Pinato, David J. MD MRes, PhD*; Pirisi, Mario MD†; Maslen, Lynn FInstAM MAUA*; Sharma, Rohini FRACP, PhD*
Hepatocellular carcinoma (HCC) accounts as the sixth most common neoplasm on a global scale, and the third most lethal with >600,000 deaths per year worldwide.1,2 The causal association with inflammation of the liver parenchyma has long been established as a key factor in the pathogenesis of HCC.3 In fact, >80% of newly diagnosed HCCs arise in a context of liver cirrhosis, secondary to alcoholic liver disease, chronic infection by hepatotropic viruses, or metabolic derangements like α-1 antitrypsin deficit and hemochromatosis. Despite the major advances achieved in the diagnostic workup of HCC, only one third of the newly diagnosed patients are presently eligible for curative treatments.4 Even in the curative setting, 5-year survival rates after resection for early-stage HCC ranges between 17% and 53%,5 and recurrence rates can be as high as 70%.6 Early recurrence occurring within 2 years of surgery can be explained by the proliferation of microscopic neoplastic foci within the residual organ, whereas late recurrence is deemed to reflect the establishment of a new neoplastic clones from a chronically damaged liver parenchyma.7 In patients treated with liver transplantation, overall survival (OS) rates approach 75% at 4 years8 and recurrence occurs in 8% to 15% of all graft recipients fulfilling the Milan criteria.9,10 In patients with unresectable disease, overall life expectancy is not homogeneously distributed and as this patient group is the current focus for clinical studies involving molecular targeted therapies, there is a need to predict prognosis before treatment.
Although several prognostic scores have been implemented into practice, the clinical picture of HCC remains quite a composite one, where the biological features of the tumor combined with the extent of hepatic reserve dictates prognosis.11 In recent years, a number of staging systems have been applied to classify HCC patients according to different clinical and pathologic prognostic factors (Table 1). Among the various proposed models, the Cancer of the Liver Italian Program has good discriminative power after both external and prospective validation, particularly for the early stages,23 whereas the Barcelona Clinic Liver Cancer score is most widely used, having been included in official guidelines and endorsed by expert panels in HCC clinical research.4,24,25 According to the Barcelona Clinic Liver Cancer staging algorithm, patients with a diagnosis of HCC can be efficiently subclassified in accordance to the extent of spread of the tumor, residual hepatic function, and performance status.17 The advantage of this staging system over the others is to match every class with the most appropriate treatment, so that potentially curative interventions like liver resection or transplantation will only be offered to those patients for whom the benefits outweigh the risks of impaired hepatic function or potential recurrence of disease. However, the reliability of each staging system is still a debated issue as the clinical course of the disease is highly variable within each class.13 Moreover, some of the proposed staging systems do not retain prognostic value in early HCC, which has significantly increased in incidence following as the introduction of surveillance programs in cirrhotic subjects.26,27
Tissue biomarkers hold great potential in enabling a better estimate of clinically meaningful outcomes such as OS and recurrence after curative treatments, and perhaps more importantly, they may also help selecting patient subpopulations more likely to achieve clinical response to targeted therapeutic agents. This aspect is of greater consequence in clinical trial design where decreasing the intrinsic variability of the observed response may entail reduced sample size, shortened follow-up times, and optimistically decreased attrition in the drug development process.28 As sorafenib has become the standard of care for advanced HCC, this aspect of translational research has been prioritized in HCC, where prediction of response to targeted therapy on molecular grounds has not been achieved. To date, a considerable number of studies have been conducted exploring molecular markers of prognostic significance in HCC, employing both a candidate base approach and unsupervised high throughput technologies.29 Albeit having generated scientifically interesting data, none of the studies have had an impact on clinical management. Furthermore, some studies give rise to contradictory results and no one tissue biomarker has been implemented into any prognostic systems. The purpose of this paper is to review the body of knowledge that has been gathered in prognostic tissue biomarkers of HCC, to address the main criticalities associated with this area of scientific interest and to discuss the positive implications which may apply to the current clinical management of this condition.
HALLMARKS OF CANCER: MOLECULAR PREDICTORS
Tumor hypervascularity is a well-known pathologic feature of HCC, where sustained angiogenesis guides the tumor to obtain a predominantly arterial blood inflow.30 Studies aimed at discovering the prognostic relevance of angiogenesis in HCC have either focused on assessing the extent of vasculature proliferation or the activation of specific molecular pathways, some of which have proven to be effective targets for systemic therapy.31
Different methods can be used to quantify tumor neovascularization: immunohistochemical staining for endothelial-specific markers such as CD-31, CD-34, or von Willebrand Factor highlight microvessels on tissue sections, allowing for a semiquantitative assessment of the vascular component (microvessel density). The prognostic power of CD-34 microvessel density has been shown in several independent studies in resected specimens, where high CD-34 count predicts for intrahepatic recurrence, shortened disease-free,32 and OS.33,34 CD-34 immunolabeling has also been recently associated with invasion and metastasis.35 However, in contradiction with the previously mentioned studies, in a small series published by Zeng et al36 including 70 consecutive patients treated with resection or transplantation, the authors reported that lower microvascular count confers a higher risk of recurrence after resection. Such inconsistency highlights the importance of adequate statistical power as a prerequisite to any study attempting to qualify or validate new prognostic markers in HCC.
Vascular Endothelial Growth Factor
Vascular endothelial growth factor (VEGF) regulates angiogenesis through a complex network of molecular interactions played between 5 ligands (VEGF-A to VEGF-E) that bind to different VEGF receptors (VEGFRs). The interplay between isoform A and VEGFR-2 is the predominant and best characterized proangiogenic axis in cancer. Although there is consensus as to the contribution of VEGF in the pathogenesis37 and treatment of HCC,38 there is discord as to the impact of VEGF protein expression39,40 on recurrence-free41,42 and OS after curative hepatectomy43–45 when the outcomes of individual studies are considered. A recent meta-analysis extrapolating data from 8 different studies with large patient cohorts (>250 patients evaluable for OS and >400 for recurrence-free survival), confirmed high tissue expression of VEGF as a predictor of early mortality (hazard ratio, 2.15; 95% confidence interval, 1.26-3.78) and recurrence (hazard ratio, 1.69; 95% confidence interval, 1.23-2.33) following resection.46 A further study profiling 3 independently collected case series using tissue microarray technology (n=454) has reinforced the concept that VEGF overproduction is linked to reduced OS and recurrence-free survival.34 Selective inhibition of the VEGF pathway has been at the focus of intense research in the area of molecularly targeted therapy of HCC, particularly after sorafenib, a dual Raf/Map kinase inhibitor with concurrent inhibitory power on VEGF-R2, 3 and platelet-derived growth factor receptor, has become the standard of care in the treatment of advanced HCC.1 However, it still remains unclear what proportion of the clinical efficacy of sorafenib relies upon its VEGF inhibitory properties as opposed to the modulation of intracellular transduction pathways. The recent negative phase III study of sunitinib, a similar antiagiogenic/tyrosine kinase inhibitor with broader and more potent anti-VEGF-R properties but lacking specificity for Raf, seems to suggest that a combined effect on intracellular mediators is needed to exert optimal clinical efficacy.47
As low oxygen tension sensors, hypoxia-inducible (Hif) transcription factors have been shown to reprogram gene expression, enhancing the production of proangiogenic mediators in response to hypoxia. However, the high turnover rate of this protein, the high heterogeneity of oxygen distribution throughout the tumor mass, and the effect of fixation processes on tissue expression of Hifs can influence the reliability of Hifs as clinical biomarkers.48 This might contribute to the observation that positive trends with clinical outcome often do not reach statistical significance.43,45Hif-1α mRNA and protein expression in tumor specimens49 has been shown to be significantly associated with lower levels of Hif-1α transcript in their surrounding nonmalignant tissue, correlating with better 1- and 5-year disease-free survival (DFS).50 Interestingly, a connection between Hif-1α and HBx protein expression was observed in a case series of 72 hepatitis B virus (HBV)-related HCCs, where worse disease-free and OS was predicted by stronger Hif-1α immunopositivity.51 An independent effect of patients’ mortality from HCC has also been noted in patients displaying higher Hif-2α immunolabeling.52
However, another and perhaps more relevant limitation applying to the use of endogenous markers of hypoxia as prognostic determinants comes from the evidence that Hifs may be upregulated as a result of nonhypoxic stimuli such as oncogene activation or cytokine signaling, therefore limiting their role of biomarkers of tumor oxygenation status.53
Cell Cycle Regulation
The prognostic implications of cell cycle regulatory pathways have been extensively studied in HCC.
