Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease worldwide, and it emerged as the leading cause of end-stage liver disease along with the increase in metabolic disorders [1]. Especially, NAFLD is the fastest growing cause of hepatocellular carcinoma (HCC), accounting for up to 35% of HCC cases worldwide given its high prevalence [2]. The shortcomings of the original name and the metabolic aspects of NAFLD led a multisociety expert panel to propose renaming this condition as metabolic dysfunction-associated steatotic liver disease (MASLD), to improve identification of patients and awareness by emphasizing metabolic risk [3,4]. MASLD, unlike NAFLD, is diagnosed in patients with liver steatosis who exhibit one of five cardiometabolic risk factors (CMRFs): overweight/obesity, prediabetes/diabetes, prehypertension/hypertension, hypertriglyceridemia, and low high-density lipoprotein cholesterol levels [3,4]. Recent studies have reported that MASLD almost directly replaces NAFLD, with more than 95% of patients with NAFLD meeting the criteria for MASLD [5,6]. The new nomenclature is expected to improve the assessment of clinical features and disease prognosis through its focus on CMRFs. In this article, we therefore review the recent research on MASLD-related HCC, such as epidemiology, risk factors with a particular focus on CMRFs in light of the new nomenclature, surveillance strategies, and risk stratification models.

Epidemiology: incidence and trend of metabolic dysfunction-associated steatotic liver disease–related hepatocellular carcinoma

The annual incidence of HCC in cirrhotic patients with MASLD has been reported to be approximately 1.0–1.5% (10–15 per 1,000 person-years), which is similar to that in cirrhotic patients with alcoholic liver disease but lower than that in patients with hepatitis B (2%) or hepatitis C (3–8%) [7,8]. The incidence of HCC in patients with MASLD without cirrhosis and metabolic dysfunction-associated steatohepatitis (MASH) is much lower, at 0.08–0.63 per 1,000 person-years [8]. In patients with precirrhotic MASH having necroinflammation and varying stages of fibrosis F1–F3, the incidence of HCC depends on the stage of fibrosis as well as on many other risk factors, such as obesity, diabetes, and insulin resistance [8]. As MASH is considered a histological diagnosis, further studies of patients with biopsy-proven MASH may be needed to clarify the incidence of HCC in MASH.

MASLD is considered the leading cause of HCC worldwide. A large-scale study utilizing the methodological framework of the Global Burden of Disease Study reported that MASLD was the only etiology of liver cancer with rising age-standardized incidence rates between 2010 and 2021 (0.46→0.49 per 100,000 cases); those for liver cancer in alcohol-related disease (1.14→1.14 per 100,000 cases) remained stable, and those for liver cancer in hepatitis B (2.54→2.37 per 100,000 cases) and C (1.99→1.82 per 100,000 cases) declined [9]. In addition, a Markov model has been used to predict that the incidence of NAFLD-related HCC in the United States will increase by 137% between 2015 and 2030 [10]. Another study showed that the prevalence of MASLD-related HCC is expected to increase in all eight countries analyzed, by between 47% and 130% [11]. In South Korea, Chon et al. [12] reported trends in the incidence of HCC over 10-year period (2008–2018) by studying data from 127,426 individuals included in the Korean National Health Insurance Service database. The proportion of MASLD-related HCC increased steadily between 2008 and 2018, with an average annual percentage change of 2.7% (p < 0.001). A more rapid increase was observed between 2011 and 2018, with an average percentage change of 5.6% per year. These results imply that, in the future, MASLD will be the predominant etiology of HCC, and urgent measures are therefore needed to reduce the impending burden of MASLD-related HCC.

