Evidence-Based Diagnosis, Staging, and Treatment …...Evidence-Based Diagnosis, Staging, and Treatment of Patients With Hepatocellular Carcinoma Jordi Bruix,1 Maria Reig,1 and Morris

Gastroenterology 2016;150:835–853

REVIEW

SAN

DPE

RSPE

CTIVES

REVIEWS IN BASIC AND CLINICAL GASTROENTEROLOGYAND HEPATOLOGY

Robert F. Schwabe and John W. Wiley, Section Editors

Evidence-Based Diagnosis, Staging, and Treatment of PatientsWith Hepatocellular Carcinoma

1Barcelona Clinic Liver Cancer Group, Liver Unit, Hospital Clinic, IDIBAPS, University of Barcelona, Centro de InvestigaciónBiomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), Barcelona, Spain; and 2Division of Gastroenterology,Department of Medicine, Toronto General Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada

Evidence-based management of patients with hepatocel-lular carcinoma (HCC) is key to their optimal care. Forindividuals at risk for HCC, surveillance usually involvesultrasonography (there is controversy over use of bio-markers). A diagnosis of HCC is made based on findingsfrom biopsy or imaging analyses. Molecular markers arenot used in diagnosis or determination of prognosis andtreatment for patients. The Barcelona Clinic Liver Canceralgorithm is the most widely used staging system. Patientswith single liver tumors or as many as 3 nodules £3 cm areclassified as having very early or early-stage cancer andbenefit from resection, transplantation, or ablation. Thosewith a greater tumor burden, confined to the liver, and whoare free of symptoms are considered to have intermediate-stage cancer and can benefit from chemoembolization ifthey still have preserved liver function. Those with symp-toms of HCC and/or vascular invasion and/or extrahepaticcancer are considered to have advanced-stage cancer andcould benefit from treatment with the kinase inhibitorsorafenib. Patients with end-stage HCC have advanced liverdisease that is not suitable for transplantation and/or haveintense symptoms. Studies now aim to identify molecularmarkers and imaging techniques that can detect patientswith HCC at earlier stages and better predict their survivaltime and response to treatment.

Keywords: Liver Cancer; BCLC; Early Detection; Therapy.

pproximately 700,000 people die of hepatocellular

Abbreviations used in this paper: AASLD, American Association for theStudy of Liver Diseases; AFP, a-fetoprotein; BCLC, Barcelona Clinic LiverCancer; c-TACE, conventional transarterial chemoembolization; DEB-TACE, transarterial chemoembolization with drug-eluting beads; HCC,hepatocellular carcinoma; RFA, radiofrequency; RECIST, Response Eval-uation Criteria in Solid Tumors; TACE, transarterial chemoembolization.

Most current article

© 2016 by the AGA Institute0016-5085/$36.00

http://dx.doi.org/10.1053/j.gastro.2015.12.041

Acarcinoma (HCC) each year worldwide, making itthe third leading cause of cancer death.1 In the United Statesand Canada, HCC is the only cancer for which mortality isincreasing2 due to the high prevalence of chronic hepatitis C,immigration from areas where hepatitis B and hepatitis Care common, and the epidemic of nonalcoholic fatty liverdisease. The incidence of intrahepatic cholangiocarcinomamight have also increased,3 but less than 10% of patientswith primary liver cancer have this cancer type. In this re-view, we do not discuss cholangiocarcinoma or the fibrola-mellar variant of HCC, which has epidemiological featuresthat differ from those of other HCCs.4

Patients with HCC usually present with symptoms ofcancer and liver failure unless the cancer is detected at anearly stage. Very advanced HCC is untreatable, and most

patients die within 3 to 6 months. However, HCC has aprolonged subclinical growth period5–8 during which in-terventions can be performed and patients can be cured. Wereview the evidence to support current methods of sur-veillance, diagnosis, staging, and treatment of HCC as well asnew treatment approaches.

Surveillance of HCCIdentification of Patients at Risk

The most significant risk factor for HCC is cirrhosis. Notall patients with cirrhosis are at equal risk for HCC, and HCCis not always found in patients with cirrhosis. There are noreliable data on the incidence of HCC in patients withoutcirrhosis. In addition to cirrhosis, other factors associatedwith increased risk include male sex, older age, persistentincrease in alanine aminotransferase level, increaseda-fetoprotein (AFP) level, and progressive impairment ofliver function.9 However, knowing that a patient has a riskfactor does not aid in the decision of whether to offer sur-veillance, because risk varies within the population identi-fied by any one risk factor. Increased risk is not sufficient tomake surveillance worthwhile; the decision to offer surveil-lance must also consider the patient’s likelihood of receivingtreatment if he or she is found to have HCC. If the severity ofliver disease and/or comorbidities indicates that effectivetreatment is impossible, there is no benefit of surveillance.

Guidelines from the American Association for the Studyof Liver Diseases (AASLD)10 were developed on the basis ofcost-effectiveness analyses and the risk of HCC in definedpopulations. More sophisticated models have since pro-duced a number of risk scoring systems11–21 (Table 1).However, these are not yet ready for general use. Most havenot been validated, and many were developed in defined

Table 1.Factors That Affect Risk of HCC

Population Variables Validation Reference

Chronic hepatitis B Age, sex, HBV DNA, cirrhosis,core promoter mutation

No Yuen et al (GAG-HCC)11

Chronic hepatitis B Age, albumin, bilirubin, HBV DNA,cirrhosis (yes or no)

Variable results in European andNorth American populations

Wong et al (CU-HCC)12

Chronic hepatitis B Age, albumin, HBV DNA, liverstiffness by transient elastography

No Wong et al13

Chronic hepatitis B Age, ALT level, HBeAg status,sex, HBV DNA

Yes (only in Asia) Yang et al (REACH-B)14

Chronic hepatitis B Age, sex, ALT level, HBV DNA,quantitative HBsAg,HBV genotype, HBeAg status

No Lee et al15

Chronic hepatitis C F3and F4

Age, race, alkaline phosphataselevel, esophageal varices,smoking, platelet count

No Lok et al16

Hepatitis C cirrhosis ALT level, AFP level, age, plateletcount

No El-Serag et al17

Liver transplant waiting list Age, diabetes, race, etiology of liverdisease, sex, severity (CTP score)

Yes Flemming et al(ADRESS-HCC)18

General population Age, sex, ALT level, liver disease,family history of HCC, cumulativesmoking history

No Hung et al19

General population Age, sex, alcohol consumption, bodymass index, diabetes (yes or no),coffee consumption, hepatitis B,hepatitis C

No Michikawa et al20

HCV post SVR Age, sex, platelet count, AFP level,fibrosis stage, HCV genotype

No Chang et al21

ALT, alanine aminotransferase; CTP, Child-Turcotte-Pugh; HBV, hepatitis B virus; HBsAg, hepatitis B surface antigen; HBeAg,hepatitis B e antigen; HCV, hepatitis C virus; SVR, sustained virologic response.

836 Bruix et al Gastroenterology Vol. 150, No. 4

REVIEWSAND

PERSPECTIVES

populations. Only one report has translated degree of riskinto a decision of whether or not to provide surveillance.19

Other studies presume that patients at highest risk requiresurveillance; however, among those with lower levels ofrisk, there is no reliable cutoff value below which surveil-lance is unnecessary. Finally, risk scoring systems were alldeveloped in untreated populations and may not performequally well in treated patients.

Liver stiffness has also been used to predict risk of HCC,either individually or in combination with a risk score,22–24

but a threshold for institution of surveillance has not beenadequately defined. The AASLD criteria for surveillance arevery broad; in the absence of a defined risk threshold, it isprobably wise to err on the side of being more inclusive andapply the AASLD incidence cutoff of 1.5% – 2% for patientswith cirrhosis and 0.2% for patients with chronic hepatitis B.

Evidence to Support Surveillanceand Its Methodology

Surveillance of HCC is controversial.25 The evidence tosupport surveillance primarily comprises demonstration ofstage migration and more frequent application of potentialcurative treatment in screened populations. There havebeen 2 randomized trials of HCC surveillance, and both wereperformed in China.26,27 One used AFP level at screening,26

and the other used AFP level plus findings on ultrasonog-raphy.27 The first study failed to show a benefit of

surveillance. The second study found a 37% reduction inmortality with surveillance, but this study has been heavilycriticized. Nonetheless, in balancing potential benefit versuspotential harm, the equation clearly tilts toward surveil-lance. The most impressive data from a prospective studythat supported surveillance came from an analysis of aTaiwanese population in which subjects were selected forsurveillance based on a risk score.28 Mortality in the groupthat received surveillance was reduced compared with thecontrol group and the general population. Virtually all cost-effectiveness analyses of surveillance find it to be effectiveand cost-effective according to accepted definitions.29–36

Patients with HCC identified by surveillance present withsmaller tumors and are more likely to undergo a curativeprocedure.37–39 These cohort studies are subject to lead timeand length bias, which cannot be completely avoided40

(Figure 1). A recent meta-analysis41 concluded that despitepoor-quality evidence, HCC surveillance increased the lifeexpectancy of patientswith cirrhosis.41 However, a systematicreview concluded that there was insufficient evidence torecommend surveillance.42 Because clinicians who care forpatients with liver disease all too often see unscreened pa-tients presentingwith advancedHCC, an a priori argument canbe made that patients at risk should undergo surveillance, atleast until there is evidence that surveillance is inefficient.

Ultrasonography is the recommended method of sur-veillance for HCC.9,10,43 There is controversy over use of as-says that measure levels of AFP, des-g-carboxy prothrombin,

Figure 1. Changes in diagnosis and treatment of HCC over time. Years ago, techniques such as ultrasonography, computedtomography, and magnetic resonance imaging were not available. Therefore, HCC was almost always diagnosed when thecancer produced symptoms. Progression registration, either symptomatic or at mere palpation or imaging, would closelyprecede death. Years later, it became feasible to diagnose HCC at earlier stages, before symptoms developed. Registration ofprogression was no longer so close to death, which probably occurred at the same time point irrespective of treatment. Thisillustrates the lead-time bias induced by earlier detection in the absence of effective treatment or suboptimal treatmentintervention. Patients are now diagnosed with early-stage disease due to surveillance programs. The best therapy can now beselected for each patient, increasing survival times. Interestingly, recurrence or progression after treatment now occurs far inadvance of death, so there may no longer be a correlation between progression and death. Progression-free survival istherefore not an informative end point for clinical trials of patients with HCC; the main end point should be survival time.

