Original article
Pancreas, biliary tract, and liver
Accuracy of Noninvasive Scoring Systems in Assessing Risk of Death and Liver-Related Endpoints in Patients With Nonalcoholic Fatty Liver Disease

https://doi.org/10.1016/j.cgh.2018.11.030Get rights and content

Background and Aims

Several non-invasive scoring systems have been developed to determine risk of advanced fibrosis in non-alcoholic fatty liver disease (NAFLD). We examined the association between 4 scoring systems and incident severe liver disease and overall mortality in a large cohort of patients with biopsy-proven NAFLD.

Methods

We performed a retrospective analysis of data from 646 patients with biopsy-proven NAFLD, recruited from 2 hospitals in Sweden, from 1971 through 2009. The NAFLD fibrosis score (NFS), FIB-4, APRI, and BARD scores were calculated at the time of the liver biopsy. Based on each score, patients were assigned to categories of low, intermediate, or high risk for advanced fibrosis. Overall mortality and severe liver disease (cirrhosis, decompensated liver disease, liver failure, or hepatocellular carcinoma) were ascertained through linkage with national registers until the end of 2014. Cox regression, area under the receiver operating characteristic (AUROC) curve, and C-statistic analyses were used to study the predictive capacity of each scoring system.

Results

During a mean follow-up time of 19.9±8.7 years, there were 214 deaths and 76 cases of severe liver disease. For overall mortality, AUROC curve values were: NFS, 0.72 (95% CI, 0.68–0.76); FIB-4, 0.72 (95% CI, 0.68–0.76); BARD, 0.62 (95% CI, 0.58–0.66); and APRI, 0.52 (95% CI, 0.47–0.57). For severe liver disease, AUROC curve values were: NFS, 0.72 (95% CI, 0.66–0.78); FIB-4, 0.72 (95% CI, 0.66–0.79); BARD, 0.62 (95% CI, 0.55–0.69); APRI, 0.69 (95% CI, 0.63–0.76). C-statistics for all scores were of moderate capacity to predict outcomes.

Conclusions

In a retrospective analysis of data from 646 patients with biopsy-proven NAFLD, we found the NFS and the FIB-4 scores to most accurately determine risk of overall death or severe liver disease. However, the AUROC values for these scoring systems are not high enough for use in the clinic; new systems are needed to determine prognoses of patients with NAFLD.

Section snippets

Subjects

We conducted a retrospective cohort study including all patients diagnosed with biopsy-proven NAFLD, at the Karolinska University Hospital (Huddinge) and Linköping University Hospital, from 1971 to 2009. The methodology for generation of the cohort and the main results are available elsewhere.5 Briefly, we identified all patients that underwent liver biopsy with the finding of steatosis at our institutions. All patients’ medical charts were scrutinized in detail. Patients with causes for

Results

Baseline characteristics of the cohort are presented in Table 1. Median age was 50 years and men comprised 62% of the cohort . The majority (66%) of the participants had NASH, and 12% had advanced (stage 3–4) fibrosis. During a mean follow-up of 19.9 ± 8.7 years (range, 0.4–40 years of age), 214 (33%) persons died and 76 persons (12%) developed severe liver disease. There were 17 (2.6%) cases of death directly attributed to liver disease and there were 12 (1.9%) cases of HCC. Absolute numbers

Discussion

In this retrospective cohort study, we studied the predictive capacity of 4 commonly used noninvasive scoring systems used to identify advanced fibrosis in NAFLD for mortality and severe liver disease. We found that the scores with the highest predictive capacity, both for overall mortality and for severe liver disease, were the FIB-4 and the NFS. However, based on several analyses, all scores only predicted the endpoints to a fair degree. In multivariable modeling and after adding additional

Conclusions

In this retrospective cohort study, the NFS and FIB-4 best predicted overall mortality and severe liver disease after a mean follow-up close to 20 years. However, no score had a clinically acceptable predictive capacity, why new scores aimed directly at predicting prognoses in NAFLD are needed.

References (34)

  • P.S. Dulai et al.

    Increased risk of mortality by fibrosis stage in non-alcoholic fatty liver disease: systematic review and meta-analysis

    Hepatology

    (2017)
  • S. McPherson et al.

    Simple non-invasive fibrosis scoring systems can reliably exclude advanced fibrosis in patients with non-alcoholic fatty liver disease

    Gut

    (2010)
  • S. Treeprasertsuk et al.

    NAFLD fibrosis score: a prognostic predictor for mortality and liver complications among NAFLD patients

    World J Gastroenterol

    (2013)
  • G. Sebastiani et al.

    Prognostic value of non-invasive fibrosis and steatosis tools, hepatic venous pressure gradient (HVPG) and histology in nonalcoholic steatohepatitis

    PLoS One

    (2015)
  • M.H. Le et al.

    Prevalence of non-alcoholic fatty liver disease and risk factors for advanced fibrosis and mortality in the United States

    PLoS One

    (2017)
  • A. Unalp-Arida et al.

    Liver fibrosis scores predict liver disease mortality in the United States population

    Hepatology

    (2017)
  • D.E. Kleiner et al.

    Design and validation of a histological scoring system for nonalcoholic fatty liver disease

    Hepatology

    (2005)
  • Cited by (70)

    • Noninvasive Assessment of Liver Fibrosis in NAFLD

      2023, Clinical Gastroenterology and Hepatology
    • Development and validation of a model to predict incident chronic liver disease in the general population: The CLivD score

      2022, Journal of Hepatology
      Citation Excerpt :

      Current risk stratification focuses on assessing the amount of liver fibrosis, but simple non-invasive liver fibrosis tests, such as non-alcoholic fatty liver disease (NAFLD) fibrosis score, fibrosis-4 (FIB-4), or aspartate aminotransferase to platelet ratio index (APRI), are of limited value in individuals in the general population compared to in highly selected NAFLD cohorts from specialist centers.9,10 Significant alcohol use also impairs the performance of these tests,11 and they were not originally designed to predict clinical liver outcomes.12–14 Direct fibrosis biomarkers and special imaging methods, such as elastography, are additionally limited by cost, accessibility, and suboptimal performance in identifying early fibrosis stages and in the presence of alcoholic steatohepatitis.15,16

    View all citing articles on Scopus

    Conflicts of interest The authors disclose no conflicts.

    Funding The work was supported by the Stockholm City Council – Clinical Postdoctoral Appointment (to Hannes Hagström) and a Bengt Ihre Fellowship (to Hannes Hagström).

    View full text