Inactivation of the p53 gene by loss of heterozygosity or point mutation is a molecular event involving up to 60% of the HCCs.54p53 accumulates within the nucleus in response to DNA damage and activates transcriptional programs leading to cell cycle arrest and apoptosis. A number of studies have addressed the role of p53 and prognosis either by immunohistochemical detection or by DNA sequence analysis. Contradicting results are again are again present with the association between p53 nuclear immunolabeling and decreased OS55–61 and DFS55,61 shown by some studies but refuted by others.56,62–69 Immunohistochemical detection of p53 is limited by nonsense or frameshift mutations that do not induce protein overexpression. Such proportion of cases, falsely labeled as p53 wild type, may have led to an underestimation of the truly p53 mutant carcinomas, potentially influencing the reliability of published results. Interestingly, studies assessing p53 mutational status by sequence analysis identified that patients with p53 mutant tumors are characterized by poorer survival,70–73 whereas the quantification of p53 mRNA expression did not correlate with survival.74 Moreover, the prognostic importance of the p53 regulator murine double minute-2 (MDM-2), whose role in antagonizing the p53-induced apoptosis program has recently emerged in HCC, strengthens the relevance of the p53-related signaling in determining tumor aggressiveness.75,76 More recently, overexpression of gankyrin, a liver oncoprotein which interacts with MDM-2 and favours degradation of p53 and Rb proteins, has been shown to correlate with more advanced tumor stage and shortened survival in a small series of resected HCC.77
Cyclins, Cyclin-dependent Kinases, and its Regulators
Cell cycle progression is a tightly regulated process in which the differential activation of cyclins and cyclin-dependent kinases (CDK) determines the transition between subsequent phases of the cell cycle. Tissue overexpression of cyclin A,78 D1 but not E65,79 has been reported as an adverse prognostic factor after curative resection of HCC. However, it has to be noted that the prognostic impact of cyclin D1 failed independent validation in a study of 51 patients with early-stage HCC.80 Among other cell cycle regulators, phospho-Rb protein exerts permissive effects on the G1/S phase transition stimulating the transcription of cyclins via E2F. Tissue expression of phospho-Rb 281 is a predictor of outcome after surgery, whereas immunopositivity for the total Rb protein has been found to correlate with intrahepatic metastasis without influencing clinical outcomes.59
Different protein subfamilies have been identified as negative regulators of cyclin/CDK complexes. The nk4/Arf locus encodes 4 different tumor suppressors, namely p16 (Ink4a), p15 (Ink4b), p18 (Ink4c), and p19 (Arf) which are constitutively expressed in differentiated cells. Promoter methylation of p15 and p16 resulting in unopposed cyclin D-CDK6 activation is a molecular mechanism conferring adverse prognosis in HCC.82 A confirmation of the detrimental effect of p16 and p18 loss on patient prognosis comes from studies analyzing their immunohistochemical expression.83,84 On the contrary, p16 gene sequence alterations do not seem to influence survival85 and in an earlier study, methylation-specific PCR analysis of p16 inactivation did not correlate with prognosis.86
Equal efforts have been spent in studying the prognostic relevance of a second subgroup of CDK inhibitors: p27, p21, and p57, also known as the KIP/CIP family. Compelling evidence has established that p27 loss is a molecular event affecting up to 75% of HCC cases.87 Interestingly, there is a strong consensus in the literature showing a survival advantage for patients with intact p27 function. A detailed description of these studies has been included in a previously published systematic review.88 Similarly, strong evidence supports the adverse prognostic role of p57 dysregulation, which can affect up to 60% of HCC.89–91 With regards to p21, 2 recent studies conclude that conserved p21 expression predicts better outcome,69,92 and in other reports p21 loss was associated with a number of adverse pathologic features but no significant difference in OS again reporting heterogeneity of results dependent on techniques used and sample size.59,93
Invasion and Metastasis
Epithelial to Mesenchymal Transition
Epithelial to mesenchymal transition (EMT) represents a series of phenotypic changes governed by specific transcription factors such as Slug, Snail. and Twist through which transforming cells progressively lose key adhesion molecules including E-cadherin and acquire unrestrained cell motility and metastatic potential. Other molecular events are related to the EMT transcription process such as upregulation of Vimentin, N-cadherin, and nuclear translocation of β-catenin. The upregulation of EMT markers Snail, Twist, but not Slug have been shown to correlate with metastatic potential and poor prognosis in HCC.94,95 In keeping with the hypothesis that suppression of E-cadherin reduces cell-cell adhesion and increases invasive potential, tumors displaying reduced expression of E-cadherin tend to have a more aggressive clinical course.95–98 Consistently, upregulation of the mesenchymal markers Vimentin and N-cadherin, 2 genes which are positively regulated during EMT, is also associated with poor outcome.99,100 No definite role can be inferred for β-catenin in the prognostic assessment of HCC. Evidence obtained by high throughput studies has confirmed that the involvement of the β-catenin/Wnt pathway in the progression of HCC.101,102 However, several discrepancies have emerged from other studies, some showing that β-catenin activation is a predictor of either favorable103–105 or poor prognosis.80,96,98,106 Differences in the proposed endpoints such as the choice of analyzing β-catenin sequence as opposed to immunoreactivity can account as potential confounders. In contrast, the expression of Wnt and its inhibitory factor 1 seems to have a definite role in dictating the prognosis of HCC after curative surgery.107,108 Wnt is functionally linked to β-catenin nuclear translocation as the activation of the Wnt signaling pathway favors cytosolic accumulation of β-catenin via Glycogen Synthase Kinase 3-β (GSK3b)-mediated phosphorylation.
Identified as major effectors of tumor cell invasion, matrix metalloproteinase (MMPs) are a group of >20 zinc-endopeptidases whose primary function is to degrade the extracellular matrix. Among all the components of the MMP family, MMP-2 and MMP-9 have been more closely related with neoplastic invasion due to their affinity to collagen type-IV, the major constituent of basement membranes. This proposed physiological role is proven in HCC, where increased MMP-2 and MMP-9 expression correlates with recurrence and OS after liver transplantation109 and resection.35,110–112 Similar prognostic roles have been elucidated for MMP-1, MMP-7, and MMP-12.112–114
It is well recognized that the proportion of proliferating cells within a tumor mass is an important hallmark of its biological aggressiveness. Immunohistochemical detection of nuclear antigens such as Mib-1 (Ki-67) and PCNA are validated and reliable methods of assessing cell proliferation. Despite some variation in the cut-offs used to categorize high-proliferating versus low-proliferating tumors, >15 independent studies have confirmed the concept that tumors with increased growth rate have an increased risk of recurrence and shorter survival times. A systematic review of all these studies can be found elsewhere.88
Avoidance of apoptosis is a key feature in the malignant progression of HCC. Bcl-2 related proteins can regulate mitochondrial membrane permeability to cytochrome C, whose cytoplasmic translocation is a major trigger of the apoptotic cascade. Bcl-2 and Bcl-xL are both suppressors of apoptosis, however, only Bcl-xL expression holds prognostic value.60,115–117 Increasing evidence shows that insensitivity to proapoptotic signals is a common trait in cancer cells. Interestingly, HCC samples conserving death-associated protein kinase expression, a positive regulator of apoptosis, are associated with earlier stage, better differentiation, and prolonged survival.118 In contrast, deregulation of the death receptors, a heterogeneous group of apoptosis inducers, confers resistance to apoptosis in HCC, with expression of Fas and its ligand predicting worse or better disease-free and OS.119,120 Loss of Trail 1 and 2 death receptors121 in conjunction with the activation of apoptosis inhibitors as measured by increased Xiap/Xaf1 ratio and stronger Xiap immunoreactivity is associated with poor survival.122 Survivin, another member of the IAP family of antiapoptic proteins, is upregulated in tumor nodules compared with adjacent non-neoplastic tissue and is associated with adverse prognosis.123–126 Taken together, these studies suggest that HCC cells fail to engage apoptosis either by losing proapoptotic signals or by reinforcing the expression of endogenous antiapoptotic mediators.
Epithelial Growth Factor Family Receptors
Epithelial growth factor (EGF) is one of the best characterized mitogens in epithelial cancer cells, acting on a family of receptors which include the EGF receptor (EGF-R or c-erbB-1), Her-2 (c-erbB-2), Her-3 (c-erbB-3), and Her-4 (c-erbB-4). A comprehensive analysis of erbB receptors family expression in 100 surgical samples showed that EGF-R, Her-3, and Her-4 were expressed in more than half of the analyzed specimens, whereas Her-2 expression was present in only 21%.127 This study failed to show an independent prognostic role on OS for EGF-R and Her-3, whose expression correlated with worse outcome only on univariate analysis. At a transcriptional level, EGF-R but not EGF mRNA levels were related to the progression and recurrence of HCC.128,129 The prognostic role of Her-2 overexpression remains uncertain as only fluorescence in situ hybridization, but not immunohistochemical-based analysis has been found to correlate with tumor relapse and postoperative survival time.130
Insulin-like Growth Factor
The insulin-like growth factor (IGF) pathway consists of 2 transmembrane receptors (IGF-1R and IGF-2R) which interact with 2 ligands (IGF-1 and IGF-2) and at least 6 different IGF-binding proteins, that are known to regulate the activation of the IGF receptors by influencing the half-life of circulating IGF. Interest into IGF signaling in HCC has strengthened since the transcriptomic classification proposed by Boyault et al131 underlined that a specific subclass of tumors depends upon IGF axis overactivation. Interestingly, overactivation of the IGF axis is a peculiar trait in at least one fifth of the cases of HCC and a composite series of molecular events including IGF-2R allelic loss, epigenetically driven IGF-2 reactivation, and downregulation of the IGF-binding proteins characterizes this peculiar subfamily of tumors.132 Some of these molecular events such as loss of heterozygosity of the IGF-2R gene,133 IGF-2 promoter hypomethylation134 are adverse prognostic features, whereas conserved expression of IGF-binding protein 3 is associated with better survival in HCC.135
Mammalian Target of Rapamycin Pathway
At least 50% of the HCCs display an aberrant activation of the mammalian target of rapamycin (mTOR) pathway, a master regulator of proliferation and survival programs currently under intense scrutiny in HCC for its prognostic136 and therapeutic implications.137 In a comprehensive analysis of mTOR pathway function, Villanueva et al136 showed that phosphoribosomal protein S6 immunoreactivity was an independent determinant of patient survival. However, a second study profiling 166 liver explants for mTOR and its related proteins did not observe a prognostic role for Phosphatase-Tensin Homolog (PTEN), phospho-AKT, phospho-mTOR, phospho-Serine 6 Kinase (S6K), and phospho-4EBP-1.138
This is in partial contradiction with other studies that report a correlation between mTOR activation and adverse clinicopathologic variables such as tumor grade, stage, and vascular invasion.139 Recently, a large study including 528 specimens confirmed that stronger p-S6K, p-AKT, and PTEN immunoreactivity predicts shortened OS.140 The prognostic role of p-S6K has been further validated in large and independent patient cohort composed of 196 patients.141 There is some preliminary evidence in the literature alluding to S6K phosphorylation as a determinant of tumor sensitivity to everolimus, suggesting a potential role for the mTOR transduction pathway as a source stratifying efficacy predictors in HCC.142
Other Growth Factor Receptors Pathways
Target qualification research is rapidly growing in HCC, partially encouraged by the survival benefit achieved by the multityrosine kinase inhibitor sorafenib in phase III clinical trials. Among the signaling pathways investigated in HCC, aberrant activation of the Met/hepatocyte growth factor pathway delineates a prognostically relevant signature which correlates with shorter survival.143,144 Also, fibroblast growth factor receptor 2, a known regulator of cell proliferation and tumor invasiveness, has been shown to determine tumor aggressiveness.145 Preliminary evidence has confirmed the therapeutic relevance of some of these signaling pathways,146 thus increasing the importance of a comprehensive dissection of the molecular tissue biomarkers of HCC in the drug development process.