April 2016 Diagnosis, Staging, and Treatment of HCC 837

REVIEW

SAN

DPE

RSPE

CTIVES

or the L3 fraction of AFP to detect HCC because there is littleevidence that they contribute to early detection or improvepatient outcomes. Increased concentrations of these markersare associated with features of poor prognosis, includinglarge tumors, invasion of the portal vein, and poor differen-tiation of HCC cells.44–46 Surveillance aims to find smalllesions in patients with a good prognosis profile if treated,so biomarkers of poor outcomes are unlikely to be usefulfor early detection of HCC. Quantification of AFP identifiespatients with HCC with only approximately 60%sensitivity47–54; AFP levels are increased in some patientswith cirrhosis, chronic hepatitis, or cholangiocarcinoma,55 sorates of false-positive results are high. The data available fortests to measure des-g-carboxy prothrombin and the L3fraction of AFP are not sufficient to recommend them. Serialassessment of levels of AFP,56 des-g-carboxy prothrombin,and the L3 fraction of AFP57,58 has been proposed, but little isknown about the sensitivity and specificity of this strategy fordetection of HCC.17

Ultrasonography detects small HCCs with a high level ofsensitivity, although findings vary among operators andpatients. With the best operators, ultrasonography identifiespatients with HCC with approximately 80% sensitivity59,60;for most operators, the level is closer to 65%. The dome ofthe liver can be hard to view, and obesity or fatty liver makevisualization difficult. Finally, differentiating a small HCCfrom a regenerative cirrhotic nodule is also difficult. Sur-veillance by computed tomography or magnetic resonanceimaging is not recommended because of radiation risks,high cost, and high rates of false-positive results.

Tumor doubling time and findings from cohort studiesindicate that the best screening interval is 6 months.57–60

Examining patients every 3 months did not increase sur-vival time but led to more diagnostic investigations.61 Astudy in Asia compared 4-month versus 12-month surveil-lance intervals and found no difference in patient survivalafter 4 years.62 However, there were too few deaths to beinformative. A large study showed that a higher proportionof patients survived when they underwent surveillanceevery 6 months compared with every year.63

DiagnosisPatients can be diagnosed with HCC based on imaging or

biopsy analyses. The specific imaging patterns observed bymagnetic resonance imaging or computed tomographyanalyses in nodules >10 mm in cirrhotic livers or livers ofpatients at high risk for HCC are intense uptake of contrastduring the arterial phase followed by contrast washout inthe venous or delayed phases.9,43,64–68 HCCs contain mostlyarterial blood and are therefore brighter during the arterialphase than the surrounding liver, which contains arterialand venous blood. In the venous phase, the HCC is lessbright because it contains contrast-free arterial blood,whereas the liver contains venous blood with contrast. Thisdynamic pattern identifies HCCs with limited levels ofsensitivity but with close to 100% specificity.64 In theabsence of this typical appearance, a biopsy is required todetect HCC. The specificity values are not quite as high forcontrast-enhanced ultrasonography.69,70

838 Bruix et al Gastroenterology Vol. 150, No. 4

REVIEWSAND

PERSPECTIVES

In histological analyses, small nodules detected bysurveillance may be very early-stage HCCs or low- or high-grade dysplastic nodules. Very early HCC is morphologicallyand radiologically different than progressed HCC.71 It isusually hypovascular because the arterial blood supply hasnot fully developed. It is as well differentiated as high-gradedysplasia. Fat may be present, and microvascular invasion israre. Malignancy is recognized by the presence of stromalinvasion (hepatocytes in the portal tract), which a needlebiopsy may miss. More advanced or progressed HCC is lessoften well differentiated and can have a well-definedcapsule and microvascular invasion.71 Immunohistochem-ical analyses of heat shock protein 70, glypican 3, glutaminesynthetase, and clathrin heavy chain72,73 can be used toconfirm a diagnosis of HCC, but they do not replace the needfor an expert pathologist.74 Assays to measure expression oftelomerase, glypican 3, LYVE1, or survivin have been pro-posed but not validated.75

Radiological analyses detect HCC with high levels ofspecificity, so biopsy analysis is no longer necessary formany patients. Some have advocated that biopsies should beperformed on all HCCs for research purposes.76 This re-quires informed consent.77

Lesions identified by surveillance must be characterized.In any cancer screening program, there is the possibility ofoverdiagnosis.78 To minimize this risk, a recall algorithm hasbeen developed and validated.9,65,66 Lesions <10 mm areunlikely to be HCC and thus require only a short-intervalfollow-up period, such as every 3 months for at least 2 years.

Lesions �10 mm are more likely to be HCC. If computedtomography or magnetic resonance imaging evaluationsshow typical features of HCC, no further investigation isrequired. If the features are not typical, alternate imagingprocedures can be performed or a biopsy specimen can becollected and analyzed. A negative finding from biopsyanalysis does not exclude HCC because it is difficult todistinguish between early-stage HCC and dysplastic nodulesand also because of sampling errors. If suspicion is suffi-ciently strong, a second biopsy specimen can be collectedfrom the lesion and analyzed. If not, the lesion should bewatched for any change that might prompt further analysis.

StagingThe Barcelona Clinic Liver Cancer (BCLC) system has

been widely validated and is the most commonly usedstaging system for HCC (Figure 2).79–82 It determines cancerstage and patient prognosis based on tumor burden,severity of liver disease, and the patient’s performancestatus. Very early and early-stage HCC (BCLC 0 or BCLC A)include patients with a solitary lesion or up to 3 nodules �3cm (without macrovascular invasion or extrahepaticspread) with preserved liver function. Patients can benefitfrom potentially curative treatments (resection, trans-plantation, or ablation). Patients with intermediate-stageHCC (BCLC B) do not have symptoms, but they have large,multifocal tumors without vascular invasion or spreadbeyond the liver. If liver function is preserved, they may becandidates for transarterial chemoembolization (TACE). The

current definition of intermediate-stage HCC encompasses awide range of patients that can be stratified. However, pa-tients with large solitary HCCs (>5 cm) without vascularinvasion are assigned to the BCLC A group82,83; if technicallyfeasible, these patients benefit from resection.79–81,84 Also, itis important to remember that Child–Pugh class A or B caninclude patients with ascites,85 in whom bacterial perito-nitis, renal failure, or recurrent encephalopathy are pre-dictive of poor outcomes.86,87 These patients should beconsidered for transplant unless tumor burden exceedslisting criteria. If so, the patient’s disease should be classi-fied as end stage.

Patients with advanced-stage HCC (BCLC C) have tumorsthat have spread beyond the liver and/or vascular invasionand/or mild cancer-related symptoms (Eastern CooperativeOncology Group [ECOG] grades 1–2). The tyrosine kinaseinhibitor sorafenib is the only treatment found to prolongsurvival.9,43,88 Patients with end-stage disease (BCLC D)have poor liver function (Child–Pugh class C, high Model forEnd-Stage Liver Disease [MELD] scores89) and are notcandidates for transplantation and/or have marked cancer-related symptoms (Performance Status >2). They have apoor prognosis90 and require supportive care.

Measurements of AFP might be useful for prognosis ofpatients in research trials, but in practice, an increased levelof AFP will not affect treatment strategy; prognostic factorsare not used to determine therapy. It was recently proposedthat liver function be evaluated using individual parametersof the Child–Pugh score (albumin and bilirubin level).91 Amore complex model, based on findings from a retrospec-tive analysis of a large cohort of treated patients (mostlywith hepatitis B), combined outcome analysis with treat-ment preferences. It has not been validated, and theoutcome stratification is suboptimal.

Molecular Markers of Risk,Diagnosis, and Determinationof Prognosis and Treatment

Recent insights into mechanisms of HCC pathogenesis aswell as studies of large sets of tumors and patients haveidentified factors that might be used in determiningpatients’ risk of HCC, their prognosis, or the best treatment.However, translation of findings from studies of cells andexperimental models requires clinical trials. Most analyseshave been based on retrospective studies of tumor and patientsamples from tissue banks. Clinical information has beenlimited and does not incorporate variables that are currentlyused to estimate risk43,92 (such as portal hypertension, whichpredicts tumor development93) or predict survival.94 Geneexpression signatures from surrounding liver95,96 can indicatedifferent stages of cirrhosis, along with activation of stellatecells,97 rather than risk of tumor development.

In most studies, tumor samples were collected duringresection and then compared with patient outcomes; thesesamples do not represent the tumors of patients withadvanced-stage disease, who do not undergo surgery. Inaddition, there is much heterogeneity within each tumor

Figure 2. Staging and treatment according to the BCLC system. The first step in evaluation of patients combines prognosis(upper section) with selection of treatment (lower section). Prognoses are made based on clinical and tumor parameters. As inall recommendations, the treatment should be selected based on a detailed evaluation of characteristics such as the patient’sage and comorbidities. As mentioned in the text, the Child–Pugh classification is not sensitive enough to accurately identifypatients with advanced liver failure who would deserve consideration for liver transplantation. Some patients corresponding toChild–Pugh class B, and even Child–Pugh class A, could have poor outcomes because of events such as spontaneousbacterial peritonitis, recurrent variceal bleeding, refractory ascites, hepatorenal syndrome, recurrent encephalopathy, or severemalnutrition, which are not registered in the Child–Pugh classification. Patients with end-stage cirrhosis due to heavilyimpaired liver function (Child–Pugh class C or earlier stages with predictors of poor outcome, or high Model for End-StageLiver Disease [MELD] scores) should be considered for liver transplantation. In these patients, HCC may become a contra-indication if it exceeds the enlistment criteria. Modified with permission from Forner et al81 and Reig et al.82

April 2016 Diagnosis, Staging, and Treatment of HCC 839

REVIEW

SAN

DPE

RSPE

CTIVES

nodule and among separate nodules.98,99 This poses a majorchallenge to the use of molecular analyses of biopsy speci-mens to determine prognosis or select treatment (see Pinyolet al100 and Borel et al101 for reviews on this subject).

TreatmentThe end point of treatment is to increase survival.

Treatments should not be offered because they are techni-cally possible.102 Treatment indications have been refined,and if patients are not candidates for first-line therapy asper stage, they can be given the treatment for a moreadvanced-stage tumor (treatment stage migration; seeFigure 3).82

BCLC Stage 0Surgery is no longer the only first-line treatment.

Resection, transplantation, and ablation provide excellentresults for single lesions �2 cm (T1 stage) in patients withpreserved liver function.82,102–104 Rates of 5-year survivalrange from 60% to 80%.105,106

Although there has been no robust trial to compare theefficacy of surgery versus ablation,107 case-control andmodeling studies have shown ablation to be noninferior andmore cost-effective for patients with very early-stageHCCs.108,109 There are no data to guide decision makingfor small tumors when patients may be candidates for alloptions. Some have reserved liver transplantation for pa-tients with recurrence of cancer after treatment.110 Others

Figure 3. Development of HCC and first-line treatment options for each stage. Disease evolution is complex in the clinicalsetting. Several factors affect treatment of advanced stage. Each patient with HCC should be carefully evaluated and treat-ment selected based on individual features. ECOG, Eastern Cooperative Oncology Group; PS, Performance Status.

