ABSTRACT
Metabolic dysfunction-associated steatotic liver disease (MASLD) and metabolic dysfunction and alcohol-related liver disease (MetALD) exist on a continuous spectrum in which fibrosis stage remains the dominant predictor of liver‑related complications. Contemporary non‑invasive risk stratification uses simple serum scores (e.g., Fibrosis‑4 index [FIB‑4]) as a first step, followed by vibration‑controlled transient elastography (VCTE) or liver‑specific blood tests (enhanced liver fibrosis [ELF], procollagen III N‑terminal propeptide [PRO‑C3]) to refine risk and triage referrals. Magnetic resonance elastography (MRE) provides higher accuracy than ultrasound‑based elastography, though limited by its availability. Composite models (Agile 3+/4, steatosis‑associated fibrosis estimator [SAFE]) can reduce indeterminate classifications but may require additional calibration for alcohol exposure for use in MetALD. Prognosis should incorporate dynamic, longitudinal measurement of non‑invasive tests, and integrate extrahepatic risks, particularly cardiovascular and kidney disease. In MetALD, quantitative alcohol assessment can be improved by incorporating objective biomarkers such as phosphatidylethanol (PEth) as part of a triangulation approach that integrates self‑reported alcohol intake, the Alcohol Use Disorders Identification Test‑Consumption (AUDIT‑C), and clinical history. Challenges remain, particularly for MetALD, and future research will focus on integrating dynamic biomarkers, predicting treatment response, and utilizing artificial intelligence (AI) for personalized risk and treatment prediction, combining genetic modifiers and polygenic risk scores with clinical and imaging data, with validation across diverse populations.
-
Keywords: MASLD; MetALD; Fibrosis; Artificial intelligence; Alcohol exposure
INTRODUCTION
Transition from non-alcoholic fatty liver disease (NAFLD)/metabolic dysfunction-associated fatty liver disease (MAFLD) to metabolic dysfunction-associated steatotic liver disease (MASLD)/metabolic dysfunction and alcohol-related liver disease (MetALD) terminology
Steatotic liver disease (SLD) is a prevalent chronic liver condition, affecting roughly one in three adults [
1]. In recent years, fatty liver disease nomenclature has been redefined, shifting from the exclusion-based construct of NAFLD to SLD as an overarching term. Within this framework, MASLD denotes hepatic steatosis accompanied by at least one cardiometabolic risk factor (CMRF), whereas MetALD captures individuals who meet MASLD criteria but also report alcohol intake above “low-risk” levels. Specifically, the consensus defines MetALD as alcohol consumption of 140–350 g/week in females and 210–420 g/week in males, which is above the “low-risk” thresholds used for MASLD (<140 g/week for females; <210 g/week for males) but below levels typically used to define alcohol-related liver diseases (ALD).
This terminology update was formalized through a multisociety Delphi consensus to improve diagnostic clarity, reduce stigma embedded in prior labels, and better reflect underlying biology as well as the real-world overlap between metabolic risk and alcohol exposure [
2].
MASLD constitutes a major component of the global chronic liver disease burden under the updated SLD umbrella, and current population estimates show that SLD is highly prevalent and predominantly driven by CMRFs. A recent systematic review and meta-analysis (44 studies; 11.3 million participants) reported a pooled global SLD prevalence of 37.5% and MASLD of 33.6%, while MetALD accounted for 4.1% of the general population [
1]. These figures imply that MetALD represents a meaningful SLD spectrum share in the community (approximately 11% of all SLD cases), supporting the concept that “pure” metabolic versus alcohol-only categories often underrepresent real-world overlap [
1]. Consistently, in a general-population cohort with available liver magnetic resonance imaging (MRI), SLD prevalence was 27.0%, and MetALD comprised 7.9% among individuals with SLD, with MASLD accounting for over 90% of SLD cases [
3].
Across MASLD and MetALD, fibrosis stage is the dominant long-term prognostic determinant, showing the strongest and most consistent association with liver-related outcomes, including hepatic decompensation, HCC, liver transplantation, and liver-related mortality [
4]. This evidence base has led to the prioritization of advanced fibrotic and cirrhotic identification as the key step for downstream management [
5]. In parallel, MASLD and MetALD are tightly linked to systemic metabolic dysfunction, and patients at highest hepatic risk frequently also carry substantial extrahepatic risks, particularly cardiovascular disease (CVD)6 and chronic kidney disease (CKD), further strengthening the rationale for structured risk assessment and integrated care [
7,
8].
Operationally, risk assessment serves three complementary purposes: first, to detect or exclude advanced fibrosis efficiently in broad, low-prevalence settings; second, to identify individuals likely to have advanced fibrosis or cirrhosis warranting specialist referral and etiologic refinement, including careful alcohol exposure quantification for MetALD classification; and third, to stratify future risk of hepatic decompensation and HCC to guide the intensity of follow-up, selection of non-invasive testing strategies, planning of variceal screening and HCC surveillance, and healthcare resource allocation, consistent with guideline-based approaches to non-invasive tests (NIT) in chronic liver disease [
9].
Finally, emerging syntheses focused on MetALD indicate that the coexistence of metabolic dysfunction and higher alcohol exposure is common and is associated with a higher adverse liver-related and extrahepatic outcome burden, supporting the rigorous risk stratification need across the full MASLD–MetALD spectrum rather than management based on etiology labels alone [
10,
11].
An overview of the current risk assessment framework across the MASLD–MetALD spectrum, including alcohol exposure assessment, sequential non-invasive fibrosis stratification, and future directions, is summarized in
Figure 1.
CURRENT RISK STRATIFICATION PARADIGM
Clinical and metabolic risk factors
Obesity, type 2 diabetes mellitus (T2DM), dyslipidemia, and metabolic syndrome in MASLD and MetALD
Obesity and T2DM are consistently associated with both prevalent and progressive fibrosis in MASLD, and therefore, are key components of modern risk-stratified care pathways. In people with T2DM, meta-analyses show a very high steatosis and clinically important fibrotic burden: MASLD prevalence is approximately 55.5%, with advanced fibrosis approximately 17.0% [
12]. An updated meta-analysis reported MASLD prevalence of 65.3%, with advanced fibrosis prevalence of 15.5% among individuals with T2DM [
13,
14]. In overweight and obese populations, a large systematic review and meta-analysis reported MASLD prevalence near 70.0% in overweight and 75.3% in obese individuals, with advanced fibrosis in approximately 6.7–6.9% among those with MASLD [
13].
Considering this high yield, current major society guidelines recommend systematic non-invasive fibrotic risk assessment in individuals with high CMRF burden, particularly those with T2DM or obesity with any additional CMRFs [
15,
16].
Dyslipidemia and the broader metabolic syndrome cluster frequently coexist with MASLD and contribute to overall risk, with MASLD commonly accompanied by an atherogenic lipid profile and increased cardiovascular risk [
17,
18]. Because MetALD is defined by metabolic dysfunction plus moderate alcohol exposure, these same cardiometabolic conditions are also central to MetALD; emerging studies emphasize that metabolic risk factors such as obesity, T2DM, hypertension, and dyslipidemia are common in MetALD and may act in concert with alcohol exposure to accelerate steatohepatitis and fibrosis progression [
19,
20].
Alcohol consumption spectrum and metabolic–alcoholic overlap
Alcohol exposure is associated with SLD outcomes in a dose-, duration-, and pattern-dependent manner, and its clinical impact becomes particularly relevant when metabolic dysfunction coexists. Systematic longitudinal data reviews consistently suggest that alcohol intake is not benign in SLD: a systematic review focusing on MASLD progression found that alcohol consumption, including levels within commonly recommended limits, was associated with worsening liver outcomes [
21]. Similarly, a meta-analysis evaluating light-to-moderate intake reported that cohort studies showed an increased advanced fibrotic risk and suggested higher cirrhotic and HCC risks associated with alcohol use in MASLD [
22]. At a population level, syntheses also emphasize that CMRFs and alcohol act as independent drivers interacting biologically and epidemiologically to increase the likelihood of severe liver disease, supporting a spectrum model rather than mutually exclusive categories [
23].
Mechanistic reviews broadly agree that metabolic dysfunction increases hepatic alcohol susceptibility, implying that the same alcohol intake may carry greater hepatic risk in people with obesity or insulin resistance. Proposed pathways include overlapping oxidative stress and mitochondrial injury, gut barrier disruption with inflammatory activation, and amplified immune and fibrogenic signaling [
24]. Accordingly, MetALD-oriented position statements recommend quantifying both recent and longer-term alcohol intake in all patients with SLD and reassessing exposure over time because changes in drinking and metabolic risk jointly influence prognosis and management decisions [
25]. In practice, the MetALD construct is intended to capture this overlapping phenotype and inform counseling and follow-up intensity, as well as future risk prediction models that integrate cardiometabolic and alcohol-related determinants [
25,
26].
Objective alcohol biomarkers: focus on phosphatidylethanol (PEth)
PEth is a direct alcohol biomarker formed exclusively in the presence of ethanol via phospholipase D‑mediated transphosphatidylation, accumulating in red blood cell membranes and providing a detection window of approximately 4 weeks [
27]. Its elimination half‑life ranges from 4 to 10 days (≈5 days on average) [
28].
The most commonly referenced PEth cut‑offs come from the Basel consensus: <20 ng/mL compatible with abstinence or low alcohol intake, 20–199 ng/mL indicating alcohol consumption, and ≥200 ng/mL suggestive of chronic excessive drinking. However, this consensus has been criticized for limited evidence integration [
29], and these thresholds should not be used in isolation to assign individuals to drinking categories. They are best interpreted alongside self‑reported alcohol intake and clinical context. Alternative thresholds exist, such as the Swedish harmonized cut‑off (<35 ng/mL for sobriety, >210 ng/mL for harmful drinking) and the Norwegian HUNT4 study cut‑offs (e.g., >40 ng/mL corresponds to >12 g alcohol/day) [
27,
30]. In a large individual patient data meta‑analysis, PEth ≥8 ng/mL showed 81.8% sensitivity for detecting unhealthy alcohol use, and sensitivity was influenced by biological factors [
31]. In a recent prospective, multi‑ethnic cohort study of 374 overweight or obese individuals with SLD, PEth demonstrated robust diagnostic accuracy for differentiating MetALD from MASLD (area under the receiver operating characteristic curve [AUROC] 0.81; 95% confidence interval 0.73–0.89), with an optimal cut‑off of 25 ng/mL based on the Youden index. This threshold may help identify individuals with alcohol consumption in the MetALD range [
32].
Several factors influence PEth interpretation. Higher body mass index (BMI) lowers PEth for the same alcohol intake. Anemia and accelerated red cell turnover reduce PEth sensitivity, whereas advanced fibrosis increases it [
31]. In dried blood spots, a 5% rise in hematocrit increases PEth by ≈70 ng/mL. Medications, endogenous ethanol production, and use of alcohol‑containing mouthwash or hand sanitizer have no clinically significant impact on PEth levels [
27].
PEth kinetics differ between drinking patterns: repeated moderate drinking leads to stepwise accumulation, whereas a single binge episode produces a distinct kinetic curve, and serial measurements can differentiate chronic daily intake from episodic heavy drinking [
27,
28]. Underreporting of alcohol intake is common; in a US study, only one‑third of patients had self‑reported use consistent with PEth levels (57.7% underreporting) [
33]. To address discrepancies, a triangulation approach combining PEth, self‑reported alcohol intake, and Alcohol Use Disorders Identification Test‑Consumption (AUDIT‑C) is recommended. A five‑step process includes: (1) measure PEth; (2) triangulate alcohol intake; (3) interpret results; (4) resolve discordance through non‑judgmental patient dialogue and exploration of causes; and (5) consider repeat testing [
27]. For positive PEth with patient denial, a constructive communication strategy is advised [
34].
In MetALD, PEth objectively quantifies alcohol intake. When used as part of a triangulation approach combining self‑reported alcohol intake, AUDIT‑C, and clinical history, it can help corroborate the differentiation between MASLD, MetALD, and ALD, reveal underreporting, monitor response to alcohol reduction interventions, and inform treatment decisions – for example, an expert panel recommends against resmetirom in patients with PEth >200 ng/mL, which suggests MetALD/ALD rather than pure MASLD [
27,
35]. However, PEth alone should not be used as a standalone rule to exclude patients from therapy; the decision should integrate the full triangulation of alcohol exposure data. Limitations include that PEth cannot quantify long‑term alcohol consumption in grams per week, reflects only the preceding 4 weeks of drinking without capturing long‑term drinking patterns, and requires validation of cut‑offs in MetALD. It cannot reliably assign individuals to the consensus categories of MASLD, MetALD, or ALD on its own, nor can a single PEth threshold serve as a standalone classification rule. Therefore, interpretation must always integrate clinical context, self‑reported alcohol intake, drinking pattern, AUDIT-C score, and repeat testing. Triangulation with validated questionnaires and clinical assessment remains essential [
27,
36].
Genetic predisposition (PNPLA3, TM6SF2, MBOAT7, HSD17B13, and beyond)
Common inherited variants contribute to heterogeneity in MASLD and MetALD.
PNPLA3 I148M (rs738409) is most strongly associated with severe steatosis, steatohepatitis, advanced fibrosis, and HCC [
37-
39].
TM6SF2 E167K (rs58542926) and
MBOAT7 (rs641738) are linked to fibrotic progression [
38,
40]. These effects extend to the metabolic–alcohol overlapping phenotype:
PNPLA3 also worsens ALD outcomes [
41,
42], while loss-of-function variants in
HSD17B13 are protective [
43]. Together, these findings support incorporating common variants into polygenic risk scores (PRSs) and other precision risk stratification approaches across the MASLD-MetALD spectrum [
38,
42,
44].
