Distinctive features of hepatocellular carcinoma in non-alcoholic fatty liver disease

Transcription

1 Distinctive features of hepatocellular carcinoma in non-alcoholic fatty liver disease Elisabetta Degasperi, Massimo Colombo Lancet Gastroenterol Hepatol 2016; 1: AM and A Migliavacca Center for Liver Disease, Division of Gastroenterology and Hepatology, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, Università degli Studi di Milano, Milan, Italy (E Degasperi MD, Prof M Colombo MD) Correspondence to: Dr Elisabetta Degasperi, Division of Gastroenterology and Hepatology, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, Università degli Studi di Milano, Milan, Italy elisabetta.degasperi@unimi.it Hepatocellular carcinoma is the fifth most common cancer and the second leading cause of cancer-related death worldwide. Non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis have been identified as emergent risk factors for this primary liver cancer. Incidence of NAFLD is increasing as a consequence of the epidemic spread of metabolic syndrome, which can result in progressive liver disease, leading to cirrhosis and its complications. Most data about the prevalence and incidence of hepatocellular carcinoma in patients with NAFLD are from a few population and cohort studies; its incidence is increasing and it is likely to become a leading indication for liver transplantation, especially in industrialised countries. In patients with NAFLD, hepatocellular carcinoma can arise in the context of non-cirrhotic liver, suggesting a specific carcinogenic pathway. Pathology studies have also described steatohepatitic hepatocellular carcinoma as a specific histological variant. NAFLD is underdiagnosed as causative liver disease, and patients are often diagnosed with hepatocellular carcinoma in the advanced stage because of the absence of efficient surveillance policies in this patient population. Management of hepatocellular carcinoma in patients with NAFLD is also complicated by comorbidities, mainly cardiac disease and diabetes, which negatively affect eligibility for radical treatments, including hepatic resection and, especially, liver transplantation. Finally, the effect of hepatocellular carcinoma treatments on postoperative morbidity, mortality, and disease-free survival remains to be precisely defined. Introduction Hepatocellular carcinoma is currently the fifth most common cancer and the second leading cause of cancer-related death worldwide. 1 Chronic infection with hepatitis C virus and hepatitis B virus and alcohol-related disease are the most important risk factors for this primary liver cancer; however, concern is growing about the pathogenic role of non-alcoholic fatty liver disease (NAFLD). 2,3 NAFLD has a vast disease spectrum, from simple steatosis to the combined association of steatosis with hepatocellular inflammation and fibrosis, known as non-alcoholic steatohepatitis (NASH), which has a greater propensity to progress to cirrhosis and eventually to hepatocellular carcinoma. 4 Because of the increasing burden of metabolic syndrome and its hallmarks type 2 diabetes, dyslipidaemia, and arterial hypertension in the setting of overweight and obesity, NAFLD resulting from liver involvement in the context of metabolic syndrome is becoming a new epidemic disease, especially in North American and western European countries. 5,6 A growing incidence has also been reported in east Asia and developing countries, particularly in those with high income. 7 Although increased short-term mortality in metabolic syndrome is mainly due to cardiovascular diseases, the liver disease burden is largely underestimated. This underestimation is mainly caused by absence of reliable epidemiological data about the incidence or prevalence of NAFLD and the prospective long-term outcomes of liver-related complications and hepatocellular carcinoma. The scarcity of data reflects the vast underdiagnosis of NAFLD, and consequently, reduced patient access to specialised tertiary centres for targeted surveillance and effective therapy for hepatocellular carcinoma. 8 Indeed, early diagnosis is important in determining treatment outcome in patients with hepatocellular carcinoma. 9 Since NAFLD is also increasing in younger generations who are more prone to development of long-term complications, hepatocellular carcinoma in patients with NAFLD is likely to become the first indication for liver transplantation in future decades. Thus, increased concern and awareness is needed to improve patient management and care. 10 Epidemiology The incidence and prevalence of hepatocellular carcinoma in patients with NAFLD are not well defined. Most data are from a few population-based studies and large cohort studies of hepatocellular carcinoma management and treatment, and from large liver transplant databases. High-income countries A population study by Younossi and colleagues 11 using the Surveillance, Epidemiology and End Results (SEER)-Medicare linked database between 2004 and 2009, assessed incidence, prevalence, and hepatocellular carcinoma risk factors in 4979 patients with hepatocellular carcinoma with matched controls. NAFLD was the underlying liver disease in 701 patients with hepatocellular carcinoma, representing 14 1% of the total hepatocellular carcinoma cases, and was the third most common liver disease after hepatitis C and alcoholrelated disease. During the 6-year study period, an overall 11% annual increase in the number of cases of hepatocellular carcinoma was reported, with an increase of 9% in patients with NAFLD compared with a 13% increase in patients with hepatitis C. This trend can probably be explained by increasing awareness of the need for hepatocellular carcinoma screening in clinical practice. The main bias in population studies assessing Vol 1 October 2016