840 Bruix et al Gastroenterology Vol. 150, No. 4

REVIEWSAND

PERSPECTIVES

have proposed resection as a first approach, reserving livertransplantation for patients with microvascular invasion orsatellites detected by pathological analysis.111

BCLC Stage AResection and transplantation produce the best outcomes

for well-selected candidates with HCC of BCLC stage A; 60%to 80% of patients survive for 5 years. These approachescompete as the first treatment option for patients with earlytumors, well-preserved liver function, and no clinically sig-nificant portal hypertension.9,43,112 Patients should beassessed for resection and transplantation to identify thosewho would have better outcomes than they would with othertreatments. Treatment should be chosen based on predictedimmediate-term and long-term survival. Liver function is notaccurately evaluated by Child�Pugh score. Clinically signifi-cant portal hypertension predicts poor outcomes for patientswith or without HCC87 and has negative effects on mortalityand morbidity.94,113–115 Transplantation should be consid-ered for patients with clinically significant portal hyperten-sion; 70% of patients survive for 5 years with adherence totheMilan criteria.116 Patientswho are not suitable for surgerybecause of comorbidities should be considered for treatmentwith ablation.9,43

The Milan criteria (single HCC �5 cm or �3 noduleseach �3 cm and no macrovascular invasion on imaging) areused to select patients for liver transplantation.117 A meta-analysis found that patients who met the Milan criteriahad longer survival times than patients with larger tumorburdens.116 Nevertheless, the Milan criteria are oftenconsidered to be restrictive, so expanded criteria have beenproposed. The University of California San Francisco criteriahave been partially validated but have much overlap withthe Milan criteria and at best would increase the proportionof patients available for liver transplantation by approxi-mately 5%.118

Outcomes after liver transplantation can be predicted asa continuous function based on different combinations of

tumor size and number (see http://www. hcc-olt-metro-ticket.org/calculator).119,120 Patients with tumors within theup-to-7 rule without microvascular invasion at explantachieve competitive outcomes with respect to conventionalcriteria. This pathology-based proposal requires prospectivevalidation using radiological findings collected before livertransplantation.119 The AFP level has been reported toimprove the predictive ability of the Milan criteria121–123 orthe combination of AFP level (especially >400 ng/mL) andtotal tumor volume rather than tumor size and number,122

with a total tumor volume cutoff of 115 cm3. In general,patients with HCC on the waiting list for transplantationwith a baseline serum AFP level of >200 ng/mL havesignificantly worse outcomes, although the most significantadverse determinant is a steady increase of AFP level >15ng/mL per month.123 Cutoff AFP levels of 300 ng/mL, 400ng/mL, and 1000 ng/mL have been proposed for removal ofpatients from the waiting list for liver transplantation.124,125

The major limitation for successful transplantation isorgan shortage. Resection, ablation, chemoembolization, andradioembolization are commonly used to avoid unaccept-able progression (bridge therapy). However, there is littleevidence to support this approach.9,43,126 Liver donationsfrom living, related donors and deceased donors have nothad significant effects on transplant rates except in somejurisdictions. The heterogeneity of tumor presentation andthe variability of response to treatment make accuratepredictions of progression, effective transplantation, andsurvival difficult for patients with HCC.127 No equitablepriority approach for all patients on the waiting list for livertransplantation is available.128

Down-staging is defined as the reduction of the HCCburden to meet acceptable criteria, based on expectedsurvival after liver transplantation126 that is equal to thatof patients who meet transplant criteria without down-staging.129,130 Most programs use TACE to down-stage toMilan Criteria (MC), and this status should be maintainedfor at least 3 to 6 months.125 The lack of a validatedapproach for staging, assessment of down-staging and

April 2016 Diagnosis, Staging, and Treatment of HCC 841

REVIEW

SAN

DPE

RSPE

CTIVES

delisting, and robust intention-to-treat analysis has pre-vented the endorsement of this strategy.130,131

Radiofrequency (RFA) is now the first-line ablationtechnique, and microwave ablation is becoming a compet-itor. Ethanol injection provides less local control but still hasa role in achieving a complete response when there isminimal residual viable tissue. All techniques achieve thesame effectiveness and survival in solitary HCC �2 cm,32,108

with portal hypertension as the main determinant ofoutcome. Survival of patients with HCC �3 cm treated byRFA is similar to that offered by resection.132 Eitherapproach could therefore be considered first-line therapy,and consideration should be given to age, comorbidities, andtumor location. In patients with HCCs >3 cm or multifocalHCCs, the rate of failure of ablation increases,133,134 in whichcase resection (for a single HCC) may offer a better outcome.Combined treatment with chemoembolization and ablationhas been proposed,135 but robust conclusions cannot bemade from the studies performed. The rate of recurrenceafter ablation of tumors �3 cm is the same as after surgicalresection, although anatomic resection may achieve betterlocal control.

Validated predictors of recurrence are tumor size, multi-focality, macroscopic and microscopic vascular invasion, andpoor differentiation.9,43,136 Early recurrence (<2 years) islikely due to intrahepatic metastasis, whereas later recur-rence is supposedly due to metachronous HCC.136 This divi-sion has not been established; late recurrences could alsoarise due to a lower capacity of the tumor cells for dissemi-nation or proliferation. There is no effective approach toreduce the rate of recurrence,9,43,137 and there is no preferredimmunosuppressive regimen after transplant.126,138

BCLC Stage BPatients with HCC of BCLC stage B should be considered

for TACE. Conventional TACE (c-TACE) involves intra-arterialinfusion of chemotherapy (usually doxorubicin or cisplatin),frequently mixed with Lipiodol (ethiodized oil) to increaseexposure of the tumor to the drug, followed by embolizationof the feeding vessel(s) with agents such as gelatinsponge.139 This causes cytotoxicity and ischemia. Random-ized controlled trials have shown that TACE increases sur-vival time.140,141 The median survival time was 28.7 monthsfor patients receiving c-TACE versus 17.9 months for controlpatients in a BCLC trial140 and 18 months versus 9.2 monthsin a trial in Hong Kong141 in which patients with portal veininvasion were excluded. A cumulative meta-analysis showedthat c-TACE increased the proportions of patients who sur-vived 2 years142 and established it as the treatment forintermediate-stage HCC.9,10,43,143,144

Not all patients with intermediate-stage HCC can beconsidered for TACE. Some absolute and relative contra-indications exist,83 including tumor burden (size >10 cm)and impairment of liver function. The best candidates forTACE are asymptomatic patients with a solitary or limitedmultifocal HCC without vascular invasion or extrahepaticspread and with well-preserved liver function (Child–Pughclass A or B-7 points without ascites).9,43,144 In these

patients, selective TACE is well tolerated (serious adverseevents such as liver failure or abscess affect <5% of pa-tients) and achieves a high rate of objective responses.Favorable outcomes are achieved with c-TACE or TACEwith drug-eluting beads (DEB-TACE), which slowly releasechemotherapy after embolization145 to increase tumorexposure to chemotherapy and reduce systemic drugexposure. Survival after DEB-TACE is not different fromthat after c-TACE, but the treatment is more reproducibleand better tolerated. A randomized phase 2 trial showed asignificant reduction in doxorubicin-related adverseevents146; a later trial in Italy confirmed the better toler-ance of patients to DEB-TACE.147

Cohort studies with an adequate selection of candi-dates148,149 have reported median survival beyond 40months after DEB-TACE. The same survival figures are re-ported with c-TACE,150 and rates of complications are thesame with both techniques. Less stringent selection criteriaaffect outcomes.151,152 If TACE is properly applied, mediansurvival times should not less than 30 months.

There are some unanswered questions about TACE. Thecombination of TACE and RFA might provide a therapeuticbenefit for patients’ tumor burdens that are unlikely to havea complete response to either therapy alone.153 Trials thatcombine TACE with RFA have either recruited patients withearly-stage HCC and/or compared TACE and RFA with RFAalone instead of TACE and RFA compared with TACEalone.135

It is important to remember that the efficacy of TACEdepends on complete obstruction of the tumor vascula-ture.139 Treatment that aims to merely achieve selectivedelivery of chemotherapy using any type of carrier and notcomplete vascular obstruction is likely to be less effec-tive.9,142 Similarly, because angiogenic factors peak afterTACE,154 treating first one lobe and the other 1 month latercould potentially stimulate tumor proliferation in theinitially untreated lobe. No study has properly evaluatedwhether on-demand TACE, according to tumor response,increases survival compared with TACE at regular intervals.Adjuvant treatment with sorafenib155 or brivanib156 has notimproved outcomes.

One important decision is when to stop TACE. Inoncology, disease progression is taken as treatment failure.However, with locoregional therapies, some forms of pro-gressing disease (regrowth of an initially responsive tumoror appearance of a new hepatic nodule) can be successfullyre-treated.83 This justifies the concept of untreatable pro-gression (progression in which re-treatment is contra-indicated).83 TACE should not be repeated when substantialnecrosis is not achieved after 2 initial rounds of TACE, whenfollow-up treatment fails to induce marked necrosis at sitesthat have progressed after initial response, when majorprogression (substantial liver involvement, vascular inva-sion, or extrahepatic spread) occurs after an initial response,and when re-treatment is unsafe because of deterioration ofliver function. Poor tolerance to TACE, based on clinical orbiochemical findings or a scoring system (such as theAssessment for Retreatment with TACE score),151 is acontraindication for additional TACE. However, it may be

842 Bruix et al Gastroenterology Vol. 150, No. 4

REVIEWSAND

PERSPECTIVES

that patients identified as poor candidates for additionalTACE were already poor candidates for TACE in the firstplace according to guidelines.9,43,144

In some settings, resection will be offered irrespective ofportal hypertension, degree of impairment of liver function,or multifocal disease115,157,158 because surgery has the po-tential to cure the patient. No studies have directlycompared outcomes of patients after surgery (or TACE)beyond current recommendations with nonstandard treat-ments. Some patients may benefit from more aggressivetreatment, but there are no data on whether those who donot benefit have been at a disadvantage because they havenot been treated per guidelines112; findings from studies ofinternational registries indicate this to be the case.158

BCLC Stage CConventional chemotherapy, administered intravenously

or intra-arterially, is ineffective for patients with HCC.9,43,144

In some centers, the intra-arterial injection of chemotherapyis performed using an emulsion in ethiodized oil, aiming toincrease tumor exposure and reduce systemic toxicity.However, the emulsion is unstable and the survival benefitshave not been established.

Sorafenib is an inhibitor of many kinases that reducestumor cell proliferation and angiogenesis and increasestumor apoptosis.159 It also induces major changes in stro-mal cells. In a phase 3 trial performed in the West,88,160

sorafenib reduced patients’ risk of death by 30% (mediansurvival time of 10.7 months with sorafenib vs 7.9 monthswithout). In a trial performed in the East,161 patients hadshorter survival times because they entered the study withmore advanced-stage HCC; median survival times were 6.5months with sorafenib versus 4.2 months without. Themagnitude of improvement was the same in each study,indicating that the drug is active in different populations.162

There are no biomarkers for response.163 Sorafenib delaystumor progression, because patients have marginal re-ductions in tumor burden.88,161 The most frequent adverseevents are hand/food/skin reactions, asthenia, diarrhea, andarterial hypertension,88,161 the incidence of which is higherin Asian patients. Up to 30% of the patients have to dis-continue treatment because of adverse events,164 butadverse events correlate with a better outcome.164 Carefulmanagement of patients and appropriate dose adjustmentsare therefore needed.