MOLECULAR PROGNOSTIC FACTORS FROM HIGH THROUGHPUT APPROACHES
Gene Expression Profiling
Transcriptomic analysis of tumor specimens is increasingly used as a tool to subcategorize disease and determine associations between mRNA expression and clinical outcomes.147 This approach has significantly deepened our knowledge regarding the pathogenesis and progression of HCC, showing that primary dysregulation of cell circuitries such as Wnt, β-catenin, AKT are involved in the pathophysiology of HCC.131 Genome-wide exploration of prognostic traits can be carried out either by comparing independent classes of tissue samples grouped according to predefined histopathologic features known to have a prognostic impact (eg, vascular invasion or extrahepatic spread) or by allowing gene expression signatures to subclassify disease without any a priori class definition.148 As summarized in Table 2, associations between HCC gene expression signatures and clinical outcome have increasingly been sought by independent groups. Statistical validation on independent patient cohorts has been performed by most groups. However, study vary in their clinical endpoints including OS143,161,162 or DFS. Other prognostic endpoints such as the presence of vascular invasion or extrahepatic spread have also been used to identify unique transcriptomic signatures.
From a clinical standpoint, OS estimate in HCC does not only reflect the biological behavior of the tumor but also the confounding effect of the hepatic reserve.11 Nevertheless, several studies have derived signatures predictive of survival from tumor sample profiling. A few studies have identified that tumors expressing progenitor cell markers such as cytokeratin 7 and 19 are associated with reduced OS.101,163 Tumors bearing this prognostic signature seem to dysregulate the expression of EpCAM, an adhesion molecule with a known prognostic role in HCC.102 Interestingly, in a recent retrospective study including 210 patients, immunohistochemical expression of cytokeratin 19 remained a significant predictor of survival,70 favoring its clinical use. The enrichment of TGF-β-related transcripts stands as another transcriptomic trait of aggressive HCC with shortened overall patient survival.143,161,163 On the basis of a retrospective analysis of >200 specimens included in a training and validation subset, it has been shown that angiogenic signals such as Hif-1α or VEGF, and EMT markers such as Vimentin, are overexpressed in TGF-β-driven tumors, potentially justifying the prognostic impact of such a transcriptomic signature.
The choice of tumor recurrence as a meaningful prognostic endpoint in most of the referenced studies is justified by its major influence on patient mortality after resection of the primary HCC.164 With the sole exception of the study published by Budhu et al,151 early recurrence of HCC has been extensively studied on primary tumor specimens in at least 8 independent studies, where different technologies encompassing cDNA, PCR-based, and oligonucleotide arrays have been used for the purpose. The number of prognostically informative genes in these studies varies from 3 to 628, with predictive accuracy ranging between 73% and 95% leading to inherent difficulties in drawing any definite conclusions. Although time to relapse is the key element in differentiating early versus late recurrence,165 the cut-off value used to differentiate these 2 events is not uniform throughout the cited studies. For instance, the time interval qualifying early intrahepatic recurrence ranges from 1149,154,155 to 2 years.150,156 The 12-gene transcriptional signature reported by Iizuka et al149 has been shown to correlate with intrahepatic recurrence with increasing accuracy in more advanced stages. Although gene ontology of the proposed cluster is composite and the number of assayed samples is limited to 50, the presence of several immune response-implied genes, such as HLADRA and TRIM22 has suggested that immunoevasion may have a role in promoting recurrence. In the attempt to optimize outcome prediction by decreasing the dimensionality of the assay, the same group subsequently qualified and validated the prognostic value of 3 sole transcripts including HLADRA, DDX17, and LAPTM5 in a larger series of tumor specimens (n=89), achieving a predictive value of 81%.154 In their exclusively HBV-oriented study, Woo et al155 outlined 2 networks that emerged as common regulators of a poor prognostic subset of 628 differentially expressed genes in 65 HCC samples: the downregulation of SP1, a protein involved in the epigenetic regulation of gene expression and the loss of PPAR-α, a nuclear receptor guiding several aspect of cell growth and metabolism. However, these results were not consistent with the report published by Wang et al,153 which shows no overlap in the genes identified by Woo and instead identified a new subset of 22 recurrence-associated transcripts in 48 HCC samples divided in a training and validation set. Among the novel candidates, the USH1C transcript, encoding for a PDZ-containing protein called harmonin and Rac GTPase activating protein 1, were identified as the most upregulated genes in recurrent HCC specimens. The relevance of the Rho GTPase family, of which the aforementioned protein is part, has been confirmed by the upregulation of other downstream components of this pathway, such as CDC42 small effector 1 and casein kinase within the same study. These findings have also been identified by other groups where these negative regulators of cell adhesion have been found to correlate with vascular invasion.166,167
After selecting 14 differentially expressed transcripts in 10 HCCs with venous invasion and 8 without, Ho et al152 provided evidence that thrombin inhibitor (THIN), G-CSF receptor 1, 8-oxoguanine DNA glycosylase (ODG), and MAFA, a member of the melanoma antigen family are associated with tumor angiotropism, again identifying novel markers consistent with those identified by previous authors despite the equal distribution of hepatitis virus B and C positive patients.
Perhaps a more interesting approach has been taken by Budhu and colleagues, who investigated the influence of tumor microenvironment in early recurrence in a series of 115 HBV-related HCCs. Compelling evidence has highlighted the functional cross-talk between tumor cells and the surrounding proinflammatory stroma is a prognostic determinant in HCC.168
Consistent with this hypothesis, a functionally coherent subset of genes codifying for a Th-2-oriented cytokine microenvironment predicts for recurrence in these patients.151 As mentioned, true primary HCC relapse is difficult to distinguish from de novo cancerization on clinical grounds, in absence of clonality information. Under the assumption that native liver explant should abolish the preneoplastic milieu favoring the appearance of novel tumors, Mas et al160 applied transcriptomic profiling to a cohort of HCV-positive HCC patients awaiting liver transplantation and identified a cluster of 10 genes predictive of recurrence. The clinical implications of this study are particularly relevant to the routine management of explanted patients. However, the limited sample size (n=38) makes independent and possibly prospective validation mandatory. Should the prognostic power of this gene subset be confirmed, subjects at high risk of recurrence could be offered a more intensive follow-up. Moreover, a precise definition of the molecular pathways driving tumor recurrence as opposed to de novo cancerization may possibly suggest novel molecular targets which could be exploited in the adjuvant setting.
Studies investigating late recurrence, which corresponds to a true second primary tumor developing from clonal divergence of chronically damaged hepatocytes as discussed previously, utilized genomic information from noncancerous tissue, in accordance with the hypothesis of “field cancerization.”169 Okamoto and colleagues compared cDNA microarray signatures of nontumorous tissue obtained from single nodular versus multicentric HCV-related HCC patients, demonstrating the accuracy of a recurrence prediction score based on a 36-gene signature. Despite the limited sample size (n=40), the authors reported the discovery of previously unidentified molecular actors in HCC such as the angiogenesis-related c-Fes/Fps oncogene, whose potential role as a therapeutic target warrants further research.159 A second study by Hoshida et al158 has importantly proven the feasibility of gathering genomic information from paraffin-embedded tissue, overcoming the technical difficulties in RNA purification from fixed tissues. However, despite the rigorous approach adopted, consisting of a multicentric validation phase carried out on 225 independent patients from the United States and Europe, the 132-gene signature identified in a training set composed of 92 Japanese patients failed validation. Whether this discrepancy is a result of differences in frozen versus paraffin-embedded samples, heterogeneous patient features or dependent on statistical constraints still needs to be clarified. As a general rule, the lack of homogeneity in the published transcriptomic studies is one of the most important limitations of this approach. Technical differences applied to the analysis platform used and variations across the patient population (eg, etiology of the underlying liver disease) and in the clinical outcomes considered makes the validation of a prognostic subset of transcripts particularly difficult. In an attempt to address the inconsistencies found in previously published studies, a recent meta-analysis has used an integrative approach to generate 3 robust HCC transcriptomic subclasses named S1, S2, and S3. These categories were generated taking advantage of the previously published gene expression data sets, and their subclassification ability was further tested on 118 new patients. To ensure maximal generalizability of the results, the total amount of 603 profiled samples were collected from both eastern and western countries, where viral epidemiology is notoriously different.170 Although activation of the Wnt/β-catenin pathway is typical of S1 tumors, in S2 carcinomas, AKT-mediated and c-Myc-mediated cell proliferation was a common trait.
On the contrary, the third transcriptomic subclass of tumors, namely S3, encompasses those tumors which retain mature hepatocyte phenotype, better differentiation, and a generally favorable prognosis compared with S1 and S2. This study is significant as it coherently organizes the molecular heterogeneities of HCC into a robust prognostic classification which are valid across different patient subpopulations. Although not yet used in clinical practice, the advantages of a common transcriptomic reference framework will certainly help with the characterization of the molecular pathways underlying the pathogenesis and progression of HCC.
MicroRNAs have increasingly been investigated in cancer biology as a novel mechanism of posttranscriptional gene expression fine-tuning.171 At least 1000 different noncoding RNA sequences, approximately 22 nucleotides long, have been identified to interact with the 3′-UTR of specific target mRNAs repressing their translation.172 Depending on the functional nature of the targeted genes, a given miRNA can be assigned the alternative role of oncogene or tumor suppressor gene.173 The interest towards miRNAs in HCC is rapidly growing, with clinical studies increasing in number,174 and potential positive implications foreseen in the diagnosis and therapy.175 A discrete number of published reports have assessed miRNA expression as a prognostic biomarker (Table 3).
Loss of mir-15b182 and high expression of mir-221176 and is a predictor of high risk of HCC recurrence after curative surgery. Furthermore, mir-221 expression is enriched in patients with multifocal lesions and correlates with a shorter time to disease relapse.176 Survival prediction by means of tumor miRNA profiling has been tested by different groups: 2 different signatures encompassing 19 and 20 different miRNAs have shown to have good correlation with OS,186 DFS, and metastatic spread.185 Interestingly, 3 other studies have confirmed miRNA 125b, 26b, 29 as predictors of OS.177–179
Cell migration and adhesion processes have been confirmed to be influenced by Let7g, a miRNA whose underexpression has been shown to enhance type I collagen α-2 chains transcription and correlates with extrahepatic diffusion.183 Mir-30d-mediated silencing of the inhibitory G protein αi2 is associated with metastatic potential, qualifying this molecular marker as part of a metastasis suppressor axis.180 Research into miRNA signatures in HCC progressively integrates with previously identified transcriptomic traits. For example, mir-122 underexpression is a common feature in tumors bearing the poor prognosis transcriptomic feature identified by Coulouarn et al,181 whereas mir-181 overexpression is a frequent molecular event in EpCAM-positive HCC harboring progenitor cell features.184 Besides holding prognostic power miRNA have also emerged as predictors of response to systemic treatment in HCC, with expression of mir-26b being predictive of response to adjuvant interferon,178 suggesting potential therapeutic implications in the differential regulation of its target genes, which include interleukin-6 and nuclear factor kB.