2 NAFLD-related complications is the difficulty of a differential diagnosis between the lower-risk NAFLD and the higher-risk NASH. This differential diagnosis should ideally be obtained by histological assessment, which is often not done in clinical practice. To reduce the risk of underestimating disease burden, many studies expand the NAFLD definition to include patients with cryptogenic liver disease and patients who are overweight or obese, or who have type 2 diabetes. Using this clinical-based definition of NAFLD, the global prevalence of NAFLD varies from 3% to 35% according to different studies, whereas NASH is limited to about 3 5% of patients in the general population, with large differences according to geographical areas. 12 Another important bias of population studies is the risk of underestimating both the incidence and prevalence of hepatocellular carcinoma because diagnosis is often not reported in cancer registries used as source databases for these studies. Together with population-based studies, cohort studies are a source of data about the incidence of hepatocellular carcinoma and long-term outcomes. The BRIDGE study, 13 a large multicentre observational study of hepatocellular carcinoma management in a real-world setting ( ), assessed patients treated for hepatocellular carcinoma in 14 countries. NAFLD was reported as one of the main risk factors for hepatocellular carcinoma in Europe and North America, after hepatitis C, hepatitis B, and alcoholic liver disease, and accounting for about 10 12% of underlying liver diseases in patients with hepatocellular carcinoma. According to different cohort studies, 1-year cumulative incidence of hepatocellular carcinoma in patients with NAFLD has been estimated at around 2 5% compared with 4% in patients with hepatitis C. The 5-year incidence has been reported as 11% in patients with NAFLD compared with 30% in patients with hepatitis C, suggesting a lower risk of hepatocellular carcinoma in patients with NAFLD. 14 A prospective study assessing 195 liver transplantation candidates for NAFLD-related cirrhosis in Cleveland (OH, USA) during , reported development of hepatocellular carcinoma in 25 (12 8%) patients over a median follow-up of 3 2 years (IQR years), with a yearly cumulative incidence of 2 6%. 15 In a cohort of 315 patients with hepatitis C, the reported hepatocellular carcinoma incidence was 64 (20%) at the end of follow-up (3 2 years), with a yearly incidence of 4%. 15 In multivariate analysis, any grade of alcohol consumption was independently associated with risk of hepatocellular carcinoma development (hazard ratio 3 8 [95% CI ], p=0 002) in the NAFLD cohort. When the cumulative incidence of hepatocellular carcinoma in the entire study population was assessed, social alcohol intake was associated with an increased risk of hepatocellular carcinoma both in the NAFLD cohort (p=0 001) and in the hepatitis C cohort (p=0 002). 15 Because of the specific subset of patients referred for liver transplant evaluation, the study could not provide data about compensated liver disease and could potentially overestimate or underestimate the risk of hepatocellular carcinoma. Despite these methodological uncertainties, published data concur that hepatocellular carcinoma in patients with NAFLD is an emerging problem and is rapidly growing. A retrospective study conducted in Newcastle upon Tyne (UK) between 2000 and 2010, reported that NAFLD was the cause of liver disease in 136 (21 5%) of 632 patients with hepatocellular carcinoma. 16 When the investigators prospectively considered the referrals for hepatocellular carcinoma during the study period, NAFLD-related hepatocellular carcinoma had the greatest increase, accounting for 41 (35%) of 118 cases at the end of the study period. Moreover, hepatocellular carcinoma in patients with NAFLD occurred in the absence of histological or radiological hallmarks of cirrhosis in 31 (23%) patients. The cause of liver disease remained undefined in 195 (31%) patients with hepatocellular carcinoma in the Newcastle cohort. However, these patients had an increased prevalence of metabolic risk factors (type 2 diabetes, obesity, hypertension, dyslipidaemia), suggesting that the link between metabolic disorders and hepatocellular carcinoma could be even stronger than that found when the classic NAFLD diagnosis criteria are used. 16 The disease is also growing as an indication for liver transplantation in most industrialised countries: according to the United Network for Organ Sharing (UNOS) registry, it is the most rapidly increasing indication for liver transplantation due to hepatocellular carcinoma in the USA. 10 In patients receiving liver transplantation for hepatocellular carcinoma in the USA from 2002 to 2012, NAFLD-related hepatocellular carcinoma was the indication that showed the greatest increase (364%), becoming the second leading cause of liver transplantation after hepatitis C. NAFLD-related hepatocellular carcinoma is predicted to further increase in future decades, whereas widespread use of direct-acting antivirals could achieve an efficient cure for hepatitis C and lead to a decline in complications associated with the virus, including hepatocellular carcinoma. 10 Growing burden in eastern countries Data from the BRIDGE study 13 report a negligible NAFLD prevalence in east Asian countries: about 2 6% in China, Taiwan, and Japan. However, other studies suggest that NAFLD is becoming a major cause of liver disease in Asia. 7 Indeed, because of the lifestyle changes associated with a rapidly growing economy, the prevalence of NAFLD in Asian countries has increased from 13% before 1990 to 30% by 2008, with NAFLD accounting for about 50% of liver disease in obese adolescents. 17,18 By contrast, Asian countries also seem to report the highest prevalence of a specific NAFLD phenotypic variant the lean NAFLD, defined as disease in patients with a body-mass index (BMI) less than 25 kg/m² Vol 1 October

3 A prospective study by Arase and colleagues 20 in 1600 Japanese patients with NAFLD reported an 8-year hepatocellular carcinoma incidence of 0 63%. In another Japanese study in the higher-risk subset of patients with NASH and cirrhosis, the 5-year incidence of hepatocellular carcinoma in creased to 11 3%. The 5-year incidence was 30 5% in patients with hepatitis C-related cirrhosis, suggesting that hepatocellular carcinoma risk in patients with NAFLD is lower than in patients with viral hepatitis. 21 Studies of the epidemiology of hepatocellular carcinoma in patients with NAFLD are summarised in table 1. Risk factors In addition to the classic risk factors associated with hepatocellular carcinoma, such as age, being male, advanced fibrosis, and smoking cigarettes, individual features of metabolic syndrome, mainly diabetes and obesity, have been identified as independent risk factors for hepatocellular carcinoma development in patients with NAFLD Indeed, in cohort and case-controlled studies, patients with diabetes had a two to four times increased risk of developing hepatocellular carcinoma When evidence for this association was systematically reviewed, the results of diabetes increasing the risk of hepatocellular carcinoma were consistent between various patient populations and control groups; some studies also adjusted for potential confounding factors, such as alcohol intake and viral hepatitis. 