Sorafenib is now the standard systemic therapy for HCC.It is the first treatment option for patients with HCC of BCLCstage C9,43,144,165 and for patients with HCC of BCLC stages Aor B who are not candidates for curative or locoregionaltreatments due to treatment stage migration and/oruntreatable progression because of tumor burden.82,83,166

Several agents have been evaluated in phase 3 trials asfirst- and second-line treatments for HCC.167–173 None haveexceeded the benefits of sorafenib, or placebo in second-linetreatment, despite suggestive findings from early-stagestudies.174–184 This indicates that inappropriate markersof efficacy were used in the earlier trials. The criteria toassess preclinical and early clinical studies should be

improved. Drugs could have activities that do not produceconventional responses such as tumor shrinkage or lack ofgrowth. Time to progression may also be equivocal becauseprogression at imaging may not represent treatment failure.This was the reason that treatment was continued beyondprogression in the sorafenib trials160,185 with refinement ofthe Response Evaluation Criteria in Solid Tumors (RECIST)criteria.82 Table 2 shows that differences in time to pro-gression have not always translated in improved survival.This may reflect the existence of a heterogeneous progres-sion pattern (Figure 4) and/or a potential drug toxicity thatcounterbalances the activity of the drug. Assessment ofcontrast uptake by computed tomography and magneticresonance imaging, as proposed by modified RECISTcriteria,186 has not been validated yet as a surrogate ofsorafenib efficacy. Interobserver or intraobserver differ-ences may become an issue in large multifocal disease.Similarly, definition of progression, based on modifiedRECIST criteria, requires recognition of washout and notjust the hypervascularity of nodules >10 mm. Thus, in theabsence of validation studies that establish the most accu-rate criteria, reported data on time to progression must befiltered according to the criteria used. Finally, there hasbeen much effort in selecting patients for trials based onmolecular features of tumors. The concept is sound and hasbeen used in treatment of other neoplasms with factorsshown to promote tumor development (breast, colorectal,lung) but has not yet been applied to patients with HCC. Theheterogeneity of HCC99,187 poses a major challenge for thisapproach.

BCLC Stage DPatients with HCC of BCLC stage D have no treatment

options, either because of their extensive tumor or advancedliver disease.90 Their prognosis is poor, and they shouldreceive the best supportive care.90

Treatments in DevelopmentRadioembolization

Radioembolization is the intra-arterial injection ofmicrospheres loaded with yttrium-90, a pure b-emitter witha short half-life (2.67 days) and a limited capacity topenetrate tissues (mean depth of penetration of 2.5 mm;maximum of 11 mm). Two types of yttrium-90 micro-spheres are commercially available: glass (TheraSphere;BTG, London, United Kingdom) and resin microspheres(SIR-Spheres; Sirtex SIR-Spheres Pty Ltd, Sidney,Australia).188 The efficacy of radioembolization has beenassessed in prospective studies with promising results intumor response189 and survival.190–192 Because yttrium-90is attached to small beads, the technique does not causesubstantial ischemia, minimizing the incidence of the post-embolization syndrome.193 Although radioembolization isusually well tolerated, there may be severe effects due toirradiation of other organs.194 Radioembolization has beenlinked to a form of sinusoidal obstruction syndrome knownas radioembolization-induced liver disease, which develops

Table 2.Results of Phase 3 Trials of First and Second-Line Treatments Assessing the Efficacy of Targeted Therapies in HCC

Drug Phase Author Year Drug n

Populationa

Drug scheme

Tumor response

Treatmentdiscontinuation

criteria

Time toprogression

(mo)P

valueOS(mo)

Pvalue

Child–Pughclass

A/B/C (%)PS 0/1/2

(%)

BCLCB/C(%) Criteria Scheme

First lineSorafenib 2 Abou-Alfa et al174 2006 Sorafenib 137 72/28/0 50/50/0 NR Sorafenib 800 mg

orally twice dailyModified

WHO criteria8 wk Stop at

diseaseprogression

5.5 9.2

3 Llovet et al88 2008 Sorafenib vsplacebo

299/303 95/5/0 54/38/8 18/82 RECIST þamendment

6 wk Treatmentcould continuebeyondradiographicprogression

5.5 vs 2.8 <.001 10.7 vs 7.9 <.001Cheng et al161 2009 150/76 97/3/0 25/70/5 5/95 2.8 vs 1.4 <.001 6.5 vs 4.2 .01

Erlotinib 2 Thomas et al175 2007 Erlotinib 40 80/20 40/55/5 NR Erlotinib 150 mgorally once daily

RECIST 8 wk NR NR 25 wk

2 Philip et al176 2005 38 71/26/3 26/63/11 Erlotinib 150 mgorally once daily

4–8 wk NR NR 13

2 Not available

Sorafenibpluserlotinib

3 Zhu et al170 2012 Sorafenib þerlotinib vssorafenib

362/358 96/4/0 60/40 13/24 Sorafenib 400 mgorally twice daily þerlotinib 150 mgorally once dailyor sorafenib400 mgorally twice dailyplus placebo150 mgorally once daily

RECIST 6 wk NR 3.2 vs 4 NS 9.5 vs 8.5 NS

Linifanib 2 Toh et al177 2012 Linifanib 44 86/14/0 52/39/9 NR Linifanib 0.25 mg/kgorally daily topatients withChild–Pughclass A andevery other dayto patients withChild–Pughclass B

8 wk Stop at diseaseprogression

5.4b 9.7

3 Cainap et al169 2012 Linifanib vssorafenib

514/521 95/5/0 66/34 Linifanib 17.5 mgorally oncedaily orsorafenib400 mgorally twice daily

NR 5.4 vs 4 .001 9.1 vs 9.8 NS

April2016Diagnosis,Staging,and

Treatment

ofHCC

843

REVIEWS ANDPERSPECTIVES

Table 2.Continued

Drug Phase Author Year Drug n

Populationa

Drug scheme

Tumor response

Treatmentdiscontinuation

criteria

Time toprogression

(mo)P

valueOS(mo)

Pvalue

Child–Pughclass

A/B/C (%)PS 0/1/2

(%)

BCLCB/C(%) Criteria Scheme

Sunitinib 2 Barone et al178 2013 Sunitinib 34 79/21/0 62/29/9 Sunitinib 50 mgorally oncedaily; 4-wkon/2-wk offscheduleof 6-wk cycles

12 wk Stop at diseaseprogression

2.8 5.8

Koeberle et al179 2010 45 87/13/0 47/53/0 Sunitinib 37.5 mgorally once daily

6 wk untilweek18 and every8 wkthereafter

1.5 9.3

Zhu et al180 2009 34 97/3 44/56/0 15/85 Sunitinib37.5 mg rallyonce daily; 4-wkon/2-wk offscheduleof 6-wk cycles

12 wk 4.1 9.8

Faivre et al181 2009 37 84/16 51/49 92/5 Sunitinib 50 mg/dayorally once daily;4-wk on/2-wkoff schedule of6-wk cycles

At 4 wk oftreatmentand thenevery 6 wk

5.3 8

3 Cheng et al167 2013 Sunitinib vssorafenib

530/544 99/1 53/47/0a 16/84a Sunitinib 37.5 mgorally once dailyor sorafenib400 mg twice daily

Every 6 wkfor the first24 wk andevery 8 wkthereafter

3.6 vs 3.6 NS 7.9 vs 10.2 NS

Brivanib 2 Park et al182 2011 Brivanib 55 91/9 45/49/5 11/89 Brivanib 800 mgorally once daily

ModifiedWHO criteriab

6 wk 2.8 10

3 Johnson et al168 2013 Brivanib vssorafenib

577/578 92/8/0 61/39/0 92/8 Sorafenib 400 mgorally twice dailyor brivanib 800 mgonce daily

mRECIST Treatmentcouldcontinuebeyondradiographicprogression

4.2 vs 4.1 NS 9.5 vs 9.9 NS

844Bruix

etal

GastroenterologyVol.150,No.4

REVIEWSANDPERSPECTIVES

Table 2.Continued

Drug Phase Author Year Drug n

Populationa

Drug scheme

Tumor response

Treatmentdiscontinuation

criteria

Time toprogression

(mo)P

valueOS(mo)

Pvalue

Child–Pughclass

A/B/C (%)PS 0/1/2

(%)

BCLCB/C(%) Criteria Scheme

FOLFOX-4 2 Qin et al183 2012 FOLFOX-4 20 NR NR NR Oxaliplatin 85 mg/m2

IV on day 1; calciumfolinate 200 mg/m2

IV on days 1 and 2;5-fluorouracil400 mg/m2 followedby continuous IVinfusion of5-fluorouracilon days 1 and 2

RECIST NR 5

3 Qin et al183 2013 FOLFOX-4 vsdoxorubicin

184/187 NA Every 6 wk, 1 weekduring the studytreatment phaseand every 2 months1 wk during thefollow-up phase atthe patients’respective medicalcenters

Every 6 wkduringthe studytreatmentand every2 mo duringthe follow-upphase at thepatients’respectivemedicalcenters

Diseaseprogressionor eligibilityforsurgicalresection

2.9 vs 1.8 NS 6.4 vs 4.9 NS

Second lineBrivanib 2 Finn et al184 2012 Brivanib 46 91/9 26/72/2 0/4/96/0 Brivanib 800 mg

orally once dailyModified WHO

criteria6 wk Stop at

diseaseprogression

6.6 (mRECIST) 9.8

3 Llovet et al171 2013 Brivanib vsplacebo

263/132 92/7/1 57/39/4 3/9/87/1 Brivanib 800 mgorally once daily þBSC or placebo þBSC daily

Initially modifiedWHO criteriaandsubsequentlymRECIST forHCC perprotocolamendment

Treatmentcouldcontinuebeyondradiographicprogression

4.2 vs 2.7 .001 9.4 vs 8.2 NS

Everolimus 3 Zhu et al173 2014 Everolimus vsplacebo

362/184 NR Everolimus 7.5 mgorally or placebodaily

RECIST Stop atdiseaseprogression

2.9 vs 2.6 NS 7.6 vs 7.3 NS

Ramucirumab 3 Zhu et al173 2014 Ramucirumabvs placebo

283/282 NR 56/44/0 NR Ramucirumab DP(IMC-1121B)

8 mg/kg IV every 2 wk

RECIST Stop atdiseaseprogression

3.5 vs 2.6 <.001 9.2 vs 7.6 NS

NOTE. Results with the agents in phase 2 studies is also provided.PS, Performance Status; NR, not reported; WHO, World Health Organization; NS, not significant; IV, intravenous; BSC, best supportive care.aThe information in the phase 3 trials exposes the patients’ data (Child–Pugh class, PS, and BCLC stage) in the sorafenib arm in first-line trials and active drug in second-line trials.bThe median TTP radiology.Modified with permission from Reig et al.82

April2016Diagnosis,Staging,and

Treatment

ofHCC

845

REVIEWS ANDPERSPECTIVES

Figure 4. Prognosis of patients with HCC based on imagingfindings.166 The pattern of HCC progression due to growth orknown sites or new intrahepatic sites has to be differentiatedfrom the appearance of new extrahepatic sites or vascularinvasion. While in BCLC B patients such distinction is obvi-ously not relevant if progression does not move them toBCLC C, the pattern of progression is key to properly stratifyBCLC C patients according to expected prognosis. Thisconcept has to be taken into account in trial design andanalysis. PS, Performance Status.

