximal Youden index were generated with AUROC = 0.811 (95% CI: 0.736-0.870), 0.812 (95% CI: 0.758-0.856) and 0.810 (95% CI: 0.694-0.888) to detect significant fibrosis, advanced fibrosis or cirrhosis, respectively. Diagnostic accuracy of the ELF test varied between different liver diseases and cut-offs to detect each stage with 95% sensitivity or specificity were also generated.

CONCLUSIONS

Meta-analysis revealed considerable variability in the ability of ELF to stage fibrosis across disease aetiologies. Research has mostly focused on viral hepatitis and NAFLD. There is currently a lack of data on the value of the ELF test in Alcohol-related liver disease and patients in primary care settings.

Diagnostic accuracy of enhanced liver fibrosis test for nonalcoholic steatohepatitis-related fibrosis: Multicenter study

AIM

The enhanced liver fibrosis (ELF) test is a noninvasive method for diagnosing hepatic fibrosis in patients with nonalcoholic fatty liver disease (NAFLD). This multicenter cohort study aimed to evaluate the accuracy of the ELF test and compare it with other noninvasive tests in Japan.

METHODS

We analyzed 371 Japanese patients with biopsy-proven NAFLD. We constructed area under the receiver operator characteristic curves (AUROC) to determine the diagnostic accuracies of the ELF test, the Mac-2-binding protein glycosylation isomer (M2BPGi), the Fibrosis-4 (FIB-4) index, and combinations of these indices.

RESULTS

In patients with F0/F1/F2/F3/F4 fibrosis, the median values of the ELF test were 8.98/9.56/10.39/10.92/11.41, respectively. The AUROCs of the ELF test for patients with F0 versus F1-4, F0-1 versus F2-4, F0-2 versus F3-4, and F0-3 versus F4 fibrosis were 0.825/0.817/0.802/0.812, respectively. The AUROCs of the ELF test were greater than those of the FIB-4 index and M2BPGi at each fibrosis stage. Respective low and high cut-off values yielded sensitivities and specificities for predicting advanced fibrosis (≥F3) of 91.1% and 50.8%, and 38.5% and 92.8%, respectively. For F3 or F4 fibrosis, the combined values from the ELF test and FIB-4 index showed a sensitivity of 98.5%, and the combined values from the ELF test and M2BPGi assay showed a specificity of 97.5%.

CONCLUSIONS

In Japan, the ELF test predicts NAFLD-related fibrosis from its early stages. The diagnostic ability of the ELF test was not inferior to that of other indices, and the combined values of ELF plus other indices were more accurate.

Comparison of various strategies to define the optimal target population for liver fibrosis screening: A population-based cohort study

BACKGROUND & AIMS

Liver fibrosis screening is recommended in high-risk populations, but the optimal definition of "high risk" remains to be established. We compared the performance of several risk-stratification strategies in a population-based setting.

METHODS

Data were obtained from the Finnish population-based health examination surveys Health 2000 and FINRISK 2002-2012. The Chronic Liver Disease Risk Score (CLivD) was compared to previously published risk-stratification strategies based on elevated liver enzymes, alcohol use, diabetes, fatty liver index, body mass index, and/or metabolic risk factors for their ability to detect either advanced liver fibrosis or incident severe liver events. Advanced fibrosis was defined as an Enhanced Liver Fibrosis (ELF^TM ) score >9.8 in the Health 2000 study (n = 6084), and incident liver events were ascertained from registry linkage in the combined FINRISK 2002-2012 and Health 2000 cohort (n = 26,957).

RESULTS

Depending on the cohort, 53%-60% of the population was considered at risk using the CLivD strategy (low-intermediate-high risk, excluding the minimal-risk category), compared to 30%-32% according to the other risk-stratification strategies. The CLivD captured 85%-91% of cases in the population with advanced liver fibrosis and 90% of incident severe liver events within 10 years from baseline. This compares to 33%-44% and 56%-67% captured by the other risk-stratification strategies, respectively. The 10-year cumulative incidence of liver events varied by risk-stratification strategy (1.0%-1.4%).

CONCLUSIONS

Compared to previously reported traditional risk factor-based strategies, use of the CLivD captured substantially more cases with advanced liver disease in the population and may be superior for targeting further fibrosis screening.

Performance of the Enhanced Liver Fibrosis Test to Estimate Advanced Fibrosis Among Patients With Nonalcoholic Fatty Liver Disease

Importance

The most important surrogate for increased risk of adverse clinical outcomes among patients with nonalcoholic fatty liver disease (NAFLD) is the patient's stage of liver fibrosis. There is a significant barrier to risk-stratifying patients in clinical practice owing to the need for liver biopsy.

Objective

To determine the performance of the enhanced liver fibrosis (ELF) test as a noninvasive test for assessment of liver fibrosis among patients with NAFLD.

Design, Setting, and Participants

This retrospective cross-sectional study was conducted among patients recruited from a large, community-based hospital system's outpatient liver clinic from 2001 to 2020. Patients with NAFLD defined as steatosis greater than 5% without evidence of other liver disease or excessive alcohol use were included. Data were analyzed from August 2020 through February 2021.

Intervention

Enhanced liver fibrosis score was calculated.

Main Outcomes and Measures

Advanced fibrosis was identified by liver biopsy or transient elastography.

Results

Among 829 patients with NAFLD, the mean (SD) age was 53.1 (14.0) years, there were 363 (43.8%) men, 294 patients (35.5%) had type 2 diabetes, and the mean (SD) fibrosis-4 (fib-4) score was 1.34 (0.97). There were 463 patients with liver biopsy, among whom 113 individuals (24.4%) had bridging fibrosis or cirrhosis; among 462 patients with transient elastography data, 79 individuals (17.1%) had liver stiffness results of 9.6 kPa or more (ie, advanced fibrosis). Patients with advanced fibrosis had statistically significantly increased mean (SD) ELF scores compared with patients without advanced fibrosis as determined by biopsy (10.1 [1.3] vs 8.6 [1.0]; P < .001) or transient elastography (10.0 [1.1] vs 9.0 [0.8]; P < .001). Among all patients with NAFLD, the area under the receiver operating characteristic curve (AUROC) for ELF in identifying patients with advanced fibrosis was 0.81 (95% CI, 0.77-0.85) for patients diagnosed by biopsy and 0.79 (95% CI, 0.75-0.82) for those diagnosed by transient elastography. Performance of the ELF score was similar among patients with NAFLD who were aged 65 years or older (AUROC, 0.74; 95% CI, 0.58-0.87) or had type 2 diabetes (AUROC, 0.78; 95% CI, 0.71-0.84). The combination of an ELF score of 7.2 or greater with a fib-4 score of 0.74 or greater was associated with a negative predictive value of 95.1% (95% CI, 91.8%-98.4%) and a sensitivity of 92.5% (95% CI, 87.4%-97.5%), which can reliably rule out advanced fibrosis. An ELF score of 9.8 or greater with a fib-4 score of 2.9 or greater was associated with a positive predictive value of 95.0% (95% CI, 85.5%-100%) and a specificity of 99.7% (95% CI, 99.1%-100%), which can be used to rule in advanced fibrosis.

Conclusions and Relevance

These findings suggest that the ELF test performs well in identifying patients with NAFLD who are at increased risk of advanced fibrosis and that this test combined with fib-4 score may be reliably used in clinical practice to assess the presence or absence of advanced fibrosis among patients with NAFLD.