A significant contribution to the modulation of gene expression is derived from the epigenetic changes, including gene promoter methylation and chromatin rearrangements that affect transcription without modifying the original DNA sequence.187 Demethylation has long time been described as an hallmark of cancer cell genomes,188 along with a concomitant aberrant hypermethylation affecting specific 1 to 2 nucleotide long sequences within gene promoters known as CpG islands.189 Although global hypomethylation status is related to genomic instability,190 methylation of CpG islands is a common mechanism leading to inactivation of tumor suppressor genes. Of note, these 2 epigenetic traits impact on the pathogenesis and progression of HCC, with the extent of methylation correlating with patient survival.191 Other prognostically meaningful epigenetic traits have also been identified. In 1 study, carcinomas harboring a global DNA hypomethylation profile and evidence of genomic instability were associated with shortened survival,192 whereas hypomethylation of repetitive DNA sequences has been associated with postoperative recurrence.193 As shown in Table 4, several independent groups have successfully derived prognostic information from promoter hypermethylation signatures, which have been identified as predictors of OS194,195 and early recurrence of disease after curative resection.82,196 In contrast, studies focusing on single candidates have proven that methylation status of certain loci such as E-cadherin and M-cadherin,197,198 Dickkopf-3199 are associated with patient survival, whereas epigenetically driven Tip30 downregulation is associated with recurrence.200 Interestingly, the extent of activation of DNA methyltransferases (DNMTs), enzymes involved in catalyzing genomic methylation processes, has been found to correlate with clinicopathologic features of HCC: a >4-fold increase in DNMT3b transcript is found in cases with poor OS and progression-free survival.201 Shorter recurrence-free and OS accompany tumors with high DNMT3a transcriptional levels201 and higher DNMT1 immunoreactivity.202
The regulation of the chromatin structure by acetyl groups turnover has been identified as a key mechanism utilized by cancer cells to promote the transcription of proangiogenic and antiapoptotic targets and epigenetically silence tumor suppressor genes such as p53 or von Hippel Lindau (VHL), proapoptotic genes of the BCL and Fas family and several cell cycle regulators. The acetylation status of nucleosomes depends upon histone acetyl transferases and histone deacetylases (HDACs), 2 classes of opposing acting enzymes which dynamically regulate protein transcription. Interestingly, the expression of several members of the HDAC family including isoforms 1, 3, and 4 has been reported to predict tumor recurrence.203–205 These findings support a potential therapeutic role for HDAC inhibitors in advanced HCC, warranting further research in this area.
CONTROVERSIES AND RECOMMENDATIONS
The study of prognostic molecular markers in HCC suffers from a number of criticalities. Although significant advances have been made in the dissection of the various molecular pathways influencing the clinical course of HCC, tissue biomarkers have proven to be of somewhat marginal utility in guiding decisions about disease management. The genomic complexity of HCC makes the translation of reliable, reproducible, and cost-effective prognostic biomarkers to the clinic a difficult process. Therefore, a process of cross-validation in independent patient cohorts is mandatory before any biomarker can enter the clinical arena. Despite the myriad of biomarkers qualified in HCC, only a limited number has undergone a rigorous validation process. In some cases, inconsistencies emerging during independent validation have questioned their utility. Different factors may account for this. For instance, the diverse ethnicity and distribution of etiologic factors between eastern and western populations is known to independently impact on survival outcomes.206 The retrospective nature of most of the analyzed studies may also imply reduced accuracy in the estimation of clinical outcomes. However, increasing efforts are being addressed by researchers so that quality of data and of study design, adequate statistical power, and reproducible methodology can guarantee for a global generalizability of the prognostic information obtained.207 Finally, a coherent positioning in the treatment algorithms of HCC is needed for any given biomarker deemed qualified to cross the border between a scientifically interesting feature of the disease to clinically a useful test. Given that the range of prognostic molecular traits is wide and likely to expand in the next future, priority should be given to those biomarkers that could potentially enable patient stratification in HCC, making the provision of currently available treatment more effective and personalized. Experience accumulated in other clinical contexts suggests that the role of molecular pathology becomes predominant when the detection of a specific molecular alteration is somehow predictive of a favorable response to treatment, a far more practice-changing property than prognostication alone. The best example of transition to a molecularly driven tumor classification can perhaps be found in non–small cell lung cancer, where the presence of epithelial growth factor receptor (EGF-R) activating mutations, found in approximately 50% of Asian patients and 10% of non-Asians, routinely helps clinicians to identify a patient subpopulation that is more likely to respond to EGF-R tyrosine kinase inhibitors. Ongoing work on novel targets such as anaplastic lymphoma kinase (ALK) rearrangements, which identify 2% to 7% of lung adenocarcinomas with favorable response rates to ALK inhibitors, has further strengthened the concept of molecular classification of non–small cell lung cancer and a number of emerging targets may further refine this in the next future.208
Given that pathway-based classification of HCC still remains an unmet clinical need, with only small evidence from early phase trials209 or case reports142 suggesting the possibility of a stratified approach, research efforts should be prioritized to the qualification of the biological determinants underlying sensitivity to molecularly targeted agents. In addition, biomarker discovery should ideally parallel the earliest stages of drug development, where the use of surrogate markers is increasingly being used in clinical trial design to identify bioactive compounds and eventually guide therapeutic selection in the individual patient.210
HCC is a molecularly composite disease in which a number of molecular actors contribute to unrestrained cell proliferation, sustained angiogenesis, limitless replicative potential, avoidance of apoptosis, peritumoral tissue invasion, and distant metastasis. In this review paper, we have summarized how such molecular mechanisms may impact on the natural history of the disease. Estimating clinically meaningful endpoints such as OS and the likelihood of tumor relapse is an undeniably important aim guiding research into prognostic markers. However, the foreseeable benefits of implementing molecular markers into the clinical area extend far beyond prognostic prediction, potentially implying a paradigm shift in the proposed treatment allocation guidelines. For instance, if a biomarker could reliably predict high risk of late relapse, in the presence of adverse prognostic factors the risk/benefit assessment between hepatic resection and local-regional treatments would favor the latters.7 Selection criteria for liver transplantation led to a dramatic improvement in recurrence-free and OS,8 however, it is controversial whether patients with less aggressive and better-differentiated tumors could equally benefit even if exceeding the Milan criteria by size or number of lesions.211,212 Increasing evidence shows that optimizing patient selection for liver transplantation by means of tissue biomarkers is a clinically achievable aim.213 Moreover, as the armamentarium of systemic targeted therapies is likely to expand in the next future, potential predictive power stands as a further advantage which could facilitate patient stratification and maximize the overall efficacy of each treatment.142
1. Llovet JM, Ricci S, Mazzaferro V, et al .Sorafenib in advanced hepatocellular carcinoma.N Engl J Med. 2008; 359:378–390.
2. Parkin DM, Bray F, Ferlay J, et al .Global cancer statistics, 2002.CA Cancer J Clin. 2005; 55:74–108.
3. Berasain C, Castillo J, Perugorria MJ, et al .Inflammation and liver cancer: new molecular links.Ann N Y Acad Sci. 2009; 1155:206–221.
4. Llovet JM, Burroughs A, Bruix J .Hepatocellular carcinoma.Lancet. 2003; 362:1907–1917.
5. Chen XP, Qiu FZ, Wu ZD, et al .Long-term outcome of resection of large hepatocellular carcinoma.Br J Surg. 2006; 93:600–606.
6. Grazi GL, Ercolani G, Pierangeli F, et al .Improved results of liver resection for hepatocellular carcinoma on cirrhosis give the procedure added value.Ann Surg. 2001; 234:71–78.
7. Sherman M .Recurrence of hepatocellular carcinoma.N Engl J Med. 2008; 359:2045–2047.
8. Mazzaferro V, Regalia E, Doci R, et al .Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis.N Engl J Med. 1996; 334:693–699.
9. Klintmalm GB .Liver transplantation for hepatocellular carcinoma: a registry report of the impact of tumor characteristics on outcome.Ann Surg. 1998; 228:479–490.
10. Yao FY, Ferrell L, Bass NM, et al .Liver transplantation for hepatocellular carcinoma: expansion of the tumor size limits does not adversely impact survival.Hepatology. 2001; 33:1394–1403.
11. Sala M, Forner A, Varela M, et al .Prognostic prediction in patients with hepatocellular carcinoma.Semin Liver Dis. 2005; 25:171–180.
12. Vauthey JN, Lauwers GY, Esnaola NF, et al .Simplified staging for hepatocellular carcinoma.J Clin Oncol. 2002; 20:1527–1536.
13. Camma C, Cabibbo G .Prognostic scores for hepatocellular carcinoma: none is the winner.Liver Int. 2009; 29:478–480.
14. Okuda K, Obata H, Nakajima Y, et al .Prognosis of primary hepatocellular carcinoma.Hepatology. 1984; 4:3S–6S.
15. Kudo M, Chung H, Osaki Y .Prognostic staging system for hepatocellular carcinoma (CLIP score): its value and limitations, and a proposal for a new staging system, the Japan Integrated Staging Score (JIS score).J Gastroenterol. 2003; 38:207–215.
16. . Investigators TCotLIPC .A new prognostic system for hepatocellular carcinoma: a retrospective study of 435 patients: the Cancer of the Liver Italian Program (CLIP) investigators.Hepatology. 1998; 28:751–755.
17. Llovet JM, Bru C, Bruix J .Prognosis of hepatocellular carcinoma: the BCLC staging classification.Semin Liver Dis. 1999; 19:329–338.
18. op den Winkel M, Nagel D, Sappl J, et al .Prognosis of patients with hepatocellular carcinoma. Validation and ranking of established staging-systems in a large western HCC-cohort.PLoS One. 2012; 7:e45066
19. Leung TW, Tang AM, Zee B, et al .Construction of the Chinese University Prognostic Index for hepatocellular carcinoma and comparison with the TNM staging system, the Okuda staging system, and the Cancer of the Liver Italian Program staging system: a study based on 926 patients.Cancer. 2002; 94:1760–1769.
20. Kinoshita A, Onoda H, Imai N, et al .Comparison of the prognostic value of inflammation-based prognostic scores in patients with hepatocellular carcinoma.Br J Cancer. 2012; 107:988–993.