Fibrosis-4 index (FIB-4), NAFLD fibrosis score (NFS), aspartate aminotransferase to platelet ratio index (APRI): strengths, limitations, and cutoff optimization
Simple serum-based scores including FIB-4, NFS, and APRI are widely used for initial fibrotic risk assessment in SLD due to their low cost and accessibility. FIB-4 is endorsed as the preferred first-line test in stepwise algorithms because of its high negative predictive value at lower cutoffs, enabling efficient exclusion of advanced fibrosis in at-risk populations [
15,
16,
45]. However, these scores have limitations: age directly influences FIB-4 and NFS, increasing false positives in older adults; therefore, age-adjusted cutoffs have been proposed and validated to improve classification accuracy [
9,
46]. Performance also varies across metabolic phenotypes (e.g., severe obesity, T2DM), and discordance with elastography is common [
47]. Therefore, confirmatory second-tier tests such as vibration-controlled transient elastography (VCTE) or enhanced liver fibrosis (ELF) are recommended [
48].
Regarding MetALD, a dedicated evaluation is particularly important. Alcohol exposure can influence aminotransferase levels and platelet counts, potentially affecting score calibration. Emerging studies specifically examining MetALD populations indicate that FIB-4 retains acceptable sensitivity and negative predictive value for advanced fibrotic detection and can be incorporated into two-step strategies with elastography for risk stratification and prognostication [
36,
49]. These data support continued simple fibrotic score use across the MASLD–MetALD spectrum, while highlighting the importance of threshold optimization and sequential testing to account for age, metabolic phenotype, and alcohol exposure.
Agile 3+/4, steatosis-associated fibrosis estimator (SAFE) score, and newer composite models
Composite noninvasive models integrating liver stiffness measurement (LSM) with routine clinical and laboratory variables have been developed to improve advanced fibrotic and cirrhotic determination and reduce the proportion of patients classified as indeterminate compared with simple serum scores alone. Within VCTE-based approaches, Agile 3+ (targeting advanced fibrosis) and Agile 4 (targeting cirrhosis) were developed and validated in MASLD populations to improve positive predictive value and reduce indeterminate classifications compared with LSM alone or simple serum-based scores [
50]. Subsequent external evaluations in MASLD cohorts supported their diagnostic performance and potential utility for risk stratification and outcome prediction in routine practice [
51].
In parallel, the SAFE score was designed to improve discriminatory performance compared with conventional serum fibrotic scores to detect significant fibrosis and enhance triage performance in primary care and population screening settings, where disease prevalence and referral capacity are key constraints [
52,
53]. Additional validation studies and comparative analyses suggested that SAFE can improve classification efficiency relative to traditional serum fibrosis scores in some settings, although calibration and performance may vary by age structure, ethnicity, and underlying fibrosis prevalence, underscoring the need for region- and setting-specific validation before widespread adoption [
53,
54]. Although the SAFE score outperforms conventional serum fibrosis scores, it requires a greater number of input variables, limiting its clinical utility and hindering broader implementation. Other emerging composite models, including FAST, liver stiffness-based scores, and biomarker-integrated approaches, have also been developed to improve fibrosis risk stratification, although their performance requires further external validation across diverse populations (
Table 1) [
55-
65].
Evidence for composite models in MetALD is limited. The SAFE score is the only NIT directly validated in MetALD. In NHANES III (n=9,939, 230 with MetALD), it effectively stratified long‑term survival (log‑rank
P<0.0001). Compared with low‑risk (SAFE<0), intermediate‑ and high‑risk MetALD patients had all‑cause mortality HRs of 1.31 and 1.90 (both
P<0.05). These data indicate that the SAFE score is applicable for long‑term risk stratification across MASLD, MetALD and ALD, and can identify low‑risk individuals (SAFE<0) in primary care who have a survival similar to those without SLD. Limitations include the dated nature of the dataset (NHANES III 1988–1994), the modest sensitivity of ultrasound for diagnosing steatosis, potential underreporting of alcohol intake, lack of direct fibrosis validation (no liver biopsy), and a relatively small MetALD subgroup, although statistical power was adequate [
66].
Intermediate zones and metabolic confounding limitations
A persistent operational limitation of simple serum fibrosis scores is the large intermediate-risk category, which is particularly common in T2DM and obesity-enriched populations and substantially increases downstream testing and referral burden in stepwise pathways [
9,
67]. This intermediate zone reflects the trade-off between sensitivity at lower thresholds and specificity at higher thresholds and becomes more problematic as pre-test probability rises and comorbidity profiles diverge from the populations in which the scores and the cutoffs were originally derived [
9,
68].
Metabolic comorbidities can confound score inputs and contribute to misclassification. Severe obesity and related metabolic states may alter transaminase patterns and platelet counts and reduce the specificity of some composite scores; for example, performance degradation and the need for BMI-adjusted thresholds were highlighted for NFS in obese individuals [
69]. As a result, pathway refinements increasingly emphasize threshold adaptation (including age-adjusted cutoffs where appropriate), earlier use of second-line liver-specific tests in intermediate-to-high-risk phenotypes, and alternative or additive markers to reduce indeterminate classifications while maintaining safety [
68,
70]. Sequential strategies that deploy a second-tier blood test in intermediate cases were proposed to reduce unnecessary elastography and referrals while preserving case detection, illustrating one pragmatic approach to managing intermediate zones within resource-constrained systems [
71].
Intermediate zone challenges may be amplified in MetALD because alcohol exposure influences aminotransferases and platelet counts. Therefore, efficient pathways for populations spanning MASLD and MetALD place additional emphasis on quantitative alcohol assessment, confirmatory intermediate/high-risk strata testing, and model calibration to local prevalence and care setting [
9,
71].
VCTE
VCTE is the most widely used second-step imaging tool after initial FIB-4 triage in SLD risk stratification. Guideline-recommended sequential pathways use VCTE to refine the likelihood of advanced fibrosis or cirrhosis and guide referral decisions [
9]. VCTE-based LSM also provides prognostic value for liver-related outcomes [
72], and combining serum scores with VCTE achieves performance similar to performing VCTE in all individuals [
73,
74].
Evidence directly validating FIB-4 followed by VCTE pathways in MetALD is emerging but remains less mature than in MASLD. A recent prospective study in individuals with suspected MetALD evaluated VCTE against magnetic resonance elastography (MRE) and reported good performance of VCTE for significant fibrosis, and a stepwise strategy using FIB-4 followed by VCTE achieved a low false-negative rate for significant fibrotic detection [
75]. These data support the feasibility of extending established sequential pathways to MetALD, while reinforcing the need for careful consideration of alcohol-related effects on aminotransferases and platelets and additional validation across regions and care settings [
75-
80].
MRE
MRE provides high diagnostic accuracy for fibrotic SLD staging and is more robust than ultrasound-based elastography in individuals with obesity, as BMI has limited impact on its performance in MASLD cohorts [
81]. Meta-analyses consistently report higher AUROC for MRE than VCTE in detecting significant and advanced fibrosis [
82,
83], and an individual participant data pooled analysis confirmed its diagnostic superiority across fibrotic stages [
84]. Practical stiffness thresholds for advanced fibrosis cluster around 3.6–3.8 kPa in MASLD populations, though cutoffs remain context-dependent (
Table 2) [
85,
86].
Dedicated MRE validation data specific to MetALD remain comparatively limited, but MetALD-focused reviews emphasize that fibrotic risk assessment should use the same structured, noninvasive framework as MASLD, with MRE serving as a valuable option when obesity or other factors compromise ultrasound-based measurements and precise staging is required [
19,
26].
Limitation of ultrasound-based elastography
Ultrasound-based elastography, while highly effective for fibrosis assessment, has several limitations affecting its accuracy and clinical utility. Inflammatory activity independently influences VCTE-based LSM in MASLD, potentially leading to fibrotic severity overestimation in some patients (e.g., inflammation drives higher LSM values even when fibrosis stage is constant) [
87,
88]. Obesity remains a notable factor affecting the technical performance of ultrasound-based elastography: high BMI can interfere with ultrasound wave transmission and contribute to higher failure rates or unreliable measurements, particularly if appropriate probes are not used [
9,
89-
92]. In practice, these challenges may necessitate alternative or complementary diagnostic methods (e.g., MRI‑based elastography), especially in individuals with severe obesity or other factors that limit successful ultrasound-based elastography acquisition.
Role and limitations of biopsy in the MASLD-MetALD era
Liver biopsy remains the reference standard for fibrotic staging and diagnosing metabolic dysfunction-associated steatohepatitis (MASH) and trial eligibility, but its population-level use is limited by invasiveness, cost, and scalability, and sampling variability. Therefore, it is reserved for discordant noninvasive tests or high-stakes decisions [
9,
93].
In MetALD, biopsy can be particularly useful because metabolic dysfunction and alcohol exposure frequently coexist and histologic patterns may overlap, making etiologic attribution and staging more challenging. Expert consensus and MetALD-focused reviews emphasize integrating biopsy findings with quantitative alcohol assessment and the broader clinical context, and using biopsy selectively when noninvasive tools cannot answer clinically important questions [
25,
26].
Cytokeratin-18, ELF test, PRO-C3, and fibrosis panels
Patented or specialized serum tests can improve fibrotic classification beyond simple scores, particularly when used as a second/third step in structured pathways [
9]. The ELF test was evaluated in diagnostic-accuracy meta-analyses, showing good performance for advanced fibrotic staging across chronic liver diseases, including MASLD [
94,
95]. Cytokeratin-18 (CK18; M30/M65) identified MASH and fibrosis within MASLD with variable performance across studies [
96,
97]. For fibrogenesis markers, PRO-C3 has an overall diagnostic accuracy of around 0.8 for diagnosing significant or advanced fibrosis in MASLD [
98]. PRO-C3-derived panels, such as ADAPT, which combines PRO-C3 with clinical variables, were developed to improve advanced fibrosis identification [
99].
American Gastroenterological Association (AGA) and European Association for the Study of the Liver (EASL)
AGA and EASL-oriented pathways emphasize targeted case-finding in at-risk groups for detecting advanced fibrosis and a sequential two-step strategy anchored by FIB-4 followed by elastography or ELF, with further confirmatory tests when results are discordant [
5]. Beyond cross-sectional fibrotic detection, the prognostic performance of this two-step approach has been validated for long-term liver-related events (LREs): a large MASLD study showed that FIB-4 followed by VCTE-based LSM effectively stratified future LRE risk, and using LSM to further classify the intermediate FIB-4 group reduced the intermediate-risk zone while maintaining predictive performance [
74]. For HCC surveillance, one study proposed noninvasive risk-based thresholds and explicitly evaluated a two-step strategy, concluding that surveillance should be offered to MASLD patients with FIB-4 ≥3.25 or LSM ≥20 kPa, and that when a two-step approach is used, LSM ≥15 kPa among those with increased FIB-4 identifies a subgroup with annual HCC incidence exceeding 1% [
100].
In MetALD, the same structured two-step fibrosis risk stratification is applicable, but accurate alcohol exposure quantification is integral to classification and longitudinal management [
25]. Prospective data in suspected MetALD support the feasibility of stepwise algorithms incorporating FIB-4 followed by elastography for clinically significant fibrosis detection [
75], and additional MetALD-focused evaluations support FIB-4 as a practical initial triage tool [
49]. Baseline FIB‑4 stratification provides robust long‑term risk discrimination: among MetALD patients, the 5‑year cumulative incidence of adverse liver outcomes is approximately 10‑fold higher in those with FIB‑4 >2.67 than in those with FIB‑4 <1.3 (18.7% vs. 1.8%), and the 10‑year incidence is 27.0% vs. 3.5% [
101]. These data validate that widely available non‑invasive tests and their sequential application are applicable to MetALD, using established cut‑offs to diagnose advanced fibrosis and predict clinical outcomes (
Fig. 2) [
36,
102]. For patients aged ≥65, an age‑adjusted FIB‑4 threshold of <2.0 is recommended to define low risk, while higher thresholds should prompt further evaluation with VCTE [
15,
103,
104].
Korean Association for the Study of the Liver (KASL) and other algorithms
Regional societies have converged on sequential, resource-scalable pathways similar to AGA/EASL. The KASL MASLD guideline and the later KASL noninvasive test guideline endorse stepwise risk stratification that begins with a simple serum fibrosis score (commonly FIB-4) and proceeds to elastography for those above a low-risk threshold, with escalation to additional tests when needed [
103,
105]. Real-world validation of the KASL two-step strategy (FIB-4 followed by VCTE) in a large MASLD cohort showed graded separation of LRE risk across strata, supporting feasibility for routine care implementation [
74].
Discordance management has also been evaluated. In a tertiary-center MASLD cohort, discordant FIB-4 and VCTE liver stiffness results were common in approximately 30% of patients, and outcomes aligned more closely with LSM than with FIB-4 in discordant pairs, supporting structured follow-up and confirmatory testing when serum-based tools and elastography disagree [
47]. Discordant FIB-4/LSM patterns were further highlighted as a practical challenge for coherent risk stratification, emphasizing the need for clinically interpretable resolution strategies within two-step pathways [
106].
Comparison of different care pathways and management of low‑risk patients
Table 3 compares the latest guidelines from American Association for the Study of Liver Diseases (AASLD), AGA, EASL, and KASL. All four endorse a sequential strategy with similar thresholds: FIB‑4 <1.3 (<2.0 for age ≥65) defines low risk, VCTE <8 kPa excludes advanced fibrosis, and VCTE ≥20 kPa suggests cirrhosis. However, several differences exist. For second‑tier low‑risk definition, AASLD and KASL use VCTE <8 kPa or ELF <9.8, while AGA adopts a lower ELF threshold (<9.2) and explicitly allows ELF as a substitute when VCTE is unavailable; EASL emphasizes VCTE <8 kPa as the core criterion, with ELF only as an alternative.