846 Bruix et al Gastroenterology Vol. 150, No. 4

REVIEWSAND

PERSPECTIVES

4 to 8 weeks after treatment. Patients present with jaundice,mild ascites, and moderate cholestasis.194

In a retrospective study, Salem et al compared the out-comes of patients who did not have portal vein thrombosisor extrahepatic disease and were treated with either TACEor radioembolization.195 Survival was not significantlydifferent between the cohorts, whether analyzing the wholecohort or just the intermediate-stage patients. Sangro et alreported a median survival among patients with HCC ofBCLC class B with either bilobar disease or more than 5nodules of 15.4 months and 16.6 months, respectively.192

Survival times were even longer in patients who wouldhave been ideal candidates for TACE. Mazzaferro et al re-ported a median overall survival of 18 months among pa-tients with intermediate-stage HCC.196 Therefore,radioembolization may be an effective approach in patientsnot eligible for TACE or sorafenib. However, no randomizedcontrolled trials have assessed the survival benefit of radi-oembolization compared with established treatment, soradioembolization is still considered experimental.9,69,144

New ApproachesThere is an active search for agents that are more

effective than sorafenib against HCC (see https://clinicaltrials.gov/). Mostly phase 1 and 2 trials but somephase 3 trials have compared other agents as first-linetherapies. Agents that block PD1 receptors have beeneffective against other cancers, and preliminary findingsfrom patients with HCC have been encouraging.197 However,

HCC develops in livers with ongoing or past inflammation,associated with an altered immune profile.198

HCCs have abnormal patters of epigenetic modification,so clinical studies are under way to study these.199 Highlyspecific agents that block kinases and growth factorreceptors have been tested in a small number of patients,some selected according to molecular features of the tumor,and several other proposals are under way.200 Analyses ofdata from a study of ramucirumab as a second-line therapyhave found activity in patients with increased levels ofAFP201; further trials are under way in these patients.

There have been many trials of second-line agents aftersorafenib; these include novel chemotherapy formulationsthat aim to increase entry of the drug into cells, agents thatmight replace cellular components, kinase inhibitors (whichenrich the population according to c-Met positivity, such asin the trial of tivantinib202), and oncolytic viruses.203 Theoncolytic and immunotherapeutic vaccinia virus JX-594 caninfect tumor cells to generate an anti-tumor immuneresponse and also disrupt the tumor vasculature.203 Thisagent was not found to be effective in a phase 2 trial,204 butanother trial, with an improved design, has started. Hope-fully, some of these efforts will yield positive outcomes andincrease the therapeutic options for patients with HCC.

References

2.El-Serag HB, Davila JA, Petersen NJ, et al. Thecontinuing increase in the incidence of hepatocellularcarcinoma in the United States: an update. Ann InternMed 2003;139:817–823.

3.Bridgewater J, Galle PR, Khan SA, et al. Guidelines forthe diagnosis and management of intrahepatic chol-angiocarcinoma. J Hepatol 2014;60:1268–1289.

4.Honeyman JN, Simon EP, Robine N, et al. Detection of arecurrent DNAJB1-PRKACA chimeric transcript in fibro-lamellar hepatocellular carcinoma. Science 2014;343:1010–1014.

5.Sheu JC, Sung JL, Chen DS, et al. Growth rate ofasymptomatic hepatocellular carcinoma and its clinicalimplications. Gastroenterology 1985;89:259–266.

6.Barbara L, Benzi G, Gaiani S, et al. Natural history of smalluntreated hepatocellular carcinoma in cirrhosis: a multi-variate analysis of prognostic factors of tumor growth rateand patient survival. Hepatology 1992;16:132–137.

7. Taouli B, Goh JSK, Lu Y, et al. Growth rate of hepato-cellular carcinoma: evaluation with serial computed to-mography or magnetic resonance imaging. J ComputAssist Tomogr 2005;29:425–429.

8. Furlan A, Marin D, Agnello F, et al. Hepatocellular carci-noma presenting at contrast-enhanced multi-detector-row computed tomography or gadolinium-enhancedmagnetic resonance imaging as a small (�2 cm), inde-terminate nodule: growth rate and optimal interval timefor imaging follow-up. J Comput Assist Tomogr 2012;36:20–25.

April 2016 Diagnosis, Staging, and Treatment of HCC 847

REVIEW

SAN

DPE

RSPE

CTIVES

9.Bruix J, Sherman M. Management of hepatocellularcarcinoma: an update. Hepatology 2011;53:1020–1022.

10.Bruix J, Sherman M. Management of hepatocellularcarcinoma. Hepatology 2005;42:1020–1022.

11.Yuen M-F, Tanaka Y, Fong DY-T, et al. Independent riskfactors and predictive score for the development of he-patocellular carcinoma in chronic hepatitis B. J Hepatol2009;50:80–88.

12.Wong VW-S, Chan SL, Mo F, et al. Clinical scoringsystem to predict hepatocellular carcinoma in chronichepatitis B carriers. J Clin Oncol 2010;28:1660–1665.

13.Wong GL-H, Chan HL-Y, Wong CK-Y, et al. Liverstiffness-based optimization of hepatocellular carcinomarisk score in patients with chronic hepatitis B. J Hepatol2014;60:339–345.

14.Yang H-I, Yuen M-F, Chan HL-Y, et al. Risk estimationfor hepatocellular carcinoma in chronic hepatitis B(REACH-B): development and validation of a predictivescore. Lancet Oncol 2011;12:568–574.

15.Lee M-H, Yang H-I, Liu J, et al. Prediction models oflong-term cirrhosis and hepatocellular carcinoma risk inchronic hepatitis B patients: risk scores integrating hostand virus profiles. Hepatology 2013;58:546–554.

16.Lok AS, Seeff LB, Morgan TR, et al. Incidence of hepa-tocellular carcinoma and associated risk factors in hep-atitis C-related advanced liver disease. Gastroenterology2009;136:138–148.

17.El-Serag HB, Kanwal F, Davila JA, et al. A newlaboratory-based algorithm to predict development ofhepatocellular carcinoma in patients with hepatitis C andcirrhosis. Gastroenterology 2014;146:1249–1255.e1.

18.Flemming JA, Yang JD, Vittinghoff E, et al. Risk predic-tion of hepatocellular carcinoma in patients withcirrhosis: the ADRESS-HCC risk model. Cancer 2014;120:3485–3493.

19.Hung Y-C, Lin C-L, Liu C-J, et al. Development of riskscoring system for stratifying population for hepatocel-lular carcinoma screening. Hepatology 2015;61:1934–1944.

20.Michikawa T, Inoue M, Sawada N, et al. Development ofa prediction model for 10-year risk of hepatocellularcarcinoma in middle-aged Japanese: the Japan PublicHealth Center-based Prospective Study Cohort II. PrevMed 2012;55:137–143.

21.Chang K-C, Wu Y-Y, Hung C-H, et al. Clinical-guide riskprediction of hepatocellular carcinoma development inchronic hepatitis C patients after interferon-based ther-apy. Br J Cancer 2013;109:2481–2488.

22.Jung KS, Kim SU, Ahn SH, et al. Risk assessment ofhepatitis B virus-related hepatocellular carcinomadevelopment using liver stiffness measurement (Fibro-Scan). Hepatology 2011;53:885–894.

23.Masuzaki R, Tateishi R, Yoshida H, et al. Prospective riskassessment for hepatocellular carcinoma development inpatients with chronic hepatitis C by transient elastog-raphy. Hepatology 2009;49:1954–1961.

24.Shin SH, Kim SU, Park JY, et al. Liver stiffness-basedmodel for prediction of hepatocellular carcinoma inchronic hepatitis B virus infection: comparison with his-tological fibrosis. Liver Int 2015;35:1054–1062.

25.Lederle FA, Pocha C. Screening for liver cancer: the rushto judgment. Ann Intern Med 2012;156:387–389.

26.Chen J-G, Parkin DM, Chen Q-G, et al. Screening for livercancer: results of a randomised controlled trial in Qidong,China. J Med Screen 2003;10:204–209.

27.Zhang BH, Yang BH, Tang ZY. Randomized controlledtrial of screening for hepatocellular carcinoma. J CancerRes Clin Oncol 2004;130:417–422.

28.Yeh Y-P, Hu T-H, Cho P-Y, et al. Evaluation of abdominalultrasonography mass screening for hepatocellular car-cinoma in Taiwan. Hepatology 2014;59:1840–1849.

29.Sarasin FP, Giostra E, Hadengue A. Cost-effectivenessof screening for detection of small hepatocellular carci-noma in western patients with Child-Pugh class Acirrhosis. Am J Med 1996;101:422–434.

30.Arguedas MR, Chen VK, Eloubeidi MA, et al. Screeningfor hepatocellular carcinoma in patients with hepatitis Ccirrhosis: a cost-utility analysis. Am J Gastroenterol2003;98:679–690.

31.Saab S, Ly D, Nieto J, et al. Hepatocellular carcinomascreening in patients waiting for liver transplantation: adecision analytic model. Liver Transpl 2003;9:672–681.

32.Lin SM, Lin CJ, Lin CC, et al. Randomised controlled trialcomparing percutaneous radiofrequency thermal abla-tion, percutaneous ethanol injection, and percutaneousacetic acid injection to treat hepatocellular carcinoma of3 cm or less. Gut 2005;54:1151–1156.

33.Patel D, Terrault NA, Yao FY, et al. Cost-effectiveness ofhepatocellular carcinoma surveillance in patients withhepatitis C virus-related cirrhosis. Clin GastroenterolHepatol 2005;3:75–84.

34.Nouso K, Tanaka H, Uematsu S, et al. Cost-effectivenessof the surveillance program of hepatocellular carcinomadepends on the medical circumstances. J GastroenterolHepatol 2008;23:437–444.

35.Thompson Coon J, Rogers G, Hewson P, et al. Surveil-lance of cirrhosis for hepatocellular carcinoma: a cost-utility analysis. Br J Cancer 2008;98:1166–1175.

36.Andersson KL, Salomon JA, Goldie SJ, et al. Costeffectiveness of alternative surveillance strategies forhepatocellular carcinoma in patients with cirrhosis. ClinGastroenterol Hepatol 2008;6:1418–1424.

