Hepatocellular carcinoma in non-alcoholic fatty liver disease: An emerging threat
Abstract
Introduction
Hepatocellular carcinoma (HCC) is the fifth most common cancer and the second leading cause of cancer-related mortality worldwide1. HCC develops mainly in chronic liver disease and cirrhosis, among which the risk factors includes hepatitis B virus (HBV), hepatitis C virus (HCV) and alcoholic liver disease. However, there is growing evidence suggesting non-alcoholic fatty liver disease (NAFLD) as one of the leading cause of HCC2.
NAFLD is a spectrum of liver pathology ranging from simple steatosis to the aggressive form, non-alcoholic steatohepatitis (NASH) as evidence by histology or imaging, in the absence of secondary causes for hepatic fat accumulation such as excessive alcohol consumption, chronic viral hepatitis, drug-induced liver injury and other liver diseases. Despite the common occurrence of NAFLD globally, only a subgroup of these patients progressed into NASH. NASH is defined as presence of hepatic steatosis with evidence of hepatocytes damage3. Studies have shown that NASH has greater tendency to develop into advanced liver fibrosis and cirrhosis, thus eventually progressing to HCC4-6.
A systematic review of the association between NAFLD and risk for HCC has reported a cumulative HCC incidence rate of 2.4% to 12.8% in patients with NASH-related cirrhosis7. As NAFLD emerges as the commonest cause of chronic liver disease in both Western and Eastern population due to increasing incidence of metabolic syndrome, the prevalence of NAFLD-related HCC is expected to rise8-11.
Hence, the purpose of this paper is to review the current state of knowledge on NAFLD-related HCC with regards to its epidemiology, pathogenesis and clinical features.
Methods
We performed a PubMed search using MeSH terms “non-alcoholic fatty liver disease” or “fatty liver” and “hepatocellular carcinoma” on May 2017. The search yielded 758 articles. Of these, 491 are original articles. The abstracts of the articles were examined and where doubt existed as to the relevance of an article to the review, the full paper was examined. In total, 44 articles were deemed relevant and were included in the review.
Results
Epidemiology of NAFLD-related HCC
Several large population-based studies and cohort studies were carried out to identify the incidence and prevalence of NAFLD-related HCC. Prevalence of NAFLD-related HCC in the general population differs across region as illustrated by Fig. 1. Large studies in the Western countries have estimated the prevalence of NAFLD-HCC to be in the range of 4-14.1%, while in the Asia-Pacific region, the prevalence of NAFLD-HCC ranges from 5.4-11.2%.
A global HCC BRIDGE study recruiting patients in 14 countries from 2005 to 2012 confirmed previously reported regional trends in HCC risk factors, in which NASH was considerably more prevalent in North America and Europe (10-12%) as compared with Asia counterparts (1-6%)12. This can be accounted by the difference in socio-economic development and lifestyle factors. However, recent meta-analysis13 has shown that prevalence of NAFLD is now comparable between the West and East and NAFLD-HCC is expected to become increasingly more common in Asia-Pacific region.
Figure 1: Prevalence of NAFLD-HCC across different regions
The incidence of NAFLD-related HCC is on the rise globally. A large, population-based study utilizing Surveillance, Epidemiology and End Results (SEER) registries in United States14 identified that there has been a 9% annual increase among NAFLD-HCC cases from 2004 to 2009 (P < 0.0001). Another retrospective cohort study using data from the United Network for Organ Sharing (UNOS) registry15 noted a fourfold increase in prevalence of NASH-related HCC from 2002 to 2012.
Similarly, a large multicentre retrospective study in Japan16 reported that the proportion of patients with non-viral etiologies has increased from 10.0 % in 1991 to 24.1 % in 2010. They attributed this finding to the increasingly obese population and rising proportion of patients with diabetes due to changes in the diet in Japan17. Studies in Korea have also highlighted the increasing trend in prevalence of NAFLD-related HCC across the decade18, 19.
In contrast, Beste et al.20 and Mittal et al.21 reported that although HCC incidence has increased by approximately 2.5 fold over the past decade, it is driven primarily by HCV, while incidence of NAFLD has increased only slightly. Their results are echoed by several studies who noted that annual proportions of NAFLD-related HCC are relatively stable22-24. However, the studies acknowledged that the impact of NAFLD epidemic on HCC has not been realized yet due to the relative recency of obesity epidemic and lag period of several decades required for NAFLD to develop into cirrhosis or HCC.
With the increase in NAFLD-HCC, NAFLD has also become one of the leading cause for liver transplantation25-27. A study using data from a large national liver transplant registry in the U.S.15 concluded that NASH-related HCC is the most rapidly growing indication for liver transplant. Another retrospective single-centre study from UK reported that the incidence of HCC in NASH recipients in UK was nearly as high as HCV recipients28. In contrast, Leary et al.29 reported that patients with NASH/CC were less likely to be transplanted than patients with HCV (48% vs. 62%; P<0.001) due to their multiple comorbids and slow progression of the disease.
