Can advanced fibrosis be reliably excluded by simple non-invasive fibrosis scoring systems? Diagnostic performance of FIB-4 and NFS in Turkish patients with biopsy-proven non-alcoholic fatty liver disease

INTRODUCTION: Histopathological examination of a liver biopsy specimen is currently the reference standard for staging hepatic fibrosis. However, the invasive nature of liver biopsy – which portends a significant risk of sampling errors, complications, and costs – has led to a continuous quest for non-invasive diagnostic methods. Simple blood-based fibrosis scores based on routine parameters – including the FIB-4 and non-alcoholic fatty liver disease (NAFLD) fibrosis score (NFS) – can be easily obtained and thus are suitable for screening purposes. Several cutoffs for the identification of advanced fibrosis have been proposed. For example, FIB-4 scores of 2.67 or 3.25 have been used to identify patients with a low and high risk for advanced fibrosis, respectively. Similarly, NFS scores 0.676 indicate a low and high risk for advanced fibrosis, respectively. In this study, we sought to investigate the diagnostic performances of FIB-4 and NFS in the detection of advanced fibrosis in patients with biopsy-proven NAFLD.  METHODS: This is a retrospective analysis of prospectively collected data. The study sample consisted of 463 consecutive patients with biopsy-proven NAFLD patients who were followed-up in our outpatient clinic over a 4-year period (between 2009−2010 and 2017−2018). The diagnostic performances of published cutoffs of FIB-4 and NFS in the detection of advanced fibrosis were compared with those reported in the literature. In addition, the optimal cutoff points in our population were determined.  RESULTS: The general characteristics of the study patients are summarized in Table 1. On histopathological examination, 34.1% of patients had an F0 fibrosis stage, 31.1% had F1, 17.3% had F2, 13.6% had F3, and 3.9% had F4. With regard to the FIB-4 cutoffs for low risk for advanced fibrosis, the sensitivity values were 64% and 48%, with positive predictive values (PPVs) of 0.342 and 0.358, respectively. A cutoff of 2.67 and 3.25 for high risk of advanced fibrosis had a negative predictive value (NPV) of 0.843 and 0.839, with specificity values of 97% and 98%, respectively. In our study population, the optimal cutoff for advanced fibrosis was 1.275 – which produced an area under ROC curve of 0.731 (95% confidence interval, 0.672–0.790) and was associated with a high kappa value. With regard to the NFS cutoff identifying low risk of advanced fibrosis, the sensitivity and the PPV were 71% and 0.291, respectively. Conversely, the NFS cutoff for high risk of advanced fibrosis had a specificity 96% and NPV of 0.842, respectively. In our sample, the optimal NFS cutoff for identifying patients with high risk of advanced fibrosis was -1.485 (Table 2). CONCLUSION: Our results indicate that NFS and FIB-4 have an acceptable clinical usefulness for excluding advanced fibrosis, whereas their diagnostic utility for identifying patients with this condition remains suboptimal.