Nonalcoholic fatty liver disease and hepatocellular carcinoma: A weighty connection


  • Brad Q. Starley,

    1. Department of Medicine, Division of Gastroenterology and Hepatology, Brooke Army Medical Center, Fort Sam Houston, TX
    Search for more papers by this author
    • All authors contributed equally to this article.

  • Christopher J. Calcagno,

    1. Department of Medicine, Brooke Army Medical Center, Fort Sam Houston, TX
    Search for more papers by this author
    • All authors contributed equally to this article.

  • Stephen A. Harrison

    Corresponding author
    1. Department of Medicine, Division of Gastroenterology and Hepatology, Brooke Army Medical Center, Fort Sam Houston, TX
    • Department of Medicine, Division of Gastroenterology and Hepatology, Brooke Army Medical Center, 3851 Roger Brooke Drive, Fort Sam Houston, TX 78234
    Search for more papers by this author
    • fax: 210-916-3195

  • Potential conflict of interest: Nothing to report.

  • Disclaimer statement: The opinion of ascertains contained herein are the private views of the authors and are not to be construed as official or reflecting the view of the Department of the Army or the Department of Defense


Hepatocellular carcinoma (HCC) is a common and deadly malignancy that is increasing in incidence in developed countries. The emergence of hepatitis C virus (HCV) accounts for about half of this increase in HCC, although the etiology of HCC in 15%-50% of new HCC cases remains unclear. The most common form of chronic liver disease in developed countries is nonalcoholic fatty liver disease (NAFLD), which encompasses a broad spectrum of histopathology. The prevalence of NAFLD, including the more aggressive nonalcoholic steatohepatitis (NASH), is increasing with the growing epidemics of diabetes and obesity. NASH can progress to cirrhosis and its related complications. Growing evidence suggests that NASH accounts for a large proportion of idiopathic or cryptogenic cirrhosis, which is associated with the typical risk factors for NASH. HCC is a rare, although important complication of NAFLD. Diabetes and obesity have been established as independent risk factors for the development of HCC. New evidence also suggests that hepatic iron deposition increases the risk of HCC in NASH-derived cirrhosis. Multiple case reports and case reviews of HCC in the setting of NASH support the associations of diabetes and obesity with the risk of HCC, as well as suggest age and advanced fibrosis as significant risks. Insulin resistance and its subsequent inflammatory cascade that is associated with the development of NASH appear to play a significant role in the carcinogenesis of HCC. The complications of NASH, including cirrhosis and HCC, are expected to increase with the growing epidemic of diabetes and obesity. (HEPATOLOGY 2010;51:1820–1832)

Primary liver cancer is one of the most common and deadly cancers worldwide. Globally, it represents the fifth most common cancer and the third most common cause of cancer death, behind only lung and stomach cancer.1-4 Hepatocellular carcinoma (HCC) accounts for the majority of these primary cancers of the liver. More than 80% of HCC cases occur in less developed countries, particularly East Asia and sub-Saharan Africa, and are typically associated with chronic hepatitis B and C, although the incidence in these countries is decreasing.3, 4 Interestingly, the incidence of HCC in developed countries including Japan, Australia, Europe, Canada, and the United States has been increasing over the last 20 years.5, 6 In the United States alone, the annual incidence of HCC has increased about 80% during the last 2 decades.1 The emergence of hepatitis C virus (HCV) in developed countries accounts for about half of this increase in HCC.1, 6 The etiology of HCC in 15%-50% of new HCC cases remains unclear, which suggests that other risk factors likely account for the increase.7 The most common form of chronic liver disease in these developed countries is nonalcoholic fatty liver disease (NAFLD), which encompasses a clinicopathologic spectrum of disease ranging from isolated hepatic steatosis to nonalcoholic steatohepatitis (NASH), the more aggressive form of fatty liver disease, which can progress to cirrhosis and its associated complications, including hepatic failure and HCC.8 NASH may account for a large proportion of idiopathic or cryptogenic cirrhosis (CC), which predisposes these patients to the development of HCC.7


BMI, body mass index; CC, cryptogenic cirrhosis; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; IGF, insulin-like growth factor; IL-6, interleukin-6; IRS-1, insulin receptor substrate-1; JNK1, c-Jun amino-terminal kinase 1; M6P/IGF2R, mannose 6-phospate/insulin-like growth factor-2 receptor; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; NF-κB, nuclear factor kappa B; Nrf1, nuclear respiratory factor-1; OR, odds ratio; ROS, reactive oxygen species; RR, relative risk; TNF-α, tumor necrosis factor-alpha.

