Molecular epidemiology of hepatocellular carcinoma

Watch a video presentation of this article

Watch the interview with the author

Answer questions and earn CME

Abbreviations

AFP

alpha‐fetoprotein

GH

genetic hemochromatosis

GWAS

genome‐wide association study

HBV

hepatitis B virus

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

SNP

single‐nucleotide polymorphism

TGF

transforming growth factor

Hepatocellular carcinoma (HCC) is the most rapidly increasing cause of cancer mortality with poor prognosis (5‐year survival <12%) in the United States.1 Although it is known that advanced liver fibrosis/cirrhosis is the high‐risk condition that rationalizes regular HCC surveillance, the extremely high prevalence of cirrhosis (1%‐2% of the general population) as well as other factors such as immigration from developing world make it difficult to adhere to the surveillance protocol: only 17% of new HCC cases were diagnosed through regular surveillance.2, 3 On the other hand, whereas annual cancer incidence in cirrhotic patients is very high (1%‐7%), a certain proportion of patients with cirrhosis do not develop HCC during their lifetime.4 In addition, there is great diversity in the clinical course of HCC. However, our ability to predict these outcomes is still limited or lacking, especially in patients with early stage disease. A great expectation has been placed on molecular biomarkers to fill this unmet need and enable effective, personalized patient management with limited medical resources. Furthermore, such biomarkers will help HCC prevention strategies by enriching at‐risk patients for clinical trials as well as by predicting and monitoring treatment response.

HCC development is a multistep process that involves establishment of chronic liver injury, progressive liver fibrosis that results in cirrhosis, initiation of neoplastic clone, and stepwise malignant transformation and dissemination of the clone (Fig. 1). Molecular mechanisms involved in each of the steps have been studied extensively with a goal of identifying the key drivers and gatekeepers, yielding many candidates of biomarkers and/or therapeutic targets.5

Figure 1.

Multistep process of HCC development in chronic liver disease. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Molecular Risk Factors of HCC Development

To date, numerous genetic polymorphisms have been reported as host genetic factors that determine susceptibility to HCC development (Table 1).6 Large‐scale case‐control or cohort studies, as well as systematic reviews, have identified HCC‐associated single‐nucleotide polymorphisms in genes involved in immune response (TNF, IL10), oxidative stress (GSTM1, GSTT1), growth signaling (EGF), cell cycle (MDM2), DNA damage repair (XPC), and iron metabolism (HFE) in viral hepatitis– or alcohol‐related HCC.7 Recent genome‐wide association studies have identified the DEPDC5 gene as well as MICA and 1p36.22 regions as risk loci in viral hepatitis–related HCC.15 These associations are generally modest (odds or hazard ratios below 2), and therefore the combination of multiple loci representing independent mechanisms may yield a more powerful polygenic signature of HCC risk variants. A panel of seven variants in immune‐related genes has been tested for its association with fibrosis progression, but not yet for HCC development.18 One caveat is that the case‐control design employed in most of these studies often failed to control several key confounding factors. Gene expression signatures in stromal liver tissue, another class of biological information assumed to capture functional molecular deregulation, were shown to be predictive of disseminative or de novo HCC recurrence after surgical resection.19, 20 Etiological agent–related molecular factors could also influence HCC risk. A high serum hepatitis B virus (HBV) DNA level, which is indicative of increased viral replication, is associated with elevated risk of HCC.21 Some studies have suggested that HBV genotype could affect HCC risk.22, 23

Table 1.

Molecular Risk Factors of HCC Development

Molecular Feature	Type of Information	Biological Pathway	Etiology	Reference
TNF, G308A	SNP	Immune response	HBV, HCV	7
IL10, A592C	SNP	Immune response	HBV, HCV	8
GSTM1/GSTT1	Deletion	Oxidative stress	HBV, HCV	9
EGF, A61G	SNP	Growth signaling	HBV, HCV, alcohol	10, 11
MDM2, G309T	SNP	Cell cycle	HBV, HCV	12
XPC, L939G	SNP	DNA damage repair	HBV, HCV, aflatoxin	13
HFE, C282Y	SNP	Iron metabolism	Alcohol, GH	14
DEPDC5	SNP	GWAS	HCV	15
MICA	SNP	GWAS	HCV	16
1p36.22	SNP	GWAS	HBV	17
Th1/Th2 signature	Gene expression	Venous metastasis	HBV	19
186‐Gene signature	Gene expression	Field effect	HCV, HBV	20

Case‐control or cohort studies enrolling >500 cases are included.

Abbreviations: SNP, single‐nucleotide polymorphism; GWAS, genome‐wide association study; HCV, hepatitis C virus; GH, genetic hemochromatosis.

Molecular Subclasses of Aggressive HCC Tumor

Genomics technology has revealed heterogeneous molecular features of HCC tumors associated with biological aggressiveness and poorer clinical outcome, especially early recurrence after surgical or ablative therapies (Table 2). Recurrent TP53 inactivation mutations and CTNNB1 activation mutations have been observed in multiple patient cohorts24, 25 and were recently confirmed by next‐generation sequencing together with other relatively frequent mutations in chromatin regulator genes such as ARID.26 Global transcriptional profiling identified subsets of HCC tumors characterized by progenitor cell–like features, transforming growth factor‐β activation, Myc activation, Met activation, and poor prognosis (Fig. 2).24, 29 MicroRNA expression adds another layer of molecular prognostic information.37 These data will collectively provide the basis for the identification and establishment of etiology‐specific or independent driver events and prognostic biomarkers.29 Future discovery of molecular‐targeted therapies may lead to treatment‐based molecular classification of HCC tumors.

Table 2.

Molecular Subclasses and Signatures of HCC Tumor

Molecular Feature	Clinical Outcome	Etiology	Reference
TP53 mutations	Poor prognosis	HBV, HCV, Alcohol	25, 26
CTNNB1 mutations	Good prognosis	HBV, HCV, Alcohol	25, 26
Hepatoblast signature	Poor survival	HBV, HCV, Alcohol	30
EpCAM signature	Poor survival	HBV	35
TGF‐β signature	Poor survival	HBV, HCV, Alcohol	31, 33
Met signature	Poor survival	HBV, HCV, Alcohol	36
Meta‐analysis transcriptome subclass	Early recurrence	HBV, HCV	31
153‐Gene signature	Poor survival	HBV	32
65‐Gene signature	Poor survival	HBV	34
Low miR‐26 expression	Poor survival	HBV	37

Abbreviations: TGF, transforming growth factor; HBV, hepatitis B virus; HCV, hepatitis C virus.

Figure 2.

Molecular subclasses of HCC. Abbreviation: TGF‐β, transforming growth factor‐β. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Clinical Translation of Molecular Indicators of HCC risk and Poor Prognosis

The clinical use and applicability of molecular information remain to be determined in future clinical studies. The establishment of a framework and resources for such evaluation will be the key issue, given that very few of the molecular prognostic factors in the literature have had successful validation for clinical deployment.38 Biomolecules involved in any of the steps of chronic liver disease progression (e.g., viral life cycle, fibrogenesis, and cellular transformation), could theoretically be considered as markers or targets of HCC prevention. A clinical framework to evaluate antiviral or fibrotic therapies (which may have unsatisfactory antiviral or fibrotic effects) in the context of HCC prevention will accelerate the development of HCC prevention therapies. Furthermore, preclinical animal models recapitulating a broader spectrum of the natural history of HCC development in cirrhosis will greatly enhance our capability to test the antiviral or fibrotic targets in the context of HCC prevention.