28 A review of 11 cohort studies reported a 17% increased risk for hepatocellular carcinoma in patients who were overweight, and an 89% increased risk in patients who were obese. A study of 659 patients undergoing liver transplantation for hepatocellular carcinoma found that obesity was an independent predictor of hepatocellular carcinoma in patients with cryptogenic cirrhosis (odds ratio [OR] 11 1; 95% CI ; p=0 02); however, patients might be affected by NAFLD but are not correctly diagnosed. 29,30 However, the relative contribution of obesity independent from diabetes is difficult to ascertain since these two conditions are strongly associated, even from a pathogenic perspective. With this unknown contribution as the main rationale, a multicentre case-control study conducted in Italy tried to separately explore the association between the risk of hepatocellular carcinoma and the single features of metabolic syndrome. When 185 cases of hepatocellular carcinoma were compared with 404 controls, the OR for diabetes was 4 33 (95% CI ), and no association was observed with hypertension and hypercholesterolaemia. The OR for overweight (BMI 25 kg/m²) was 1 25 (95% CI ), and for obesity (BMI 30 kg/m²) was 1 97 (95% CI ). When combining multiple hallmarks Group of interest Comparison group (if applicable) Study period NAFLD prevalence NAFLD-related HCC incidence Population-based studies Younossi et al; 11 Surveillance, Epidemiology and End Results-Medicare USA Database Cohort studies Mittal et al; 8 USA Veterans Wong et al; 10 United Network for Organ Sharing Registry Park et al; 13 multicentre; 14 countries in Europe, Asia, and North America 4979 with HCC; median age 68 years; 71% male; cirrhosis NA 1500 with HCC; median age 64 years; 99 8% male; cirrhosis in 58% underwent transplantation, of whom 807 had NAFLD-related HCC; median age 59 years; 71% male; body-mass index=33 kg/m² with HCC; median age 61 years; 76% male; 81% Child-Pugh class A disease Oda et al; 14 review of cohort studies 247 with NAFLD ( ); 195 with NAFLD ( ); 68 with NAFLD ( ) Ascha et al; 15 USA (Cleveland) Dyson et al; 16 UK (Newcastle upon Tyne); retrospective Arase et al; 20 Japan Yatsuji et al; 21 Japan 195 with NASH with cirrhosis; median age 57 years; 44% male 632 with HCC, of whom 136 had NAFLD-related HCC; median age 71 years; 78% male; cirrhosis in 77% 1600 with NAFLD; >60 years old; 75% male; cirrhosis in 10% 68 with NAFLD with cirrhosis; median age 62 years; 43% male controls; median age NA; 44% male; cirrhosis NA % in patients with HCC; 8% in controls 9% yearly increase vs 13% in HCV % 2% per 5 years % per 10 years; 364% increase in transplantation % in Europe, 2 6% in east Asia ; ; % per 7 1 years ( ); 2 6% per year ); 11 3% per 5 years ( ) 315 with HCV with cirrhosis; median age 48 years; 76% male % per year vs 4% in HCV % 10-times increase to 35% in with HCV; >60 years old; 75% were males; cirrhosis incidence NA 69 with HCV with cirrhosis; median age 61 years; 43% male % per 8 2 years % NAFLD vs 30 5% HCV at 5 years NAFLD=non-alcoholic fatty liver disease. HCC=hepatocellular carcinoma. NA=not available. HCV=hepatitis C virus. NASH=non-alcoholic steatohepatitis. Table 1: Epidemiology of hepatocellular carcinoma in patients with non-alcoholic fatty liver disease Vol 1 October 2016

4 of metabolic syndrome, the risk of hepatocellular carcinoma increased in parallel with the number of features considered, reaching an OR of 4 75 (95% CI ) in patients who were overweight and had diabetes. 31 Although international guidelines state that the diagnosis of NAFLD requires exclusion of significant alcohol intake, a precise definition of significant alcohol quantity differs widely across studies, so that the potential role of moderate alcohol intake is difficult to ascertain. 32 Furthermore, some studies have even suggested that moderate alcohol intake (<30 g/day) is beneficial for reducing cardiovascular risk and NAFLD development in metabolic syndrome, but these data have not been confirmed in other studies. Taken together, evidence is still mixed and could be hindered by multiple biases resulting from poor quality of alcohol assessment, being overweight or obese as concomitant factors, and the observational design of the studies, which means that they cannot ascertain causality. Indeed, a large cohort study done in Taiwan reported a synergistic association of alcohol and obesity in determining the hepatocellular carcinoma risk. 36 Development of hepatocellular carcinoma in the context of NAFLD is the result of a complex interplay between genetic and environmental risk factors. Although the precise molecular pathways are still not fully understood, gene expression studies and the availability of robust animal models has provided deeper insight into some molecular mechanisms in NAFLD hepatocarcinogenesis. These carcinogenic pathways in NAFLD hepatocellular carcinoma are discussed in the appendix. Clinical presentation Published data depict patients with NAFLD and hepatocellular carcinoma as mainly white, male, and significantly older than patients with hepatocellular carcinoma of other origins. Moreover, as expected, patients with NAFLD have increased cardiovascular comorbidities resulting from coexisting metabolic syndrome and diabetes complications compared with patients with hepatocellular carcinoma without NAFLD. These data come from population-based studies and from large cohort studies. In the SEER Registry, the mean age of patients with NAFLD and hepatitis C was 73 years (SD 8 11) versus 66 years (SD 10 65); white patients accounted for 532 (76%) of 701 cases in NAFLD versus 1560 (57%) of 2536 in patients with hepatitis C. 11 A cohort study conducted between 2005 and 2010, in 1500 American Veterans with hepatocellular carcinoma replicated these findings. The study reported a 95% prevalence of arterial hypertension and 89% prevalence of diabetes in the NAFLD group, compared with 70% prevalence of arterial hypertension and 33% prevalence of diabetes in patients with hepatitis C. 8 Patients with NAFLD also had significantly higher rates of cardiovascular diseases, such as congestive heart failure and myocardial infarction, and peripheral vascular disease. In that study, 8 68 (57%) of 120 patients with NAFLD diagnosed with hepatocellular carcinoma had not been included in any surveillance programme, compared with 13% in patients with hepatitis C, resulting in higher rates of intermediate and advanced hepatocellular carcinoma according to the Barcelona Clinic Liver Cancer staging system in patients with NAFLD. Indeed, working against the application of potentially curative treatments for hepatocellular carcinoma was the finding that only 6% of patients with NAFLD had an early detected hepatocellular carcinoma compared with 16% in the hepatitis C population. Further worsening the prognosis in patients with NAFLD is the finding of the burden of comorbidities in patients with NAFLD that resulted in limited access to radical treatments such as hepatic resection or transplantation. Pathological features In several patients with NAFLD, a distinctive histological variant called steatohepatitic hepatocellular carcinoma has been described: these tumours are characterised by histological hallmarks resembling steatohepatitis, such as steatosis, hepatocyte ballooning, Mallory bodies, and perihepatocellular fibrosis. Steatohepatitic hepatocellular carcinoma was first described by pathologists at Columbia University (New York, NY, USA) in a series of patients with hepatitis C undergoing liver transplantation for hepatocellular carcinoma. 