Epidemiology of NAFLD

NAFLD is the most common etiology of chronic liver disease in the United States and other developed countries.8, 9-11 The annual incidence of NAFLD has been estimated to be as high as 10% with the development of NAFLD associated most directly with the metabolic syndrome and preceding weight gain.12-14 Worldwide, the prevalence of NAFLD in the general population ranges from 9%-37%.15-22 In the United States, recent estimates suggest that NAFLD affects 30% of the general population and as high as 90% of the morbidly obese.23 The histopathologic entity known as NASH has been estimated to affect 5%-7% of the general population and as many as 34%-40% of patients who have elevated liver enzymes in the setting of negative serologic markers for liver disease.8, 24-26 An ongoing prevalence study at Brooke Army Medical Center in San Antonio, TX, suggests a much higher prevalence of NAFLD and NASH than previously estimated. A total of 255 asymptomatic patients enrolled predominantly during routine colon cancer screening visits underwent a right upper quadrant ultrasound, to look for NAFLD as defined by ultrasound criteria. One hundred patients who demonstrated NAFLD on ultrasound underwent liver biopsy to determine the prevalence of NASH. The overall prevalence of NAFLD by ultrasound criteria was found to be 46.3%. In the same population, the overall prevalence of histopathologically confirmed NASH was 13.1% rising to 31% among patients identified with NAFLD on ultrasound.27 These findings suggest that the prevalence of NAFLD and NASH may be much higher than previously estimated in the general population, although it should be noted that this is not a true population-based study. NAFLD has been described in persons of all ages, although the prevalence increases with age.15, 28 The highest prevalence of NAFLD has been found among Hispanics, followed by non-Hispanic whites, and lower prevalence in African-Americans.10, 11, 27, 29, 30 Obesity, diabetes mellitus, hyperlipidemia, metabolic syndrome, and insulin resistance have been established as risk factors for primary NAFLD.8, 31-35 With the progressive epidemics of obesity and diabetes mellitus, particularly in developed countries, the prevalence of NAFLD and its associated complications is expected to increase.8

Natural History of NAFLD

NAFLD encompasses a wide spectrum of disease. The natural history of the disease appears to be linked to the histology at the time of presentation. Patients with isolated steatosis on presentation generally have a benign prognosis, although patients with NASH may develop progressive fibrosis leading to cirrhosis and its complications8 (Fig. 1). Despite a lack of long-term prospective data, we do have some insight into the natural history of this disease. Some 26%-37% of patients with NASH demonstrate progression of fibrosis over time periods up to 5.6 years, with up to 9% progressing to cirrhosis.36-39 Of these same patients with NASH, the disease remains stable in 34%-50% of patients, and histology improves in 18%-29%.36-39 Body mass index (BMI) and diabetes have been found to be independent risk factors associated with progression of fibrosis.37 Previous reviews have demonstrated that one-third to one-half of NASH patients have progressive hepatic fibrosis over 3.5-5 years, and as many as 20% progress to advanced fibrosis over the same time period. This progression rate may be an overestimate due to selection bias.40, 41 Some 40%-62% of patients with NASH-related cirrhosis develop a complication of cirrhosis, including HCC, after 5-7 years of follow-up.42, 43 Retrospective data suggest that as many as 4%-27% of cases of NASH transform to HCC after the development of cirrhosis, although the overall occurrence of HCC in the setting of NAFLD remains a rare complication.44, 45 Longitudinal outcome studies report the prevalence of HCC in NAFLD to be 0%-0.5%, and the prevalence of HCC in NASH to be 0%-2.8% over time periods of up to 19.5 years43, 46-48 (Table 1). The development of cirrhosis in NASH typically occurs at an older age than in other liver diseases, although once cirrhosis does develop in patients with NASH, their clinical course is comparable to patients with other causes of cirrhosis.40, 44

Figure 1.

Natural history of nonalcoholic fatty liver disease.

Table 1. Longitudinal Outcome Studies: HCC Prevalence in NAFLD
AuthorDateNAge of cohortType of NAFLDYears follow-upNumber of HCC casesHCC prevalence
  1. NNFL, non-NASH fatty liver; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; NR, not reported.