21. Pinato DJ, Stebbing J, Ishizuka M, et al .A novel and validated prognostic index in hepatocellular carcinoma: the inflammation based index (IBI).J Hepatol. 2012; 57:1013–1020.
22. Pinato DJ, North BV, Sharma R .A novel, externally validated inflammation-based prognostic algorithm in hepatocellular carcinoma: the prognostic nutritional index (PNI).Br J Cancer. 2012; 106:1439–1445.
23. Daniele B, Annunziata M, Barletta E, et al .Cancer of the Liver Italian Program (CLIP) score for staging hepatocellular carcinoma.Hepatol Res. 2007; 37:suppl 2 S206–S209.
24. Llovet JM, Di Bisceglie AM, Bruix J, et al .Design and endpoints of clinical trials in hepatocellular carcinoma.J Natl Cancer Inst. 2008; 100:698–711.
25. Bruix J, Sherman M .Management of hepatocellular carcinoma.Hepatology. 2005; 42:1208–1236.
26. Llovet JM, Bruix J, Fuster J, et al .Liver transplantation for small hepatocellular carcinoma: the tumor-node-metastasis classification does not have prognostic power.Hepatology. 1998; 27:1572–1577.
27. Cillo U, Bassanello M, Vitale A, et al .The critical issue of hepatocellular carcinoma prognostic classification: which is the best tool available? J Hepatol. 2004; 40:124–131.
28. Kuhlmann J, Wensing G .The applications of biomarkers in early clinical drug development to improve decision-making processes.Curr Clin Pharmacol. 2006; 1:185–191.
29. Hoshida Y, Villanueva A, Llovet JM .Molecular profiling to predict hepatocellular carcinoma outcome.Expert Rev Gastroenterol Hepatol. 2009; 3:101–103.
30. Yang ZF, Poon RT .Vascular changes in hepatocellular carcinoma.Anat Rec (Hoboken). 2008; 291:721–734.
31. Tanaka S, Arii S .Current status of molecularly targeted therapy for hepatocellular carcinoma: basic science.Int J Clin Oncol. 2010; 15:235–241.
32. Tanigawa N, Lu C, Mitsui T, et al .Quantitation of sinusoid-like vessels in hepatocellular carcinoma: its clinical and prognostic significance.Hepatology. 1997; 26:1216–1223.
33. Nanashima A, Nakayama T, Sumida Y, et al .Relationship between microvessel count and post-hepatectomy survival in patients with hepatocellular carcinoma.World J Gastroenterol. 2008; 14:4915–4922.
34. Yang XR, Xu Y, Yu B, et al .High expression levels of putative hepatic stem/progenitor cell biomarkers related to tumour angiogenesis and poor prognosis of hepatocellular carcinoma.Gut. 2010; 59:953–962.
35. Yang P, Yuan W, He J, et al .Overexpression of EphA2, MMP-9, and MVD-CD34 in hepatocellular carcinoma: Implications for tumor progression and prognosis.Hepatol Res. 2009; 39:1169–1177.
36. Zeng W, Gouw AS, van den Heuvel MC, et al .Hepatocellular carcinomas in cirrhotic and noncirrhotic human livers share angiogenic characteristics.Ann Surg Oncol. 2010; 17:1564–1571.
37. Park YN, Kim YB, Yang KM, et al .Increased expression of vascular endothelial growth factor and angiogenesis in the early stage of multistep hepatocarcinogenesis.Arch Pathol Lab Med. 2000; 124:1061–1065.
38. Shen YC, Hsu C, Cheng AL .Molecular targeted therapy for advanced hepatocellular carcinoma: current status and future perspectives.J Gastroenterol. 2010; 45:794–807.
39. Jeng KS, Sheen IS, Wang YC, et al .Is the vascular endothelial growth factor messenger RNA expression in resectable hepatocellular carcinoma of prognostic value after resection? World J Gastroenterol. 2004; 10:676–681.
40. Sheen IS, Jeng KS, Shih SC, et al .Clinical significance of the expression of isoform 165 vascular endothelial growth factor mRNA in noncancerous liver remnants of patients with hepatocellular carcinoma.World J Gastroenterol. 2005; 11:187–192.
41. Cui J, Dong BW, Liang P, et al .Effect of c-myc, Ki-67, MMP-2 and VEGF expression on prognosis of hepatocellular carcinoma patients undergoing tumor resection.World J Gastroenterol. 2004; 10:1533–1536.
42. Guo RP, Zhong C, Shi M, et al .Clinical value of apoptosis and angiogenesis factors in estimating the prognosis of hepatocellular carcinoma.J Cancer Res Clin Oncol. 2006; 132:547–555.
43. Huang GW, Yang LY, Lu WQ .Expression of hypoxia-inducible factor 1alpha and vascular endothelial growth factor in hepatocellular carcinoma: Impact on neovascularization and survival.World J Gastroenterol. 2005; 11:1705–1708.
44. El-Assal ON, Yamanoi A, Soda Y, et al .Clinical significance of microvessel density and vascular endothelial growth factor expression in hepatocellular carcinoma and surrounding liver: possible involvement of vascular endothelial growth factor in the angiogenesis of cirrhotic liver.Hepatology. 1998; 27:1554–1562.
45. Wada H, Nagano H, Yamamoto H, et al .Expression pattern of angiogenic factors and prognosis after hepatic resection in hepatocellular carcinoma: importance of angiopoietin-2 and hypoxia-induced factor-1 alpha.Liver Int. 2006; 26:414–423.
46. Schoenleber SJ, Kurtz DM, Talwalkar JA, et al .Prognostic role of vascular endothelial growth factor in hepatocellular carcinoma: systematic review and meta-analysis.Br J Cancer. 2009; 100:1385–1392.
47. Cheng A, Kang Y, Lin D, et al .
Phase III Trial of Sunitinib (Su) Versus Sorafenib (So) in Advanced Hepatocellular Carcinoma (HCC). ASCO Annual Meeting. 2011; Chicago
48. Le QT, Courter D .Clinical biomarkers for hypoxia targeting.Cancer Metastasis Rev. 2008; 27:351–362.
49. Dai CX, Gao Q, Qiu SJ, et al .Hypoxia-inducible factor-1 alpha, in association with inflammation, angiogenesis and MYC, is a critical prognostic factor in patients with HCC after surgery.BMC Cancer. 2009; 9:418
50. Simon F, Bockhorn M, Praha C, et al .Deregulation of HIF1-alpha and hypoxia-regulated pathways in hepatocellular carcinoma and corresponding non-malignant liver tissue—influence of a modulated host stroma on the prognosis of HCC.Langenbecks Arch Surg. 2010; 395:395–405.
51. Xie H, Song J, Liu K, et al .The expression of hypoxia-inducible factor-1alpha in hepatitis B virus-related hepatocellular carcinoma: correlation with patients’ prognosis and hepatitis B virus X protein.Dig Dis Sci. 2008; 53:3225–3233.
52. Bangoura G, Liu ZS, Qian Q, et al .Prognostic significance of HIF-2alpha/EPAS1 expression in hepatocellular carcinoma.World J Gastroenterol. 2007; 13:3176–3182.
53. Zhong H, Chiles K, Feldser D, et al .Modulation of hypoxia-inducible factor 1alpha expression by the epidermal growth factor/phosphatidylinositol 3-kinase/PTEN/AKT/FRAP pathway in human prostate cancer cells: implications for tumor angiogenesis and therapeutics.Cancer Res. 2000; 60:1541–1545.
54. Teramoto T, Satonaka K, Kitazawa S, et al .p53 gene abnormalities are closely related to hepatoviral infections and occur at a late stage of hepatocarcinogenesis.Cancer Res. 1994; 54:231–235.
55. Lee TK, Man K, Poon RT, et al .Disruption of p53-p21/WAF1 cell cycle pathway contributes to progression and worse clinical outcome of hepatocellular carcinoma.Oncol Rep. 2004; 12:25–31.
56. Hu TH, Huang CC, Lin PR, et al .Expression and prognostic role of tumor suppressor gene PTEN/MMAC1/TEP1 in hepatocellular carcinoma.Cancer. 2003; 97:1929–1940.
57. Qin LX, Tang ZY, Ma ZC, et al .P53 immunohistochemical scoring: an independent prognostic marker for patients after hepatocellular carcinoma resection.World J Gastroenterol. 2002; 8:459–463.
58. Mise K, Tashiro S, Yogita S, et al .Assessment of the biological malignancy of hepatocellular carcinoma: relationship to clinicopathological factors and prognosis.Clin Cancer Res. 1998; 4:1475–1482.
59. Naka T, Toyota N, Kaneko T, et al .Protein expression of p53, p21WAF1, and Rb as prognostic indicators in patients with surgically treated hepatocellular carcinoma.Anticancer Res. 1998; 18:555–564.
60. Stroescu C, Dragnea A, Ivanov B, et al .Expression of p53, Bcl-2, VEGF, Ki67 and PCNA and prognostic significance in hepatocellular carcinoma.J Gastrointestin Liver Dis. 2008; 17:411–417.
61. Hu TH, Wang CC, Huang CC, et al .Down-regulation of tumor suppressor gene PTEN, overexpression of p53, plus high proliferating cell nuclear antigen index predict poor patient outcome of hepatocellular carcinoma after resection.Oncol Rep. 2007; 18:1417–1426.
62. Qin HX, Nan KJ, Yang G, et al .Expression and clinical significance of TAp73alpha, p53, PCNA and apoptosis in hepatocellular carcinoma.World J Gastroenterol. 2005; 11:2709–2713.
63. Kobayashi T, Sugawara Y, Shi YZ, et al .Telomerase expression and p53 status in hepatocellular carcinoma.Am J Gastroenterol. 2002; 97:3166–3171.
64. Endo K, Terada T .Protein expression of CD44 (standard and variant isoforms) in hepatocellular carcinoma: relationships with tumor grade, clinicopathologic parameters, p53 expression, and patient survival.J Hepatol. 2000; 32:78–84.
65. Ito Y, Matsuura N, Sakon M, et al .Expression and prognostic roles of the G1-S modulators in hepatocellular carcinoma: p27 independently predicts the recurrence.Hepatology. 1999; 30:90–99.
66. Terris B, Laurent-Puig P, Belghitti J, et al .Prognostic influence of clinicopathologic features, DNA-ploidy, CD44H and p53 expression in a large series of resected hepatocellular carcinoma in France.Int J Cancer. 1997; 74:614–619.
