Early detection of individuals with metabolic dysfunction–associated steatotic liver disease (MASLD) at risk of progression to advanced liver disease is crucial for interventions and appropriate management. MASLD is a slowly progressive condition with only a small subset of patients progressing to clinically significant fibrosis over time, thus early detection of significant fibrosis is required to allow for appropriate care. As such, several risk models have been developed to identify those at an increased risk of liver fibrosis.^1-4 The models differ in their derivation cohort and predictors and the distribution of fibrosis in their population. Even though multiple models are available, relatively few studies directly compare their performance within the same cohort for identifying significant fibrosis.

The advent of pharmacologic therapies for MASLD has changed risk stratification approaches in clinical practice. Prior to the availability of drug treatments, therapeutic intervention was based on lifestyle advice irrespective of disease stage. The main aim of risk stratification was to identify the individuals with cirrhosis who could then be entered into surveillance programmes for hepatocellular cancer and screened for the development of portal hypertension. In this context, identifying those at precirrhotic stages of fibrosis (≤F3) would enable closer follow‑up for the development of cirrhosis. Consequently, much of the earlier literature focused on biomarkers and thresholds for diagnosing advanced fibrosis (F3–F4). The focus of risk stratification is now changing. Because pharmacotherapies are generally available for noncirrhotic metabolic dysfunction–associated steatohepatitis (F2–F3), some studies have examined dual‑threshold approaches, so that the patients with cirrhosis can also be identified, as MASLD therapeutic treatments would not be indicated in this group.⁵ There is now more scope to identify the individuals with F2 fibrosis who would be eligible for treatment with medication.⁶

In their article published in this issue of Polish Archives of Internal Medicine, Werel‑Ołdziejewska et al⁷ have sought to compare the diagnostic accuracy of 9 noninvasive tests (NITs) for identifying significant fibrosis (≥F2) in a cohort of 134 patients with MASLD. They provided useful comparative data on NITs, particularly given the focus on significant fibrosis. In the patients with significant fibrosis, the majority (30/52) had F2 fibrosis, indicating a skewed distribution toward milder disease. Also, the overall majority of participants had milder disease (F0–F1). Furthermore, the implementation of single‑threshold approaches to identify individuals with at least F2 fibrosis does not determine how well the tests distinguish F2–F3 specifically from F4; F2–F3 being the population targeted for treatment of noncirrhotic MASLD.

The authors conclude that the Hepamet score may be considered an alternative to Fibrosis‑4 (FIB‑4) index for ruling out significant fibrosis. However, due to the relatively small sample size, most of the areas under the receiver operating characteristic curves were clustered around 0.8, and the differences between the scores did not reach significance. Interpretations indicating superior performance should therefore be made with caution. Furthermore, and as acknowledged by the authors, the Hepamet score requires additional investigations that are not routinely available in many clinical settings, something that can limit applicability of this test. Taking these 2 factors into account (no statistical superiority in diagnostic performance and applicability concerns), the incremental clinical value of adopting the Hepamet score may be limited. Moreover, for the Hepamet score to be a valid alternative to FIB‑4, validation in independent cohorts— particularly those with a lower disease burden—is required, given the small cohort size of this study. The relatively young median age of the cohort (38 y) further limits generalizability. In this context, the proposed lower FIB‑4 thresholds appear premature, as age‑specific thresholds for younger populations have already been widely reported, as discussed by the authors themselves.

Werel‑Ołdziejewska et al⁷ applied Youden‑derived thresholds across multiple NITs and demonstrated improved negative predictive value for the detection of significant fibrosis (≥F2). As this target condition differs from the advanced fibrosis (≥F3) end point used in other derivation studies, the optimal cutoffs identified here are lower than those evaluated so far.⁸ The authors suggest a FIB‑4 threshold of 0.86 for ruling out significant fibrosis, which sits substantially below the standard 1.3 cutoff derived for advanced fibrosis. Others have suggested a FIB‑4 cutoff of 0.96⁹ or 1¹⁰ as optimal when testing for significant fibrosis, both of which are much closer to the cutoff of 0.86 being proposed by Werel‑Ołdziejewska et al.⁷ Taking into account a slowly progressive disease such as MASLD and the small number of patients who go on to develop major adverse liver outcomes, it is important to consider the downstream clinical consequences of lowering thresholds based on measures such as the Youden index. False‑negative and false‑positive classifications do not carry equivalent clinical risks. The Youden index weighs false negatives and false positives equally. Changes to thresholds must therefore be considered in the context of clinical outcomes, unnecessary referrals, and patient harm, especially if pharmacologic therapies are to target individuals with F2 fibrosis. Further analysis is needed to determine the clinical relevance of the discriminative ability of the alternate thresholds.

The clinical setting in which these thresholds would be applied is therefore crucial. If intended for use in the community, external validation in such settings is essential, as the thresholds were derived in hospital‑based cohorts that likely include patients with more advanced disease than would typically be encountered in primary care. Conversely, in hepatology clinics (as in the patients in this cohort), using lower diagnostic thresholds may offer limited incremental benefit for detecting significant fibrosis. For example, in the United Kingdom, a 2‑step primary care pathway incorporating FIB‑4 followed by enhanced liver fibrosis test has been shown to change referral patterns substantially.¹¹ By excluding patients at a low risk of fibrosis, these pathways have a high proportion of individuals with more advanced disease in secondary and tertiary care. Consequently, hepatology cohorts are likely to represent a higher‑risk population than that of the present sample.

With the burden of MASLD cirrhosis expected to increase 2–3‑fold from 2015 to 2030, a clear care pathway including the identification and treatment of significant disease is required.¹² Werel‑Ołdziejewska et al⁷ add to this discussion by highlighting the need to reconsider thresholds when the diagnostic target shifts. However, implementation and uptake of alternative NITs in different clinical settings will require extensive validation and health economic analysis considering the slow progression of the disease and the changing landscape of potential therapeutics.