37.Izzo F, Piccirillo M, Albino V, et al. Prospective screeningincreases the detection of potentially curable hepato-cellular carcinoma: results in 8,900 high-risk patients.HPB (Oxford) 2013;15:985–990.

38.Ando E, Kuromatsu R, Tanaka M, et al. Surveillanceprogram for early detection of hepatocellular carcinomain Japan: results of specialized department of liver dis-ease. J Clin Gastroenterol 2006;40:942–948.

39.Lu S-N, Wang J-H, Chen P-F, et al. Community-basedmass ultrasonographic screening of hepatocellular car-cinoma among thrombocytopenic adults. Cancer Epi-demiol Biomarkers Prev 2008;17:1813–1821.

40.Cucchetti A, Trevisani F, Pecorelli A, et al. Estimation oflead-time bias and its impact on the outcome of sur-veillance for the early diagnosis of hepatocellular carci-noma. J Hepatol 2014;61:333–341.

41.Singal A, Volk ML, Waljee A, et al. Meta-analysis: sur-veillance with ultrasound for early-stage hepatocellular

848 Bruix et al Gastroenterology Vol. 150, No. 4

REVIEWSAND

PERSPECTIVES

carcinoma in patients with cirrhosis. Aliment PharmacolTher 2009;30:37–47.

42.Kansagara D, Papak J, Pasha AS, et al. Screening forhepatocellular carcinoma in chronic liver disease: asystematic review. Ann. Intern. Med 2014;161:261–269.

43.EASL-EORTC clinical practice guidelines: managementof hepatocellular carcinoma. J Hepatol 2012;56:908–943.

44.Asaoka Y, Tateishi R, Nakagomi R, et al. Frequency ofand predictive factors for vascular invasion after radio-frequency ablation for hepatocellular carcinoma. PLoSOne 2014;9:e111662.

45.Kang SH, Kim DY, Jeon SM, et al. Clinical characteristicsand prognosis of hepatocellular carcinoma with differentsets of serum AFP and PIVKA-II levels. Eur J Gastro-enterol Hepatol 2012;24:849–856.

46.Chon YE, Choi GH, Lee MH, et al. Combined measure-ment of preoperative a-fetoprotein and des-g-carboxyprothrombin predicts recurrence after curative resectionin patients with hepatitis-B-related hepatocellular carci-noma. Int J Cancer 2012;131:2332–2341.

47.Trevisani F, D’Intino PE, Morselli-Labate AM, et al. Serumalpha-fetoprotein for diagnosis of hepatocellular carci-noma in patients with chronic liver disease: influence ofHBsAg and anti-HCV status. J Hepatol 2001;34:570–575.

48.Pateron D, Ganne N, Trinchet JC, et al. Prospectivestudy of screening for hepatocellular carcinoma inCaucasian patients with cirrhosis. J Hepatol 1994;20:65–71.

49.Sherman M, Peltekian KM, Lee C. Screening for hepa-tocellular carcinoma in chronic carriers of hepatitis Bvirus: incidence and prevalence of hepatocellular carci-noma in a North American urban population. Hepatology1995;22:432–438.

50.Lok AS, Sterling RK, Everhart JE, et al. Des-gamma-carboxy prothrombin and alpha-fetoprotein as bio-markers for the early detection of hepatocellularcarcinoma. Gastroenterology 2010;138:493–502.

51.Zoli M, Magalotti D, Bianchi G, et al. Efficacy of a sur-veillance program for early detection of hepatocellularcarcinoma. Cancer 1996;78:977–985.

52.Bolondi L, Sofia S, Siringo S, et al. Surveillance pro-gramme of cirrhotic patients for early diagnosis andtreatment of hepatocellular carcinoma: a cost effective-ness analysis. Gut 2001;48:251–259.

53.Tong MJ, Blatt LM, Kao VW. Surveillance for hepato-cellular carcinoma in patients with chronic viral hepatitisin the United States of America. J Gastroenterol Hepatol2001;16:553–559.

54.Marrero JA, Feng Z, Wang Y, et al. Alpha-fetoprotein,des-gamma carboxyprothrombin, and lectin-boundalpha-fetoprotein in early hepatocellular carcinoma.Gastroenterology 2009;137:110–118.

55.Rimola J, Forner A, Reig M, et al. Cholangiocarcinoma incirrhosis: absence of contrast washout in delayedphases by magnetic resonance imaging avoids misdi-agnosis of hepatocellular carcinoma. Hepatology 2009;50:791–798.

56.Biselli M, Conti F, Gramenzi A, et al. A new approach tothe use of a-fetoprotein as surveillance test for

hepatocellular carcinoma in patients with cirrhosis. Br JCancer 2015;112:69–76.

57.Oka H, Saito A, Ito K, et al. Multicenter prospectiveanalysis of newly diagnosed hepatocellular carcinomawith respect to the percentage of Lens culinarisagglutinin-reactive alpha-fetoprotein. J GastroenterolHepatol 2001;16:1378–1383.

58.Sterling RK, Jeffers L, Gordon F, et al. Utility of Lensculinaris agglutinin-reactive fraction of alpha-fetoproteinand des-gamma-carboxy prothrombin, alone or in com-bination, as biomarkers for hepatocellular carcinoma.Clin Gastroenterol Hepatol 2009;7:104–113.

59.Singal AG, Conjeevaram HS, Volk ML, et al. Effective-ness of hepatocellular carcinoma surveillance in patientswith cirrhosis. Cancer Epidemiol Biomarkers Prev 2012;21:793–799.

60.Pocha C, Dieperink E, McMaken KA, et al. Surveillancefor hepatocellular cancer with ultrasonography vs.computed tomography—a randomised study. AlimentPharmacol Ther 2013;38:303–312.

61.Trinchet J-C, Chaffaut C, Bourcier V, et al. Ultrasono-graphic surveillance of hepatocellular carcinoma incirrhosis: a randomized trial comparing 3- and 6-monthperiodicities. Hepatology 2011;54:1987–1997.

62.Wang J-H, Chang K-C, Kee K-M, et al. Hepatocellularcarcinoma surveillance at 4- vs. 12-month intervals forpatients with chronic viral hepatitis: a randomized studyin community. Am J Gastroenterol 2013;108:416–424.

63.Han K-H, Kim DY, Park JY, et al. Survival of hepatocel-lular carcinoma patients may be improved in surveillanceinterval not more than 6 months compared with morethan 6 months: a 15-year prospective study. J ClinGastroenterol 2013;47:538–544.

64.Forner A, Vilana R, Ayuso C, et al. Diagnosis of hepaticnodules 20 mm or smaller in cirrhosis: Prospective vali-dation of the noninvasive diagnostic criteria for hepato-cellular carcinoma. Hepatology 2008;47:97–104.

65.Sangiovanni A, Manini MA, Iavarone M, et al. The diag-nostic and economic impact of contrast imaging tech-niques in the diagnosis of small hepatocellular carcinomain cirrhosis. Gut 2010;59:638–644.

66.Khalili K, Kim TK, Jang H-J, et al. Optimization of imagingdiagnosis of 1-2 cm hepatocellular carcinoma: an anal-ysis of diagnostic performance and resource utilization.J Hepatol 2011;54:723–728.

67.Mitchell DG, Bruix J, Sherman M, et al. LI-RADS (LiverImaging Reporting and Data System): Summary, dis-cussion, consensus of the LI-RADS ManagementWorking Group and future directions. Hepatology 2015;61:1056–1065.

68.Darnell A, Forner A, Rimola J, et al. Liver imagingreporting and data system with MR imaging: evaluationin nodules 20 mm or smaller detected in cirrhosis atscreening US. Radiology 2015;275:698–707.

69.Galassi M1, Iavarone M, Rossi S, et al. Patterns ofappearance and risk of misdiagnosis of intrahepaticcholangiocarcinoma in cirrhosis at contrast enhancedultrasound. Liver Int 2013;33:771–779.

70.Vilana R, Forner A, Bianchi L, et al. Intrahepatic peripheralcholangiocarcinoma in cirrhosis patients may display a

April 2016 Diagnosis, Staging, and Treatment of HCC 849

REVIEW

SAN

DPE

RSPE

CTIVES

vascular pattern similar to hepatocellular carcinoma oncontrast-enhanced ultrasound. Hepatology 2010;51:2020–2029.

71.Pathologic diagnosis of early hepatocellular carcinoma: areport of the International Consensus Group for Hepa-tocellular Neoplasia. Hepatology 2009;49:658–664.

72.Di Tommaso L, Destro A, Seok JY, et al. The applicationof markers (HSP70 GPC3 and GS) in liver biopsies isuseful for detection of hepatocellular carcinoma.J Hepatol 2009;50:746–754.

73.Di Tommaso L, Destro A, Fabbris V, et al. Diagnosticaccuracy of clathrin heavy chain staining in a markerpanel for the diagnosis of small hepatocellular carci-noma. Hepatology 2011;53:1549–1557.

74.Tremosini S, Forner A, Boix L, et al. Prospective valida-tion of an immunohistochemical panel (glypican 3, heatshock protein 70 and glutamine synthetase) in liver bi-opsies for diagnosis of very early hepatocellular carci-noma. Gut 2012;61:1481–1487.

75.Llovet JM, Chen Y, Wurmbach E, et al. A molecularsignature to discriminate dysplastic nodules from earlyhepatocellular carcinoma in HCV cirrhosis. Gastroenter-ology 2006;131:1758–1767.

76.Torbenson M, Schirmacher P. Liver cancer biopsy—backto the future? Hepatology 2015;61:431–433.

77.Sherman M, Bruix J. Biopsy for liver cancer: how tobalance research needs with evidence-based clinicalpractice. Hepatology 2015;61:433–436.

78.Welch HG, Black WC. Overdiagnosis in cancer. J NatlCancer Inst 2010;102:605–613.

79.Llovet JM, Bru C, Bruix J. Prognosis of hepatocellularcarcinoma: the BCLC staging classification. Semin LiverDis 1999;19:329–338.

80.Llovet JM, Burroughs A, Bruix J. Hepatocellular carci-noma. Lancet 2003;362:1907–1917.

81.Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma.Lancet 2012;379:1245–1255.

82.Reig M, Darnell A, Forner A, et al. Systemic therapy forhepatocellular carcinoma: the issue of treatment stagemigration and registration of progression using theBCLC-refined RECIST. Semin Liver Dis 2014;34:444–455.

83.Forner A, Gilabert M, Bruix J, et al. Treatment ofintermediate-stage hepatocellular carcinoma. Nat RevClin Oncol 2014;11:525–535.

84.de Lope CR, Tremosini S, Forner A, et al. Management ofHCC. J Hepatol 2012;56(suppl 1):S75–S87.

85.Pugh RN, Murray-Lyon IM, Dawson JL, et al. Transectionof the oesophagus for bleeding oesophageal varices. BrJ Surg 1973;60:646–649.