37 This histological variant was confirmed in the pure NAFLD setting by the same group 38 and in Japanese patients by another group, 39 establishing the strong link between steatohepatitic hepatocellular carcinoma and NAFLD in a population with an overall low prevalence of metabolic risk factors. In other studies, hepatocellular carcinoma in patients with NAFLD has been associated with increased iron content, although whether this finding is a consequence of advanced liver disease rather than a causal factor for hepatocellular carcinoma development remains unclear. 40 Of note, one of the main confounding factors in the histological diagnosis of NAFLD as the cause of hepatocellular carcinoma is the finding that characteristic features of NAFLD, such as steatosis, necroinflammation, and ballooning, diminish in advanced disease. 41 One noteworthy aspect of hepatocellular carcinoma occurrence in patients with NAFLD is the prevalence of tumours developing in non-cirrhotic livers: these findings, emerging from cohort studies (23% of patients with NAFLD without cirrhosis in the Newcastle study 16 and only 58% of patients with cirrhosis in the USA Veteran population 8 ), have also been confirmed by several pathological studies. A single-centre retrospective study assessing 128 patients undergoing liver resection for hepatocellular carcinoma between 1995 and 2007, showed that hepatocellular carcinoma in patients with NAFLD occurred more frequently in those without significant fibrosis. 65 5% of patients had liver staging See Online for appendix Vol 1 October

5 F0 F2 according to METAVIR score versus 35 5% with disease stage F3 and F4. By contrast, 75% of hepatocellular carcinoma developed in cirrhotic livers in patients with known liver disease of other origin. 42 Notably, a high prevalence of hepatocellular carcinoma was also found in patients with cryptogenic cirrhosis, which in other studies were shown to also include many undiagnosed patients with NAFLD. Hepatocellular carcinoma in patients with non-cirrhotic NAFLD was usually larger and better differentiated than in patients with histological evidence of cirrhosis, often as a consequence of being identified outside of screening programmes. A limitation of this study was that patient selection was skewed towards surgical indication, which could have led to underestimation of the prevalence of hepatocellular carcinoma in patients with advanced liver disease who were not candidates for radical treatments. However, the study supports the existence of a different carcinogenesis pathway in patients with NAFLD with respect to the classic fibrosis cirrhosis hepatocellular carcinoma sequence, thereby suggesting the need for updated, targeted surveillance algorithms in these patients. Access to screening and hepatocellular carcinoma surveillance International societies recommend hepatocellular carcinoma surveillance in selected target populations, including patients with cirrhosis of any cause and with chronic hepatitis B (because of the oncogenic risk related to hepatitis B virus infection). 9,43,44 Surveillance of patients with NAFLD is only recommended by the American Association for the Study of Liver Disease, although cost-effectiveness of such a policy is uncertain owing to the absence of robust data on the incidence of hepatocellular carcinoma in NAFLD, particularly in non-cirrhotic patients. 9 Surveillance in this population is also a challenge because of the difficulty of identifying patients with NASH who are at a high risk of cancer without a liver biopsy and because screening with abdominal ultrasound in obese populations is inaccurate. 45 As a result, the proportion of patients with NAFLD undergoing regular or effective surveillance is negligible and usually this surveillance is restricted to tertiary care centres. In the Veteran Affairs population, only 40% of patients with NAFLD diagnosed with hepatocellular carcinoma underwent surveillance compared with more than 80% of patients with hepatitis C. 8 Additionally, the suboptimal performance of abdominal ultrasound, which is the current standard of care for surveillance in patients with metabolic syndrome and obesity, has often resulted in under-recognition of small hepatocellular carcinoma nodules. 45,46 Accurate staging of NAFLD is hindered by the restricted access to, and suboptimal diagnostic accuracy of, a non-invasive test, such as transient elastography, and there are no reliable serological tests for liver fibrosis. As a result, many patients with NAFLD are not diagnosed and are left untreated for their underlying liver disease. 47,48 Hepatocellular carcinoma diagnosis Early diagnosis through screening is the mainstay to improve treatment outcomes in patients with hepatocellular carcinoma. Indeed, in solitary tumours with a diameter of 20 mm or less, the likelihood of vascular invasion or intrahepatic spread is very low, making curative treatment highly effective. 49 Nodules with a tumour diameter smaller than 10 mm within a cirrhotic liver are frequently not malignant and are difficult to diagnose with contrast imaging or ultrasound-guided fine-needle biopsy. 9 Diagnosis is therefore initiated using dynamic imaging with CT or MRI when the tumour diameter exceeds 10 mm. Diagnosis of hepatocellular carcinoma in patients with cirrhosis is established by the demonstration of increased contrast uptake in the arterial phase followed by contrast washout in the venous phase and delayed phases of dynamic imaging. 9 The potential of imaging based on the injection of organ parenchymalspecific contrast and the use of diffusion-weighted imaging on MRI are under investigation as standard of care to optimise hepatocellular carcinoma diagnosis in cirrhosis; meanwhile liver biopsy helps to differentiate between hepatocellular carcinoma and other malignancies, including intrahepatic cholangiocarcinoma with diagnostic accuracy greater than 90%. 