Risk factors of metabolic dysfunction-associated steatotic liver disease–related hepatocellular carcinoma

Fibrotic burden of liver

Patients with MASLD who are aged over 50 years, male, and exhibit cirrhosis are widely considered to be at a greater risk of HCC [2]. Cirrhosis is well established as the most important risk factor for HCC development in MASLD. A recent study of MASLD patients with advanced fibrosis or cirrhosis reported that baseline liver stiffness measurement (LSM) and change in LSM were associated with risk of HCC [13]. However, more than 20% of patients with MASLD-related HCC do not exhibit cirrhosis [14]. Indeed, a recent multinational multicenter cohort study of patients with MASLD found that LSM determined by transient elastography was more predictive of hepatic decompensation than HCC [15]. In the absence of cirrhosis, metabolic risk factors such as obesity, diabetes, and insulin resistance may play important roles in HCC development [8]. A recent study of 392,800 patients with MASLD reported that diabetes was a significant HCC risk factor for HCC in non-cirrhotic patients, increasing the risk of HCC 1.6 fold [16].

Cardiometabolic risk factors: overweight/obesity

Obesity has been linked to an increase in the frequency of liver cancers, including HCC. The presence, duration, and severity of obesity are associated with an increased risk of disease progression in MASLD [17]. A prospective study of 578,700 adults followed up over a median period of 12 years showed that an increased body mass index (BMI) significantly increased the risk of HCC development 1.5-fold [18]. Another case–control study comparing 622 patients newly diagnosed with HCC and 660 healthy controls reported that obesity in early adulthood was a significant risk factor for HCC; each unit increase in BMI was associated with a 3.89-month decrease in age at HCC diagnosis [19].

Several molecular mechanisms linked to obesity, such as adipose-derived inflammation, lipotoxicity, insulin resistance, and alterations in the gut microbiota, may accelerate the development of HCC in patients with MASLD. The chronic inflammation that accompanies obesity, which is induced by adipose tissue remodeling due to excess fat accumulation, results in the secretion of adipokines such as leptin, adiponectin, resistin, interleukin-1beta and -6, tumor necrosis factor-alpha, and plasminogen activator inhibitor-1 by adipocytes and recruited macrophages [20,21]. These adipokines may activate oncogenic pathways in non-adipose liver tissues. Lipid accumulation in the liver is also directly associated with lipotoxicity, which may contribute to hepatocarcinogenesis, inducing endoplasmic reticulum stress [22] and the generation of reactive oxygen species [23], and interfering with cellular signaling mechanisms and transcriptional regulation [24]. In addition, obesity-related increases in the levels of insulin and insulin-like growth factor may promote HCC development through the activation of various oncogenic pathways [25]. Finally, alterations in the gut microbiota of obese patients have been implicated in hepatocarcinogenesis [26].

Cardiometabolic risk factors: prediabetes/diabetes

The presence and duration of diabetes are major risk factors for fibrosis progression and HCC development in patients with MASLD [17,27]. A recent study of 271,906 patients with MASLD showed that the risk of HCC development was significantly higher in patients with diabetes (8.36) than in patients without diabetes (1.07) [28]. Another study reported a 4-fold increased risk of HCC in individuals with diabetes and MASH-related cirrhosis during a 47-months follow-up period [29]. In addition, a meta-analysis of 16 studies involving a total of 891,426 individuals found that prediabetes was associated with an increased risk of cancer, particularly liver cancer [30].

Various biological mechanisms may be responsible for the association between prediabetes/diabetes and HCC. Insulin resistance and hyperinsulinemia due to prediabetes/diabetes increase the serum level and biological activity of insulin-like growth factor 1, triggering downstream cellular pathways such as phosphatidylinositol-3 kinase, protein kinase B, and mitogen-activated protein kinase. These pathways induce the proliferation of HCC cells and inhibit their apoptosis, ultimately promoting tumorigenesis [31,32]. Chronic inflammation caused by the hyperglycemia-induced generation of glycosylated hemoglobin and reactive oxygen species can also lead to cellular damage and oxidative stress in hepatocytes, resulting in carcinogenesis [33,34]. In addition, endoplasmic reticulum stress-induced damage due to diabetes may impair cellular functions and lead to cell death [35]. Finally, the overactivation of platelet-derived growth factor signaling [36] and increased lipopolysaccharide levels induced by changes in the gut microbiota [37] may be associated with the development of HCC.