Adams (43)200543549 ± 15NAFLD7.6 ± 420.5%
Ekstedt (46)20065847 ± 12NNFL13.7 ± 1.300%
Ong (47)2008817NRNAFLD8.400%
Rafiq (48)20097453 ± 25NNFL19.5 (median)00%
NAFLD/NNFL overall HCC prevalence = 0.14%
Ekstedt (46)20067155 ± 12NASH13.7 ± 1.322.8%
Rafiq (48)20095754 ± 12NASH14.98 (median)00%
NASH overall HCC prevalence = 1.6%

Cryptogenic Cirrhosis and NASH

NASH has been proposed as a probable cause of idiopathic or cryptic cirrhosis even though most of the histologic hallmarks of NASH are not present in CC.49-51 Patients with CC have a prevalence of diabetes and obesity similar to that of patients with NASH, and a significantly higher prevalence than in patients with cirrhosis from viral and autoimmune disease.50 Patients with CC also have a significantly higher prevalence of diabetes and obesity than age and sex matched patients with cirrhosis of well-defined etiology.51 The histologic findings of NASH, fatty deposition, and necroinflammation may disappear when the disease progresses to cirrhosis.51-53 These findings make a definitive diagnosis of NASH difficult when patients present with advanced disease, although the significant association between diabetes, obesity, and CC is very convincing. In addition, patients who undergo liver transplantation for CC frequently develop NAFLD and NASH after transplant. One study demonstrated that 25% of patients developed NAFLD and 16% showed histological evidence of NASH within 26 months of transplant.54 A large proportion of CC, therefore, likely represents end-stage NASH.

Cryptogenic Cirrhosis and HCC

Multiple retrospective studies have been done evaluating HCC in the setting of CC, which support the notion that NASH accounts for a large proportion of CC and can progress to HCC.42, 49, 55, 56 In 2002, Bugianesi et al. reviewed 641 patients with HCC.49 A total of 6.9% of the 641 patients developed HCC in the setting of CC, and these patients were compared to patients with HCC from HCV-related cirrhosis, hepatitis B virus (HBV)-related cirrhosis, and alcoholic cirrhosis.49 Analysis from this comparison confirmed that features associated with NASH, including obesity, diabetes, dyslipidemia, elevated glucose, and insulin resistance, were all significantly associated with CC.49 Another review of a little more than 100 patients with HCC found a much higher prevalence of 29% with underlying CC.55 This study confirms the significant association of obesity, diabetes, and hypertriglyceridemia with CC when compared to other causes of liver disease.55 In this review, 20% of patients in the cryptogenic liver disease group had evidence of NASH on liver biopsies prior to developing HCC, whereas half of the patients with CC had prior NASH or suspected NAFLD. The authors concluded that NAFLD was the underlying liver disease in 13% of the patients with HCC.55 Another retrospective review found CC in 27% of patients with HCC and again confirmed the correlation with diabetes, insulin resistance, and dyslipidemia with CC and HCC.42 Patients with CC were, on average, 8 years older than patients with HCV at the time of development of cirrhosis and about 3 years older at the time when HCC was detected.42

Risk Factors for HCC in NASH

Emerging evidence has established multiple independent risk factors for the development of HCC including obesity, diabetes, and iron deposition (Table 2). These factors also increase the risk for the development of NASH, a probable precursor to CC. It is well established that HCC develops in the presence of chronic liver disease, typically associated with cirrhosis from HBV, HCV, and/or alcoholic liver disease. Cirrhosis is the most important single risk factor for HCC and is present in about 80% of patients with HCC, regardless of underlying liver disease.57 As noted previously, NASH likely accounts for a large proportion of the idiopathic cirrhosis that makes up 6.9%-50% of underlying liver disease in patients with HCC in developed countries.7 This conclusion is further supported by evidence of linking common risk factors for NASH with risk factors for HCC.