50 By contrast with Asian Pacific Association for the Study of the Liver recommendations, the American Association for the Study of Liver Diseases and European Association for the Study of the Liver have dismissed contrast-enhanced ultrasound as a standard of care for the radiological diagnosis of hepatocellular carcinoma in patients with cirrhosis of any cause, owing to the risk of misclassification with intrahepatic cholangiocarcinoma. 51,52 These societies also maintain differing positions about the use of serum α-fetoprotein as a diagnostic marker of hepatocellular carcinoma, since this marker can also be increased in patients with intrahepatic cholangiocarcinoma. 9,43,44 Finally, international societies recommend contrast-imaging techniques to diagnose hepatocellular carcinoma in patients with established cirrhosis of any cause; however, imaging diagnosis of hepatocellular carcinoma is not recommended for patients with chronic hepatitis, except for those with chronic hepatitis B. This recommendation implies that, in patients with NAFLD without evidence of cirrhosis, diagnosis of hepatocellular carcinoma can be established by histology only. Outcome of curative treatments for hepatocellular carcinoma In population-based studies using the SEER registry, NAFLD was independently associated with increased 11-year mortality (61% of patients with NAFLD vs 51% of patients with hepatitis C, p<0 0001). 11 By contrast, cohort studies provided opposing data. In a retrospective study investigating short-term and long-term outcomes of liver resection for hepatocellular carcinoma, in which 96 patients with metabolic syndrome were matched for Vol 1 October 2016

6 relevant demographic and clinical variables with 96 patients with hepatitis C, the groups showed similar short-term morbidity in terms of surgical, cardiovascular, and infectious complications, and mean time in hospital. Cirrhosis, major hepatectomy, and MELD score greater than 8 were independently associated with morbidity and liver failure; however, NAFLD and metabolic disorders did not affect short-term outcomes, with an overall 90-day mortality rate of 2 1%. At 5 years, patients with metabolic syndrome showed better overall survival (65 6%) compared with patients with hepatitis C (61 4%; p=0 031), although recurrence-free survival was not significantly different (37% vs 27 5%; p=0 07). The recurrence rate was lower in patients with metabolic syndrome and hepatocellular carcinoma (44 6%) than in patients with hepatitis C (65 2%; p=0 005), because metabolic syndrome and hepatocellular carcinoma emerged in the multivariate analysis as a protective factor against early recurrence. 53 These findings corroborate the retrospective observations of a cohort in Pittsburgh (PA, USA) assessing the outcome of curative treatments for hepatocellular carcinoma, such as percutaneous radiofrequency, liver resection, and liver transplantation, in patients with NAFLD, hepatitis C, or alcohol-related liver disease, between 2000 and The study population included 52 patients with NASH compared with 162 patients with hepatitis C or alcohol-related liver disease, or both. As expected, patients with NAFLD were older, with higher BMI, and greater prevalence of diabetes and metabolic syndrome at hepatocellular carcinoma diagnosis, whereas patients with hepatitis C or alcohol-related liver disease, or both, showed poorer hepatic synthetic function and were more likely to undergo liver transplantation instead of resection. Rates of liver resection were also higher in the NAFLD group (41%) than in the hepatitis C and alcohol-related liver disease group (13%; p=0 002) because of increased prevalence of non-cirrhotic disease. Short-term postoperative mortality and recurrence-free survival did not significantly differ between the two cohorts, and patients with NAFLD had longer overall survival than did patients with hepatitis C or alcohol-related liver disease (p=0 009). In a multivariate analysis, primary tumour stage and liver transplantation were independently associated with recurrence-free survival, and NAFLD was an independent predictor of overall survival. 54 Outcome data from studies of radical treatments for hepatocellular carcinoma in patients with NAFLD are summarised in table 2. Better outcomes in patients with NAFLD: true or false? Better outcomes in patients enrolled in cohort studies compared with other liver diseases can partly be accounted for by a selection bias inherent to clinical studies. Patients who are candidates for radical treatments are enrolled, and patients with various comorbidities, such as cardiovascular diseases (which are common in patients with NAFLD and negatively affect survival) are excluded. This is also true for liver transplantation, since advanced age, obesity, and cardiovascular diseases are the main exclusion criteria for liver transplant evaluation. Cohort studies are therefore not entirely representative of the hepatocellular carcinoma burden and outcomes in patients with NAFLD. Moreover, because of the scarcity of hepatocellular carcinoma surveillance, the disease is likely to be diagnosed in the advanced stages, thereby preventing many patients from access to radical treatments. These restrictions are only partly counterbalanced by the increased eligibility for resection of patients with non-cirrhotic NAFLD with hepatocellular carcinoma, since this procedure is often contraindicated by adverse predictors of outcome, such as ageing, obesity, and diabetes. These challenges have been recently Patients Features of hepatocellular carcinoma Treatment Overall survival Recurrence-free survival Mittal et al with HCC Barcelona Clinic Liver Cancer staging system A: 65 8% for patients with NAFLD vs 15 7% for patients with HCV Wong et al 10 Viganò et al 53 Reddy et al with NAFLD, with HCV, with ALD, 7111 with HCV or ALD, or both 96 with MS-HCC matched with 96 HCV-HCC; retrospective 52 with NAFLD vs 162 with HCV or ALD, or both; retrospective Resection ablation, liver transplant, TACE, chemotherapy 46 7% for NAFLD vs 50 5% for HCV at 1 year Not assessed Liver transplant 1-year survival for ALD vs NAFLD OR 0 74 (95% CI , p<0 001); 1-year probability of liver transplant for HCV vs NAFLD OR 1 54 (95% CI , p<0 001) No differences between patients with MS-HCC and HCV-HCC; number of nodules=1 (1 4); >1 nodule=10%; size >50 mm=50% No differences between NAFLD and HCV or ALD, or both; mean HCC number=1 (1 3); largest tumour diameter=3 cm ( ); T3 4 stages=11 18% Resection Resection, ablation, and liver transplant 65 6% for MS vs 61 4% for HCV at 5 years; p=0 031 NAFLD vs HCV or ALD, or both; OR=0 509 (95% CI ; p=0 018) % for MS vs 27 5% for HCV; p=0 07 NAFLD>HCV or ALD, or both; p=0 303 MS=metabolic syndrome. HCC=hepatocellular carcinoma. HCV=hepatitis C virus. NAFLD=non-alcoholic fatty liver disease. ALD=alcohol-related liver disease. OR=odds ratio. TACE=transarterial chemoembolisation. Table 2: Treatment outcomes in patients with non-alcoholic fatty liver disease-related hepatocellular carcinoma Vol 1 October