67. Ng IO, Lai EC, Chan AS, et al .Overexpression of p53 in hepatocellular carcinomas: a clinicopathological and prognostic correlation.J Gastroenterol Hepatol. 1995; 10:250–255.
68. Soini Y, Virkajarvi N, Lehto VP, et al .Hepatocellular carcinomas with a high proliferation index and a low degree of apoptosis and necrosis are associated with a shortened survival.Br J Cancer. 1996; 73:1025–1030.
69. Zhang MF, Zhang ZY, Fu J, et al .Correlation between expression of p53, p21/WAF1, and MDM2 proteins and their prognostic significance in primary hepatocellular carcinoma.J Transl Med. 2009; 7:110
70. Yuan RH, Jeng YM, Hu RH, et al .Role of p53 and beta-catenin mutations in conjunction with CK19 expression on early tumor recurrence and prognosis of hepatocellular carcinoma.J Gastrointest Surg. 2011; 15:321–329.
71. Honda K, Sbisa E, Tullo A, et al .p53 mutation is a poor prognostic indicator for survival in patients with hepatocellular carcinoma undergoing surgical tumour ablation.Br J Cancer. 1998; 77:776–782.
72. Park NH, Chung YH, Youn KH, et al .Close correlation of p53 mutation to microvascular invasion in hepatocellular carcinoma.J Clin Gastroenterol. 2001; 33:397–401.
73. Sugo H, Takamori S, Kojima K, et al .The significance of p53 mutations as an indicator of the biological behavior of recurrent hepatocellular carcinomas.Surg Today. 1999; 29:849–855.
74. Guo C, Liu QG, Zhang L, et al .Expression and clinical significance of p53, JunB and KAI1/CD82 in human hepatocellular carcinoma.Hepatobiliary Pancreat Dis Int. 2009; 8:389–396.
75. Endo K, Ueda T, Ohta T, et al .Protein expression of MDM2 and its clinicopathological relationships in human hepatocellular carcinoma.Liver. 2000; 20:209–215.
76. Schoniger-Hekele M, Hanel S, Wrba F, et al .Hepatocellular carcinoma—survival and clinical characteristics in relation to various histologic molecular markers in Western patients.Liver Int. 2005; 25:62–69.
77. Umemura A, Itoh Y, Itoh K, et al .Association of gankyrin protein expression with early clinical stages and insulin-like growth factor-binding protein 5 expression in human hepatocellular carcinoma.Hepatology. 2008; 47:493–502.
78. Chao Y, Shih YL, Chiu JH, et al .Overexpression of cyclin A but not Skp 2 correlates with the tumor relapse of human hepatocellular carcinoma.Cancer Res. 1998; 58:985–990.
79. Tannapfel A, Anhalt K, Hausermann P, et al .Identification of novel proteins associated with hepatocellular carcinomas using protein microarrays.J Pathol. 2003; 201:238–249.
80. Inagawa S, Itabashi M, Adachi S, et al .Expression and prognostic roles of beta-catenin in hepatocellular carcinoma: correlation with tumor progression and postoperative survival.Clin Cancer Res. 2002; 8:450–456.
81. Claudio PP, Russo G, Kumar CA, et al .pRb2/p130, vascular endothelial growth factor, p27(KIP1), and proliferating cell nuclear antigen expression in hepatocellular carcinoma: their clinical significance.Clin Cancer Res. 2004; 10:3509–3517.
82. Li B, Liu W, Wang L, et al .CpG island methylator phenotype associated with tumor recurrence in tumor-node-metastasis stage I hepatocellular carcinoma.Ann Surg Oncol. 2010; 17:1917–1926.
83. Matsuda Y, Ichida T, Genda T, et al .Loss of p16 contributes to p27 sequestration by cyclin D(1)-cyclin-dependent kinase 4 complexes and poor prognosis in hepatocellular carcinoma.Clin Cancer Res. 2003; 9:3389–3396.
84. Morishita A, Masaki T, Yoshiji H, et al .Reduced expression of cell cycle regulator p18(INK4C) in human hepatocellular carcinoma.Hepatology. 2004; 40:677–686.
85. Anzola M, Cuevas N, Lopez-Martinez M, et al .p16INK4A gene alterations are not a prognostic indicator for survival in patients with hepatocellular carcinoma undergoing curative hepatectomy.J Gastroenterol Hepatol. 2004; 19:397–405.
86. Li X, Hui AM, Sun L, et al .p16INK4A hypermethylation is associated with hepatitis virus infection, age, and gender in hepatocellular carcinoma.Clin Cancer Res. 2004; 10:7484–7489.
87. Tannapfel A, Grund D, Katalinic A, et al .Decreased expression of p27 protein is associated with advanced tumor stage in hepatocellular carcinoma.Int J Cancer. 2000; 89:350–355.
88. Mann CD, Neal CP, Garcea G, et al .Prognostic molecular markers in hepatocellular carcinoma: a systematic review.Eur J Cancer. 2007; 43:979–992.
89. Ito Y, Takeda T, Sakon M, et al .Expression of p57/Kip2 protein in hepatocellular carcinoma.Oncology. 2001; 61:221–225.
90. Nakai S, Masaki T, Shiratori Y, et al .Expression of p57(KIP2) in hepatocellular carcinoma: relationship between tumor differentiation and patient survival.Int J Oncol. 2002; 20:769–775.
91. Nan KJ, Guo H, Ruan ZP, et al .Expression of p57(kip2) and its relationship with clinicopathology, PCNA and p53 in primary hepatocellular carcinoma.World J Gastroenterol. 2005; 11:1237–1240.
92. Kao JT, Chuah SK, Huang CC, et al .P21/WAF1 is an independent survival prognostic factor for patients with hepatocellular carcinoma after resection.Liver Int. 2007; 27:772–781.
93. Qin LF, Ng IO .Expression of p27(KIP1) and p21(WAF1/CIP1) in primary hepatocellular carcinoma: clinicopathologic correlation and survival analysis.Hum Pathol. 2001; 32:778–784.
94. Yang MH, Chen CL, Chau GY, et al .Comprehensive analysis of the independent effect of twist and snail in promoting metastasis of hepatocellular carcinoma.Hepatology. 2009; 50:1464–1474.
95. Zhang CH, Xu GL, Jia WD, et al .Activation of STAT3 signal pathway correlates with twist and E-cadherin expression in hepatocellular carcinoma and their clinical significance.J Surg Res. 2012; 174:120–129.
96. Endo K, Ueda T, Ueyama J, et al .Immunoreactive E-cadherin, alpha-catenin, beta-catenin, and gamma-catenin proteins in hepatocellular carcinoma: relationships with tumor grade, clinicopathologic parameters, and patients’ survival.Hum Pathol. 2000; 31:558–565.
97. Iso Y, Sawada T, Okada T, et al .Loss of E-cadherin mRNA and gain of osteopontin mRNA are useful markers for detecting early recurrence of HCV-related hepatocellular carcinoma.J Surg Oncol. 2005; 92:304–311.
98. Zhai B, Yan HX, Liu SQ, et al .Reduced expression of E-cadherin/catenin complex in hepatocellular carcinomas.World J Gastroenterol. 2008; 14:5665–5673.
99. Na DC, Lee JE, Yoo JE, et al .Invasion and EMT-associated genes are up-regulated in B viral hepatocellular carcinoma with high expression of CD133-human and cell culture study.Exp Mol Pathol. 2011; 90:66–73.
100. Cho SB, Lee KH, Lee JH, et al .Expression of E- and N-cadherin and clinicopathology in hepatocellular carcinoma.Pathol Int. 2008; 58:635–642.
101. Yamashita T, Forgues M, Wang W, et al .EpCAM and alpha-fetoprotein expression defines novel prognostic subtypes of hepatocellular carcinoma.Cancer Res. 2008; 68:1451–1461.
102. Yamashita T, Ji J, Budhu A, et al .EpCAM-positive hepatocellular carcinoma cells are tumor-initiating cells with stem/progenitor cell features.Gastroenterology. 2009; 136:1012–1024.
103. Hsu HC, Jeng YM, Mao TL, et al .Beta-catenin mutations are associated with a subset of low-stage hepatocellular carcinoma negative for hepatitis B virus and with favorable prognosis.Am J Pathol. 2000; 157:763–770.
104. Mao TL, Chu JS, Jeng YM, et al .Expression of mutant nuclear beta-catenin correlates with non-invasive hepatocellular carcinoma, absence of portal vein spread, and good prognosis.J Pathol. 2001; 193:95–101.
105. Fujito T, Sasaki Y, Iwao K, et al .Prognostic significance of beta-catenin nuclear expression in hepatocellular carcinoma.Hepatogastroenterology. 2004; 51:921–924.
106. Liu L, Zhu XD, Wang WQ, et al .Activation of beta-catenin by hypoxia in hepatocellular carcinoma contributes to enhanced metastatic potential and poor prognosis.Clin Cancer Res. 2010; 16:2740–2750.
107. Lee HH, Uen YH, Tian YF, et al .Wnt-1 protein as a prognostic biomarker for hepatitis B-related and hepatitis C-related hepatocellular carcinoma after surgery.Cancer Epidemiol Biomarkers Prev. 2009; 18:1562–1569.
108. Huang L, Li MX, Wang L, et al .Prognostic value of Wnt inhibitory factor-1 expression in hepatocellular carcinoma that is independent of gene methylation.Tumour Biol. 2011; 32:233–240.
109. Nart D, Yaman B, Yilmaz F, et al .Expression of matrix metalloproteinase-9 in predicting prognosis of hepatocellular carcinoma after liver transplantation.Liver Transpl. 2010; 16:621–630.
110. Xu Y, Chen Q, Li W, et al .Overexpression of transcriptional coactivator AIB1 promotes hepatocellular carcinoma progression by enhancing cell proliferation and invasiveness.Oncogene. 2010; 29:3386–3397.
111. Liu WB, Xu GL, Jia WD, et al .Prognostic significance and mechanisms of patterned matrix vasculogenic mimicry in hepatocellular carcinoma.Med Oncol. 2011; 28:Suppl 1 S228–S238.
112. Ip YC, Cheung ST, Fan ST .Atypical localization of membrane type 1-matrix metalloproteinase in the nucleus is associated with aggressive features of hepatocellular carcinoma.Mol Carcinog. 2007; 46:225–230.