Other cardiometabolic risk factors: (pre)hypertension and dyslipidemia

A high proportion of patients with MASLD exhibit hypertension and dyslipidemia [17]. However, the association among hypertension, dyslipidemia, and HCC development in these patients remains controversial. A recent retrospective study reported that patients with MASLD, hypertension, and dyslipidemia had a 1.8-fold higher risk of progression to cirrhosis or HCC than those without CMRFs [28]. However, a prospective study found that mean blood pressure and cholesterol or triglyceride levels were not independently associated with HCC development after adjusting for cofounding factors [18]. Therefore, further well-designed studies are required to determine whether an association exists between hypertension and dyslipidemia and the development of HCC.

Genetic predisposition

Genetic predisposition plays a role in the development of HCC in MASLD. The expression of genetic variants such as patatin-like phospholipase domain-containing protein 3 [38] and transmembrane 6 superfamily member 2 [39] are associated with an increased risk of MASLD-related HCC. A recent study reported that MBOAT7 variant rs641738 is associated with an increased risk of HCC, especially in patients with MASLD without cirrhosis [40], and a meta-analysis found that the HSD17B13 TA allelic variant rs72613567 reduced the risk of HCC in patients with MASLD [41]. Further studies are needed to fully elucidate the contribution of genetic predisposition to HCC risk in patients with MASLD.

Race and ethnicity

Determining the impact of cultural and ethnic factors on MASLD progression and HCC development is challenging. However, a meta-analysis of 34 studies comprising 368,569 individuals identified significant racial and ethnic disparities in MASLD prevalence and severity in the United States, with the highest burden in Hispanic individuals and the lowest in African Americans [42].

Alcohol and smoking

Alcohol consumption increases risk of liver cancer dramatically, especially in those with heavy alcohol use in patients with liver disease from any etiology: 46% for 50 g/day of ethanol and 66% for 100 g/day of ethanol [43]. However, the influence of mild to moderate alcohol exposure on the development of HCC in patients with MASLD is still controversial. Recent study showed that low-to-moderate alcohol consumption increased the risk of significant fibrosis in patients with MASLD, with a dose-dependent supra-additive interaction with CMRFs [44]. In contrast, a meta-analysis reported that modest alcohol consumption resulted in a 41% lower risks for MASH and advanced fibrosis in patients with MASLD [45]. Another study showed that a low alcohol consumption (< 20 g/day of ethanol) was a risk factor for HCC development in MASLD patients, especially those with advanced fibrosis [46]. Smoking is a known risk factor in many types of cancer, and smoking may be associated with advanced fibrosis in patients with MASLD through its effect on insulin resistance [47]. Additional data would be necessary to elucidate the association of smoking with HCC risk in MASLD.

Hepatocellular carcinoma surveillance strategies in patients with metabolic dysfunction-associated steatotic liver disease

current guidelines recommend HCC surveillance for patients with cirrhosis at Child-Pugh stage A and B, and stage C awaiting liver transplantation [48-50]. In terms of cost-effectiveness, an HCC incidence of 1.5%/year or greater would warrant surveillance in patients with cirrhosis [48,49]. However, there is currently no consensus regarding HCC surveillance in patients without cirrhosis. The European Association for the Study of the Liver guidelines recommend considering HCC surveillance in patients with stage F3 fibrosis, following an individual risk assessment [48]. The American Association for the Study of Liver Diseases guidelines do not recommend HCC surveillance in patients with MASLD in the absence of cirrhosis [49]. The American Gastroenterological Association recommends HCC surveillance in patients with MASLD positive for noninvasive markers suggestive of advanced liver fibrosis [50]. Because HCC develops in a substantial proportion of patients with MASLD but without cirrhosis [14], surveillance strategies are needed in this population. It has been suggested that surveillance would be warranted by an HCC incidence o f at least 0.2%/year in patients without cirrhosis [48,49]. However, no randomized trials on HCC surveillance have been performed to investigate all-cause and HCC-related mortality [51]. Further well-designed studies are needed to identify groups at a high risk of HCC development in non-cirrhotic patients with MASLD.