7 confirmed in a prospective multicentre observational study in Italy, 55 comparing clinical features and outcomes of 145 patients with NAFLD-related hepatocellular carcinoma with those of 611 patients with hepatitis C virus-related hepatocellular carcinoma between 2010 and Patients with NAFLD were more likely to miss regular surveillance (63% in hepatitis C vs 48% in NAFLD for patients not undergoing regular hepatocellular carcinoma surveillance; p<0 0001), resulting in more advanced hepatocellular carcinoma burden at diagnosis. However, because of preserved liver function, more patients with NAFLD (19%) were eligible for liver resection compared with patients with hepatitis C (11%), although the overall proportion of patients receiving curative treatments was similar in the two populations. After adjustment for potential confounding factors (surveillance, age, liver function, tumour burden, and rates of curative treatment), overall survival did not differ between the groups. 55 The actual burden of surgical complications in patients with NAFLD still needs to be precisely defined, because, in addition to surgical risk deriving from medical comorbidities, steatosis itself has been identified as an independent predictor of complications after hepatic surgery. In a Swiss cohort of 116 patients undergoing liver resection, incidence of surgical site infections increased in parallel with steatosis severity assessed through liver signal attenuation on contrast-enhanced CT scan. 56 Additionally, severity of steatosis is a known risk factor for primary non-function after liver transplantation: the proposed mechanisms are impaired microcirculation resulting from fat obstruction of sinusoidal spaces, mitochondrial dysfunction, and increased apoptosis. 57 Another problematic issue for patients with NAFLD is competition for liver graft allocation because they have lower MELD score at transplant waiting list registration compared with patients with other liver diseases. The lower score is a consequence of better liver function in patients with NAFLD, having hepatocellular carcinoma as a main indication for liver transplant, and results in reduced probability of liver transplant in the short term, and longer time on the waiting list. A retrospective analysis of UNOS registry data for liver transplant waiting list registrations in the USA between 2004 and 2013 reported that patients with NAFLD were the least likely to receive a liver graft at 90 days and 1 year after registration, resulting in higher mortality while on the waiting list compared with patients with hepatitis C and alcohol use disorders. 58 Taken together, these data suggest that more robust evidence is needed to clearly define prognosis and long-term survival of patients with NAFLD with hepatocellular carcinoma. Primary prevention: facts or myths? Published evidence about possible strategies to prevent hepatocellular carcinoma development in patients with NAFLD, including those with NASH, concentrates on lifestyle modifications to reduce progression of liver damage and use metabolically active drugs, such as metformin and statins. In type 2 diabetes and metabolic syndrome, aerobic physical exercise is known to reduce insulin resistance and improve diabetes compensation; such behavioural therapy can reduce the need for pharmacological treatment. In patients with NAFLD, studies show that regular exercise and weight loss can improve inflammatory activity, as assessed in liver biopsies, 59 but little information is available to show whether this strategy results in prevention of hepatocellular carcinoma. A large prospective study in Taiwan, assessing a cohort of patients with or without hepatitis C, showed physical exercise to be associated with a significant reduction of the risk of hepatocellular carcinoma during a median follow-up of 8 5 years (range ) according to intensity of physical activity. 60 In the PTEN-deficient mouse model, which is characterised by spontaneous development of both NAFLD and hepatocellular carcinoma, the incidence of hepatocellular carcinoma was significantly lower in exercised animals (71%) than in sedentary animals (100%; p<0 05). The biological plausibility of these findings relies on the fact that regular exercise has been shown to inhibit the mtor complex, which is involved in cell growth and proliferation. 61 mtor inhibition has also been suggested as one possible pharmacological pathway activated by metformin: in a meta-analysis, administration of this drug was associated with a 70% reduction in the risk of hepatocellular carcinoma in patients with type 2 diabetes. Other possible antitumour mechanisms of metformin could be weight loss and reduction of endogenous reactive oxygen species, together with reduction of hyperinsulinaemia, which is a known activator of cell proliferation pathways. 62 A potential antineoplastic effect has also been suggested for statins, which have been associated with a 37% reduction in the risk of hepatocellular carcinoma, independently from its lipid-lowering effect, in a meta-analysis of 4298 patients with hepatocellular carcinoma. Statins inhibit the 3-hydroxy-3-methylglutaryl coenzyme A reductase, and this enzyme activates multiple downstream proliferative pathways; statins also exhibit a proapoptotic effect via RAF/MAPK1/ERK and growth-inhibitory action by inhibition of the proteasome pathway. 63 Conclusion NAFLD-related hepatocellular carcinoma is increasing in many geographical areas, together with the epidemic spread of metabolic syndrome worldwide. However, robust epidemiological data about prevalence and incidence of this problem are still unavailable, mainly because of the scarcity of population-based studies and the inability to identify NASH, which is the real predictor of hepatocellular carcinoma in patients with NAFLD. Although patients with NAFLD-related hepatocellular carcinoma have typical clinical features of metabolic Vol 1 October 2016