113. Yamamoto H, Itoh F, Adachi Y, et al .Messenger RNA expression of matrix metalloproteinases and tissue inhibitors of metalloproteinases in human hepatocellular carcinoma.Jpn J Clin Oncol. 1999; 29:58–62.
114. Gorrin Rivas MJ, Arii S, Furutani M, et al .Expression of human macrophage metalloelastase gene in hepatocellular carcinoma: correlation with angiostatin generation and its clinical significance.Hepatology. 1998; 28:986–993.
115. Chiu CT, Yeh TS, Hsu JC, et al .Expression of Bcl-2 family modulated through p53-dependent pathway in human hepatocellular carcinoma.Dig Dis Sci. 2003; 48:670–676.
116. Garcia EJ, Lawson D, Cotsonis G, et al .Hepatocellular carcinoma and markers of apoptosis (bcl-2, bax, bcl-x): prognostic significance.Appl Immunohistochem Mol Morphol. 2002; 10:210–217.
117. Watanabe J, Kushihata F, Honda K, et al .Prognostic significance of Bcl-xL in human hepatocellular carcinoma.Surgery. 2004; 135:604–612.
118. Matsumoto H, Nagao M, Ogawa S, et al .Prognostic significance of death-associated protein-kinase expression in hepatocellular carcinomas.Anticancer Res. 2003; 23:1333–1341.
119. Ito Y, Monden M, Takeda T, et al .The status of Fas and Fas ligand expression can predict recurrence of hepatocellular carcinoma.Br J Cancer. 2000; 82:1211–1217.
120. Lee WC, Yu MC, Chen MF .Prognostic impact of Fas ligand on hepatocellular carcinoma after hepatectomy.World J Surg. 2004; 28:792–796.
121. Kriegl L, Jung A, Engel J, et al .Expression, cellular distribution, and prognostic relevance of TRAIL receptors in hepatocellular carcinoma.Clin Cancer Res. 2010; 16:5529–5538.
122. Augello C, Caruso L, Maggioni M, et al .Inhibitors of apoptosis proteins (IAPs) expression and their prognostic significance in hepatocellular carcinoma.BMC Cancer. 2009; 9:125
123. Ye CP, Qiu CZ, Huang ZX, et al .Relationship between survivin expression and recurrence, and prognosis in hepatocellular carcinoma.World J Gastroenterol. 2007; 13:6264–6268.
124. Fields AC, Cotsonis G, Sexton D, et al .Survivin expression in hepatocellular carcinoma: correlation with proliferation, prognostic parameters, and outcome.Mod Pathol. 2004; 17:1378–1385.
125. Montorsi M, Maggioni M, Falleni M, et al .Survivin gene expression in chronic liver disease and hepatocellular carcinoma.Hepatogastroenterology. 2007; 54:2040–2044.
126. Cho S, Lee JH, Cho SB, et al .Epigenetic methylation and expression of caspase 8 and survivin in hepatocellular carcinoma.Pathol Int. 2010; 60:203–211.
127. Ito Y, Takeda T, Sakon M, et al .Expression and clinical significance of erb-B receptor family in hepatocellular carcinoma.Br J Cancer. 2001; 84:1377–1383.
128. Zhao YN, Cao J, Wu FX, et al .Expression and significance of EGF mRNA and EGFR mRNA in hepatocellular carcinoma.Ai Zheng. 2004; 23:762–766.
129. Daveau M, Scotte M, Francois A, et al .Hepatocyte growth factor, transforming growth factor alpha, and their receptors as combined markers of prognosis in hepatocellular carcinoma.Mol Carcinog. 2003; 36:130–141.
130. Huang TJ, Huang BJ, Liang QW, et al .Dual fluorescence in situ hybridization in detection of HER-2 oncogene amplification in primary hepatocellular carcinoma.Hepatobiliary Pancreat Dis Int. 2004; 3:62–68.
131. Boyault S, Rickman DS, de Reynies A, et al .Transcriptome classification of HCC is related to gene alterations and to new therapeutic targets.Hepatology. 2007; 45:42–52.
132. Tovar V, Alsinet C, Villanueva A, et al .IGF activation in a molecular subclass of hepatocellular carcinoma and pre-clinical efficacy of IGF-1R blockage.J Hepatol. 2010; 52:550–559.
133. Jang HS, Kang KM, Choi BO, et al .Clinical significance of loss of heterozygosity for M6P/IGF2R in patients with primary hepatocellular carcinoma.World J Gastroenterol. 2008; 14:1394–1398.
134. Tang SH, Yang DH, Huang W, et al .Hypomethylated P4 promoter induces expression of the insulin-like growth factor-II gene in hepatocellular carcinoma in a Chinese population.Clin Cancer Res. 2006; 12:4171–4177.
135. Aishima S, Basaki Y, Oda Y, et al .High expression of insulin-like growth factor binding protein-3 is correlated with lower portal invasion and better prognosis in human hepatocellular carcinoma.Cancer Sci. 2006; 97:1182–1190.
136. Villanueva A, Chiang DY, Newell P, et al .Pivotal role of mTOR signaling in hepatocellular carcinoma.Gastroenterology. 2008; 135:1972–1983
137. Zhu AX, Abrams TA, Miksad R, et al .Phase 1/2 study of everolimus in advanced hepatocellular carcinoma.Cancer. 2011; 117:5094–102.198.
138. Sieghart W, Fuereder T, Schmid K, et al .Mammalian target of rapamycin pathway activity in hepatocellular carcinomas of patients undergoing liver transplantation.Transplantation. 2007; 83:425–432.
139. Chen JS, Wang Q, Fu XH, et al .Involvement of PI3K/PTEN/AKT/mTOR pathway in invasion and metastasis in hepatocellular carcinoma: association with MMP-9.Hepatol Res. 2009; 39:177–186.
140. Zhou L, Huang Y, Li J, et al .The mTOR pathway is associated with the poor prognosis of human hepatocellular carcinoma.Med Oncol. 2010; 27:255–261.
141. Baba HA, Wohlschlaeger J, Cicinnati VR, et al .Phosphorylation of p70S6 kinase predicts overall survival in patients with clear margin-resected hepatocellular carcinoma.Liver Int. 2009; 29:399–405.
142. Bhoori S, Toffanin S, Sposito C, et al .Personalized molecular targeted therapy in advanced, recurrent hepatocellular carcinoma after liver transplantation: a proof of principle.J Hepatol. 2010; 52:771–775.
143. Kaposi-Novak P, Lee JS, Gomez-Quiroz L, et al .Met-regulated expression signature defines a subset of human hepatocellular carcinomas with poor prognosis and aggressive phenotype.J Clin Invest. 2006; 116:1582–1595.
144. Ueki T, Fujimoto J, Suzuki T, et al .Expression of hepatocyte growth factor and its receptor c-met proto-oncogene in hepatocellular carcinoma.Hepatology. 1997; 25:862–866.
145. Harimoto N, Taguchi K, Shirabe K, et al .The significance of fibroblast growth factor receptor 2 expression in differentiation of hepatocellular carcinoma.Oncology. 2010; 78:361–368.
146. Huynh H, Ngo VC, Fargnoli J, et al .Brivanib alaninate, a dual inhibitor of vascular endothelial growth factor receptor and fibroblast growth factor receptor tyrosine kinases, induces growth inhibition in mouse models of human hepatocellular carcinoma.Clin Cancer Res. 2008; 14:6146–6153.
147. Englert CR, Petricoin EF, Krizman DB, et al .Molecular profiling of human cancer: new opportunities.Curr Opin Mol Ther. 1999; 1:712–719.
148. Simon R, Radmacher MD, Dobbin K, et al .Pitfalls in the use of DNA microarray data for diagnostic and prognostic classification.J Natl Cancer Inst. 2003; 95:14–18.
149. Iizuka N, Oka M, Yamada-Okabe H, et al .Oligonucleotide microarray for prediction of early intrahepatic recurrence of hepatocellular carcinoma after curative resection.Lancet. 2003; 361:923–929.
150. Kurokawa Y, Matoba R, Takemasa I, et al .Molecular-based prediction of early recurrence in hepatocellular carcinoma.J Hepatol. 2004; 41:284–291.
151. Budhu A, Forgues M, Ye QH, et al .Prediction of venous metastases, recurrence, and prognosis in hepatocellular carcinoma based on a unique immune response signature of the liver microenvironment.Cancer Cell. 2006; 10:99–111.
152. Ho MC, Lin JJ, Chen CN, et al .A gene expression profile for vascular invasion can predict the recurrence after resection of hepatocellular carcinoma: a microarray approach.Ann Surg Oncol. 2006; 13:1474–1484.
153. Wang SM, Ooi LL, Hui KM .Identification and validation of a novel gene signature associated with the recurrence of human hepatocellular carcinoma.Clin Cancer Res. 2007; 13:6275–6283.
154. Somura H, Iizuka N, Tamesa T, et al .A three-gene predictor for early intrahepatic recurrence of hepatocellular carcinoma after curative hepatectomy.Oncol Rep. 2008; 19:489–495.
155. Woo HG, Park ES, Cheon JH, et al .Gene expression-based recurrence prediction of hepatitis B virus-related human hepatocellular carcinoma.Clin Cancer Res. 2008; 14:2056–2064.
156. Yoshioka S, Takemasa I, Nagano H, et al .Molecular prediction of early recurrence after resection of hepatocellular carcinoma.Eur J Cancer. 2009; 45:881–889.
157. Ye QH, Qin LX, Forgues M, et al .Predicting hepatitis B virus-positive metastatic hepatocellular carcinomas using gene expression profiling and supervised machine learning.Nat Med. 2003; 9:416–423.
158. Hoshida Y, Villanueva A, Kobayashi M, et al .Gene expression in fixed tissues and outcome in hepatocellular carcinoma.N Engl J Med. 2008; 359:1995–2004.
159. Okamoto M, Utsunomiya T, Wakiyama S, et al .Specific gene-expression profiles of noncancerous liver tissue predict the risk for multicentric occurrence of hepatocellular carcinoma in hepatitis C virus-positive patients.Ann Surg Oncol. 2006; 13:947–954.
160. Mas VR, Fisher RA, Archer KJ, et al .Genes associated with progression and recurrence of hepatocellular carcinoma in hepatitis C patients waiting and undergoing liver transplantation: preliminary results.Transplantation. 2007; 83:973–981.
161. Coulouarn C, Factor VM, Thorgeirsson SS .Transforming growth factor-beta gene expression signature in mouse hepatocytes predicts clinical outcome in human cancer.Hepatology. 2008; 47:2059–2067.
