Editorial: stepwise risk stratification for both F3 and F4 in NAFLD patients—authors’ reply

EDITORS,

We read with great interest the comment of Rajab Idriss and Michelle Lai that relates to our recent article.1, 2 The manuscript highlights the clinical characteristics of patients with advanced, non-cirrhotic NAFLD and presents the diagnostic power of the commonly employed non-invasive composite scores and vibration-controlled transient elastography (VCTE) for separating advanced, non-cirrhotic NAFLD—equal to histological stage F3 according to Kleiner et  al3—from earlier stages. We chose this approach to stress the fair-to-poor performance of non-invasive composite scores in pre-cirrhotic NAFLD when relying solely on blood-based markers. To us, the importance of identifying F3-NASH relates to the incremental mortality—both liver-related and -unrelated that occurs in these patients.4 While there are not yet any approved therapies, the design of current pivotal trials—which included F3 and all but one excluded F4 patients—could also mandate this separation according to the label of potential drugs that may become available. Also, while current guidelines do not recommend screening for HCC in advanced, non-cirrhotic NALFD,5 clinical evidence is emerging that HCC may arise in the absence of cirrhosis.6 Our current analysis provides a good estimate on the potential characteristic of these patient populations. Considering the low diagnostic accuracy of non-invasive markers, to us there remains a clear indication to perform liver biopsy to diagnose and stage patients with NAFLD. However, we also believe that establishing better non-invasive markers, like the recently published composite score including markers of extracellular matrix turnover, for example, Pro-C3, will move the field forward.7

Two additional aspects stand out and should be highlighted. First, there seems to be heterogeneity between histological cohorts in different countries. The current analysis did not observe gender differences which have been inconsistently reported.8 On the other hand, arterial hypertension and diabetes were confirmed and could be validated in the current analysis. Second, we were surprised by the low rate of statin use in this high-risk cohort for cardiovascular mortality. While this has not been explored as a primary endpoint in this study, our clinical experience reflects the low use of this class of medication that may have been related to safety concerns.9 This precaution might not be warranted and could even be counterproductive. Current guidelines from the cardiology societies should be followed—also in the context of co-existing NAFLD.