Table 2. Risk Factors for HCC in NASH
Advanced fibrosis
Diabetes mellitus
Iron deposition

Obesity and HCC

Obesity has been established as a significant risk factor for the development of various malignancies, including liver cancer.49, 58-60 A large, prospective mortality study by the American Cancer Society61 demonstrated increased cancer mortality with increased body weight. The death rates from all types of cancers among the heaviest patients in the cohort (patients with a BMI > 40 kg/m2) were 52% higher for men and 62% higher for women compared with patients of normal weight. These significant mortality rates included death from esophageal, stomach, colorectal, liver, gallbladder, pancreatic, prostate, and kidney cancer as well as leukemia, non-Hodgkin's lymphoma, and multiple myeloma. Compared to patients with normal BMI, the relative risk (RR) of mortality from liver cancer was 1.68 times higher in women and 4.52 times higher in men with BMI > 35 kg/m2. Death from liver cancer among obese males demonstrated the highest RR of all the cancers in the study. This confirmed the results of another population-based study from Denmark of more than 40,000 obese patients, which showed that the RR of liver cancer was increased to 1.9 compared to the general population.62 A study from Korea published in 2005 examined the relationship between BMI and various cancers in 781,283 men without a prior diagnosis of cancer.63 The patients were followed over a 10-year period. Korean men with a BMI > 30 kg/m2 had a 26% increase in risk for all types of cancer compared to men with a normal BMI.63 An RR of 1.53 was demonstrated for HCC in obese males compared to normal controls, even after controlling for HBV infection, which is the most common cause of HCC in Korea.63, 64

A review of data from 19,271 patients who underwent orthotopic liver transplant in the United States between 1991 and 2000 showed the overall incidence of HCC was 3.4% with a slightly higher incidence among obese patients at 4.0%.65 The multivariate analysis confirmed that obesity is a statistically significant independent predictor for HCC in CC (odds ratio [OR] = 11.1) and alcoholic cirrhosis (OR = 3.2), although obesity was not a significant predictor in patients with viral hepatitis, primary biliary cirrhosis, or autoimmune hepatitis.

In a recent study, Ohki et al. followed 62 patients with HCC in the setting of non-HBV, non-HCV, nonalcoholic HCC after curative ablation. The analysis demonstrated older age and the accumulation of visceral fat as independent risk factors for recurrence of HCC. Patients with very high visceral fat areas (>130 cm2 in males and >90 cm2 in females) had significantly higher rates of recurrence of HCC (75.1% versus 43.1% at 3 years). The recurrence of HCC was also more likely to develop de novo in the setting of high visceral fat.66 Although these results are in the setting of HCC recurrence, this increased visceral fat accumulation is possibly involved in both tumor initiation and promotion of progression.

Obesity has definitively been established as a risk factor for the development of HCC, with a 1.5-4 times increased risk (Fig. 2).60-63 This risk is likely conferred by two factors: the increased risk for NAFLD with subsequent progression to NASH and the carcinogenic potential of obesity alone.7

Figure 2.

Relative risk of HCC in obesity.

Diabetes Mellitus and HCC

Large population-based cohort studies from Sweden, Denmark, and Greece demonstrate a 1.86-fold to 4-fold increase in risk of HCC among patients with diabetes (Fig. 3), which is closely associated with obesity and NAFLD.67-69 More recently, a case-control study in the United States showed that diabetes was associated with an increased risk for HCC, but only in patients with concomitant HCV-related, HBV-related, or alcohol-related cirrhosis.70 In a larger longitudinal study, the same group compared 173,643 diabetic patients with 650,620 nondiabetic controls over 10-15 years.71 The incidence of HCC increased more than two-fold among diabetic patients with higher increase among those with longer duration of follow-up. The risk of HCC with diabetes remained elevated even after excluding patients who were subsequently diagnosed with HCV, HBV, alcohol use, and/or fatty liver disease at any time during the follow-up.71 The risk for HCC was attributable to diabetes, and could not be explained by the presence of underlying liver disease or other risk factors. Diabetes is clearly established as an independent risk factor for HCC.

Figure 3.

Relative risk of HCC in diabetes mellitus.

The risk of HCC from diabetes may be decreased with the use of statins. Experimental and indirect human data suggest that statin use may reduce the progression of HCC as well as increase survival in advanced HCC.72-74 More recently, statins have been shown to significantly reduce the risk of HCC among patients with diabetes.75 A total of 1303 cases and 5212 controls were compared in a nested, matched, case-control study in patients with diabetes given the known higher risk of developing HCC. The study demonstrated a risk reduction range between 25% and 40% for the development of HCC in diabetic patients who were prescribed statins.75 This risk reduction persisted in analysis of patients with and without known liver disease or cirrhosis.75 NAFLD was found in approximately 38% of cases and possibly contributed to a higher number of cases given the fact that all of the patients had established diabetes. Future studies should be done to confirm these findings in diabetic patients as well as to evaluate the benefit of statins in NASH.