Beyond initial risk stratification, the four guidelines also provide detailed recommendations for managing patients who remain at low risk after second‑tier testing (
Table 4). While all guidelines define low risk as FIB‑4 <1.3 (or age‑adjusted <2.0) plus VCTE <8 kPa (or ELF below specified thresholds), their follow‑up schedules and triggers for earlier re‑assessment differ: AGA recommends FIB‑4 every 1–2 years or VCTE every 3 years; AASLD and KASL suggest every 1–2 years (or every 2–3 years in lower‑risk patients), whereas EASL offers a flexible 1–3-year range. Triggers for earlier re‑assessment are explicitly listed by AGA and KASL (e.g., significant weight gain, worsening glycemic control, and new‑onset T2DM), while AASLD and EASL mention more generic “new metabolic risk factors or weight increase”. Regarding ELF, AGA and AASLD provide specific rule‑out/rule‑in thresholds, EASL mentions ELF only as an alternative without a clear cut‑off, and KASL references ELF but does not make it a primary recommendation [
15,
16,
103,
104]. These differences reflect regional variations in healthcare resources and disease epidemiology, yet the core risk‑stratification logic remains consistent across all four guidelines.
Is there any MetALD algorithm?
Most structured pathways developed for MASLD are not yet fully mature for MetALD, where alcohol exposure and metabolic dysfunction jointly influence classification and risk [
25]. A pragmatic MetALD‑specific algorithm can be proposed integrating objective alcohol quantification and sequential NIT‑based risk stratification (
Fig. 3). Alcohol intake is quantified using a triangulation approach (AUDIT‑C, self‑reported alcohol intake, and PEth; PEth ≥25 ng/mL to capture underreporting) [
25,
36], enabling classification into MASLD, MetALD, or ALD. Fibrotic assessment follows a sequential strategy: FIB‑4 as the first‑tier test, with an age‑adjusted threshold of <2.0 defining low risk for patients aged ≥65 [
15,
103,
104], while those with FIB‑4 ≥1.3 proceed to VCTE [
49,
75]. High‑risk escalation criteria include VCTE ≥12 kPa, persistent ALT/AST elevation, or signs of portal hypertension [
16]. This stepwise algorithm offers a clinically implementable framework for risk‑based management of MetALD. Of note, the fibrosis risk stratification component is extrapolated from major society guidelines developed for MASLD populations [
15,
16,
103,
104], as dedicated validation of these thresholds in MetALD populations remains limited [
36,
49]. Therefore, the algorithm should be applied with caution, particularly regarding the interpretation of FIB‑4 and LSM in the context of active alcohol consumption, until MetALD‑specific data become available.
LONG-TERM OUTCOMES AND PROGNOSTIC MODELING IN MASLD AND METALD
Natural history
Most patients with MASLD and MetALD have simple steatosis, but some progress to MASH and advanced fibrosis, increasing risks of decompensation and HCC (which can also occur without cirrhosis) [
10,
107]. Underlying cardiometabolic factors also raise risks of cardiovascular events, CKD, and extrahepatic cancers, with cardiovascular and metabolic outcomes being leading causes of death.
Comparative outcomes
In MetALD, the coexistence of metabolic dysfunction and alcohol exposure accelerates fibrosis progression, resulting in a higher cirrhotic risk, hepatic decompensation, and HCC compared with MASLD alone. Although both diseases share similar pathways to advanced fibrosis and cirrhosis, the synergistic interaction of alcohol and CMRFs leads to more rapid fibrotic progression and greater susceptibility to liver‑related complications. Surveillance for HCC should be tailored to these populations, with particular attention to those with advanced fibrosis and elevated noninvasive tests [
10,
100]. Consistent evidence shows that alcohol exacerbates fibrosis in individuals with metabolic dysfunction, while metabolic syndrome increases vulnerability to alcohol‑related liver injury. Meta‑analyses further demonstrate that even moderate alcohol intake accelerates fibrotic progression in MASLD, with higher rates of liver‑related events than in those with low or no alcohol consumption [
108]. Systematic reviews confirm that patients with combined metabolic and alcohol risk factors have higher incidence of liver events and mortality, supporting MetALD as a distinct clinical entity and reinforcing the need for risk‑stratification models that account for both determinants [
10,
109]. These findings further highlight the rationale for MetALD as a distinct clinical entity and support joint risk models that consider both metabolic and alcohol factors in liver disease prognosis [
10].
Alcohol’s synergistic or multiplicative effect with metabolic stress
Metabolic dysfunction and alcohol consumption share overlapping pathogenic pathways in SLD, including oxidative stress, inflammation, and insulin resistance. Cohort evidence suggests that their coexistence amplifies liver damage and accelerates progression toward steatohepatitis and fibrosis, providing a conceptual basis for MetALD [
110].
LSM-based risk prediction
VCTE-based LSM is increasingly used to predict long-term LREs in MASLD. Studies show that baseline LSM values predict cirrhosis, hepatic decompensation, and HCC, with higher LSM thresholds linked to worse outcomes. Importantly, VCTE has demonstrated comparable prognostic accuracy to liver histology in predicting LREs, supporting its use as a noninvasive alternative for risk stratification [
111]. Sequential approaches, in which LSM is combined with simple serum scores like FIB‑4, improve risk stratification and help guide follow-up and referral [
9,
108].
Although less data is available for MetALD, emerging evidence supports LSM use in this population, where it helps assess fibrotic risk in individuals with both metabolic dysfunction and alcohol exposure [
10].
Integration with comorbidity burden (CVD, CKD, and malignancy)
MASLD is closely linked to increased risks of CVD, CKD, and extrahepatic malignancies, with liver fibrotic severity correlating with worse outcomes across these domains. Systematic reviews and cohort studies confirm higher rates of CVD and CKD in MASLD patients than in the general population, and clinical guidelines recommend integrated cardiovascular and renal risk assessment to optimize management and surveillance [
6,
9,
11]. In MetALD, the combined impact of metabolic dysfunction and alcohol exposure amplifies liver and extrahepatic complication risks, reinforcing the need for integrated assessment beyond liver-focused measures [
112].
Current HCC risk scores: PAGE-B, age–male–albumin–bilirubin–platelets score (aMAP), and MASLD-HCC score
HCC risk models like aMAP have been validated across various chronic liver diseases, demonstrating effective long‑term LRE prediction in viral and non-viral populations. aMAP combines clinical variables such as age, sex, albumin, bilirubin, and platelet count, and exhibits moderate HCC accuracy [
113]. However, existing HCC risk models (e.g., aMAP) lack MetALD‑specific factors, and no validated prediction scores exist for MetALD; calibration and transportability remain challenging in this multi‑etiology population [
9,
114].
The need for disease-specific models incorporating metabolic and alcohol factors
Patients with MetALD have a higher HCC incidence than those with MASLD, supporting a synergistic effect of combined metabolic and alcohol risk factors [
115]. Accurate alcohol quantification is challenging due to self‑report under-reporting, but biomarkers such as PEth can improve phenotyping. Future HCC risk models should integrate alcohol intake, fibrosis severity, and metabolic burden [
27].
Role of genetic and artificial intelligence (AI)-based predictors
Genetic variants, such as
PNPLA3, TM6SF2, GCKR, MBOAT7, and
HSD17B13, are strongly associated with MASLD progression and HCC risk. PRS that integrate multiple genetic loci have shown better predictive power for these outcomes compared with single variants. Machine learning (ML) models increasingly used to predict HCC and other LREs in MASLD show promising performance in stratifying patients based on clinical, imaging, and genetic data. While specific studies on MetALD are limited, similar approaches could be applied in this population [
10].
BEYOND CURRENT PRACTICE: PRECISION AND DYNAMIC MASLD AND METALD RISK ASSESSMENT
Multidimensional risk layers
Genetic, metabolic, inflammatory, microbiome, and lifestyle axes
MASLD is a multifactorial disease, where genetic variants like
PNPLA3, TM6SF2, and
MBOAT7, combined with metabolic dysfunction and systemic inflammation, drive progression to steatosis and HCC. PRS incorporating these factors shows better predictive value than single variants. The gut microbiome plays a key role in MASLD through increased permeability and systemic inflammation, contributing to liver injury. Dysbiosis represents a potential target for precision risk models. Lifestyle factors, including diet and physical activity, modulate MASLD risk and progression, highlighting the need for models that integrate behavior with genetic and metabolic factors [
9]. In MetALD, alcohol exposure compounds metabolic and genetic risks, necessitating models that combine alcohol dose, metabolic markers, and fibrosis severity to fully assess risk in this population [
10].
ML and PRS
PRS have been increasingly applied in MASLD research, integrating multiple genetic variants to assess susceptibility to disease, fibrosis progression, and HCC risk. Systematic reviews indicate that PRSs, compared with single variant associations, show better predictive potential in identifying patients with high-risk MASLD and are significantly linked to liver fibrosis and severe liver disease outcomes [
116]. These integrative genetic models improve risk stratification and provide a foundation for personalized risk prediction. Integrating genetic, metabolic and alcohol factors may improve risk stratification in MetALD [
10].
ML and PRS enhance MASLD and MetALD risk assessment across multiple domains. In imaging, deep learning automates fibrosis staging and reduces variability [
117]. For NITs, ML integrates multiple serum biomarkers to improve diagnostic accuracy [
118]. For longitudinal risk, electronic health record (EHR)‑based models can predict mortality and hepatic decompensation [
119]. In HCC prediction, ML models achieve high discrimination (AUROC 0.97) [
120]. For treatment response, ML may predict response to resmetirom or semaglutide [
121,
122]. However, limitations include insufficient external validation, poor explainability, and limited dataset diversity [
123], constraining clinical translation, especially in MetALD [
10]. PRS incorporating
PNPLA3, TM6SF2, and
MBOAT7 outperform single variants [
116], but predictive power remains modest.
Continuous risk modeling rather than dichotomized categories
In MASLD and MetALD, risk factors such as metabolic dysfunction and alcohol intake exhibit gradual, dose‑response relationships with liver injury and fibrosis progression, rather than fitting into binary categories. Similarly, alcohol exposure increases fibrotic and HCC risk in a continuous manner, with higher consumption linked to more severe liver outcomes, emphasizing the need for models that account for gradual increases in risk rather than arbitrary thresholds [
10].
Joint alcohol-metabolic risk scoring (e.g., MetALD index proposal)
While standardized MetALD‑specific indices are still under development, recent proposals integrate fibrotic severity (LSM/ELF/PRO‑C3), metabolic burden, and alcohol biomarkers to enhance classification and disease outcome prediction. PEth, a reliable marker of recent alcohol intake, helps overcome self-report bias and improves alcohol exposure assessment in MetALD [
27]. Noninvasive fibrosis tests (ELF and PRO‑C3) are valuable for staging and prognosis. When combined with alcohol biomarkers (PEth), these markers form the foundation of a MetALD-specific risk index, enabling better disease severity stratification [
124]. These models allow clinicians to integrate fibrosis, metabolic burden, and alcohol exposure into a cohesive risk profile, aiding personalized risk prediction for LREs and guiding MetALD treatment decisions [
10].
FUTURE DIRECTIONS AND UNMET NEEDS
Despite growing recognition of MetALD as a distinct clinical entity, significant knowledge gaps remain across the risk assessment pathway.
Table 5 summarizes the grading of strength and source of evidence for each major NIT and PEth specifically in MetALD populations. As shown, FIB‑4, VCTE, and ELF have the strongest direct evidence, whereas Agile scores and MRE lack dedicated MetALD‑specific validation, highlighting critical knowledge gaps that require further investigation.
Table 6 summarizes the key areas of uncertainty and unresolved questions that currently limit the implementation of evidence‑based, MetALD‑specific risk stratification. These gaps span diagnostic algorithms, prognostic modeling, therapeutic monitoring, health equity, and the integration of emerging technologies. Addressing them will be essential to advance precision care across the MASLD-MetALD spectrum. The following sections elaborate on each domain and propose specific research priorities to fill these gaps.
MetALD-specific NIT cutoff calibration
Evidence for composite non‑invasive tests specifically calibrated in MetALD remains limited, and alcohol exposure can influence aminotransferases and platelet counts, thereby affecting model calibration and optimal threshold [
49]. Intermediate‑zone challenges may be amplified in MetALD because alcohol intake may shift patients across risk thresholds [
9,
71]. Although commonly used NITs (FIB‑4, VCTE, and ELF) retain reasonable accuracy for advanced fibrosis screening in MetALD, dedicated validation and threshold refinement are needed [
49]. Furthermore, the impact of alcohol biomarkers on the performance of fibrotic scores may differ by population, metabolic burden, and amount of alcohol consumed, necessitating validation across diverse populations [
10]. Future research should focus on: (i) prospective multi‑center studies to establish MetALD‑specific cut‑offs for FIB‑4, VCTE, MRE, and ELF; (ii) evaluating the influence of drinking patterns (binge versus daily moderate intake) on NIT performance; and (iii) developing alcohol‑adjusted scoring algorithms that incorporate objective biomarkers such as PEth.
Standardization of PEth and other alcohol biomarkers
PEth cut‑offs require validation specifically in MetALD populations, as current thresholds were derived mainly from MASLD or general population cohorts [
27,
36]. Interpretation is further complicated by biological confounders (BMI, anemia, and advanced fibrosis) and the lack of standardization across analytical platforms. Future research should focus on: (i) multi‑center studies to define optimal PEth thresholds for discriminating MASLD, MetALD, and ALD across diverse ethnicities; (ii) assessing the impact of liver disease severity on PEth pharmacokinetics; and (iii) establishing harmonized laboratory protocols to ensure cross‑platform comparability.
Development of integrated alcohol metabolic risk indices
Standardized MetALD‑specific risk indices are still under development. Recent proposals have combined fibrotic severity (LSM/ELF/PRO‑C3), metabolic burden, and alcohol biomarkers (PEth) to enhance classification and outcome prediction [
27,
124]. However, prospective validation and determination of the relative weights of each component are lacking. Future research should focus on: (i) developing and externally validating a continuous risk score that integrates fibrosis markers, metabolic traits, and quantitative alcohol exposure (e.g., PEth) in large, multi‑ethnic MetALD cohorts; and (ii) establishing the proportional contribution of alcohol versus metabolic factors to overall risk.
