HCC in metabolic dysfunction-associated steatotic liver disease
Background:
Metabolic dysfunction-associated steatotic liver disease (MASLD), formerly known as nonalcoholic fatty liver disease (NAFLD), is estimated to affect 25-30% of the population worldwide. MASLD encompasses a spectrum of liver disease from hepatic steatosis, steatohepatitis, to cirrhosis. MASLD accounts for up to 35% of HCC cases worldwide, with variability between different regions of the world. It is the fastest-growing contributor to liver cancer.
The incidence of HCC in MASH cirrhosis has been reported to be around 10-15 per 1,000 person years or annual incidence around 1.5%. Incidence of HCC in patients with MASLD without cirrhosis is much lower, reported to be around 0.08 per 1,000 person years. However, it has been reported that 15- 46% of patients with MASLD related HCC do not have evidence of cirrhosis, with 27 – 38% of these patients having evidence of advanced fibrosis. The proportion of MASLD associated HCC patients without cirrhosis is higher than in other etiologies of liver disease, though estimates vary widely depending on the studied cohort.
In comparison to other etiologies of liver disease, such as viral hepatitis and alcohol-associated liver disease, MASLD is associated with a lower risk of developing HCC. In patients with hepatitis B virus infection, there is a 10-25% lifetime risk of developing HCC. Cohort studies have shown the annual risk of HCC to be 2-3% in patients with HCV cirrhosis. In patients with hepatitis C, treatment with direct acting antiviral agents decreases risk of HCC, though patients with cirrhosis still require HCC screening after sustained viral response.
Clinical risk factors
Several studies have shown that older age and male sex are associated with increased risk of HCC across different etiologies of liver disease, and in MASLD specifically. Studies showed that estrogen may be protective against inflammation and injury associated with carcinogens, and may also slow progression of HCC. Differences in adipocyte metabolism through androgen and estrogen signaling pathways likely also contribute to these differences.
Tobacco use is a known risk factor in many types of cancer. Many of the elements of tobacco are known carcinogens, such as the nitrosamines which induce mutations in genes and promote tumor progression. Specifically in MASLD patients, smoking has been associated more advanced fibrosis.
Alcohol exposure increases risk of HCC dramatically, by up to 66% in those with heavy alcohol use in patients with liver disease from any etiology. Alcohol not only results in chronic inflammation leading to cirrhosis, but also induces carcinogenesis through production of acetaldehyde and reactive oxygen species. Studies examining the effect of alcohol consumption in MASLD patients have been mostly observational studies with mixed results. Some data has shown that mild to moderate alcohol use decreases the risk of developing MASH and advanced fibrosis. However, other studies have that moderate alcohol consumption is associated with fibrosis progression in patients with MASLD, and specifically in patients with MASLD and type 2 DM. Loomba et al showed a synergistic effect between obesity and alcohol in the development of HCC, and studies have shown that alcohol is associated with increased risk of HCC in patients with MASH cirrhosis and MASLD with advanced fibrosis.
Obesity is a known risk factor for MASLD as well as multiple cancers, though the underlying mechanism is complex and not fully understood. Increased adipose tissue leads increased secretion of adipokines, chemokines, and components of extracellular matrix, which leads to an altered micro and macroenvironment. Obesity is also linked to increased concentrations of insulin and insulin-like grown factor-1, which stimulate cell proliferation and protein synthesis pathways for tumorigenesis. Additionally, elevated free fatty acids are associated with tumor initiation as well as cancer progression. In one study examining patients with MASLD and obesity, those who underwent bariatric surgery had a lower risk of HCC compared to those who did not undergo surgery. Kanwal et al found that patients with a combination of metabolic risk factors, including obesity, diabetes and hypertension, had a higher risk of HCC compared to those with obesity alone.
Diabetes mellitus (DM) has been found to be an independent risk factor in the development of HCC in MASLD as well as in other etiologies of chronic liver disease. It has been reported that having DM confers a two to four fold increase in risk of HCC, with higher risk in those with poor glycemic control compared to adequate glycemic control (HbA1c <7). Additionally, antidiabetic medications have been shown to decrease the risk of HCC. One study showed that use of metformin decreased the risk of developing HCC in patients with MASLD by 21% over 10 years. Mouse models have also shown protective effects of diabetes medications such as dipeptidyl peptidase 4 and sodium-glucose cotransporter-2 inhibitors against the development of HCC.
Genetic predisposition also plays a role in the progression of MASLD and development of HCC. The polymorphism Patatin-like phospholipase domain-containing protein 3 (PNPLA3) rs738409: C>G is associated with increased risk of MASLD associated HCC. Membrane Bound O-Acyltransferase Domain Containing 7 (MBOAT7) rs641738 polymorphism is also correlated with increased risk of HCC, particularly in patients without cirrhosis. A systematic review and meta-analysis showed that that the TA allele of HSD17B13 rs72613567 had a protective effect on development of HCC, and slows progression of liver disease in patient with MASLD. More research is needed to understand the role of genetic predisposition in the development of HCC.