Iron Deposition and HCC

Increased hepatic iron stores are being recognized as clinically significant in a number of conditions including alcoholic liver disease, NAFLD, chronic hepatitis C, and end-stage liver disease.76 Excess hepatic iron may increase the risk for NASH and its progression to include HCC, although supporting data is limited.77 Clinical studies have shown that hepatic iron overload is a risk factor for the development of HCC in hemochromatosis, alcoholic liver disease, posttransplant patients, and in patients with HCC developed in a noncirrhotic liver.78-81 A recent retrospective study by Sorrentino et al. implicates iron deposition as a risk factor for HCC development in patients with NASH-related cirrhosis.82 Fifty-one patients with HCC in the setting of NASH-related cirrhosis were compared with 102 age-matched, sex-matched, and disease-matched HCC-free patients with NASH-related cirrhosis. Patients with hemochromatosis (including heterozygotes for the C282Y or H63D mutations), significant alcohol use, viral hepatitis, and other chronic liver diseases were excluded. Hepatic iron deposits were assessed retrospectively. The iron score was significantly higher in patients with HCC-NASH than in HCC-free NASH controls. A multivariate analysis demonstrated that histologically identified sinusoidal iron deposits were more frequent and larger in patients who had developed HCC than in controls.82 In this study, the only other condition independently associated with the development of HCC was diabetes mellitus.82 Excess sinusoidal iron deposition in NASH may play a role in liver injury, as well as possible carcinogenesis.

Risk Factors for NASH and HCC in Chronic HCV

The presence of hepatic steatosis, along with obesity and diabetes mellitus, has also been shown to increase the risk of HCC in patients with chronic HCV. Hepatic steatosis is an established histopathologic feature of chronic HCV with a prevalence ranging from 31%-72%.83-88 In 2003, Ohata et al. demonstrated that hepatic steatosis increases the risk for the development of HCC in patients with chronic HCV.89 The presence of steatosis, compared with no steatosis, independently and significantly increased the risk of developing HCC by 2.81 times.89 Hepatic steatosis also directly correlated with increasing BMI.89 Obesity has also been shown to be an independent risk factor for HCC development in patients with chronic HCV.90 Ohki et al. evaluated 1431 patients with chronic HCV that were followed for up to 10 years. The patients were divided into four groups according to BMI. Of these patients, 340 subsequently developed HCC. The risk of HCC in patients with chronic HCV increased in proportion to BMI with a hazard ratio of 1.86 for overweight patients and 3.10 for obese patients as compared to underweight patients.90 Recently, Konishi et al. published a study of a cohort of chronic HCV patients without a previous diagnosis of diabetes mellitus.91 The study demonstrated that the presence of diabetes mellitus, as evidenced by a positive 75 g oral glucose tolerance test, independently increased the risk of HCC development in patients with chronic HCV.91 NAFLD and its associated risk factors appear to act synergistically with other conditions to promote HCC.92 The data suggesting that hepatic steatosis, obesity, and diabetes mellitus increase the risk of HCC development in chronic HCV strengthens the notion that they are risk factors for the development of HCC in NASH. Treatment of these conditions related to NASH may decrease the risk of carcinogenesis associated with chronic HCV, and should be further evaluated.

Case Reports and Case Series of NASH and HCC

Multiple case reports of HCC in the setting of NASH have been published and reviewed in the literature (Table 3).8, 52, 57, 93-109 Male patients make up the majority of cases with a mean age at diagnosis of 66.7 (range = 45-82). The patients are typically older at presentation than patients with HCC related to other chronic liver diseases. This older age at presentation correlates to reports of older ages at presentation in patients with CC and HCC compared to alcohol-related, HBV-related, and HCV-related HCC controls.50 The majority of the patients with HCC in the setting of NASH also have underlying diabetes (64%), obesity (58%), or other manifestations of the metabolic syndrome which is consistent with previous findings. Patients typically have large, well-differentiated tumors at the time of presentation which may be secondary to a delayed diagnosis. Up to 50% of patients have HCC at the time of initial referral, and rarely patients present with HCC in the absence of cirrhosis.8, 93, 94, 96, 101, 105-109 The fact that HCC can arise in the setting of NASH without underlying cirrhosis raises the interesting possibility that carcinogenesis can occur in NAFLD in the absence of advanced liver disease.108

Table 3. Case reports of HCC in setting of NASH
Case #AuthorAge HCC DxGenderOther DxLiver D1Tumor #Size(cm)Liver HistologyTxFollow up (mo)Outcome
  1. Reproduced from Page JM, Harrison SA. NASH and HCC. Clin Liver Dis 2009;13:631-647. Used with permission.