MetALD-specific HCC risk scores
No validated HCC prediction scores currently exist for MetALD. Existing general models (e.g., aMAP) omit MetALD‑specific factors and fail to account for the interaction between metabolic dysfunction and alcohol exposure [
114]. Although future SLD risk models should integrate alcohol consumption, fibrosis severity, metabolic burden, and genetic modifiers such as
PNPLA3, dedicated MetALD‑focused research is needed. The relative weight of alcohol versus metabolic factors in driving HCC risk should also be determined to guide surveillance intensity.
Integration of liver and extrahepatic (CVD, CKD, and malignancy) outcomes
MASLD and MetALD are increasingly recognized as multisystem conditions, with liver fibrosis severity correlating with elevated CVD, CKD, and extrahepatic cancer risks. Studies show that MASLD significantly increases the risk of major adverse cardiovascular events and cardiovascular mortality, with more severe liver fibrosis further amplifying these risks [
125]. In addition to CVD, MASLD is independently linked to CKD, driven by shared factors such as insulin resistance and chronic inflammation [
126]. Emerging evidence also highlights increased MASLD cancer risk, particularly gastrointestinal cancers, with a consistent association across diverse cohorts [
127].
For MetALD, the combined burden of metabolic dysfunction and alcohol exposure may further elevate CVD, CKD, and cancer risks, reinforcing the need for integrated risk models that incorporate both liver and extrahepatic outcomes [
10]. Therefore, even patients classified as low risk by fibrosis scores still require optimal management of CMRFs and counseling to reduce or stop alcohol use, as extrahepatic complications, particularly cardiovascular and metabolic outcomes, remain the most common morbidity and mortality causes in MASLD and MetALD.
Longitudinal validation of dynamic biomarkers
Prognostic models for MetALD should incorporate dynamic, longitudinal measurements of non‑invasive tests, yet the optimal frequency and magnitude of clinically meaningful change remain undefined. Prospective cohort studies with serial assessments (e.g., FIB‑4, VCTE‑LSM, or ELF every 1–2 years) are needed to characterize individual fibrosis trajectories and to determine thresholds that reliably predict progression to cirrhosis or hepatic decompensation. Furthermore, research should examine whether repeated NIT measurements offer incremental prognostic value over single baseline assessments, and whether integrating longitudinal alcohol biomarker data (e.g., serial PEth) with dynamic NIT changes can disentangle alcohol‑related fluctuations from true fibrotic evolution.
Therapeutic monitoring
Resmetirom and semaglutide are now approved by the US Food and Drug Administration as treatments for MASH with F2-F3 fibrosis. The pivotal phase 3 trials MAESTRO-NASH and ESSENCE demonstrated significant histological improvements with resmetirom and semaglutide, respectively, providing the regulatory basis for their approval [
128,
129]. With expanding treatment options, the ability to monitor therapeutic response non‑invasively and to define meaningful thresholds for response, non‑response, and treatment futility has become essential for clinical decision‑making. Besides GLP‑1 receptor agonists (GLP‑1RAs), SGLT2 inhibitors (SGLT2i) have been shown to improve liver enzymes and hepatic fat content on imaging, but further histological evidence is needed to support their anti-fibrotic efficacy. These therapies may help reduce LRE risk, although personalized predictions are needed for optimal use [
130].
Serial fibrotic measurements, such as VCTE-based LSM or serum markers like PRO‑C3, help track treatment response and predict long-term outcomes. Controlled attenuation parameter (CAP) is useful for quantifying hepatic steatosis but is not an established marker of treatment response to antifibrotic therapy or a validated surrogate for long‑term clinical outcomes. Dynamic biomarkers are essential for tailoring treatment and improving individualized risk prediction [
27]. Recent guidances from the AASLD and international consensus panels have proposed specific changes in NITs to define treatment response, as well as criteria for drug discontinuation or modification (
Table 7) [
131-
134]. In MetALD, the coexistence of metabolic dysfunction and alcohol exposure complicates therapeutic response, and NIT cut points may also differ in this population, making alcohol biomarkers (e.g., PEth) essential for personalized therapy and better risk stratification [
10]. However, current pharmacotherapies for MASLD across different stages have generally not been studied in patients with persistent heavy alcohol consumption.
External validation of risk models across diverse ethnicities, care settings, and diagnostic platforms is crucial. Studies show that fibrotic scores (FIB‑4, APRI, and NFS) vary in performance across racial groups, requiring tailored thresholds based on local prevalence and comorbidities [
135], and the impact of alcohol biomarkers on NIT accuracy may differ by population, metabolic burden, and amount of alcohol consumed [
10]. Therefore, external validation of risk models and NIT cut‑offs in large, multi‑ethnic MetALD cohorts is essential. Moreover, low‑resource settings often lack access to elastography or PEth testing, so simple, low‑cost algorithms (e.g., FIB‑4 combined with ELF or SAFE score) should be validated to screen for advanced fibrosis in primary care and to identify patients who would benefit from referral. Regionally tailored calibration may be needed to account for ethnic differences in alcohol metabolism and disease prevalence.
ML and PRS have shown promise for risk stratification in MASLD, and similar approaches could be extended to MetALD by integrating clinical, imaging, genetic, and alcohol‑exposure data.10 However, beyond improving predictive accuracy, future AI models must prioritize explainability, enabling clinicians to understand which features drive individual risk estimates, to foster trust and clinical adoption. Implementation research is also needed to evaluate how such models integrate into real‑world workflows, including EHR linkage and impact on decision‑making. Finally, standardized guidance for deploying these models across diverse healthcare settings should be developed, with attention to algorithmic fairness to avoid exacerbating health disparities.
CONCLUSION
Paradigm shift: from single-point static assessment
MASLD and MetALD are evolving toward dynamic risk prediction models that integrate clinical, metabolic, genetic, and fibrosis data along with alcohol biomarkers. This shift represents a departure from static assessments, enabling more personalized and accurate predictions of disease progression and liver‑related outcomes. The recognition of MetALD as a distinct clinical entity underscores the necessity of incorporating quantitative alcohol exposure, particularly through biomarkers such as PEth, into routine risk stratification, rather than relying on etiology‑based labels alone.
Future of risk stratification
Artificial intelligence and PRS will further refine risk stratification by combining clinical, imaging, and genetic data, offering the potential for real‑time, individualized risk assessments that can guide treatment decisions and surveillance intensity. However, the successful translation of these technologies into clinical practice will require rigorous external validation across diverse populations, standardization of biomarker assays, and careful attention to algorithmic fairness and explainability. Addressing these challenges will be essential to realize the promise of precision hepatology and to improve outcomes across the MASLD-MetALD spectrum.
FOOTNOTES
-
Authors’ contribution
H.L., H.W., K.Y., and M.L. contributed equally to this work by conducting the literature review, synthesizing the evidence, and drafting the manuscript. Y.W. assisted with literature analysis and manuscript editing. T.C.F.Y. and S.U.K. supervised the study, critically revised the manuscript, and coordinated the overall project. All authors read and approved the final manuscript.
-
Conflicts of Interest
SUK has served as an advisory committee member for Gilead Sciences, GSK, Bayer, Novo Nordisk, Sysmex, AstraZeneca, Roche Diagnostics, and Eisai. He is a speaker for Gilead Sciences, GSK, Bayer, Eisai, Abbive, Echo- Sens, MSD, Bristol-Myers Squibb, Novo Nordisk, Hanhwa, Yuhan, Samil, PharmaKing, Celltrion, Daewoong, Chong Kun Dang, Boryung, and Bukwang. He has also received research grants from Abbive, Bristol-Myers Squibb, and Gilead Sciences. TCFY has served as an advisory committee member for Altimmune, Gilead Sciences, GSK, and as a speaker for Gilead Sciences. He received a research grant from Gilead Sciences.
Figure 1.Risk assessment across the MASLD-MetALD spectrum. The figure illustrates the three core components of risk assessment across the MASLD–MetALD spectrum. Left: Patients range from purely metabolic-driven MASLD to Met-ALD with moderate alcohol intake. Alcohol exposure is assessed using a triangulation approach that integrates self‑reported intake (grams/week), drinking pattern, AUDIT‑C score, clinical history, and, when clinically indicated, the objective biomarker PEth. Center: Sequential non‑invasive risk stratification begins with the triangulation of alcohol consumption, followed by a two-step examination using FIB‑4 and VCTE. Most cutoffs derive from MASLD; in Met-ALD, recent drinking or alcoholic hepatitis may transiently elevate LSM, and thresholds may need recalibration. Right: Low‑risk patients receive lifestyle advice and periodic follow‑up; intermediate/high‑risk patients require specialist referral, HCC surveillance, and consideration of pharmacotherapy (resmetirom, semaglutide). Future directions include dynamic biomarkers, AI/PRS models, and personalized treatment monitoring. AI, artificial intelligence; AUDIT‑C, Alcohol Use Disorders Identification Test‑Consumption; FIB‑4, Fibrosis‑4 index; HCC, hepatocellular carcinoma; LSM, liver stiffness measurement; MASLD, metabolic dysfunction‑associated steatotic liver disease; MetALD, metabolic dysfunction and alcohol‑related liver disease; PEth, phosphatidylethanol; PRS, polygenic risk score; VCTE, vibration‑controlled transient elastography.
Figure 2.Sequential non‑invasive risk stratification for MASLD and MetALD using FIB‑4 and VCTE‑based LSM. The algorithm begins with identification of at‑risk individuals, followed by quantification of alcohol intake to classify patients into MASLD or MetALD. FIB‑4 serves as the initial triage test; patients with low risk are managed in primary care, while those with intermediate/high risk proceed to VCTE. The two‑step pathway stratifies patients into low‑ (<8 kPa), intermediate‑ (8–12 kPa), and high‑risk (>12 kPa) groups. For MASLD, 5‑year cumulative incidences of liver‑related events are 0–0.5%, 0.4–1%, and 6–11% across the three groups; corresponding HCC incidences are 0–0.3%, 0.2–0.6%, and 2–5%. For MetALD, emerging data show substantially higher event rates, with hepatic decompensation occurring in 0%, 20%, and 63% of low‑, intermediate‑, and high‑risk patients over 10–44 months of follow‑up [
36]. Patients classified as low risk should be managed in primary care, as they remain at risk for extrahepatic outcomes including cardiovascular disease and extrahepatic cancers [
102]. Longitudinal reassessment is suggested for low-risk patients: FIB‑4 every 1–2 years for patients with T2DM/prediabetes or ≥2 metabolic risk factors, or every 2–3 years for those without [
15,
16,
103,
104]. CVD, cardiovascular disease; FIB‑4, Fibrosis‑4 index; HCC, hepatocellular carcinoma; LRE, liver-related event; LSM, liver stiffness measurement; MASLD, metabolic dysfunction‑associated steatotic liver disease; MetALD, metabolic dysfunction and alcohol-related liver disease; T2DM, type 2 diabetes mellitus; VCTE, vibration‑controlled transient elastography.
Figure 3.Stepwise diagnostic and risk stratification algorithm for MetALD. The algorithm integrates objective alcohol quantification with sequential non‑invasive testing. Alcohol intake is assessed using a triangulation approach combining AUDIT‑C, self‑reported grams per week, and PEth (threshold ≥25 ng/mL to capture underreporting) [
25,
36]. Based on these assessments, patients are classified as MASLD, MetALD, or ALD, with repeat assessment recommended when self‑reported alcohol intake and PEth are discordant. Fibrotic risk stratification follows a sequential strategy: FIB‑4 serves as the first‑tier test (cut‑offs: <1.3 low risk, 1.3–2.67 intermediate, >2.67 high risk) [
49], with an age‑adjusted threshold of <2.0 defining low risk for patients aged ≥65. For patients aged ≥65 with FIB‑4 ≥2.0, the same intermediate/ high‑risk categorization applies, and further evaluation with VCTE is recommended [
15,
103,
104]. Patients with FIB‑4 ≥1.3 proceed to VCTE; values <8 kPa confirm low risk, 8–12 kPa indicate intermediate risk, and ≥12 kPa suggest advanced fibrosis warranting specialist referral [
75]. High‑risk escalation criteria include VCTE ≥12 kPa, persistently elevated ALT/AST, or signs of portal hypertension [
16]. This pragmatic algorithm provides a clinically implementable framework for risk‑based management of MetALD. ALD, alcohol-related liver disease; ALT, alanine aminotransferase; AST, aspartate aminotransferase; AUDIT‑C, Alcohol Use Disorders Identification Test‑Consumption; FIB‑4, Fibrosis‑4 index; LSM, liver stiffness measurement; MASLD, metabolic dysfunction‑associated steatotic liver disease; MetALD, metabolic dysfunction and alcohol‑related liver disease; PEth, phosphatidylethanol; SLD, steatotic liver disease; VCTE, vibration‑controlled transient elastography. *The disease classification component (alcohol triangulation and SLD subtyping) is supported by multi‑society Delphi consensus [
2,
25] and emerging MetALD‑specific data [
27,
36]. **The fibrosis risk stratification component (FIB‑4 → VCTE sequential testing and associated cutoffs) is extrapolated from major society guidelines developed for MASLD populations [
15,
16,
103,
104]. Dedicated validation of these thresholds in MetALD populations remains limited [
36,
49], and the algorithm should be applied with caution, particularly regarding the interpretation of FIB‑4 and LSM in the context of active alcohol consumption.