Cirrhosis is a known risk factor for development of HCC in MASLD and in other liver disease etiologies. In the absence of cirrhosis, metabolic factors and degree of steatosis play a role in the progression to HCC, with a higher incidence of HCC in MASLD patients with a higher grade of steatosis. In fact, data suggests that most patients with non-cirrhotic HCC have advanced fibrosis. Pinyopornpanish et al showed that age > 65, male sex, elevated alanine aminotransferase, diabetes and smoking were risk factors in the development of HCC in MASLD patients without cirrhosis. One study showed that in patients with MASLD, MASH, or at risk for MASH, Fib-4 > 2.67 was found to be an independent risk factor for development of HCC. Baseline liver stiffness measurement (LSM) and increases in LSM from baseline are also associated with risk of HCC in patients with MASLD. In patients without cirrhosis, HCC is more likely to be larger at presentation, likely due to lack of screening recommendations.
Risk Stratification
Several scoring systems have been studied for early detection of HCC (Table 1). The Toronto HCC risk index (THRI) uses age, sex, etiology of cirrhosis and platelet count to stratify patients into low-, intermediate-, and high-risk groups. The high-risk group had a 5 year incidence of HCC of 15.4%, with a modified c-statistic of 0.76. The aMAP score involves age, sex, albumin, bilirubin and platelets. Patients were stratified into low-, medium- and high-risk groups with overall sensitivity 85.7%-100% and negative predictive value 99.2%-100% in the validations cohorts. Another study of MASH patients used an HCC prediction model including liver stiffness, age, platelet count and AST, with AUC 0.777, 0.781, and 0.784 at 2, 3, and 5 years, respectively, in the validation cohort.
There is emerging interest in moving beyond clinical risk factors and validating potential biomarkers for risk stratification. Fujiwara et al defined a 133-gene signature PLS-NAFLD, which predicted incident HCC over 15 years, with incidence rates of 22.7% in high-risk patients and 0% in low-risk patients with MASLD. Genetic risk factors have also been proposed as part of risk stratification models. A combination of these genes, including PNPLA3, MBOAT7, transmembrane 6 superfamily member 2 (TM6SF2) and glucokinase regulator (GCKR) were used to develop a polygenic risk score of hepatic fat content. A modified polygenic risk score was associated with higher risk of HCC in MASLD patients with and without severe fibrosis, with an AUROC 0.74. Nanon et al showed that addition of a genetic risk score to clinical models for HCC risk prediction only modestly improved performance. Further studies are needed for validation of risk prediction models for HCC.
HCC Surveillance
Guidelines recommend ultrasound (US) screening +/- AFP every 6 months for patients with Child Pugh A or B cirrhosis, or Child Pugh C cirrhosis on the transplant waiting list. AASLD guidelines specifically do not recommend surveillance in non-cirrhotic MASLD patients. EASL guidelines specify that non-cirrhotic F3 patients may be considered for surveillance based on individual risk assessment. AGA guidelines recommend surveillance for MASLD patients with non-invasive markers suggesting advanced liver fibrosis or cirrhosis. Cost-effectiveness studies suggest that patient populations with incidence of HCC of ~1.0% per year or greater should undergo HCC surveillance, which is higher than currently estimated in MASLD patients without cirrhosis.
Studies have shown that US evaluation can be limited, potentially missing up to 41% of HCC in patients with MASLD, with significantly lower sensitivity when compared to other etiologies of chronic liver disease. Studies have shown that high BMI is associated with inadequate ultrasound quality, with a thicker subcutaneous fat layer causing attenuation of the US beam.
CT and MRI are alternative modalities; however, both have limitations including use of contrast and cost. One prospective study showed improved detection of early-stage HCC when compared to US, though patients mostly had HBV-related cirrhosis. Smaller observational have shown that abbreviated magnetic resonance imaging (AMRI) for HCC screening results in improved detection of HCC when compared to US. Additionally, non-contrasted AMRI has been shown to have improved HCC detection regardless of presence of hepatic steatosis. A randomized control trial in patients with various etiologies of cirrhosis showed that biannual US was comparable to annual CT for detection of HCC, with lower costs and without use of radiation. However, only 3.1% of patients had MASLD in this study. Per AASLD guidelines, CT and MRI can be used as alternative imaging methods for HCC screening when US evaluation is limited. However, these modalities are not recommended as first-line screening modalities in guidelines. Further studies are needed to validate these findings and determine utility of CT and MRI for HCC screening.
Biomarkers for Early Detection
Several biomarkers have been proposed and studied for HCC surveillance and identification of patients with early stage HCC. Given heterogeneity in HCC lesions, there is increasing recognition that a single biomarker is unlikely to be sufficiently sensitive and a combination of biomarkers is likely required to improve early-stage HCC detection.