8 Search strategy and selection criteria We identified references for this Review through a search of PubMed with the terms hepatocellular carcinoma, steatosis, non-alcoholic fatty liver disease, and steatohepatitis from Jan 1, 2000, to Dec 31, Only papers published in English were reviewed. References were selected on the basis of originality and relevance to the scope of this Review. syndrome, such as type 2 diabetes, cardiovascular comorbidities, older age, and obesity, no molecular features have been recognised. However, the steatohepatitic hepatocellular carcinoma variant has been identified as a pathological feature of NAFLD-related hepatocellular carcinoma. Because hepatocellular carcinoma surveillance is not standard care in NAFLD, many patients do not undergo screening, and hepatocellular carcinoma is thus often diagnosed in advanced stages, negatively affecting survival. Survival is also compromised by the association with metabolic syndrome, which restricts access to radical treatments including transplantation because of the increased surgical risk resulting from concurrent cardiovascular diseases. Although lifestyle modifications and some drugs, such as metformin and statins, have been suggested as potential strategies for primary prevention, the main goal is to improve secondary prevention by expanding access to hepatocellular carcinoma screening, a strategy that calls for better identification of at-risk patients. Risk stratification of patients with NAFLD should take into account simple demographic and clinical variables to make the process reproducible in non-expert centres. For example, a scoring system developed in Taiwan, which accounted for age, sex, medical history (including presence of diabetes), viral hepatitis status, aminotransferases, and α-fetoprotein, enabled identification of almost all hepatocellular carcinoma cases by ultrasound in high-risk patients. 60 However, reproducing this experience in other geographical settings requires risk stratification scores to be based on local disease epidemiology, whereas to have an impact on hepatocellular carcinoma burden, patient selection for targeted surveillance should be moved to the community to overcome the limitation of patient access by general practitioners or geographical barriers. Accordingly, raising awareness of the burden of NAFLD and implementation of screening pro grammes are the primary goals to be pursued by health authorities worldwide. Contributors ED analysed and interpreted the data, and drafted the manuscript. MC critically revised the manuscript. Declaration of interests MC received grant and research support from Bristol-Myers Squibb (BMS) and Gilead Science, and is on the advisory committees for Merck, Roche, Novartis, Bayer, BMS, Gilead Science, Tibotec, Vertex, Janssen Cilag, Achillion, Lundbeck, GlaxoSmithKline, GenSpera, AbbVie, AlfaWasserman, and Jennerex. MC is a speaker and teaches for Tibotec, Roche, Novartis, Bayer, BMS, Gilead Science, Vertex, Merck, Janssen, Sanofi, and AbbVie. ED declares no competing interests. References 1 Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA Cancer J Clin 2011; 61: El Serag HB. Epidemiology of viral hepatitis and hepatocellular carcinoma. Gastroenterology 2012; 142: Adams LA, Lindor KD. Nonalcoholic fatty liver disease. Ann Epidemiol 2007; 17: Younossi ZM, Koenig AB, Abdelatif D, et al. Global epidemiology of non-nonalcoholic fatty liver disease meta-analytic assessment of prevalence, incidence and outcomes. Hepatology 2016; 64: Ng M, Fleming T, Robinson M, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during : a systematic analysis for the Global Burden of Disease Study Lancet 2014; 384: Flegal KM, Carrol MD, Kit BK, et al. Prevalence of obesity and trends in the distribution of body mass index among US adults, JAMA 2012; 307: Wong VW. Nonalcoholic fatty liver disease in Asia: a story of growth. J Gastroenterol Hepatol 2013; 28: Mittal S, Sada Y, El-Serag HB, et al. Temporal trends of nonalcoholic fatty liver disease-related hepatocellular carcinoma in the Veteran Affairs Population. Clin Gastroenterol Hepatol 2015; 13: Bruix J, Sherman M. Management of hepatocellular carcinoma: an update. Hepatology 2011; 53: Wong RJ, Cheung R, Ahmed A. Nonalcoholic steatohepatitis is the most rapidly growing indication for liver transplantation in patients with hepatocellular carcinoma in the US. Hepatology 2014; 59: Younossi ZM, Otgonsuren M, Henry L, et al. Association of nonalcoholic fatty liver disease (NAFLD) with hepatocellular carcinoma (HCC) in the United States from 2004 to Hepatology 2015; 62: Vernon G, Baranova A, Younossi ZM. Systematic review: the epidemiology and natural history of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults. Aliment Pharmacol Ther 2011; 34: Park JW, Chen M, Colombo M, et al. Global patterns of hepatocellular carcinoma management from diagnosis to death: the BRIDGE study. Liver Int 2015; 35: Oda K, Uto H, Matawari S, et al. Clinical features of hepatocellular carcinoma associated with non-alcoholic fatty liver disease: a review of human studies. Clin J Gastroenterol 2015; 8: Ascha M, Hanouneh IA, Lopez R, et al. The incidence and risk factors of hepatocellular carcinoma in patients with alcoholic steatohepatitis. Hepatology 2010; 51: Dyson J, Jacques B, Chattopadyhay D, et al. Hepatocellular cancer: the impact of obesity, type 2 diabetes and a multidisciplinary team. J Hepatol 2014; 60: Farrell GC, Wong VW, Chitturi S. NAFLD in Asia as common and important as in the West. Nat Rev Gastroenterol Hepatol 2013; 10: Fan JG, Farrell GC. Epidemiology of non-alcoholic fatty liver disease in China. J Hepatol 2009; 50: Xu C, Yu C, Ma H, Xu L, Miao M, Li Y. Prevalence and risk factors for the development of nonalcoholic fatty liver disease in a nonobese Chinese population: the Zhejiang Zhenhai Study. Am J Gastroenterol 2013; 108: Arase Y, Kobayashi M, Suzuki F, et al. Difference in malignancies of chronic liver disease due to non-alcoholic fatty liver disease or hepatitis C in Japanese elderly patients. Hepatol Res 2012; 42: Yatsuji S, Hashimoto E, Tobari M, Taniai M, Tokushige K, Shiratori K. Clinical features and outcomes of cirrhosis due to non-alcoholic steatohepatitis compared with cirrhosis caused by chronic hepatitis C. J Gastroenterol Hepatol 2009; 24: Lee SS, Jeong SH, Byoun YS, et al. Clinical features and outcome of cryptogenic hepatocellular carcinoma compared to those of viral and alcoholic hepatocellular carcinoma. BMC Cancer 2013; 13: Vol 1 October