Table 1.Newer composite noninvasive models for fibrosis prediction
Table 1.
|
Models |
Target |
Key variables |
Advantages |
Limitations / Notes |
Performance metrics |
Training and validation cohort |
Independent external validation cohort |
|
Agile 3+ |
Advanced fibrosis (≥F3) |
LSM, age, sex, AST, ALT, platelets, diabetes status |
High positive predictive value, reduces indeterminate classifications |
Requires elastography; cohort-specific calibration may be needed |
Training AUROC 0.90 for ≥F3; external validation AUROC 0.90–0.91; [50] other external validation AUROC 0.82–0.916 [55-58] |
Training set (North America/Europe/Asia, n=2,134, mixed ethnicity); internal validation (n=700); external validation: NASH CRN (USA, n=585, mixed ethnicity) and French NAFLD cohort (France, n=1,042, predominantly White) [50] |
(i) Romania, n=246, biopsy-proven MASLD (Caucasian); [55] |
|
(ii) Europe, n=912, MASLD (multi-center, predominantly Caucasian); [56] |
|
(iii) Russia, n=43, biopsy-proven MASLD (Caucasian); [57] |
|
(iv) China, n=NR, patients with CHB and concurrent MASLD (Asian) [58] |
|
Agile 4 |
Cirrhosis (F4) |
LSM, sex, AST, ALT, platelets, diabetes status |
High positive predictive value, reduces indeterminate classifications |
Same as Agile 3+ |
Training AUROC 0.91; external validation AUROC 0.93 for F4; [50] other external validation AUROC 0.875–0.968 [55-57] |
Same as Agile 3+; validated in NASH CRN (USA, n=585, mixed ethnicity) and French NAFLD cohort (France, n=1,042, predominantly White) [50] |
(i) Romania, n=246, biopsy-proven MASLD (Caucasian); [55] |
|
(ii) Europe, n=912, MASLD (multi-center, predominantly Caucasian); [56] |
|
(iii) Russia, n=43, biopsy-proven MASLD (Caucasian) [57] |
|
FAST |
Fibrotic MASH (≥F2+ MASH) |
CAP+LSM+AST |
Moderate accuracy for fibrotic MASH |
Focused on MASH; less precise for isolated fibrosis |
Derivation AUROC 0.80 (95% CI 0.76–0.85); pooled external validation AUROC 0.85 (95% CI 0.83–0.87); [59] other external validation AUROC 0.83 [57] |
Training set: UK (n=350, predominantly White); external validation: French bariatric (France, n=110, White), USA screening (USA, n=242, mixed), China Hong Kong (Hong Kong, n=83, Asian Chinese), China Wenzhou (China, n=104, Asian Chinese), French NAFLD (France, n=182, White), Malaysian (Malaysia, n=176, Malay/Chinese/ Indian), and Turkish (Turkey, n=129, Caucasian) [59] |
Russia, n=43, biopsy-proven MASLD (Caucasian) [57] |
|
SAFE score |
Significant fibrosis (≥F2) |
Age, BMI, diabetes status, AST, ALT, globulins, platelets |
Better discrimination than traditional serum scores; suitable for primary care and population screening |
Requires more input variables; performance varies by age, ethnicity, baseline fibrosis prevalence |
Training AUROC 0.80; test set #1 0.80; test set #2 0.83; Hong Kong external validation AUROC 0.773; low‑risk (SAFE<0) survival similar to no NAFLD; [52] other external validation AUROC 0.75–0.753 [60,61] |
Training set: NASH CRN (USA, n=676, mixed ethnicity); test #1: FLINT trial (USA, n=280, mixed); test #2: Stanford MRE (USA, n=130, mixed); external validation: Hong Kong biopsy cohort (Hong Kong, n=279, Asian Chinese) and Hong Kong EHR cohort (Hong Kong, n=4,603, Asian Chinese); survival validation: NHANES III (USA, n=11,953, multi‑ethnic) [52] |
(i) South Korea, n=6,229, MASLD (Asian, TE reference); [60] |
|
(ii) Argentina, n=219, biopsy‑proven MASLD (Latin American) [61] |
|
LSS model |
Advanced fibrosis (≥F3) |
Body weight, race, diabetes status, AST, platelets, albumin |
Outperforms several traditional scores |
Requires elastography; population-specific thresholds |
F3+F4: AUROC 0.78 (95% CI 0.74–0.82); F4: AUROC 0.79 (95% CI 0.73–0.86) [62] |
Training set: NHANES 2017–2020 (USA, n=2,414, multi‑ethnic); validation: NHANES 2017–2018 (USA, n=1,562, multi‑ethnic) [62] |
No independent external validation identified |
|
Biomarker panel (IGFBP7, SSc5D, Sema4D) |
Fibrosis stage prediction |
GFBP7, SSc5D, Sema4D |
Outperforms traditional serum-only tests |
Early-stage; requires lab biomarker assays; ML model complexity |
F0/F1: AUROC 0.82; F2: AUROC 0.89; F3/F4: AUROC 0.87; independent validation: F0/F1 and F3/F4 AUROC 0.84 [63] |
Training set: Dutch cohort (Netherlands, n=128, European Caucasian); validation: Danish prospective cohort (Denmark, n=156, European Caucasian) [63] |
No independent external validation identified |
Table 2.Imaging-based assessment for fibrosis
Table 2.
|
Characteristics |
Vibration-controlled transient elastography |
Magnetic resonance elastography |
|
Purpose / Use |
Noninvasive fibrosis assessment; stepwise use after blood-based triage |
Noninvasive fibrosis staging; high-accuracy reference method |
|
Key features / Parameters |
Liver stiffness measurement (kPa), controlled attenuation parameter (dB/m) |
Liver stiffness measurement (kPa) |
|
Fibrosis-stage specific cut-off values [86] |
Exclusion of advanced fibrosis: <8 kPa; Indication of significant fibrosis (≥F2): ≥8 kPa; |
Exclusion of significant fibrosis (≥F2): <3.3 kPa (in combination with FIB-4); |
|
Indication of advanced fibrosis (≥F3), specialist referral recommendewd: ≥12 kPa; |
Identification of significant fibrosis (≥F2): ≥3.3 kPa (in combination with FIB-4); |
|
Diagnosis of cirrhosis (F4), requiring specialist management: ≥15 kPa; |
Diagnosis of advanced fibrosis (≥F3): ≥3.6 kPa; |
|
Threshold for predicting risk of portal hypertension or clinical decompensation events, and triggering HCC surveillance: ≥20 kPa |
Diagnosis of cirrhosis (F4): ≥4.7 kPa |
|
Factors affecting measurement reliability |
Obesity / narrow intercostal space – use XL probe or 2D-SWE; [86] |
Obesity – minimal effect; |
|
Ascites – may cause failure; MRE preferred; [9] |
Ascites – no interference; |
|
Right-sided heart failure – passive congestion raises LSM; [88] |
Iron deposition – may reduce signal quality; [90] |
|
Inflammation / cholestasis / steatosis – transient LSM elevation; [9] |
Right‑sided heart failure – similar effect as VCTE; [91] |
|
Postprandial state – fast ≥2 h recommended [89] |
Postprandial state – fast ≥4 h recommended [92] |
|
Advantages |
Widely available; sequential pathways reduce unnecessary tests; prognostic value for liver-related outcomes |
High diagnostic accuracy across fibrosis stages; less influenced by obesity; superior discrimination performance |
|
Limitations / Notes |
Performance affected by obesity, ascites; thresholds may require cohort-specific calibration |
Less accessible; higher cost; requires specialized equipment/software; limited MetALD-specific validation |
|
Validation status |
Extensively validated in MASLD; emerging prospective validation in MetALD; sequential FIB-4 → VCTE pathway shows low false-negative rate |
Validated in MASLD biopsy-proven cohorts; meta-analyses support superior performance vs. VCTE; limited but emerging data in MetALD |
Table 3.Comparison of key recommendations from AASLD, AGA, EASL and KASL guidelines for MASLD risk stratification and management
Table 3.
|
Aspect |
AASLD (2023) [16] |
AGA (2026) [104] |
EASL (2024) [15] |
KASL (2025) [103] |
|
Target population for screening |
T2DM, obesity with metabolic risk factors, family history of cirrhosis |
T2DM, overweight/obesity + ≥1 additional metabolic risk factor, imaging or liver enzyme abnormalities |
T2DM, abdominal obesity + ≥1 additional metabolic risk factor, persistently elevated liver enzymes |
T2DM, obesity, components of metabolic syndrome, imaging or liver enzyme abnormalities |
|
Screening tools |
First-line: FIB-4 |
First-line: FIB-4 |
First-line: FIB-4 |
First-line: FIB-4 |
|
Second-line: VCTE, ELF, MRE |
Second-line: VCTE (preferred), ELF (when VCTE unavailable), MRE |
Second-line: VCTE, ELF, MRE, SWE |
Second-line: VCTE, ELF, MRE, SWE |
|
FIB-4 thresholds |
Low risk: <1.3; for age ≥65 y use <2.0 (with caution) |
Low risk: <1.3; for age ≥65 y use <2.0 |
Low risk: <1.3; for age ≥65 y use <2.0 |
Low risk: <1.3; for age ≥65 y use <2.0 |
|
High risk: ≥1.3 (or ≥2.67 – indicates an increased probability of advanced fibrosis, refer directly) |
High risk: ≥1.3 (or ≥2.67 – refer directly) |
High risk: ≥1.3 (or >2.67 – hepatology referral) |
High risk: ≥1.3 |
|
VCTE thresholds |
Exclude advanced fibrosis: <8 kPa (high NPV for ruling out F3–F4); |
Exclude advanced fibrosis: <8 kPa; |
Exclude advanced fibrosis: <8 kPa |
Exclude advanced fibrosis: <8 kPa |
|
Suggest advanced fibrosis: ≥12 kPa (rule-in advanced fibrosis for specialist referral); |
Suggest advanced fibrosis: ≥8 kPa (prompts referral or further assessment); |
Suggest advanced fibrosis: 8–12 kPa (requires further assessment) |
Suggest advanced fibrosis: 8–12 kPa (requires further assessment) |
|
Suggest cirrhosis: ≥20 kPa (also serves as a trigger for HCC surveillance) |
Suggest cirrhosis: ≥15–20 kPa (combined with platelet count for portal hypertension risk assessment) |
Suggest cirrhosis: ≥15–20 kPa (combined with Baveno VII criteria; ≥20 kPa suggests high probability of cirrhosis and risk of decompensation) |
Suggest cirrhosis: ≥15–20 kPa (combined with Baveno VII criteria; ≥20 kPa suggests high probability of cirrhosis) |
|
ELF thresholds |
Exclude advanced fibrosis: <9.8 |
Exclude advanced fibrosis: <9.2 |
Exclude advanced fibrosis: <9.8 |
Exclude advanced fibrosis: <9.8 (reference) |
|
Suggest advanced fibrosis: ≥9.8 |
Suggest advanced fibrosis: ≥9.2 (9.2–9.7 requires individualized assessment) |
Suggest advanced fibrosis: ≥9.8 |
Suggest advanced fibrosis: ≥9.8 |
|
Core treatment recommendations |
Lifestyle modification, weight loss ≥7–10%, vitamin E (non-diabetic), pioglitazone (diabetic), semaglutide (obesity/T2DM), metabolic surgery |
Lifestyle modification, semaglutide, tirzepatide, resmetirom (F2-F3), pioglitazone (T2DM), SGLT2i |
Lifestyle modification, weight loss ≥7–10%, resmetirom (F2-F3), semaglutide/tirzepatide (T2DM/ obesity), metabolic surgery |
Lifestyle modification, resmetirom (F2-F3), vitamin E, pioglitazone, statins (CVD prevention), GLP-1 RA/SGLT2i (metabolic comorbidities) |
|
HCC surveillance recommendations |
F3-F4 cirrhosis: monitor; FIB-4 ≥3.25 or LSM ≥20 kPa may be considered |
F3 cirrhosis: individualized; cirrhosis: monitor |
F3 cirrhosis: individualized; cirrhosis: monitor |
Cirrhosis: monitor; F3 high-risk individuals may be considered |
Table 4.Management of low‑risk patients after second‑tier non‑invasive testing
Table 4.
|
Aspect |
AASLD (2023) [16] |
AGA (2026) [104] |
EASL (2024) [15] |
KASL (2025) [103] |
|
Definition of low risk |
FIB-4 <1.3 and VCTE <8 kPa or ELF <9.8 |
FIB-4 <1.3 (or <2.0 if age ≥65 y) and VCTE <8 kPa or ELF <9.2 |
FIB-4 <1.3 (or <2.0 if age ≥65 y) and VCTE <8 kPa |
FIB-4 <1.3 (or <2.0 if age ≥65 y) and VCTE <8 kPa |
|
Follow-up recommendations |
Repeat FIB-4 every 1–2 years |
Repeat FIB-4 every 1–2 years or repeat VCTE every 3 years |
Repeat FIB-4 every 1–3 years |
Repeat FIB-4 every 1–2 years |
|
Details of follow-up |
Primary care/endocrinology follow-up. For T2DM or ≥2 metabolic risk factors: repeat FIB-4 every 1–2 years; otherwise every 2–3 years. |
Primary care/endocrinology follow-up. For T2DM or multiple risk factors: repeat FIB-4 annually. If prior VCTE <8 kPa, can repeat VCTE every 3 years. |
Primary care follow-up. Repeat FIB-4 every 1–3 years; shorter interval (1–2 years) for T2DM or high metabolic risk. |
Primary care/endocrinology follow-up. For T2DM or ≥2 metabolic risk factors: repeat FIB-4 every 1–2 years; otherwise every 2–3 years. If FIB-4 remains low, continue follow-up; if increases, consider repeat VCTE or referral. |
|
Triggers for earlier re-assessment |
“Appearance of new metabolic risk factors or weight increase” |
Significant weight gain, worsening glycemic control, new-onset T2DM |
New metabolic risk factors or substantial weight gain |
Significant weight gain, worsening glycemic control, new-onset T2DM |
Table 5.Summary of evidence strength for NITs and PEth in MetALD
Table 5.