86.D’Amico G, Garcia-Tsao G, Pagliaro L. Natural historyand prognostic indicators of survival in cirrhosis: a sys-tematic review of 118 studies. J Hepatol 2006;44:217–231.

87.Tsochatzis EA, Bosch J, Burroughs AK. Liver cirrhosis.Lancet 2014;383:1749–1761.

88.Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib inadvanced hepatocellular carcinoma. N Engl J Med 2008;359:378–390.

89.Malinchoc M, Kamath PS, Gordon FD, et al. A model topredict poor survival in patients undergoing transjugularintrahepatic portosystemic shunts. Hepatology 2000;31:864–871.

90.Cabibbo G, Enea M, Attanasio M, et al. A meta-analysisof survival rates of untreated patients in randomizedclinical trials of hepatocellular carcinoma. Hepatology2010;51:1274–1283.

91.Johnson PJ, Berhane S, Kagebayashi C, et al. Assess-ment of liver function in patients with hepatocellularcarcinoma: a new evidence-based approach-the ALBIgrade. J Clin Oncol 2015;33:550–558.

92.Llovet JM, Di Bisceglie AM, Bruix J, et al. Design andendpoints of clinical trials in hepatocellular carcinoma.J Natl Cancer Inst 2008;100:698–711.

93.Ripoll C, Groszmann RJ, Garcia-Tsao G, et al. Hepaticvenous pressure gradient predicts development of he-patocellular carcinoma independently of severity ofcirrhosis. J Hepatol 2009;50:923–928.

94.Berzigotti A, Reig M, Abraldes JG, et al. Portal hyper-tension and the outcome of surgery for hepatocellularcarcinoma in compensated cirrhosis: a systematic reviewand meta-analysis. Hepatology 2015;61:526–536.

95.Hoshida Y, Villanueva A, Kobayashi M, et al. Geneexpression in fixed tissues and outcome in hepatocellularcarcinoma. N Engl J Med 2008;359:1995–2004.

96.Nault JC, de Reynies A, Villanueva A, et al.A hepatocellular carcinoma 5-gene score associatedwith survival of patients after liver resection. Gastroen-terology 2013;145:176–187.

97.ZhangDY, Goossens N, Guo J, et al. A hepatic stellate cellgene expression signature associated with outcomes inhepatitis C cirrhosis and hepatocellular carcinoma aftercurative resection. Gut 2015 Jun 4 [Epub ahead of print].

98.Zucman-Rossi J, Villanueva A, Nault J-C, et al. The ge-netic landscape and biomarkers of hepatocellular carci-noma. Gastroenterology 2015;149:1226–1239.

99.Fox EJ, Loeb LA. Cancer: one cell at a time. Nature 2014;512:143–144.

100.Pinyol R, Nault JC, Quetglas IM, et al. Molecular profilingof liver tumors: classification and clinical translation fordecision making. Semin Liver Dis 2014;34:363–375.

101.Borel F, Konstantinova P, Jansen PLM. Diagnostic andtherapeutic potential of miRNA signatures in patientswith hepatocellular carcinoma. J Hepatol 2012;56:1371–1383.

102.Romagnoli R, Mazzaferro V, Bruix J. Surgical resectionfor hepatocellular carcinoma: moving from what can bedone to what is worth doing. Hepatology 2015;62:340–342.

103.Esnaola NF, Lauwers GY, Mirza NQ, et al. Predictors ofmicrovascular invasion in patients with hepatocellularcarcinoma who are candidates for orthotopic livertransplantation. J Gastrointest Surg 2002;6:224–232.

104.Wald C, Russo MW, Heimbach JK, et al. New OPTN/UNOS policy for liver transplant allocation: standardiza-tion of liver imaging, diagnosis, classification, andreporting of hepatocellular carcinoma. Radiology 2013;266:376–382.

850 Bruix et al Gastroenterology Vol. 150, No. 4

REVIEWSAND

PERSPECTIVES

105.Livraghi T, Meloni F, Di Stasi M, et al. Sustained com-plete response and complications rates after radio-frequency ablation of very early hepatocellular carcinomain cirrhosis: Is resection still the treatment of choice?Hepatology 2008;47:82–89.

106.Roayaie S, Obeidat K, Sposito C, et al. Resection ofhepatocellular cancer �2 cm: results from two Westerncenters. Hepatology 2013;57:1426–1435.

107.Majno PE, Mentha G, Mazzaferro V. Partial hepatectomyversus radiofrequency ablation for hepatocellular carci-noma: confirming the trial that will never be, and somecomments on the indications for liver resection. Hep-atology 2010;51:1116–1118.

108.Cho YK, Kim JK, Kim WT, et al. Hepatic resection versusradiofrequency ablation for very early stage hepatocel-lular carcinoma: a Markov model analysis. Hepatology2010;51:1284–1290.

109.CucchettiA,PiscagliaF,CesconM,etal.Cost-effectivenessof hepatic resection versus percutaneous radiofrequencyablation for early hepatocellular carcinoma. J Hepatol 2013;59:300–307.

110.Roberts JP, Venook A, Kerlan R, et al. Hepatocellularcarcinoma: Ablate and wait versus rapid transplantation.Liver Transpl 2010;16:925–929.

111.Ferrer-Fàbrega J, Forner A, Liccioni A, et al. Prospectivevalidation of “ab initio” liver transplantation in hepato-cellular carcinoma upon detection of risk factorsfor recurrence after resection. Hepatology 2016;63:839–849.

112.Bruix J, Gores GJ, Mazzaferro V. Hepatocellular carci-noma: clinical frontiers and perspectives. Gut 2014;63:844–855.

113.Bruix J, Castells A, Bosch J, et al. Surgical resection ofhepatocellular carcinoma in cirrhotic patients: prognosticvalue of preoperative portal pressure. Gastroenterology1996;111:1018–1022.

114.Llovet JM, Fuster J, Bruix J. Intention-to-treat analysis ofsurgical treatment for early hepatocellular carcinoma:resection versus transplantation. Hepatology 1999;30:1434–1440.

115.Ishizawa T, Hasegawa K, Aoki T, et al. Neither multipletumors nor portal hypertension are surgical contraindi-cations for hepatocellular carcinoma. Gastroenterology2008;134:1908–1916.

116.Mazzaferro V, Bhoori S, Sposito C, et al. Milan criteria inliver transplantation for hepatocellular carcinoma: anevidence-based analysis of 15 years of experience. LiverTranspl 2011;17(suppl 2):S44–S57.

117.Mazzaferro V, Regalia E, Doci R, et al. Liver transplantationfor the treatment of small hepatocellular carcinomas inpatients with cirrhosis. N Engl J Med 1996;334:693–699.

118.Prasad KR, Young RS, Burra P, et al. Summary ofcandidate selection and expanded criteria for livertransplantation for hepatocellular carcinoma: a reviewand consensus statement. Liver Transpl 2011;17(suppl 2):S81–S89.

119.Mazzaferro V, Llovet JM, Miceli R, et al. Predicting sur-vival after liver transplantation in patients with hepato-cellular carcinoma beyond the Milan criteria: a

retrospective, exploratory analysis. Lancet Oncol 2009;10:35–43.

120.Raj A, McCall J, Gane E. Validation of the “Metroticket”predictor in a cohort of patients transplanted for pre-dominantly HBV-related hepatocellular carcinoma.J Hepatol 2011;55:1063–1068.

121.Duvoux C, Roudot-Thoraval F, Decaens T, et al. Livertransplantation for hepatocellular carcinoma: a modelincluding alpha-fetoprotein improves the performance ofMilan criteria. Gastroenterology 2012;143:985–986.

122.Toso C, Asthana S, Bigam DL, et al. Reassessing se-lection criteria prior to liver transplantation for hepato-cellular carcinoma utilizing the Scientific Registry ofTransplant Recipients database. Hepatology 2009;49:832–838.

123.Vibert E, Azoulay D, Hoti E, et al. Progression of alpha-fetoprotein before liver transplantation for hepatocellularcarcinoma in cirrhotic patients: a critical factor. Am JTranspl 2010;10:129–137.

124.Duvoux C, Roudot-Thoraval F, Decaens T, et al. Livertransplantation for hepatocellular carcinoma: a modelincluding a-fetoprotein improves the performance ofMilan criteria. Gastroenterology 2012;143:986–994.e3;quiz e14–e15.

125.Pomfret EA, Washburn K, Wald C, et al. Report of anational conference on liver allocation in patients withhepatocellular carcinoma in the United States. LiverTranspl 2010;16:262–278.

126.Clavien PA, Lesurtel M, Bossuyt PM, et al. Recommen-dations for liver transplantation for hepatocellular carci-noma: an international consensus conference report.Lancet Oncol 2012;13:e11–e22.

127.Navasa M, Bruix J. Multifaceted perspective of thewaiting list for liver transplantation: the value of phar-macokinetic models. Hepatology 2010;51:12–15.

128.Toso C, Mazzaferro V, Bruix J, et al. Toward a better livergraft allocation that accounts for candidates with andwithout hepatocellular carcinoma. Am J Transplant 2014;14:2221–2227.

129.Majno PE, Adam R, Bismuth H, et al. Influence of pre-operative transarterial lipiodol chemoembolization onresection and transplantation for hepatocellular carci-noma in patients with cirrhosis. Ann Surg 1997;226:688–701; discussion 701–703.

130.Yao FY, Mehta N, Flemming J, et al. Downstaging ofhepatocellular cancer before liver transplant: long-termoutcome compared to tumors within Milan criteria.Hepatology 2015;61:1968–1977.

131.Ravaioli M, Grazi GL, Piscaglia F, et al. Liver trans-plantation for hepatocellular carcinoma: results of down-staging in patients initially outside the Milan selectioncriteria. Am J Transplant 2008;8:2547–2557.

132.Mazzaferro V, Lencioni R, Majno P, et al. Early hepato-cellular carcinoma on the procrustean bed of ablation,resection, and transplantation. Semin Liver Dis 2014;34:415–426.

133.Cho YK, Kim JK, Kim MY, et al. Systematic review ofrandomized trials for hepatocellular carcinoma treatedwith percutaneous ablation therapies. Hepatology 2009;49:453–459.

April 2016 Diagnosis, Staging, and Treatment of HCC 851

REVIEW

SAN

DPE

RSPE

CTIVES

134.Germani G, Pleguezuelo M, Gurusamy K, et al. Clinicaloutcomes of radiofrequency ablation, percutaneousalcohol and acetic acid injection for hepatocelullar car-cinoma: a meta-analysis. J Hepatol 2010;52:380–388.

135.Peng Z-W, Zhang Y-J, Chen M-S, et al. Radiofrequencyablation with or without transcatheter arterial chemo-embolization in the treatment of hepatocellular carci-noma: a prospective randomized trial. J Clin Oncol 2013;31:426–432.

136.Imamura H, Matsuyama Y, Tanaka E, et al. Risk factorscontributing to early and late phase intrahepatic recur-rence of hepatocellular carcinoma after hepatectomy.J Hepatol 2003;38:200–207.