The GALAD score includes age, sex, AFP-LCA-bound fraction (L3), AFP, and des-gamma-carboxy prothrombin (DCP) and has been studied and validated in various patient cohorts. In MASLD patients, it has high accuracy in early detection of HCC, in those with and without cirrhosis (AUC 0.93 and 0.98, respectively). The HCC early detection screening (HES) algorithm includes age, AFP, platelets, rate of change in AFP, and etiology of cirrhosis. It identified HCC with specificity 90% and sensitivity 56% with improved performance when combined with US; however, it has not been studied in the MASLD population. The APAC score consists of age, soluble platelet-derived growth factor receptor beta (sPDGFRβ), AFP, and creatinine. It identified HCC in cirrhotic patients with AUC 0.95, and specifically in MASLD patients with AUC 0.95.
DNA fragments from tumor cells can be detected in the serum and plasma by analysis of circulating cell free DNA (cfDNA). CfDNA contains a varying amount of circulating tumor DNA (ctDNA), from which mutations and methylation have been studied in the context of developing a diagnostic markers for early detection of HCC. Xu et al developed a model of 10 DNA methylation markers which yielded high sensitivity (83.3%) and specificity (90.5%) for detection of HCC in validation cohorts. In a phase II trial a methylated DNA marker panel with 6 markers resulted in an AUC of 0.96, and stage 0 and stage A HCC were detected at 75% and 93%, respectively. In a subsequent study, Chalasani et al combined 3 methylation markers with AFP and sex in a multitarget HCC blood test (mt-HBT) which had 88% sensitivity and 87% specificity in detection of HCC in an independent cohort. These HCC early detection models have not been studied specifically in MASLD cohorts. Research is ongoing to validate the use of cfDNA for early detection of HCC.
Conclusions:
MASLD is a leading cause of liver disease worldwide, and the fastest growing cause of HCC. Several clinical risk factors have been associated with the development of HCC in MASLD patients include age, sex, diabetes, obesity, and tobacco and alcohol use. Cirrhosis is a major risk factor, though over 20% of MASLD patients with HCC do not have cirrhosis. Guidelines currently recommend screening for HCC in MASLD patients with cirrhosis, with EASL and AGA also recommending consideration of screening in patients with advanced fibrosis. The current screening approach of US +/- AFP may be less sensitive in patients with MASLD especially in those with obesity, and cross-sectional imaging may be considered in cases with severe visualization limitations on ultrasound. Further research is needed to validate emerging HCC risk stratification and early detection strategies in patients with MASLD.
Take-Home Points:
- MASLD continues to be an increasing issue worldwide and is the fastest growing contributor to HCC.
- The overall risk of HCC in MASLD patients is lower than in other etiologies of liver disease; however, is responsible for up to 35% HCC cases given its high prevalence.
- A significant proportion of MASLD patients with HCC do not have cirrhosis. Current AASLD guidelines do not recommend HCC screening in MASLD patients without advanced fibrosis or cirrhosis.
- Further research is needed for risk stratification and early detection of HCC, particularly in patients with MASLD.
Table 1: HCC Risk Stratification and Early Detection Models
Risk stratification | Patient population | Variables | Notes |
(Sharma et al) |
2,079 cirrhotic patients -111 with MASLD |
Age, sex, cirrhosis etiology, platelet count | Has not been studied in MASLD cohorts |
(Fan et al) |
3,688 patients with chronic hepatitis B | Age, sex, albumin, bilirubin, platelets | Validated in a cohort with non-viral hepatitis, most of which attributable to MASLD. |
Bianco et al |
2566 non-cirrhotic MASLD patients - 266 with HCC |
PNPLA3, TM6SF2, MBOAT7, GCKR, rs72613567 HSD17B13 | Developed and validated in MASLD cohort |
Lee et al |
2666 MASLD patients -22 with HCC |
Age > 60, platelet count <150, liver stiffness, AST > 34 | Developed and validated in MASLD cohort |
PLS-NAFLD (Fujiwara et al) |
48 MASLD-related HCC patients | 133 gene signature with 80 high- and 53 low-risk genes | Developed and validated in MASLD cohort |
(Pelusi et al) |
72 MASLD-related HCC patients, without cirrhosis | Combination of rare genetic variants | Developed and validated in MASLD cohort |
Early Detection | |||
(Johnson et al) |
331 HCC patients (17 MASLD) 339 chronic liver disease patients without HCC (12 MASLD) |
Sex, age, AFP, AFP-L3, DCP | External validation in MASLD cohort |
HCC Early Detection Screening Algorithm (El-Serag et al) |
11,721 HCC cirrhosis | Age, AFP, ALT, platelet count | Has not been studied in MASLD cohorts |
Xu et al |
715 HCC, 560 control cases -Proportion of patients with MASLD not reported |
10 methylation markers (BMPR1A, PSD, ARHGAP25, KLF3, PLAC8, ATXN1, Chr 6:170, Chr 6:3, ATAD2, Chr 8:20) | Not studied specifically in MASLD cohort |
(Chalasani et al) |
136 HCC, 404 control cases - 26% with MASLD in both HCC and control groups |
3 methylation markers (HOXA1, TSPYL5, and B3GALT6), AFP, and patient sex | Similar performance in viral and non-viral liver disease subgroups, but not yet validated in MASLD cohort. |