Current guidelines recommend that surveillance is conducted using abdominal ultrasound (US), with or without testing for tumor biomarkers such as serum α-fetoprotein (AFP), every six months [48-50]. However, this approach has several limitations. A recent study reported that US is much less sensitive in patients with MASH (0.59) than in those with other etiologies (0.84), and less sensitive in patients with a BMI ≥ 30 kg/m² (0.76) than in those with a BMI < 30 kg/m² (0.86) [52]. In addition, US has been shown to be inferior to cross-sectional imaging techniques such as computed tomography (CT) and magnetic resonance imaging (MRI) in patients with MASH, which may result in up to 41% of HCC cases being missed [52]. A study of 941 patients with cirrhosis showed that US was inadequate in over one-third of patients with Child-Pugh class C cirrhosis, a BMI > 35 kg/m², or MASH cirrhosis [53]. In these populations, CT and MRI are useful alternatives. A recent study used a Markov model to show that semiannual surveillance using MRI with liver-specific contrast may be more cost-effective than US in patients with virus-associated compensated cirrhosis at a sufficiently high risk of HCC, despite the higher cost of MRI [54]. A meta-analysis of HCC diagnostic accuracy showed that the sensitivity of abbreviated MRI (AMRI) was higher than that of US (82% vs. 53%), and that the sensitivity and specificity of non-contrast AMRI were comparable to those of contrast-enhanced AMRI (86% and 94% vs. 87% and 94%, respectively) [55]. In contrast, a randomized study of 163 patients with various etiologies of compensated cirrhosis reported that biannual US was marginally more sensitive for the detection of early HCC than annual CT, supporting the use of less costly US for HCC surveillance in this population [56]. Further studies are required to validate the clinical utility of CT and MRI for HCC surveillance in patients with MASLD.

Hepatocellular carcinoma risk prediction models in patients with metabolic dysfunction-associated steatotic liver disease

Risk models for cirrhotic patients

Because HCC risk may not be uniform even in patients with cirrhosis [57], several HCC risk prediction models have been developed to identify high-risk patients (Table 1) [58-61]. The ADRESS-HCC score was developed using a cohort of 34,932 cirrhotic patients and is composed of six factors: age, diabetes, race, etiology, sex, Child-Pugh score [58]. The C-indices in the training and internal validation cohorts were 0.70 and 0.69, respectively. The score stratified patients correctly according to the risk of developing HCC within 5 years (quartile 1: low risk, quartiles 2 and 3: indicate intermediate risk, quartile 4: high risk). The Toronto HCC risk index was developed using a cohort of 2,079 cirrhotic patients and is composed of four factors: age, sex, etiology, and platelets [59]. It had a C-index of 0.77 in both the training and validation cohorts, and divided HCC risk into three categories: low, < 120; medium, 120–240; and high, > 240. The 10-year cumulative HCC incidences were 3%, 10%, and 32% in the low-, medium-, and high-risk groups, respectively [59]. Ioannou et al. [60] also developed a model to predict the risk of HCC in patients with MASLD or alcohol-related cirrhosis, which comprised seven predictors: age, sex, presence of diabetes, BMI, platelet count, serum albumin level, and aspartate aminotransferase to √alanine aminotransferase ratio. The model showed good performance, with an area under the receiver operating characteristic curve (AUROC) of 0.75 for MASLD-cirrhosis. The annual risk of HCC in patients in the low-, medium-, and high-risk groups was 0–1%, 1–3%, and 3%, respectively [60]. The APAC score, which consists of age and levels of soluble platelet-derived growth factor receptor beta, AFP, and creatinine, was developed using 267 patients with liver cirrhosis [61]. The score identified HCC with an AUROC of 0.95, which was superior to the GALAD score for patients with MASLD (AUROC: 0.90). In particular, the APAC score demonstrated a greater ability to identifying early stage HCC (BCLC stage 0/A) than the GALAD score (AUROC: 0.93 vs. 0.81).