  2. DM, diabetes mellitus; Dx, diagnosis; NR not reported; OB, obesity; PEI, percutaneous ethanol injection; RFA, radiofrequency ablation; TAE, transarterial embolization; TAI, transarterial infusion chemotherapy; Tx, treatment; Unk, unknown.

1Powell57FDM4Mult CirrhosisSurgery/Died
4Orikasa67FDMNRSingle2,6CirrhosisSurgery NR
6Shimada/Yoshioka69F 2Mult1.5CirrhosisTAE26Died
9Shimada63MOB0Single2CirrhosisSurgery48No Recurr
11Benchcqroun68MOB0UnkUnkNo CirrUnkUnkUnk
12Mori76MOB/DM10Single1.9CirrhosisRFA8No Recurr
13Bullock74MOB/DM0Single4No CirrSurgeryNRNR
14Bullock64MOB/DM0Single“large”No CirrTAE/Sur12No Recurr
15Cuadrado74MOB/DM0SingleNRNo CirrRFA12Recurred
17Sato64MOB0Single13No CirrNone12Died
19Hai72MOB/DM0Single4No CirrSurgeryNRNR
20Hai65M NRSingle6CirrhosisSurgeryNRNR
21Ichikawa66FOBNRSingle2.5No CirrSurgery14Recurred
22Ichikawa60M NRSingle1.5No CirrSurgery5No Reairr
25Hashizume74MDMUNKSingle2.4CirrhosisSurgery30No Recurr
26Hashizume45MOB/DMUNKSingle3.5CirrhosisSurgery22No Recurr
27Hashizume54MOB/DMUNKMult7.0/1.5No CirrSur/Chemo13No Recurr
28Hashizume82MOBUNKSingle3.0No CirrSurgery22No Recurr
30Hashizume75M UNKSingle3.0CirrhosisSurgeryNRNo Recurr
31Hashizume67FDMLINKSingle2.0CirrhosisRFANRNo Recurr
32Hashizume63FOB/DMUNKSingle1.5CirrhosisRFANRNo Recurr
33Hashizume72MDMUNKSingle4.0No CirrSurgery50No Recurr
34-67HashimotoMedianM (21)OB 62%NRNRNRFl/2 22%NRNRNR
 Study (N = 34)70F (13)DM 74%   F3/4 88%   

Three recent case studies of patients with HCC in the setting of NASH support the findings of previous case reports.57, 107, 109 Hashizume et al. reviewed nine patients with HCC in NASH. The majority of patients were male and the median age of diagnosis was 71.5 years. All of the patients had diagnoses of diabetes, hypertension, or hypertriglyceridemia, and the majority showed clinical evidence of insulin resistance and the metabolic syndrome. Three of nine patients developed HCC without evidence of underlying cirrhosis. The review supports the associations of age, cirrhosis, diabetes, and obesity with the risk for HCC.107 Chagas et al. identified seven cases of HCC in histologically defined NASH patients including one case without cirrhosis.109 The mean age at diagnosis was 63 years compared to 57 in patients with HCC associated with HBV and HCV infection. All seven of the patients were overweight, 57% of the patients had diabetes mellitus, and 28.5% had dyslipidemia. The histologic features were predominantly well-differentiated HCC similar to features of isolated case reports of HCC in NASH.109 A larger, case-controlled study from Japan reviewed 34 patients with NASH who had HCC and compared them to patients with NASH without HCC. Of the patients with HCC, the median age was 70 years compared to 50 years in the case of patients without HCC. Male sex, diabetes mellitus, and hypertension were more common in the NASH patients with HCC. Advanced fibrosis was significantly higher in NASH patients with HCC (88% versus 31%). Significant risk factors for HCC in the setting of NASH included older age, low level of aspartate aminotransferase, low grade of histological activity, and advanced stage of fibrosis. Older age and advanced fibrosis were the strongest risk factors for the development of HCC, and HCC was the major cause of mortality in NASH patients with advanced fibrosis.57

Pathophysiology of HCC in NASH

The majority of basic and clinical evidence regarding the pathogenesis of HCC arise in the setting of chronic viral hepatitis.110 It is clear that cirrhosis is linked to the development of HCC regardless of the underlying etiology of liver disease. The exact mechanism behind the development of HCC in NASH remains unclear, although the pathophysiologic mechanisms behind the development of NASH related to insulin resistance and the subsequent inflammatory cascade likely contribute to the carcinogenic potential of NASH (Fig. 4).