|
NITs / PEth |
Population |
Evidence strength (Diagnostic) |
Evidence strength (Prognostic) |
|
FIB-4 |
MetALD |
★★★ |
★★★ |
|
VCTE |
★★★ |
★★★ |
|
ELF |
★★★ |
★★★ |
|
MRE |
☆☆☆ |
☆☆☆ |
|
SAFE score |
★★☆ |
★★☆ |
|
Agile 3+/4 |
☆☆☆ |
☆☆☆ |
|
PEth |
★★★ |
☆☆☆ |
Table 6.Key areas of uncertainty and unresolved clinical questions in MetALD risk assessment
Table 6.
|
Domain |
Unresolved questions / Research needs |
|
Diagnosis & classification |
• Optimal PEth cut‑offs for discriminating MASLD, MetALD, and ALD across ethnicities and BMI strata |
|
• Harmonization of PEth laboratory protocols and cross‑platform standardization |
|
• Clinical management when self‑reported alcohol intake and PEth are discordant (common in approximately 60% of cases) |
|
• Influence of drinking pattern (binge vs. daily moderate) on PEth kinetics and interpretation |
|
• Effect of advanced fibrosis / cirrhosis on PEth pharmacokinetics and sensitivity |
|
Non-invasive testing (NIT) cut-offs |
• MetALD‑specific cut‑offs for FIB‑4, VCTE, MRE, and ELF (currently extrapolated from MASLD) |
|
• Impact of alcohol‑induced AST/ALT elevation and thrombocytopenia on model output and calibration |
|
• Performance and re‑calibration of composite scores (Agile 3+/4, FAST, SAFE) in MetALD |
|
• Influence of recent alcohol consumption or active alcoholic hepatitis on LSM overestimation |
|
Prognosis & outcomes |
• Validated MetALD‑specific HCC risk scores (current models like aMAP lack alcohol‑metabolic interaction terms) |
|
• Relative weight of alcohol vs. metabolic factors in driving HCC, decompensation, and mortality |
|
• Longitudinal validation of dynamic biomarkers (serial FIB‑4, VCTE, ELF, PEth) to differentiate fibrotic progression from alcohol‑related fluctuation |
|
• Optimal thresholds for HCC surveillance in MetALD (e.g., should LSM ≥15 kPa trigger annual imaging?) |
|
Therapeutic monitoring |
• NIT‑based definitions of treatment response in MetALD: what magnitude of change in VCTE, MRE, or ELF constitutes meaningful response (20%/25%/30%)? |
|
• Stopping rules for pharmacotherapy (resmetirom, semaglutide) based on NIT thresholds or lack of response |
|
• Impact of concurrent changes in alcohol consumption (assessed by serial PEth) on NIT interpretation during treatment |
|
• Applicability of current phase 3 trial eligibility and response criteria to MetALD patients (mostly excluded) |
|
Care pathways & low-risk management |
• Re‑evaluation interval for low‑risk MetALD patients: FIB‑4 alone at 1–2 years or FIB‑4+VCTE at 2–3 years? |
|
• Clinical triggers for earlier re‑assessment (e.g., weight gain ≥5%, new T2DM, ALT doubling, PEth rise) |
|
• External validation of sequential pathways (FIB‑4→VCTE→referral) in dedicated MetALD cohorts |
|
• Cost‑effectiveness of incorporating PEth into routine primary care pathways |
|
Health equity & generalizability |
• Multi‑ethnic, multi‑regional validation of existing NIT cut‑offs and risk models in MetALD |
|
• Performance of simplified algorithms (e.g., FIB‑4+ELF or SAFE) in low‑resource settings without elastography or Peth |
|
• Applicability of MetALD‑specific pathways in primary care vs. tertiary hepatology clinics |
|
AI & precision medicine |
• Development of multi‑modal predictive models integrating clinical, imaging, genetic (PNPLA3, TM6SF2), and alcohol‑biomarker data |
|
• Explainability of AI‑based risk estimates to foster clinical trust and adoption |
|
• Implementation barriers (EHR integration, algorithm fairness, prospective validation) |
|
• Role of PRS in MetALD‑specific HCC and fibrosis progression prediction |
Table 7.NIT‑based definitions of treatment response, stopping rules, and considerations for MetALD in patients receiving resmetirom or semaglutide
Table 7.
|
Parameter |
Resmetirom (after ≥12 months) [131] |
Semaglutide (after ≥72 weeks) |
|
Response thresholds |
|
|
|
VCTE-LSM |
≥25% reduction |
≥30% reduction (52% of treated patients achieve this in ESSENCE) [132-134] |
|
MRE-LSM |
≥20% reduction |
≥20% reduction [132,133] |
|
ELF score |
– |
≥0.5-point decrease (55.8% achieve this) [132,134] |
|
ALT |
≥17 U/L decrease, or reduction to ≤40 U/L plus ≥30% from baseline |
≥17 U/L or ≥20% reduction [132] |
|
MRI-PDFF |
≥30% reduction (associated with histologic response, but PPV <50%) |
– |
|
Stopping rules / futility |
|
|
|
Disease progression (confirmed by second NIT) |
VCTE-LSM increase ≥25% or MRE-LSM increase ≥20% → consider discontinuation |
VCTE-LSM increase >30% or ELF increase >0.5 → consider discontinuation or add resmetirom [134] |
|
Hepatotoxicity |
AST/ALT >5×ULN (twice), or TBil >2.5 mg/dL with elevated enzymes, or INR >1.5 → discontinue immediately |
No treatment-related liver enzyme elevations reported in ESSENCE [132] |
|
MetALD‑specific considerations |
|
Alcohol eligibility |
Exclude if ongoing intake >20 g/d (women) or >30 g/d (men) |
Excluded from ESSENCE; individualized decision using AUDIT-C and PEth [132] |
|
NIT interpretation |
Thresholds derived from MASLD; alcohol-induced AST elevation and thrombocytopenia may cause overestimation of fibrosis stage |
NIT cut-offs may differ in MetALD; confirm with objective alcohol biomarkers [132] |
|
Monitoring |
PEth recommended to detect unreported alcohol use |
Serial PEth to distinguish drug effect from alcohol-related changes; GLP-1RAs may reduce alcohol craving, but impact on NITs uncertain [132] |
Abbreviations
American Association for the Study of Liver Diseases
American Gastroenterological Association
alcohol-related liver disease
aspartate aminotransferase to platelet ratio index
Alcohol Use Disorders Identification Test‑Consumption
area under the receiver operating characteristic curve
cardiometabolic risk factor
European Association for the Study of the Liver
Korean Association for the Study of the Liver
liver stiffness measurement
metabolic dysfunction-associated fatty liver disease
metabolic dysfunction-associated steatohepatitis
metabolic dysfunction-associated steatotic liver disease
magnetic resonance elastography
magnetic resonance imaging
metabolic dysfunction and alcohol-related liver disease
non-alcoholic fatty liver disease
steatosis-associated fibrosis estimator
vibration-controlled transient elastography
REFERENCES
- 1. Ho GJK, Tan FXN, Sasikumar NA, Tham EKJ, Ko D, Kim DH, et al. High global prevalence of steatotic liver disease and associated subtypes: a meta-analysis. Clin Gastroenterol Hepatol 2025;23:2423-2432.e1.
- 2. Rinella ME, Lazarus JV, Ratziu V, Francque SM, Sanyal AJ, Kanwal F, et al. A multisociety Delphi consensus statement on new fatty liver disease nomenclature. Hepatology 2023;78:1966-1986.
- 3. Schneider CV, Schneider KM, Raptis A, Huang H, Trautwein C, Loomba R. Prevalence of at-risk MASH, MetALD and alcohol-associated steatotic liver disease in the general population. Aliment Pharmacol Ther 2024;59:1271-1281.
- 4. Taylor RS, Taylor RJ, Bayliss S, Hagström H, Nasr P, Schattenberg JM, et al. Association between fibrosis stage and outcomes of patients with nonalcoholic fatty liver disease: a systematic review and meta-analysis. Gastroenterology 2020;158:1611-1625.e12.
- 5. Younossi ZM, Zelber-Sagi S, Lazarus JV, Wong VW, Yilmaz Y, Duseja A, et al. Global consensus recommendations for metabolic dysfunction-associated steatotic liver disease and steatohepatitis. Gastroenterology 2025;169:1017-1032.e2.
- 6. Lee HH, Lee HA, Kim EJ, Kim HY, Kim HC, Ahn SH, et al. Metabolic dysfunction-associated steatotic liver disease and risk of cardiovascular disease. Gut 2024;73:533-540.
- 7. Targher G, Byrne CD, Tilg H. NAFLD and increased risk of cardiovascular disease: clinical associations, pathophysiological mechanisms and pharmacological implications. Gut 2020;69:1691-1705.
- 8. Mantovani A, Petracca G, Beatrice G, Csermely A, Lonardo A, Schattenberg JM, et al. Non-alcoholic fatty liver disease and risk of incident chronic kidney disease: an updated meta-analysis. Gut 2022;71:156-162.
- 9. European Association for the Study of the Liver. EASL Clinical Practice Guidelines on non-invasive tests for evaluation of liver disease severity and prognosis - 2021 update. J Hepatol 2021;75:659-689.
- 10. Tampaki M, Tsochatzis E, Lekakis V, Cholongitas E. Prevalence, characteristics and outcomes of patients with metabolic and alcohol related/associated liver disease (MetALD): a systematic review and meta-analysis. Metabolism 2025;163:156101.
- 11. Celsa C, Pennisi G, Tulone A, Ciancimino G, Vaccaro M, Pecorella F, et al. Risk of hepatic and extrahepatic outcomes associated with metabolic dysfunction-associated steatotic liver disease and metabolic dysfunction and alcohol-associated steatotic liver disease: a systematic review and metaanalysis. Lancet Gastroenterol Hepatol 2025;10:998-1012.
- 12. Younossi ZM, Golabi P, de Avila L, Paik JM, Srishord M, Fukui N, et al. The global epidemiology of NAFLD and NASH in patients with type 2 diabetes: a systematic review and meta-analysis. J Hepatol 2019;71:793-801.
- 13. En Li Cho E, Ang CZ, Quek J, Fu CE, Lim LKE, Heng ZEQ, et al. Global prevalence of non-alcoholic fatty liver disease in type 2 diabetes mellitus: an updated systematic review and meta-analysis. Gut 2023;72:2138-2148.
- 14. Younossi ZM, Golabi P, Price JK, Owrangi S, Gundu-Rao N, Satchi R, et al. The global epidemiology of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis among patients with type 2 diabetes. Clin Gastroenterol Hepatol 2024;22:1999-2010.e8.
- 15. European Association for the Study of the Liver (EASL); European Association for the Study of Diabetes (EASD); European Association for the Study of Obesity (EASO). EASLEASD-EASO Clinical Practice Guidelines on the management of metabolic dysfunction-associated steatotic liver disease (MASLD). J Hepatol 2024;81:492-542.
- 16. Rinella ME, Neuschwander-Tetri BA, Siddiqui MS, Abdelmalek MF, Caldwell S, Barb D, et al. AASLD Practice Guidance on the clinical assessment and management of nonalcoholic fatty liver disease. Hepatology 2023;77:1797-1835.
- 17. Chatrath H, Vuppalanchi R, Chalasani N. Dyslipidemia in patients with nonalcoholic fatty liver disease. Semin Liver Dis 2012;32:22-29.
- 18. Katsiki N, Mikhailidis DP, Mantzoros CS. Non-alcoholic fatty liver disease and dyslipidemia: an update. Metabolism 2016;65:1109-1123.
- 19. Gao B, Arab JP, Liangpunsakul S, Ding WX, Szabo G, Mehal W, et al. Metabolic dysfunction and alcohol-associated liver disease (MetALD). eGastroenterology 2025;3:e100319.
- 20. Staufer K, Stauber RE. Steatotic liver disease: metabolic dysfunction, alcohol, or both? Biomedicines 2023;11:2108.
- 21. Jarvis H, O’Keefe H, Craig D, Stow D, Hanratty B, Anstee QM. Does moderate alcohol consumption accelerate the progression of liver disease in NAFLD? A systematic review and narrative synthesis. BMJ Open 2022;12:e049767.
- 22. Magherman L, Van Parys R, Pauwels NS, Verhelst X, Devisscher L, Van Vlierberghe H, et al. Meta-analysis: the impact of light-to-moderate alcohol consumption on progressive non-alcoholic fatty liver disease. Aliment Pharmacol Ther 2023;57:820-836.
- 23. Åberg F, Byrne CD, Pirola CJ, Männistö V, Sookoian S. Alcohol consumption and metabolic syndrome: clinical and epidemiological impact on liver disease. J Hepatol 2023;78:191-206.
- 24. Acierno C, Barletta F, Caturano A, Nevola R, Sasso FC, Adinolfi LE, et al. Alcohol consumption and liver metabolism in the era of MASLD: integrating nutritional and pathophysiological insights. Nutrients 2025;17:2229.
- 25. Arab JP, Díaz LA, Rehm J, Im G, Arrese M, Kamath PS, et al. Metabolic dysfunction and alcohol-related liver disease (MetALD): position statement by an expert panel on alcohol-related liver disease. J Hepatol 2025;82:744-756.
- 26. Ayares G, Diaz LA, Idalsoaga F, Alkhouri N, Noureddin M, Bataller R, et al. MetALD: new perspectives on an old overlooked disease. Liver Int 2025;45:e70017.
- 27. Torp N, Israelsen M, Thiele M, Rinella ME, Krag A. Phosphatidylethanol in steatotic liver disease. J Hepatol 2025;83:1189-1203.
- 28. Gnann H, Weinmann W, Thierauf A. Formation of phosphatidylethanol and its subsequent elimination during an extensive drinking experiment over 5 days. Alcohol Clin Exp Res 2012;36:1507-1511.