Table 1 . HCC Risk Prediction Models in Cirrhotic Patients with MASLD.

HCC risk model (year)	Country	Patients, number	Variables (number)	Predictive performance	Validation
ADRESS-HCC (2014) [58]	USA	34,932 (MASLD: 6,113)	Age, diabetes, race, etiology, sex, Child-Pugh score (6)	C-index (training: 0.70, internal validation: 0.69)	Yes (internal, external)
Toronto HCC risk index (2018) [59]	Canada	2,079 (MASLD: 111)	Age, sex, etiology, platelets (4)	C-index (training: 0.76, external validation: 0.77)	Yes (internal, external)
Ioannou et al. (2019) [60]	USA	7,068	Age, gender, diabetes, BMI, platelets, serum albumin and AST/√ALT ratio (7)	C-index (training: 0.75, internal validation: 0.72)	Yes (internal)
APAC score (2021) [61]	Germany	267 (MASLD: 60)	Age, sPDGFRβ, AFP, creatinine (4)	AUROC (training: 0.95, internal validation: 0.94)	Yes (internal)

AFP, α-fetoprotein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; AUROC, area under the receiver operating characteristic curve; BMI, body mass index; C-index, concordance index; HCC, hepatocellular carcinoma; MASLD, metabolic dysfunction-associated steatotic liver disease; sPDGFRβ, soluble platelet-derived growth factor receptor beta..

Risk models for non-cirrhotic patients

Risk stratification for HCC in patients with MASLD but without cirrhosis is challenging. Several HCC risk prediction models have been developed for this population (Table 2) [40,62-64]. Donati et al. [40] reported that the number of PNPLA3, TM6SF2, and MBOAT7 risk variants was associated with HCC development in patients with MASLD without cirrhosis, independent of clinical factors, and developed a combined risk score considering the acquired and genetic risk factors (age, sex, obesity, Type 2 diabetes, severe fibrosis, and number of risk alleles). The score had a AUROC of 0.96 for detecting HCC development and the optimal cutoff value exhibited 96% sensitivity and 89% specificity. The GALAD score was developed using 356 patients with MASH and comprises five factors: age, sex, and serum levels of AFP, AFP isoform L3, and des-gamma-carboxy prothrombin [62]. The AUROC for identifying any stage of HCC development using the GALAD score (0.96) was significantly higher than that using AFP (0.88), AFP isoform L3 (0.86), and des-gamma-carboxy prothrombin (0.87) levels. In particular, the AUROC for non-cirrhotic patients with MASH was 0.98. The GALAD score also had a high AUROC of 0.91 for the detection of HCC using the Milan Criteria, with a sensitivity of 68% and specificity of 95% at a cutoff of –0.63. In addition, one study investigated genetic predispositions contributing to HCC development in patients with MASLD [63]. The polygenic risk score (PRS), including PNPLA3, TM6SF2, GCKR, and MBOAT7 variants, predicted HCC development more robustly than the expression of single variants in 2,566 patients with MASLD. The association between the PRS and HCC risk was independent of fibrosis and was identified in patients without severe fibrosis. Although the PRS had only a moderate AUROC of 0.65, the fact that this score is easily calculated by a simple, one-off blood test and is independent of environmental factor fluctuations, makes it an attractive potential option for HCC surveillance considering cost-effectiveness [63] Recently, Kim et al. [64] developed an HCC risk prediction model for non-cirrhotic patients with MASLD using a nationwide cohort of 409,888 individuals in South Korea. The 11-point HCC risk prediction model comprised six factors: age, sex, diabetes, obesity, serum alanine aminotransferase level, and gamma-glutamyl transferase level. The AUROCs were 0.79 at 5 years and 0.84 at 10 years in the validation cohort (n = 8,721) with good calibration. The score stratified patients into three risk groups: 0–6, low; 7 or 8, moderate; and 9–11, high (estimated HCC incidence > 0.2%/year).