9 23 Lee YC, Cohet C, Yang YC, et al. Meta-analysis of epidemiologic studies on cigarette smoking and liver cancer. Int J Epidemiol 2009; 38: El-Serag HB, Tran T, Everhart JE. Diabetes increases the risk of chronic liver disease and hepatocellular carcinoma. Gastroenterology 2004; 126: Davila JA, Morgan RO, Shaib Y, McGlynn KA, El-Serag HB. Diabetes increases the risk of hepatocellular carcinoma in the United States: a population based case control study. Gut 2005; 54: Harrison SA. Liver disease in patients with diabetes mellitus. J Clin Gastroenterol 2006; 40: Bosetti C, Rosato V, Polesel J, et al. Diabetes mellitus and cancer risk in a network of case-control studies. Nutr Cancer 2012; 64: El-Serag HB, Hampel H, Javadi F. The association between diabetes and hepatocellular carcinoma: a systematic review of epidemiologic evidence. Clin Gastroenterol Hepatol 2006; 4: Larsson SC, Wolk A. Overweight, obesity and risk of liver cancer: a meta-analysis of cohort studies. Br J Cancer 2007; 97: Nair S, Mason A, Eason J, Loss G, Perrillo RP. Is obesity an independent risk factor for hepatocellular carcinoma in cirrhosis? Hepatology 2002; 36: Turati F, Talamini R, Pelucchi C, et al. Metabolic syndrome and hepatocellular carcinoma risk. Br J Cancer 2013; 108: Chalasani N, Younossi Z, Lavine J, et al. The diagnosis and management of non-nonalcoholic fatty liver disease: practice guideline by the American Association for the Study of Liver Diseases, American College of Gastroenterology, and the American Gastroenterological Association. Hepatology 2012; 55: Sookoian S, Castaño GO, Pirola CJ. Modest alcohol consumption decreases the risk of non-alcoholic fatty liver disease: a meta-analysis of individuals. Gut 2014; 63: Hashimoto Y, Hamaguchi M, Kojima T, et al. Modest alcohol consumption reduces the incidence of fatty liver in men: a population-based large-scale cohort study. J Gastroenterol Hepatol 2015; 30: Lau K, Baumeister SE, Lieb W, et al. The combined effects of alcohol consumption and body mass index on hepatic steatosis in a general population sample of European men and women. Aliment Pharmacol Ther 2015; 41: Loomba R, Yang HI, Su J, Brenner D, et al. Synergism between obesity and alcohol in increasing the risk of hepatocellular carcinoma: a prospective cohort study. Am J Epidemiol 2013; 177: Salomao MS, Yu WM, Brown RS, et al. Steatohepatitic hepatocellular carcinoma (SH-HCC): a distinctive histological variant of HCC in hepatitis C virus-related cirrhosis with associated NAFLD/NASH. Am J Surg Pathol 2010; 34: Salomao M, Remotti H, Vaughan R, et al. The steatohepatitic variant of hepatocellular carcinoma and its association with underlying steatohepatitis. Hum Pathol 2012; 43: Shibahara J, Ando S, Sakamoto Y, et al. Hepatocellular carcinoma with steatohepatitic features: a clinicopathological study of Japanese patients. Histopathology 2014; 64: Sorrentino P, D Angelo S, Ferbo U, et al. Liver iron excess in patients with hepatocellular carinoma developed on non-alcoholic steato-hepatitis. J Hepatol 2009; 50: Caldwell SH, Lee VD, Kleiner DE, et al. NASH and cryptogenic cirrhosis: a histological analysis. Ann Hepatol 2009; 8: Paradis V, Zalinski S, Chelbi E, et al. Hepatocellular carcinomas in patients with metabolic syndrome often develop without significant liver fibrosis: a pathological analysis. Hepatology 2009; 49: Omata M, Lesmana LA, Tateishi R, et al. Asian Pacific Association for the Study of the Liver consensus recommendations on hepatocellular carcinoma. Hepatol Int 2010; 18: European Association for the Study of the Liver, European Organisation for Research and Treatment of Cancer. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol 2012; 56: Singal AG, Yopp AC, Gupta S, et al. Failure rates in the hepatocellular carcinoma surveillance process. Cancer Prev Res 2012; 5: Della Corte C, Colombo M. Surveillance for hepatocellular carcinoma. Semin Oncol 2012; 39: Musso G, Gambino R, Cassader M, Pagano G. Meta-analysis: natural history of non-alcoholic fatty liver disease (NAFLD) and diagnostic accuracy of non-invasive tests for liver disease severity. Ann Med 2011; 43: Kwok R, Tse YK, Wong GL, 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: Roskams T, Kojiro M. Pathology of early hepatocellular carcinoma: conventional and molecular diagnosis. Semin Liver Dis 2010; 30: Sangiovanni A, Manini MA, Lavarone M, et al. The diagnostic and economic impact of contrast imaging techniques in the diagnosis of small hepatocellular carcinoma in cirrhosis. Gut 2010; 59: Vilana R, Forner A, Bianchi L, et al. Intrahepatic peripheral cholangiocarcinoma in cirrhosis patients may display a vascular pattern similar to hepatocellular carcinoma on contrast-enhanced ultrasound. Hepatology 2010; 51: Iavarone M, Piscaglia F, Vavassori S, et al. Contrast enhanced CT-scan to diagnose intrahepatic cholangiocarcinoma in patients with cirrhosis. J Hepatol 2013; 58: Viganò L, Conci C, Cescon M, et al. Liver resection for hepatocellular carcinoma in patients with metabolic syndrome: a multicenter matched analysis with HCV-related HCC. J Hepatol 2015; 63: Reddy SK, Steel JL, Chen HW, et al. Outcomes of curative treatment for hepatocellular cancer in nonalcoholic steatohepatitis versus hepatitis C and alcoholic liver disease. Hepatology 2012; 55: Piscaglia F, Svegliati-Baroni G, Barchetti A, et al. Clinical patterns of hepatocellular carcinoma in nonalcoholic fatty liver disease: a multicenter prospective study. Hepatology 2016; 63: Kurmann A, Wanner B, Martens F, et al. Hepatic steatosis is associated with surgical-site infection after hepatic and colorectal surgery. Surgery 2014; 156: Selzner M, Clavien PA. Fatty liver in transplantation and surgery. Semin Liver Dis 2001; 21: Wong R, Aguilar M, Cheung R, et al. Nonalcoholic steatohepatitis is the second leading etiology of liver disease among adults awaiting liver transplantation in the United States. Gastroenterology 2015; 148: Torres DM, Williams CD, Harrison SA. Features, diagnosis and treatment of nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol 2012; 10: Wen CP, Lin J, Yang CY, et al. Hepatocellular carcinoma risk prediction model for the general population: the predictive power of transaminases. J Natl Cancer Inst 2012; 104: Piguet AC, Saran U, Simillion C, et al. Regular exercise decreases liver tumors development in hepatocyte-specific PTEN-deficient mice independently of steatosis. J Hepatol 2015; 62: Zhang ZJ, Zheng ZJ, Shi R, et al. Metformin for liver cancer prevention in patients with type 2 diabetes: a systematic review and meta-analysis. J Clin Endocrinol Metab 2012; 97: Singh S, Singh PP, Singh AG, et al. Statins are associated with a reduced risk of hepatocellular cancer: a systematic review and meta-analysis. Gastroenterology 2013; 144: Vol 1 October 2016