162. Lee JS, Chu IS, Heo J, et al .Classification and prediction of survival in hepatocellular carcinoma by gene expression profiling.Hepatology. 2004; 40:667–676.
163. Lee JS, Heo J, Libbrecht L, et al .A novel prognostic subtype of human hepatocellular carcinoma derived from hepatic progenitor cells.Nat Med. 2006; 12:410–416.
164. Shah SA, Greig PD, Gallinger S, et al .Factors associated with early recurrence after resection for hepatocellular carcinoma and outcomes.J Am Coll Surg. 2006; 202:275–283.
165. Imamura H, Matsuyama Y, Tanaka E, et al .Risk factors contributing to early and late phase intrahepatic recurrence of hepatocellular carcinoma after hepatectomy.J Hepatol. 2003; 38:200–207.
166. Okabe H, Satoh S, Kato T, et al .Genome-wide analysis of gene expression in human hepatocellular carcinomas using cDNA microarray: identification of genes involved in viral carcinogenesis and tumor progression.Cancer Res. 2001; 61:2129–2137.
167. Tsunedomi R, Iizuka N, Yamada-Okabe H, et al .Identification of ID2 associated with invasion of hepatitis C virus-related hepatocellular carcinoma by gene expression profile.Int J Oncol. 2006; 29:1445–1451.
168. Yang JD, Nakamura I, Roberts LR .The tumor microenvironment in hepatocellular carcinoma: current status and therapeutic targets.Semin Cancer Biol. 2011; 21:35–43.
169. Braakhuis BJ, Tabor MP, Kummer JA, et al .A genetic explanation of Slaughter’s concept of field cancerization: evidence and clinical implications.Cancer Res. 2003; 63:1727–1730.
170. Hoshida Y, Nijman SM, Kobayashi M, et al .Integrative transcriptome analysis reveals common molecular subclasses of human hepatocellular carcinoma.Cancer Res. 2009; 69:7385–7392.
171. Lieberman J .Micromanaging cancer.N Engl J Med. 2009; 361:1500–1501.
172. Bartel DP .MicroRNAs: target recognition and regulatory functions.Cell. 2009; 136:215–233.
173. Chen CZ .MicroRNAs as oncogenes and tumor suppressors.N Engl J Med. 2005; 353:1768–1771.
174. Gramantieri L, Fornari F, Callegari E, et al .MicroRNA involvement in hepatocellular carcinoma.J Cell Mol Med. 2008; 12:2189–2204.
175. Arbuthnot P, Thompson LJ .Harnessing the RNA interference pathway to advance treatment and prevention of hepatocellular carcinoma.World J Gastroenterol. 2008; 14:1670–1681.
176. Gramantieri L, Fornari F, Ferracin M, et al .MicroRNA-221 targets Bmf in hepatocellular carcinoma and correlates with tumor multifocality.Clin Cancer Res. 2009; 15:5073–5081.
177. Li W, Xie L, He X, et al .Diagnostic and prognostic implications of microRNAs in human hepatocellular carcinoma.Int J Cancer. 2008; 123:1616–1622.
178. Ji J, Shi J, Budhu A, et al .MicroRNA expression, survival, and response to interferon in liver cancer.N Engl J Med. 2009; 361:1437–1447.
179. Xiong Y, Fang JH, Yun JP, et al .Effects of microRNA-29 on apoptosis, tumorigenicity, and prognosis of hepatocellular carcinoma.Hepatology. 2010; 51:836–845.
180. Yao J, Liang L, Huang S, et al .MicroRNA-30d promotes tumor invasion and metastasis by targeting Galphai2 in hepatocellular carcinoma.Hepatology. 2010; 51:846–856.
181. Coulouarn C, Factor VM, Andersen JB, et al .Loss of miR-122 expression in liver cancer correlates with suppression of the hepatic phenotype and gain of metastatic properties.Oncogene. 2009; 28:3526–3536.
182. Chung GE, Yoon JH, Myung SJ, et al .High expression of microRNA-15b predicts a low risk of tumor recurrence following curative resection of hepatocellular carcinoma.Oncol Rep. 2010; 23:113–119.
183. Ji J, Zhao L, Budhu A, et al .Let-7g targets collagen type I alpha2 and inhibits cell migration in hepatocellular carcinoma.J Hepatol. 2010; 52:690–697.
184. Ji J, Yamashita T, Budhu A, et al .Identification of microRNA-181 by genome-wide screening as a critical player in EpCAM-positive hepatic cancer stem cells.Hepatology. 2009; 50:472–480.
185. Budhu A, Jia HL, Forgues M, et al .Identification of metastasis-related microRNAs in hepatocellular carcinoma.Hepatology. 2008; 47:897–907.
186. Jiang J, Gusev Y, Aderca I, et al .Association of MicroRNA expression in hepatocellular carcinomas with hepatitis infection, cirrhosis, and patient survival.Clin Cancer Res. 2008; 14:419–427.
187. Jones PA, Baylin SB .The fundamental role of epigenetic events in cancer.Nat Rev Genet. 2002; 3:415–428.
188. Gama-Sosa MA, Slagel VA, Trewyn RW, et al .The 5-methylcytosine content of DNA from human tumors.Nucleic Acids Res. 1983; 11:6883–6894.
189. Tost J .DNA methylation: an introduction to the biology and the disease-associated changes of a promising biomarker.Mol Biotechnol. 2010; 44:71–81.
190. Chen RZ, Pettersson U, Beard C, et al .DNA hypomethylation leads to elevated mutation rates.Nature. 1998; 395:89–93.
191. Calvisi DF, Ladu S, Gorden A, et al .Mechanistic and prognostic significance of aberrant methylation in the molecular pathogenesis of human hepatocellular carcinoma.J Clin Invest. 2007; 117:2713–2722.
192. Calvisi DF, Simile MM, Ladu S, et al .Altered methionine metabolism and global DNA methylation in liver cancer: relationship with genomic instability and prognosis.Int J Cancer. 2007; 121:2410–2420.
193. Itano O, Ueda M, Kikuchi K, et al .Correlation of postoperative recurrence in hepatocellular carcinoma with demethylation of repetitive sequences.Oncogene. 2002; 21:789–797.
194. Hernandez-Vargas H, Lambert MP, Le Calvez-Kelm F, et al .Hepatocellular carcinoma displays distinct DNA methylation signatures with potential as clinical predictors.PLoS One. 2010; 5:e9749
195. Lee HS, Kim BH, Cho NY, et al .Prognostic implications of and relationship between CpG island hypermethylation and repetitive DNA hypomethylation in hepatocellular carcinoma.Clin Cancer Res. 2009; 15:812–820.
196. Arai E, Ushijima S, Gotoh M, et al .Genome-wide DNA methylation profiles in liver tissue at the precancerous stage and in hepatocellular carcinoma.Int J Cancer. 2009; 125:2854–2862.
197. Kwon GY, Yoo BC, Koh KC, et al .Promoter methylation of E-cadherin in hepatocellular carcinomas and dysplastic nodules.J Korean Med Sci. 2005; 20:242–247.
198. Yamada S, Nomoto S, Fujii T, et al .Frequent promoter methylation of M-cadherin in hepatocellular carcinoma is associated with poor prognosis.Anticancer Res. 2007; 27:2269–2274.
199. Yang B, Du Z, Gao YT, et al .Methylation of Dickkopf-3 as a prognostic factor in cirrhosis-related hepatocellular carcinoma.World J Gastroenterol. 2010; 16:755–763.
200. Lu B, Ma Y, Wu G, et al .Methylation of Tip30 promoter is associated with poor prognosis in human hepatocellular carcinoma.Clin Cancer Res. 2008; 14:7405–7412.
201. Oh BK, Kim H, Park HJ, et al .DNA methyltransferase expression and DNA methylation in human hepatocellular carcinoma and their clinicopathological correlation.Int J Mol Med. 2007; 20:65–73.
202. Saito Y, Kanai Y, Nakagawa T, et al .Increased protein expression of DNA methyltransferase (DNMT) 1 is significantly correlated with the malignant potential and poor prognosis of human hepatocellular carcinomas.Int J Cancer. 2003; 105:527–532.
203. Wu LM, Yang Z, Zhou L, et al .Identification of histone deacetylase 3 as a biomarker for tumor recurrence following liver transplantation in HBV-associated hepatocellular carcinoma.PLoS One. 2010; 5:e14460
204. Zhang J, Yang Y, Yang T, et al .microRNA-22, downregulated in hepatocellular carcinoma and correlated with prognosis, suppresses cell proliferation and tumourigenicity.Br J Cancer. 2010; 103:1215–1220.
205. Rikimaru T, Taketomi A, Yamashita Y, et al .Clinical significance of histone deacetylase 1 expression in patients with hepatocellular carcinoma.Oncology. 2007; 72:69–74.
206. Hsu C, Shen YC, Cheng CC, et al .Geographic difference in survival outcome for advanced hepatocellular carcinoma: implications on future clinical trial design.Contemp Clin Trials. 2010; 31:55–61.
207. Dancey JE, Dobbin KK, Groshen S, et al .Guidelines for the development and incorporation of biomarker studies in early clinical trials of novel agents.Clin Cancer Res. 2010; 16:1745–1755.
208. Nevins JR .Pathway-based classification of lung cancer: a strategy to guide therapeutic selection.Proc Am Thorac Soc. 2011; 8:180–182.
209. Abou-Alfa GK, Schwartz L, Ricci S, et al .Phase II study of sorafenib in patients with advanced hepatocellular carcinoma.J Clin Oncol. 2006; 24:4293–4300.
210. Rodon J, Saura C, Dienstmann R, et al .Molecular prescreening to select patient population in early clinical trials.Nat Rev Clin Oncol. 2012; 9:359–366.
211. D’Amico F, Schwartz M, Vitale A, et al .Predicting recurrence after liver transplantation in patients with hepatocellular carcinoma exceeding the up-to-seven criteria.Liver Transpl. 2009; 15:1278–1287.
212. Cillo U, Vitale A, Bassanello M, et al .Liver transplantation for the treatment of moderately or well-differentiated hepatocellular carcinoma.Ann Surg. 2004; 239:150–159.
213. Sieghart W, Wang X, Schmid K, et al .Osteopontin expression predicts overall survival after liver transplantation for hepatocellular carcinoma in patients beyond the Milan criteria.J Hepatol. 2011; 54:89–97.