Editorial: stepwise risk stratification for both F3 and F4 fibrosis stage in NAFLD patients

Advanced liver fibrosis, as defined by stage 3-4 fibrosis, predicts increased liver-related mortality in patients with nonalcoholic fatty liver disease (NAFLD).1 Given the high prevalence of NAFLD and the limitations of liver biopsy, non-invasive fibrosis assessment tools in NAFLD have been the focus of many studies over the last decade,2, 3 resulting in the availability of non-invasive liver fibrosis assessment tools, including general clinical scoring systems, combination biomarkers, and radiologic tools such as elastography. General clinical scoring systems such as NAFLD Fibrosis Score (NFS), FIB-4 and APRI are published formulae using commonly available variables. These scoring systems are attractive as they are free and accessible to everyone and have been validated in the NAFLD cohort for identifying patients with advanced fibrosis. Liver stiffness as measured by vibration-controlled transient elastography (VCTE) has also been validated as a surrogate marker of liver fibrosis and is excellent at ruling out advanced fibrosis. In a recent issue, Labenz et al present their results from a prospective study of the performance of these clinical scoring systems and VCTE in identifying patients with noncirrhotic advanced fibrosis (ie, stage 3) in a cohort of 261 biopsy-proven NAFLD patients in Germany.4 Patients with cirrhosis on biopsy were excluded from this study and about 15.7% had noncirrhotic advanced fibrosis. The study confirms previous findings that NAFLD patients with advanced fibrosis are older, have a higher BMI, and more likely to have DM and hypertension,5 highlighting the importance of clinical variables such as DM, age and BMI which are components of the NFS. The findings also confirmed that VCTE has better diagnostic performance over clinical and blood-based scores with APRI being the least accurate.

What is new about this study is that it excluded patients with cirrhosis and looked at the test performance in identifying just stage 3 fibrosis from stage 0-2 fibrosis. However, we question the clinical applicability of excluding cirrhosis from the analysis. The cut-offs proposed by this study (eg, 12 kPA for VCTE) does not exclude patients with cirrhosis.6-9 As there is significant sampling error with liver biopsy, a proportion of the patients identified as F3 could be actually F4 resulting in a higher liver stiffness measurement.10 Regardless, given the increased risk of liver-related mortality, patients with advanced fibrosis, both F3 and F4, should be screened for HCC and varices and monitored for the development of decompensation as well as prioritised when treatment becomes available. This study does confirm previous findings that accessible general clinical scoring systems are good at the initial risk stratification of a NAFLD patient, and supports the approach advocated by Vilar-Gomez and Chalasani in that they should be used as the first step in risk stratification with a cut-off aimed at a high negative predictive value to rule out advanced fibrosis. Those with a low-risk score can be monitored while those with intermediate and high risk scores should be further evaluated with VCTE and/or a liver biopsy.11