Nevertheless, surveillance of NAFLD patients should be advocated as evidence shows that they are at high risk of developing HCC. Edenvik et al.30 reported that NAFLD were associated with deficient surveillance with only 13% of HCC discovered through surveillance resulting in delay in diagnosis. As NAFLD is expected to increase in the coming decades due to obesity epidemic11, 31, it is critical that clinicians have higher suspicion to diagnose possible liver disease in patients with obesity and diabetes.
Selection of patients with NAFLD for screening becomes difficult. Cryptogenic cirrhosis and NAFLD share many similar features. Study even shows CC as a progression of NASH (Histology shows burnt-out NASH). Modified NASH criteria (listed by Wong RJ). A revision of existing knowledge?
NASH
Non-alcoholic fatty liver disease (NAFLD) encompasses a histological spectrum from non-alcoholic fatty liver (NAFL), which is characterized by steatosis with no or minor inflammation, to non-alcoholic steatohepatitis (NASH) where inflammation and ballooning is present, with or without fibrosis32. The natural history of NASH has been implicated as more aggressive with prospective cohort studies demonstrating a higher rate of morbidity and mortality compared to NAFL, especially with the presence of fibrosis33, 34.
Ascha et al.35 noted that over a median follow-up of 3.2 years, 25/195 (12.8%) of NASH-cirrhotic developed HCC, with a yearly cumulative incidence of HCC around 2.6% in patients with NASH-cirrhosis. A large-scale retrospective cohort study at a public hospital in Japan reported that annual incidence of NAFLD-HCC was 0.043% and the cumulative rate of HCC among NASH is higher as compared to patients without significant fibrosis (HR: 25.03; 95% CI: 9.02–69.52; P<0.001).
Ethnicity
Ethnicity appears to play a role in NAFLD-related HCC as well. Couto et al.36 noted that Hispanics were more likely to develop NAFLD-related HCC compared to non-Hispanics. Earlier studies had reported that Hispanics are more predisposed to NASH37-39, while Altekruse et al. found an increased risk of HCC among Hispanics. US-born Hispanics had higher incidence of HCC when compared to foreign-born Hispanics, suggesting that environmental, socioeconomic and cultural differences are contributing factors as well.
Development of HCC in Non-Cirrhotic Liver
HCC can develop in NAFLD patients in the absence of fibrosis. Several studies have shown that as many as 40-47%40-43 of NAFLD-HCC patients had no clinical or histological evidence of cirrhosis. A multicenter prospective study in Italy comparing between NAFLD-HCC and HCV-HCC reported that cirrhosis was present in only about 50% of NAFLD-HCC, in contrast with the totality of HCV-HCC43.
Schutte et al.44 and Ertle et al.41 reported that majority of patients with HCC in non-cirrhotic liver had metabolic syndrome and NAFLD, suggesting that NASH per se could promote development of HCC, in the absence of cirrhotic liver. Moreover, Alexander et al.45 reported a significant association between steatosis with HCC in 157 HCC patients who had no histological features of cirrhosis. This further supports the hypothesis that steatohepatitis alone could have a causative role in hepatocarcinogenesis in non-cirrhotic liver46.
Curiously, a study comparing cirrhotic and non-cirrhotic NAFLD-HCC40 reported that NAFLD-HCC among non-cirrhotic liver are less likely to be obese, less likely to be associated with metabolic syndrome and type 2 diabetes. However, this was a relatively small retrospective study.
In terms of clinicopathological findings, most studies agreed that NAFLD-HCC patients without cirrhosis were more likely to present with larger tumor and had higher rates of tumor recurrence40, 44, 47, 48. This can be accounted by the lack of screening in non-cirrhotic NAFLD-HCC patients as per AASLD guidelines.
Clinical Characteristics of NAFLD-related HCC
Obesity
According to the most recent World Health Organisation estimate in 2014, more than 1.9 billion adults were overweight (BMI 25.0-29.9 kg/m2), and of these, over 600 million adults were obese (BMI ≥ 30 kg/m2). Generally, obesity is associated with an increased risk of malignancy, particularly as a risk factor of HCC49. Among the 32 observational studies shown in Table 2, 17 studies have reported a significant relationship between obesity and HCC15, 18, 30, 35, 40, 41, 44, 45, 49-57.
A retrospective cohort study of 10,061 patients with HCC in United States reported that the NASH-related HCC patients had significantly higher BMI (33.6 ± 4.3 vs 27.3 ± 4.9, P < 0.001) than non-NASH-related HCC patients15. Moreover, the frequency of obesity was 25% to 50% among cryptogenic cirrhosis-related HCC patients15.
Interestingly, a review of data from Alexander et al45 also demonstrated that patients with NASH-related HCC were more likely to be obese (OR= 3.81, 95% CI: 1.55-9.31) compared to non-NASH-related HCC cohort. Besides, in a recent study in 2007 by Than et al52, patients with NAFLD had higher BMI (mean 32.3 vs 26.5, P < 0.001) than those who had hepatitis C virus at the time of HCC diagnosis. On a univariate analysis of factors associated with the development of HCC in NASH-cirrhosis cohort found that obesity was statistically associated with HCC (HR=0.94, 95% CI: 0.89-0.99; P=0.025)35. These studies support the positive correlation between obesity and HCC.