Figure 4.

Proposed pathogenesis of HCC in NASH.

Obesity and diabetes have clearly been established as risk factors for the development of NASH and CC, and they have been implicated in the development of multiple cancers, including HCC.7 Insulin resistance associated with obesity, metabolic syndrome, and diabetes leads to increased release of FFA from adipocytes, release of multiple proinflammatory cytokines including tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), leptin, and resistin, as well as decreased amounts of adiponectin. These processes favor the development of hepatic steatosis and inflammation within the liver.7, 110 Hyperinsulinemia up-regulates the production of insulin-like growth factor-1 (IGF1), which is a peptide hormone that stimulates growth through cellular proliferation and inhibition of apoptosis within the liver.93, 111, 112 Insulin also activates insulin receptor substrate-1 (IRS-1), which is involved in cytokine signaling pathways and has been shown to be up-regulated in HCC.113 The mannose 6-phosphate/IGF2 receptor (M6P/IGF2R) is involved in regulating cell growth by activating growth inhibitor and inactivating IGF2, a growth stimulator. The M6P/IGF2 receptor functions as a tumor suppressor. Mutations leading to loss of heterozygosity at this receptor have been found in 61% of patients with HCC.114 Adiponectin is an anti-inflammatory polypeptide specific to adipose tissue that is decreased in insulin-resistant states and has been shown to inhibit angiogenesis via modulation of apoptosis in an animal model.115, 116 These complex factors related to an insulin-resistant state, promote uninhibited cell growth and appear to play a significant role in the development of HCC in the setting of NASH.

The development of NASH is also associated with oxidative stress and the release of reactive oxygen species (ROS) which likely contributes to the development of HCC. An insulin resistant obese mouse model demonstrated that ROS production is increased in the mitochondria of hepatocytes with fatty infiltration, and that oxidative stress may be implicated in hepatic hyperplasia.7, 117 During carcinogenesis, epithelial hyperplasia and dysplasia generally precede cancer by many years.7, 118 The oxidative stress may favor tumorigenesis through steatosis, inflammation, and cell proliferation, or it may induce cancer-promoting mutations directly. Trans-4-hydroxy-2-nonenal, a product of lipid peroxidation, has been shown to cause mutations of the p53 tumor suppressor gene which is associated with more than half of human cancers including HCC.119 Nuclear respiratory factor-1 (Nrf1) is an essential transcription factor important in mediating oxidative stress. In an animal model, Xu et al. demonstrated that hepatocytes lacking transcription factor Nrf1 had increased susceptibility to oxidative stress. The hepatic histology in cases lacking Nrf1 demonstrated steatosis, apoptosis, necrosis, inflammation, and fibrosis. The specimens ultimately developed hepatic cancer related to the oxidative stress.120

Hepatocarcinogenesis in NASH may also be partially mediated by increased release of inflammatory and inhibitory cytokines such as TNF-α, IL-6, and NF-κB.79, 121-123 Evidence suggests a complex molecular interplay related to these inflammatory cytokines that leads to hepatocyte death, compensatory proliferation, and ultimately carcinogenesis.122 NF-κB regulates immune and inflammatory responses and is activated in many tumors, inhibiting apoptosis.123 A recent study by Luedde et al. demonstrated that inhibition of NF-κB in mouse livers induced steatohepatitis and ultimately HCC by sensitizing hepatocytes to spontaneous apoptosis. This chronic cycle of injury, cell death, and regeneration through compensatory cellular proliferation likely contributes to the development of hepatocellular carcinoma.123