- 29. Luginbühl M, Wurst FM, Stöth F, Weinmann W, Stove CP, Van Uytfanghe K. Consensus for the use of the alcohol biomarker phosphatidylethanol (PEth) for the assessment of abstinence and alcohol consumption in clinical and forensic practice (2022 Consensus of Basel). Drug Test Anal 2022;14:1800-1802.
- 30. Skråstad RB, Aamo TO, Andreassen TN, Havnen H, Hveem K, Krokstad S, et al. Quantifying alcohol consumption in the general population by analysing phosphatidylethanol concentrations in whole blood: results from 24,574 subjects included in the HUNT4 study. Alcohol Alcohol 2023;58:258-265.
- 31. Hahn JA, Murnane PM, Vittinghoff E, Muyindike WR, Emenyonu NI, Fatch R, et al. Factors associated with phosphatidylethanol (PEth) sensitivity for detecting unhealthy alcohol use: an individual patient data meta-analysis. Alcohol Clin Exp Res 2021;45:1166-1187.
- 32. Tavaglione F, Amangurbanova M, Yang AH, Tincopa MA, Ajmera V, Richards L, et al. Head-to-head comparison between phosphatidylethanol versus indirect alcohol biomarkers for diagnosis of MetALD versus MASLD: a prospective study. Aliment Pharmacol Ther 2025;61:1043-1054.
- 33. Scholten K, Twohig P, Samson K, Brittan K, Fiedler A, Warner J, et al. You can’t handle the truth! Comparing serum phosphatidylethanol to self-reported alcohol intake in chronic liver disease patients. Dig Liver Dis 2024;56:1215-1219.
- 34. Winder GS, Clifton EG, Denysenko L, DiChiara AM, Hathaway D, Perumalswami PV, et al. “But I didn’t drink!”: what to do with discordant phosphatidylethanol results. Liver Transpl 2024;30:213-222.
- 35. Noureddin M, Charlton MR, Harrison SA, Bansal MB, Alkhouri N, Loomba R, et al. Expert panel recommendations: practical clinical applications for initiating and monitoring resmetirom in patients with MASH/NASH and moderate to noncirrhotic advanced fibrosis. Clin Gastroenterol Hepatol 2024;22:2367-2377.
- 36. Torp N, Israelsen M, Johansen S, Semmler G, Dalby Hansen C, Bech KT, et al. MetALD: diagnosis and prognosis with noninvasive tests. Aliment Pharmacol Ther 2025;62:990-1002.
- 37. Anstee QM, Day CP. The genetics of nonalcoholic fatty liver disease: spotlight on PNPLA3 and TM6SF2. Semin Liver Dis 2015;35:270-290.
- 38. Sookoian S, Rotman Y, Valenti L. Genetics of metabolic dysfunction-associated steatotic liver disease: the state of the art update. Clin Gastroenterol Hepatol 2024;22:2177-2187 e3.
- 39. Kubina M, Prasitsumrit V, Tan J, Quek JWE, Peddu D, Mishra A, et al. Meta-analysis: effects of steatotic liver disease-associated genetic risk alleles on longitudinal outcomes. Aliment Pharmacol Ther 2025;62:244-276.
- 40. Ajmera V, Loomba R. Advances in the genetics of nonalcoholic fatty liver disease. Curr Opin Gastroenterol 2023;39:150-155.
- 41. Galvanin C, Buch S, Nahon P, Trépo E. PNPLA3 in alcohol-related liver disease. Liver Int 2025;45:e16211.
- 42. Chen VL, Kuppa A, Oliveri A, Chen Y, Ponnandy P, Patel PB, et al. Human genetics of metabolic dysfunction-associated steatotic liver disease: from variants to cause to precision treatment. J Clin Invest 2025;135:e186424.
- 43. Amangurbanova M, Huang DQ, Loomba R. Review article: the role of HSD17B13 on global epidemiology, natural history, pathogenesis and treatment of NAFLD. Aliment Pharmacol Ther 2023;57:37-51.
- 44. Pei Y, Goh GB. Genetic risk factors for metabolic dysfunction-associated steatotic liver disease. Gut Liver 2025;19:8-18.
- 45. Kanwal F, Shubrook JH, Adams LA, Pfotenhauer K, Wai-Sun Wong V, Wright E, et al. Clinical care pathway for the risk stratification and management of patients with nonalcoholic fatty liver disease. Gastroenterology 2021;161:1657-1669.
- 46. McPherson S, Hardy T, Dufour JF, Petta S, Romero-Gomez M, Allison M, et al. Age as a confounding factor for the accurate non-invasive diagnosis of advanced NAFLD fibrosis. Am J Gastroenterol 2017;112:740-751.
- 47. Rabbat J, Yang B, Lee HW, Lin H, Tsochatzis E, Petta S, et al. Histological severity and hepatic outcomes in patients with metabolic dysfunction-associated steatotic liver disease and discrepant FIB-4 and liver stiffness measurement. Clin Mol Hepatol 2026;32:289-304.
- 48. Mózes FE, Lee JA, Selvaraj EA, Jayaswal ANA, Trauner M, Boursier J, et al. Diagnostic accuracy of non-invasive tests for advanced fibrosis in patients with NAFLD: an individual patient data meta-analysis. Gut 2022;71:1006-1019.
- 49. Oh JH, Ahn SB, Cho S, Nah EH, Yoon EL, Jun DW. Diagnostic performance of non-invasive tests in patients with MetALD in a health check-up cohort. J Hepatol 2024;81:772-780.
- 50. Sanyal AJ, Foucquier J, Younossi ZM, Harrison SA, Newsome PN, Chan WK, et al. Enhanced diagnosis of advanced fibrosis and cirrhosis in individuals with NAFLD using FibroScan-based Agile scores. J Hepatol 2023;78:247-259.
- 51. Truong E, Alnimer L, Gornbein JA, Yang JD, Alkhouri N, Harrison SA, et al. Agile 3+ and 4 scores accurately predict major adverse liver outcomes, liver transplant, progression of MELD score, the development of hepatocellular carcinoma, and death in NAFLD. Dig Dis Sci 2025;70:2487-2495.
- 52. Sripongpun P, Kim WR, Mannalithara A, Charu V, Vidovszky A, Asch S, et al. The steatosis-associated fibrosis estimator (SAFE) score: a tool to detect low-risk NAFLD in primary care. Hepatology 2023;77:256-267.
- 53. van Kleef LA, de Knegt RJ, Ayada I, Pan Q, Brouwer WP. The steatosis-associated fibrosis estimator (SAFE) score: validation in the general US population. Hepatol Commun 2023;7:e0075.
- 54. Li M, Lin Y, Yu H, Lin W, Chen J, Yang Y, et al. The steatosis-associated fibrosis estimator (SAFE) outperformed the FIB-4 score in screening the population for liver disease. Ann Hepatol 2024;29:101516.
- 55. Taru MG, Tefas C, Neamti L, Minciuna I, Taru V, Maniu A, et al. FAST and Agile-the MASLD drift: validation of Agile 3+, Agile 4 and FAST scores in 246 biopsy-proven NAFLD patients meeting MASLD criteria of prevalent caucasian origin. PLoS One 2024;19:e0303971.
- 56. Papatheodoridi M, De Ledinghen V, Lupsor-Platon M, Bronte F, Boursier J, Elshaarawy O, et al. Agile scores in MASLD and ALD: external validation and their utility in clinical algorithms. J Hepatol 2024;81:590-599.
- 57. Validation of CASLI, Fibroscan-AST (FAST), and Agile3+ in a Russian cohort of patients with metabolic dysfunction-associated steatotic liver disease. New Horizons for Steatotic Liver Disease: Cutting Edge Research and Emerging Therapeutics Y2 - 2 October 2025 through 3 October 2025: p. 218.
- 58. Yu H, Wu J, Li M, Hu C, Ren Y, Jiang H, et al. Diagnostic and prognostic performance of the Agile 3+ score in patients with chronic hepatitis B and concurrent MASLD after complete viral suppression with nucleos(t)ide analogues. J Gastroenterol Hepatol 2025;40:2963-2975.
- 59. Newsome PN, Sasso M, Deeks JJ, Paredes A, Boursier J, Chan WK, et al. FibroScan-AST (FAST) score for the noninvasive identification of patients with non-alcoholic steatohepatitis with significant activity and fibrosis: a prospective derivation and global validation study. Lancet Gastroenterol Hepatol 2020;5:362-373.
- 60. Kim K, Lee HW, Lee JS, Kim MN, Park JY. Steatosis-associated fibrosis estimator score in Asian patients with metabolic dysfunction-associated steatotic liver disease. Yonsei Med J 2026;67:104-111.
- 61. Barreyro FJ, Bori M, Balderramo D, Del Valle Aubone M, Fernandez G, Soto Oca O, et al. Diagnostic accuracy of non invasive tests for fibrosis in metabolic dysfunction associated steatotic liver disease. Ann Hepatol 2025;30:102098.
- 62. Sun H. An alternative non-invasive screening model for liver fibrosis among US adults at risk of MASLD. Diseases 2024;12:150.
- 63. Verschuren L, Mak AL, van Koppen A, Özsezen S, Difrancesco S, Caspers MPM, et al. Development of a novel noninvasive biomarker panel for hepatic fibrosis in MASLD. Nat Commun 2024;15:4564.
- 64. Johnston DE. Special considerations in interpreting liver function tests. Am Fam Physician 1999;59:2223-2230.
- 65. Silczuk A, Habrat B. Alcohol-induced thrombocytopenia: current review. Alcohol 2020;86:9-16.
- 66. Sripongpun P, Kaewdech A, Udompap P, Kim WR. Characteristics and long-term mortality of individuals with MASLD, MetALD, and ALD, and the utility of SAFE score. JHEP Rep 2024;6:101127.
- 67. Graupera I, Thiele M, Serra-Burriel M, Caballeria L, Roulot D, Wong GL, et al. Low accuracy of FIB-4 and NAFLD fibrosis scores for screening for liver fibrosis in the population. Clin Gastroenterol Hepatol 2022;20:2567-2576.e6.
- 68. Stern C, Castera L. Identification of high-risk subjects in nonalcoholic fatty liver disease. Clin Mol Hepatol 2023;29:S196-S206.
- 69. Qadri S, Ahlholm N, Lønsmann I, Pellegrini P, Poikola A, Luukkonen PK, et al. Obesity modifies the performance of fibrosis biomarkers in nonalcoholic fatty liver disease. J Clin Endocrinol Metab 2022;107:e2008-e2020.
- 70. Archer AJ, Belfield KJ, Orr JG, Gordon FH, Abeysekera KW. EASL clinical practice guidelines: non-invasive liver tests for evaluation of liver disease severity and prognosis. Frontline Gastroenterol 2022;13:436-439.
- 71. Kjaergaard M, Lindvig KP, Thorhauge KH, Andersen P, Hansen JK, Kastrup N, et al. Using the ELF test, FIB-4 and NAFLD fibrosis score to screen the population for liver disease. J Hepatol 2023;79:277-286.
- 72. Lin H, Lee HW, Yip TC, Tsochatzis E, Petta S, Bugianesi E, et al. Vibration-controlled transient elastography scores to predict liver-related events in steatotic liver disease. JAMA 2024;331:1287-1297.
- 73. Boursier J, Guillaume M, Leroy V, Irlès M, Roux M, Lannes A, et al. New sequential combinations of non-invasive fibrosis tests provide an accurate diagnosis of advanced fibrosis in NAFLD. J Hepatol 2019;71:389-396.
- 74. Yip TC, Lee HW, Lin H, Tsochatzis E, Petta S, Bugianesi E, et al. Prognostic performance of the two-step clinical care pathway in metabolic dysfunction-associated steatotic liver disease. J Hepatol 2025;83:304-314.
- 75. Díaz LA, Tavaglione F, Mittal N, Bettencourt R, Amangurbanova M, Johnson A, et al. Noninvasive pathway for stratifying fibrosis in suspected metabolic dysfunction and alcohol-associated liver disease (MetALD). Hepatol Commun 2025;9:e0718.
- 76. Gelsi E, Dainese R, Truchi R, Mariné-Barjoan E, Anty R, Autuori M, et al. Effect of detoxification on liver stiffness assessed by Fibroscan® in alcoholic patients. Alcohol Clin Exp Res 2011;35:566-570.
- 77. Ichikawa S, Motosugi U, Nakazawa T, Morisaka H, Sano K, Ichikawa T, et al. Hepatitis activity should be considered a confounder of liver stiffness measured with MR elastography. J Magn Reson Imaging 2015;41:1203-1208.
- 78. Giuffrè M, Campigotto M, Colombo A, Visintin A, Budel M, Aversano A, et al. The role of elastography in alcoholic liver disease: fibrosis staging and confounding factors, a review of the current literature. Minerva Gastroenterol (Torino) 2021;67:112-121.
- 79. Trabut JB, Thépot V, Nalpas B, Lavielle B, Cosconea S, Corouge M, et al. Rapid decline of liver stiffness following alcohol withdrawal in heavy drinkers. Alcohol Clin Exp Res 2012;36:1407-1411.
- 80. Otero Sanchez L, Moreno C. Noninvasive tests in assessment of patients with alcohol-associated liver disease. Clin Liver Dis 2024;28:715-729.
- 81. Singh S, Venkatesh SK, Loomba R, Wang Z, Sirlin C, Chen J, et al. Magnetic resonance elastography for staging liver fibrosis in non-alcoholic fatty liver disease: a diagnostic accuracy systematic review and individual participant data pooled analysis. Eur Radiol 2016;26:1431-1440.
- 82. Selvaraj EA, Mózes FE, Jayaswal ANA, Zafarmand MH, Vali Y, Lee JA, et al. Diagnostic accuracy of elastography and magnetic resonance imaging in patients with NAFLD: a systematic review and meta-analysis. J Hepatol 2021;75:770-785.