Diabetes
Diabetes, being one of the key pathogenetic factor of NAFLD, was found to have a positive association with HCC according to multiple observational studies. A total of 22 studies out of 32 studies being reviewed have established a significant relationship between diabetes and HCC15, 18, 21, 23, 35, 36, 40, 41, 44, 45, 49, 50, 52-61. A prospective, multicentre study of 756 patients with HCC in Italy demonstrated that there are 73.1% of patients with diabetes in NAFLD-related HCC cohort compared to HCV-HCC cohort with only 24.9% of patients with diabetes (P < 0.0001)58. In a subsequent larger cohort study from the Veteran Affairs database in United States including 1,500 patients with HCC showed that patients with NAFLD-related HCC were more likely diabetic (89.2% vs 50.4% vs 32.9%, P <0.01) compared to alcohol abuse patients and HCV patients at the time of HCC diagnosis21.
Besides obesity, diabetes has also been identified as an independent risk factor for HCC. There is a 2-3 fold increase in relative risk of developing HCC risk with diabetes41. Hashimoto et al60 established that the difference in prevalence of diabetes between NASH patients with and without HCC was statistically significant with the odds ratio of 3.618 (95% CI: 1.632-8.016).
However, a review of study by Ascha et al35 on risk factors for developing HCC in NASH-cirrhosis cohort with univariate analysis reported that there was no statistical significance on diabetes (HR=1.00, 95% CI: 0.40-2.5; P=0.99). Moreover, according to a univariate Cox regression analysis for factors influencing survival of HCC patients with non-cirrhotic liver done by Schutte et al44, diabetes mellitus was not statistically significant in affecting the survival outcome (HR=1.246, 95% CI: 0.564-2.753) with P value of 0.586. Although there are substantial evidence indicating diabetes is strongly associated with the progression of HCC, larger multinational studies should be conducted to validate its impact in survival outcome of HCC patients.
Age and Gender
According to the data published in most of the studies, NAFLD-related HCC patients are mainly male and significantly older than patients with HCC of other aetiological factor such as alcoholic liver disease and viral hepatitis. For instance, a large retrospective study involving 7,670 patients with HCC by Beste et al20 in 2013 reported that the mean age of patients with NAFLD was 70.5 years (SD 10.2) while the patients with HBV, HCV and alcoholic liver disease have a mean age of 63.3 years (SD9.1), 61.7 years (SD 5.8) and 66.8 years (SD 8.1) respectively. A prospective cohort study in Japan demonstrated a statistically significant difference in the median age of NASH patients with HCC and NASH patients without HCC (70 years vs 50 years; OR=1.129, 95% CI: 1.082-10.179)60. In addition, older age at time of NASH-cirrhosis diagnosis was also established as one of the risk factor associated with HCC development (HR=1.07, 95% CI: 1.02-1.10; P=0.012)35.
A retrospective study involving 1,266 liver transplant recipients has found that more than half of the patients (67%) with liver disease due to various aetiologies are male36. Among the 1,266 liver transplant recipients, 292 patients were diagnosed with HCC either with or without NAFLD, and 218 out of 292 patients are male36. However, there is no statistical significance among the NAFLD-related HCC cohort and non-NAFLD-related HCC cohort36. Similar results were obtained from the study conducted by Wong et al15 which consists of 10,061 patients with HCC. The incidence of HCC is significantly high in males with 79.2% in non-NASH-related HCC cohort and 70.6% in NASH-related HCC cohort (P < 0.001)15. Although there is gender disparity in HCC of other aetiologies, male NAFLD patients are at high risk of concomitant HCC especially with other risk factors.
Dyslipidaemia and Hypertension
Dyslipidaemia is characterized by increased serum triglyceride and LDL cholesterol levels and decreased HDL cholesterol levels, all of which are strongly associated with NAFLD. However, unlike obesity and diabetes, there is no direct evidence which shows that dyslipidaemia and hypertension causes development of HCC in NAFLD patients.
Out of 32 studies shown in Table 2, only 9 studies reported a positive correlation between dyslipidaemia and HCC23, 36, 45, 55-58, 60, 61. For instance, a univariate analysis results for factors associated with cryptogenic cirrhosis in HCC patients has demonstrated an OR of 6.41 (95% CI: 2.3–17.8, P=0.0004) and 10.69 (95% CI: 2.3–48.9, P=0.0023) in hypercholesterolemia and hypertriglyceridemia respectively, however, there is no association observed with hypertension61.
Another study in Germany which involves 1,119 HCC patients established that NASH-related HCC cohort is more commonly associated with hyperlipidaemia than non-NASH-related HCC cohort (40% vs 19.6%, P=0.016)57. In contrast, hypertension is not statistically significant in the two cohorts57. Therefore, the causal link and mechanism between both dyslipidaemia and hypertension towards progression of HCC in NAFLD patients require further validation by larger observational studies.