The c-Jun amino-terminal kinase 1 (JNK1) has also recently been linked to obesity, insulin resistance, NASH, and HCC development. JNK1 is a ubiquitously expressed, mitogen-activated protein kinase. Obesity is associated with abnormally elevated JNK activity.124 Free fatty acids, TNF-α, and ROS released in the setting of hyperinsulinemia are all potent activators of JNK, which in turn phosphorylates IRS-1.124 JNK activation and the subsequent phosphorylation of IRS-1 are crucial components of the biochemical pathway responsible for obesity-induced insulin resistance.124 JNK activation is also known to increase hepatic inflammation and apoptosis.125 Puri et al. demonstrated that human patients with NASH have significantly increased phosphorylated JNK levels in comparison to patients with benign NAFLD.125 JNK activation is specifically associated with the presence of NASH, as well as the level of histologic activity.125 Mouse models have also demonstrated that JNK1 promotes the development of steatohepatitis.126 One mouse model demonstrated a protective effect with JNK1 ablation. The absence of JNK1 prevented weight gain and the development of insulin resistance, protected against the development of hepatic steatosis, and reduced hepatic injury as reflected by serum alanine aminotransferase levels compared to wild-type mice in response to a high-fat diet.127 These findings suggest that anti-JNK therapy can prevent the development of NASH as well as reverse chronic steatohepatitis, even in the setting of a persistent high-fat diet.127 JNK inhibitors have been used in treatment of human diseases, and possibly have a place in the future treatment of NASH.127

JNK activity has also previously been linked to a variety of cancer cell lines.128, 129 More recently, definitive evidence has revealed a significant relationship between sustained JNK activation and the development of HCC.129-132 JNK1 is overactivated in more than 50% of human HCC samples.129-132 In one study, 56% of HCC tissue samples demonstrated elevated JNK1 activity relative to the case-matched noncancerous liver tissue.131 This finding was supported by immunoblotting studies which demonstrated highly active JNK1 in about 55% of human HCC samples.130 JNK1 appears to be the most important kinase that is up-regulated in HCC.129 This sustained overactivation of JNK1 leads to an aberrant increase in several genes important for hepatocyte proliferation.129 With further research, these genes can potentially be defined and targeted as specific therapy.129 ROS, which are critical to the pathophysiology of NASH, are known to sustain JNK activation by inactivating JNK phosphatases and boosting JNK activity.133 As discussed previously, evidence suggests that statins significantly decrease the risk of HCC in diabetic patients, presumed secondary to the anti-inflammatory properties of the statins.72-75 Interestingly, atorvastatin therapy has been shown to acutely decrease expression of JNK and other inflammatory cells in patients with abdominal aortic aneurysms.134 This finding that statin treatment reduces JNK expression may explain, in part, the decreased risk of HCC in diabetic patients on statin therapy, although this has yet to be proven. Further studies linking statins and JNK activity with NASH and HCC may lead to important therapeutic benefit in the prevention and treatment of NASH as well as HCC secondary to NASH.


The most common cause of liver disease in developed countries is NAFLD, which includes NASH and its associated complications. The prevalence of NAFLD and NASH is likely higher than previously estimated and is associated with the growing epidemics of obesity and diabetes. Increasing evidence supports the fact that NASH can progress to HCC. The overall prevalence of HCC in patients with NAFLD remains low, although the incidence of HCC in developed countries is rising. Idiopathic, or CC, which accounts for 6.9%-50% of these cases of HCC, is clearly associated with diabetes, insulin resistance, and obesity. A large proportion of CC likely represents end stage NASH. Obesity, diabetes, iron deposition, advanced fibrosis, and age significantly increase the risk of NASH patients progressing to cirrhosis and subsequent HCC. HCC secondary to NASH typically develops in the setting of cirrhosis, although rare cases of HCC arising in NASH without cirrhosis raises the possibility that carcinogenesis secondary to NAFLD can occur in the absence of advanced liver disease. Obesity, diabetes, and hepatic steatosis are also risk factors for the development of HCC in other liver diseases such as chronic HCV. Efforts to maximize the management of these conditions should be should be considered in patients with any form of chronic liver disease. Because of the long, indolent clinical course of NASH, patients with significant disease may be overlooked, and often present with advanced age, multiple comorbidities, and larger tumor size, portending a poor prognosis. The connection between NAFLD and progression to HCC is becoming clearer, and the increasing burden of NASH, DM, and obesity is becoming heavier. Community awareness of the potential for this disease to progress to HCC is critical. Complications of NASH are expected to increase with the continuing epidemics of obesity and diabetes. Once the diagnosis of cirrhosis is made, screening for HCC should be pursued. Given recent epidemiologic data in diabetes, thought should be given to the use of statins in NASH patients, particularly those with diabetes and hyperlipidemia. Further research is urgently needed to better elucidate the signaling pathways for HCC development in the setting of insulin resistance. Studies evaluating potential targets for treatment of NASH and HCC, including targeting JNK activation, should be actively pursued.