Table 2 . HCC Risk Prediction Models in Non-cirrhotic Patients with MASLD.

HCC risk model (year)	Country	Patients, number	Variables (number)	Predictive performance	Validation
Donati et al. (2017) [40]	Italy	765	Age, sex, obesity, T2DM, severe fibrosis, and number of risk alleles (PNPLA3, TM6SF2, and MBOAT7) (6)	AUROC: 0.96	No
GALAD score (2020) [62]	Germany	356 MASH patients (non-cirrhotic: 212)	Age, sex, AFP, AFP-L3, DCP (5)	AUROC: 0.98 (AUROC: 0.94 for early HCC)	Yes (external)
Polygenic risk scores (2021) [63]	Italy, UK, Germany	2,566	PRS-HFC: PNPLA3-TM6SF2-MBOAT7-GCKR PRS-5: PRS-HFC score adjusted for rs72613567 HSD17B13 variant	AUROC (PRS-HFC: 0.64, PRS-5: 0.65)	Yes (external)
Kim et al. (2024) [64]	Korea	409,088	Age, sex, diabetes, obesity, ALT, γ-GTP (6)	AUROC (training: 0.72 [5 yr], 0.75 [10 yr]; external validation: 0.79 [5 yr], 0.84 [10 yr])	Yes (internal, external)

AFP, α-fetoprotein; AFP-L3, AFP isoform L3; ALT, alanine aminotransferase; AUROC, area under the receiver operating characteristic curve; DCP, des-gamma-carboxy prothrombin; γ-GTP, gamma-glutamyl transferase; HCC, hepatocellular carcinoma; MASH, metabolic dysfunction-associated steatohepatitis; MASLD, metabolic dysfunction-associated steatotic liver disease; PRS, polygenic risk scores; PRS-HFC, polygenic risk scores for hepatic fat; T2DM, type 2 diabetes mellitus..

MASLD progression and HCC development are expected to increase gradually worldwide, alongside the increase in the incidence of metabolic disorders. Further studies are needed to elucidate the incidence of HCC in patients with MASH at various stages of fibrosis and with various metabolic risk factors. The substantial proportion of patients with MASLD without cirrhosis that develop HCC indicates that risk factors such as CMRFs and genetic predisposition may play an important role in the development of HCC, alongside fibrosis. The unfavorable influence of CMRFs on HCC risk in MASLD suggests that the risk of HCC development should be stratified according to the presence of CMRFs. An HCC incidence of > 1.5%/year warrants HCC surveillance using abdominal US with or without measurement of serum AFP levels in patients with cirrhosis; however, there is no consensus on surveillance strategies for patients without cirrhosis. Further well-designed studies are therefore needed to stratify the risk of HCC in patients without cirrhosis. In addition, the use of cross-sectional imaging techniques such as CT and MRI to address the limitations of US for HCC detection should be further investigated. Although several HCC risk stratification models have been developed for patients with MASLD, the heterogeneity of factors associated with HCC risk in these patients requires a more refined risk prediction model for individual patients.

This study was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (grant number: 2022R1I1A1A01068809 and 2022R1I1A1A01067589), the National Research Foundation of Korea (NRF) grant funded by the Korea government (Ministry of Science and ICT) (grant number: 2020R1C1C1004112 and 2019R1A2C4070136), and The Research Supporting Program of The Korean Association for the Study of the Liver and The Korean Liver Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

No potential conflict of interest relevant to this article was reported.

Conceptualization: Ho Soo Chun, Minjong Lee. Data curation: Ho Soo Chun, Minjong Lee, Tae Hun Kim. Funding acquisition: Ho Soo Chun, Minjong Lee. Investigation: Ho Soo Chun, Minjong Lee, Tae Hun Kim. Project administration: Minjong Lee, Tae Hun Kim. Supervision: Minjong Lee, Tae Hun Kim. Writing – original draft: Ho Soo Chun. Writing – review & editing: Minjong Lee, Tae Hun Kim.