- 83. Xiao G, Zhu S, Xiao X, Yan L, Yang J, Wu G. Comparison of laboratory tests, ultrasound, or magnetic resonance elastography to detect fibrosis in patients with nonalcoholic fatty liver disease: a meta-analysis. Hepatology 2017;66:1486-1501.
- 84. Hsu C, Caussy C, Imajo K, Chen J, Singh S, Kaulback K, et al. Magnetic resonance vs transient elastography analysis of patients with nonalcoholic fatty liver disease: a systematic review and pooled analysis of individual participants. Clin Gastroenterol Hepatol 2019;17:630-637.e8.
- 85. Chon YE, Jin YJ, An J, Kim HY, Choi M, Jun DW, et al. Optimal cut-offs of vibration-controlled transient elastography and magnetic resonance elastography in diagnosing advanced liver fibrosis in patients with nonalcoholic fatty liver disease: a systematic review and meta-analysis. Clin Mol Hepatol 2024;30:S117-S133.
- 86. Sterling RK, Duarte-Rojo A, Patel K, Asrani SK, Alsawas M, Dranoff JA, et al. AASLD Practice Guideline on imaging-based noninvasive liver disease assessment of hepatic fibrosis and steatosis. Hepatology 2025;81:672-724.
- 87. Mendoza YP, Rodrigues SG, Delgado MG, Murgia G, Lange NF, Schropp J, et al. Inflammatory activity affects the accuracy of liver stiffness measurement by transient elastography but not by two-dimensional shear wave elastography in nonalcoholic fatty liver disease. Liver Int 2022;42:102-111.
- 88. Millonig G, Friedrich S, Adolf S, Fonouni H, Golriz M, Mehrabi A, et al. Liver stiffness is directly influenced by central venous pressure. J Hepatol 2010;52:206-210.
- 89. Barr RG, Ferraioli G, Palmeri ML, Goodman ZD, Garcia-Tsao G, Rubin J, et al. Elastography assessment of liver fibrosis: Society of Radiologists in Ultrasound Consensus Conference Statement. Radiology 2015;276:845-861.
- 90. Wagner M, Corcuera-Solano I, Lo G, Esses S, Liao J, Besa C, et al. Technical failure of MR elastography examinations of the liver: experience from a large single-center study. Radiology 2017;284:401-412.
- 91. Tada A, Nagai T, Kato Y, Omote K, Oyama-Manabe N, Tsuneta S, et al. Liver stiffness assessed by magnetic resonance elastography predicts clinical outcomes in patients with heart failure and without chronic liver disease. Eur Radiol 2023;33:2062-2074.
- 92. Yin M, Glaser KJ, Talwalkar JA, Chen J, Manduca A, Ehman RL. Hepatic MR elastography: clinical performance in a series of 1377 consecutive examinations. Radiology 2016;278:114-124.
- 93. Ratziu V, Charlotte F, Heurtier A, Gombert S, Giral P, Bruckert E, et al. Sampling variability of liver biopsy in nonalcoholic fatty liver disease. Gastroenterology 2005;128:1898-1906.
- 94. Sharma C, Cococcia S, Ellis N, Parkes J, Rosenberg W. Systematic review: accuracy of the enhanced liver fibrosis test for diagnosing advanced liver fibrosis and cirrhosis. J Gastroenterol Hepatol 2021;36:1788-1802.
- 95. Hinkson A, Lally H, Gibson H, Jones R, Rowe IA, Shinkins B, et al. Meta-analysis: enhanced liver fibrosis test to identify hepatic fibrosis in chronic liver diseases. Aliment Pharmacol Ther 2023;57:750-762.
- 96. Lee J, Vali Y, Boursier J, Duffin K, Verheij J, Brosnan MJ, et al. Accuracy of cytokeratin 18 (M30 and M65) in detecting non-alcoholic steatohepatitis and fibrosis: a systematic review and meta-analysis. PLoS One 2020;15:e0238717.
- 97. Kwok R, Tse YK, Wong GL, Ha Y, Lee AU, Ngu MC, et al. Systematic review with meta-analysis: non-invasive assessment of non-alcoholic fatty liver disease--the role of transient elastography and plasma cytokeratin-18 fragments. Aliment Pharmacol Ther 2014;39:254-269.
- 98. Mak AL, Lee J, van Dijk AM, Vali Y, Aithal GP, Schattenberg JM, et al. Systematic review with meta-analysis: diagnostic accuracy of Pro-C3 for hepatic fibrosis in patients with nonalcoholic fatty liver disease. Biomedicines 2021;9:1920.
- 99. Daniels SJ, Leeming DJ, Eslam M, Hashem AM, Nielsen MJ, Krag A, et al. ADAPT: an algorithm incorporating PRO-C3 accurately identifies patients with NAFLD and advanced fibrosis. Hepatology 2019;69:1075-1086.
- 100. Lai JC, Yang B, Lee HW, Lin H, Tsochatzis EA, Petta S, et al. Non-invasive risk-based surveillance of hepatocellular carcinoma in patients with metabolic dysfunction-associated steatotic liver disease. Gut 2025;74:2050-2057.
- 101. Ochoa-Allemant P, Hubbard RA, Kaplan DE, Serper M. Adverse liver outcomes, cardiovascular events, and mortality in steatotic liver disease. JAMA Intern Med 2025;185:986-995.
- 102. Armandi A, Bugianesi E, Loomba R, Caussy C. Algorithms for risk stratification in patients with metabolic dysfunction-associated steatohepatitis. Clin Liver Dis 2026;30:343-358.
- 103. Sohn W, Lee YS, Kim SS, Kim JH, Jin YJ, Kim GA, et al. KASL clinical practice guidelines for the management of metabolic dysfunction-associated steatotic liver disease 2025. Clin Mol Hepatol 2025;31:S1-S31.
- 104. Kanwal F, Bril F, Wong VW, Adams LA, Pfotenhauer K, Skolnik N, et al. Clinical care pathway for the risk stratification and management of patients with metabolic dysfunction-associated steatotic liver disease. Gastroenterology 2026;171:164-183.
- 105. Kim MN, Han JW, An J, Kim BK, Jin YJ, Kim SS, et al. KASL clinical practice guidelines for noninvasive tests to assess liver fibrosis in chronic liver disease. Clin Mol Hepatol 2024;30:S5-S105.
- 106. Zhang Z, Yu Z, Wu G. Discordant FIB-4 and liver stiffness in metabolic dysfunction-associated steatotic liver disease: Toward clinically coherent risk stratification: Letter to the Editor on “Histological severity and hepatic outcomes in patients with metabolic dysfunction-associated steatotic liver disease and discrepant FIB-4 and liver stiffness measurement”. Clin Mol Hepatol 2026;32:e297-e299.
- 107. Lekakis V, Papatheodoridis GV. Natural history of metabolic dysfunction-associated steatotic liver disease. Eur J Intern Med 2024;122:3-10.
- 108. Blomdahl J, Nasr P, Ekstedt M, Kechagias S. Moderate alcohol consumption is associated with significant fibrosis progression in NAFLD. Hepatol Commun 2023;7:e0003.
- 109. Ciardullo S, Mantovani A, Morieri ML, Muraca E, Invernizzi P, Perseghin G. Impact of MASLD and MetALD on clinical outcomes: a meta-analysis of preliminary evidence. Liver Int 2024;44:1762-1767.
- 110. Odriozola A, Santos-Laso A, Del Barrio M, Cabezas J, Iruzubieta P, Arias-Loste MT, et al. Fatty liver disease, metabolism and alcohol interplay: a comprehensive review. Int J Mol Sci 2023;24:7791.
- 111. Zhang Y, Lee HW, Lin H, Tsochatzis E, Petta S, Bugianesi E, et al. Head-to-head comparison between vibration-controlled transient elastography and histology in predicting liver-related events due to metabolic dysfunction-associated steatotic liver disease. Hepatology 2026;84:175-190.
- 112. Pennisi G, Infantino G, Celsa C, Di Maria G, Enea M, Vaccaro M, et al. Clinical outcomes of MAFLD versus NAFLD: a meta-analysis of observational studies. Liver Int 2024;44:2939-2949.
- 113. Fan R, Papatheodoridis G, Sun J, Innes H, Toyoda H, Xie Q, et al. AMAP risk score predicts hepatocellular carcinoma development in patients with chronic hepatitis. J Hepatol 2020;73:1368-1378.
- 114. Burke L, Hinkson A, Haghnejad V, Jones R, Parker R, Rowe IA. Hepatocellular carcinoma risk scores for non-viral liver disease: a systematic review and meta-analysis. JHEP Rep 2024;7:101227.
- 115. John BV, Bastaich D, Paulus E, Spector S, Dahman B. Hepatocellular carcinoma risk in metabolic dysfunction-associated steatotic liver disease with increased alcohol intake (MetALD). JHEP Rep 2025;8:101639.
- 116. Bianco C, Tavaglione F, Romeo S, Valenti L. Genetic risk scores and personalization of care in fatty liver disease. Curr Opin Pharmacol 2021;61:6-11.
- 117. Akbary K, Noureddin M, Yayun R, Tai D, Boudes P. Development of AI based fibrosis detection algorithm by SHG/TPEF microscopy for fully quantified liver fibrosis assessment in MASH. Liver Int 2025;45:e70258.
- 118. Kosick HM, McIntosh C, Bera C, Fakhriyehasl M, Shengir M, Adeyi O, et al. Machine learning models using non-invasive tests & B-mode ultrasound to predict liver-related outcomes in metabolic dysfunction-associated steatotic liver disease. Sci Rep 2025;15:24579.
- 119. Drozdov I, Szubert B, Rowe IA, Kendall TJ, Fallowfield JA. Accurate prediction of all-cause mortality in patients with metabolic dysfunction-associated steatotic liver disease using electronic health records. Ann Hepatol 2024;29:101528.
- 120. Sarkar S, Alurwar A, Ly C, Piao C, Donde R, Wang CJ, et al. A machine learning model to predict risk for hepatocellular carcinoma in patients with metabolic dysfunction-associated steatotic liver disease. Gastro Hep Adv 2024;3:498-505.
- 121. Alkhouri N, Cheuk-Fung Yip T, Castera L, Takawy M, Adams LA, Verma N, et al. ALADDIN: a machine learning approach to enhance the prediction of significant fibrosis or higher in metabolic dysfunction-associated steatotic liver disease. Am J Gastroenterol 2026;121:362-374.
- 122. Iyer JS, Juyal D, Le Q, Shanis Z, Pokkalla H, Pouryahya M, et al. AI-based automation of enrollment criteria and endpoint assessment in clinical trials in liver diseases. Nat Med 2024;30:2914-2923.
- 123. Souza NS, Vitório TCV, Souza RA, Machado MAD, Cotrim HP. Artificial intelligence for fibrosis diagnosis in metabolic-dysfunction-associated steatotic liver disease: a systematic review. Diagnostics (Basel) 2026;16:261.
- 124. Johansen S, Thiele M, Krag A. Non-invasive tests for MetALD and alcohol-related liver disease. JHEP Rep 2025;7:101569.
- 125. Targher G, Byrne CD, Tilg H. MASLD: a systemic metabolic disorder with cardiovascular and malignant complications. Gut 2024;73:691-702.
- 126. Miller DM, McCauley KF, Dunham-Snary KJ. Metabolic dysfunction-associated steatotic liver disease (MASLD): mechanisms, clinical implications and therapeutic advances. Endocrinol Diabetes Metab 2025;8:e70132.
- 127. Mantovani A, Lonardo A, Stefan N, Targher G. Metabolic dysfunction-associated steatotic liver disease and extrahepatic gastrointestinal cancers. Metabolism 2024;160:156014.
- 128. Harrison SA, Bedossa P, Guy CD, Schattenberg JM, Loomba R, Taub R, et al. A phase 3, randomized, controlled trial of resmetirom in NASH with liver fibrosis. N Engl J Med 2024;390:497-509.
- 129. Sanyal AJ, Newsome PN, Kliers I, Østergaard LH, Long MT, Kjær MS, et al. Phase 3 trial of semaglutide in metabolic dysfunction-associated steatohepatitis. N Engl J Med 2025;392:2089-2099.
- 130. Li L, Mao X, Lee CH, Wai Lui DT, Wan-Hin Hui R, Mak LY, et al. Impact of SGLT2i and GLP-1RA on liver-related events in patients with MASLD and type 2 diabetes: a network meta-analysis. JHEP Rep 2025;8:101661.
- 131. Chen VL, Morgan TR, Rotman Y, Patton HM, Cusi K, Kanwal F, et al. Resmetirom therapy for metabolic dysfunction-associated steatotic liver disease: October 2024 updates to AASLD Practice Guidance. Hepatology 2025;81:312-320.
- 132. Bansal MB, Patton H, Morgan TR, Carr RM, Dranoff JA, Allen AM. Semaglutide therapy for metabolic dysfunction-associated steatohepatitis: November 2025 updates to AASLD Practice Guidance. Hepatology 2026;83:1326-1340.
- 133. Nitze LM, Ratziu V, Sanyal AJ, Wong VW, Balendran C, Fleckner J, et al. Exploration of multiple non-invasive tests for assessing response to treatment in a semaglutide phase 2b trial in patients with MASH. Aliment Pharmacol Ther 2026;63:396-404.
- 134. Younossi ZM, Kalligeros M, Wong VW, Al-Naamani KM, Tacke F, Tsochatzis E, et al. Updated global consensus recommendations for risk stratification, treatment initiation, and response monitoring in metabolic dysfunction-associated steatotic liver disease. Clin Gastroenterol Hepatol 2026;24:2333-2347.
- 135. Bril F, Gray M. Noninvasive tests to identify liver fibrosis in metabolic dysfunction-associated steatotic liver disease are affected by race. Obesity (Silver Spring) 2024;32:612-622.
, Seung Up Kim7,8