137.Bruix J,TakayamaT,MazzaferroV, et al. Adjuvant sorafenibfor hepatocellular carcinoma after resection or ablation(STORM): a phase 3, randomised, double-blind, placebo-controlled trial. Lancet Oncol 2015;16:1344–1354.

138.Geissler EK, Schnitzbauer AA, Zülke C, et al. Sirolimususe in liver transplant recipients with hepatocellular car-cinoma: a randomized, multicenter, open-label phase 3trial. Transplantation 2016;100:116–125.

139.Bruix J, Sala M, Llovet JM. Chemoembolization for he-patocellular carcinoma. Gastroenterology 2004;127:S179–S188.

140.Llovet JM, Real MI, Montana X, et al. Arterial embolisationor chemoembolisation versus symptomatic treatment inpatients with unresectable hepatocellular carcinoma: arandomised controlled trial. Lancet 2002;359:1734–1739.

141.Lo C-M, Ngan H, Tso W-K, et al. Randomized controlledtrial of transarterial lipiodol chemoembolization for unre-sectable hepatocellular carcinoma. Hepatology 2002;35:1164–1171.

142.Llovet JM, Bruix J. Systematic review of randomizedtrials for unresectable hepatocellular carcinoma: che-moembolization improves survival. Hepatology 2003;37:429–442.

143.Poon R, Anderson BO, Chen LT, et al. Management ofhepatocellular carcinoma in Asia: consensus statementfrom the Asian Oncology Summit 2009. Lancet Oncol2009;10:1111–1118.

144.Verslype C, Rosmorduc O, Rougier P. Hepatocellularcarcinoma: ESMO-ESDO clinical practice guidelines fordiagnosis, treatment and follow-up. Ann Oncol 2012;23(suppl 7):vii41–vii48.

145.Varela M, Real MI, Burrel M, et al. Chemoembolization ofhepatocellular carcinoma with drug eluting beads: effi-cacy and doxorubicin pharmacokinetics. J Hepatol 2007;46:474–481.

146.Lammer J, Malagari K, Vogl T, et al. Prospective random-ized study of doxorubicin-eluting-bead embolization in thetreatment of hepatocellular carcinoma: results of the PRE-CISION V study. Cardiovasc Interv Radiol 2009;33:41–52.

147.Golfieri R, Giampalma E, Renzulli M, et al. Randomisedcontrolled trial of doxorubicin-eluting beads vs conven-tional chemoembolisation for hepatocellular carcinoma.Br J Cancer 2014;111:255–264.

148.Burrel M, Reig M, Forner A, et al. Survival of patients withhepatocellular carcinoma treated by transarterial che-moembolisation (TACE) using drug eluting beads.

Implications for clinical practice and trial design.J Hepatol 2012;56:1330–1335.

149.Malagari K, Pomoni M, Moschouris H, et al. Chemo-embolization with doxorubicin-eluting beads for unre-sectable hepatocellular carcinoma: five-year survivalanalysis. Cardiovasc Interv Radiol 2012;35:1119–1128.

150.Takayasu K, Arii S, Kudo M, et al. Superselective trans-arterial chemoembolization for hepatocellular carcinoma.Validation of treatment algorithm proposed by Japaneseguidelines. J Hepatol 2012;56:886–892.

151.Sieghart W, Hucke F, Pinter M, et al. The ART of decisionmaking: retreatment with transarterial chemoembolizationin patients with hepatocellular carcinoma. Hepatology2013;57:2261–2273.

152.Kadalayil L, Benini R, Pallan L, et al. A simple prognosticscoring system for patients receiving transarterial em-bolisation for hepatocellular cancer. Ann Oncol 2013;24:2565–2570.

153.Lencioni R, Crocetti L, Petruzzi P, et al. Doxorubicin-eluting bead-enhanced radiofrequency ablation of he-patocellular carcinoma: a pilot clinical study. J Hepatol2008;49:217–222.

154.Poon RT-P, Lau C, Yu W-C, et al. High serum levels ofvascular endothelial growth factor predict poor responseto transarterial chemoembolization in hepatocellularcarcinoma: a prospective study. Oncol Rep 2004;11:1077–1084.

155.Lencioni R, Llovet JM, Han G, et al. Sorafenib or placeboplus TACE with doxorubicin-eluting beads for interme-diate stage HCC: The SPACE trial. J Hepatol 2016 Jan22. pii: S0168-8278(16)00018-0. http://dx.doi.org/10.1016/j.jhep.2016.01.012. [Epub ahead of print].

156.Kudo M, Han G, Finn RS, et al. Brivanib as adjuvanttherapy to transarterial chemoembolization in patientswith hepatocellular carcinoma: a randomized phase IIItrial. Hepatology 2014;60:1697–1707.

157.Torzilli G, Belghiti J, Kokudo N, et al. A snapshot of theeffective indications and results of surgery for hepato-cellular carcinoma in tertiary referral centers: is itadherent to the EASL/AASLD recommendations? Anobservational study of the HCC East-West study group.Ann. Surg 2013;257:929–937.

158.Roayaie S, Jibara G, Tabrizian P, et al. The role of hepaticresection in the treatment of hepatocellular cancer.Hepatology 2015;62:440–451.

159.Wilhelm S, Carter C, Lynch M, et al. Discovery anddevelopment of sorafenib: a multikinase inhibitor fortreating cancer. Nat Rev Drug Discov 2006;5:835–844.

160.Raoul JL, Bruix J, Greten TF, et al. Relationship betweenbaseline hepatic status and outcome, and effect of sor-afenib on liver function: SHARP trial subanalyses. JHepatol 2012;56:1080–1088.

161.Cheng AL, Kang YK, Chen Z, et al. Efficacy and safety ofsorafenib in patients in the Asia-Pacific region withadvanced hepatocellular carcinoma: a phase III rando-mised, double-blind, placebo-controlled trial. LancetOncol 2009;10:25–34.

162.Bruix J, Raoul J-L, Sherman M, et al. Efficacy and safetyof sorafenib in patients with advanced hepatocellular

852 Bruix et al Gastroenterology Vol. 150, No. 4

REVIEWSAND

PERSPECTIVES

carcinoma: subanalyses of a phase III trial. J Hepatol2012;57:821–829.

163.Llovet JM, Peña CEA, Lathia CD, et al. Plasma bio-markers as predictors of outcome in patients withadvanced hepatocellular carcinoma. Clin Cancer Res2012;18:2290–2300.

164.Reig M, Torres F, Rodriguez-Lope C, et al. Early dermato-logic adverse events predict better outcome in HCC pa-tients treated with sorafenib. J Hepatol 2014;61:318–324.

165.Reig M, Gazzola A, Di Donato R, et al. Systemic treat-ment. Best Pract Res Clin Gastroenterol 2014;28:921–935.

166.Reig M, Rimola J, Torres F, et al. Postprogression sur-vival of patients with advanced hepatocellular carcinoma:Rationale for second-line trial design. Hepatology 2013;58:2023–2031.

167.ChengAL,KangYK, LinDY, et al. Sunitinibversus sorafenibin advanced hepatocellular cancer: results of a randomizedphase III trial. J Clin Oncol 2013;31:4067–4075.

168.JohnsonPJ,QinS, Park JW, et al. Brivanib versussorafenibas first-line therapy inpatientswith unresectable, advancedhepatocellular carcinoma: results from the randomizedphase IIIBRISK-FLstudy. JClinOncol 2013;31:3517–3524.

169.Cainap C, Qin S, Huang W-T, et al. Phase III trial of lini-fanib versus sorafenib in patients with advanced hepa-tocellular carcinoma (HCC) (abstr 249). J Clin Oncol2013;31(suppl 4):249.

170.Zhu AX, Rosmorduc O, Evans TRJ, et al. SEARCH: aphase III, randomized, double-blind, placebo-controlledtrial of sorafenib plus erlotinib in patients with advancedhepatocellular carcinoma. J Clin Oncol 2015;33:559–566.

171.Llovet JM, Decaens T, Raoul JL, et al. Brivanib in patientswith advanced hepatocellular carcinoma who wereintolerant to sorafenib or for whom sorafenib failed:results from the randomized phase III BRISK-PS study.J Clin Oncol 2013;31:3509–3516.

172.Zhu AX, Kudo M, Assenat E, et al. Effect of everolimus onsurvival in advanced hepatocellular carcinoma after fail-ure of sorafenib: the EVOLVE-1 randomized clinical trial.JAMA 2014;312:57–67.

173.Zhu AX, Park JO, Ryoo B-Y, et al. Ramucirumab versusplaceboassecond-line treatment inpatientswithadvancedhepatocellular carcinoma following first-line therapy withsorafenib (REACH): a randomised, double-blind, multi-centre, phase 3 trial. Lancet Oncol 2015;16:859–870.

174.Abou-Alfa GK, Schwartz L, Ricci S, et al. Phase II studyof sorafenib in patients with advanced hepatocellularcarcinoma. J Clin Oncol 2006;24:4293–4300.

175.Thomas MB, Chadha R, Glover K, et al. Phase 2 study oferlotinib in patients with unresectable hepatocellularcarcinoma. Cancer 2007;110:1059–1067.

176.Philip PA, Mahoney MR, Allmer C, et al. Phase II study oferlotinib (OSI-774) in patients with advanced hepatocel-lular cancer. J Clin Oncol 2005;23:6657–6663.

177.Toh HC, Chen P-J, Carr BI, et al. Phase 2 trial of linifanib(ABT-869) in patients with unresectable or metastatichepatocellular carcinoma. Cancer 2013;119:380–387.

178.Barone C, Basso M, Biolato M, et al. A phase II study ofsunitinib in advanced hepatocellular carcinoma. Dig LiverDis 2013;45:692–698.

179.Koeberle D, Montemurro M, Samaras P, et al. Contin-uous sunitinib treatment in patients with advanced he-patocellular carcinoma: a Swiss Group for ClinicalCancer Research (SAKK) and Swiss Association for theStudy of the Liver (SASL) multicenter phase II trial (SAKK77/06). Oncologist 2010;15:285–292.

180.Zhu AX, Sahani DV, Duda DG, et al. Efficacy, safety, andpotential biomarkers of sunitinib monotherapy inadvanced hepatocellular carcinoma: a phase II study.J Clin Oncol 2009;27:3027–3035.

181.Faivre S, Raymond E, Boucher E, et al. Safety and effi-cacy of sunitinib in patients with advanced hepatocellularcarcinoma: an open-label, multicentre, phase II study.Lancet Oncol 2009;10:794–800.

182.Park J-W, Finn RS, Kim JS, et al. Phase II, open-labelstudy of brivanib as first-line therapy in patients withadvanced hepatocellular carcinoma. Clin Cancer Res2011;17:1973–1983.

183.Qin SK, Cao MR, Qian J, et al. Clinical study of FOLFOXregimen for the patients with advanced hepatocellu-lar carcinoma [in Chinese]. Chin Clin Oncol 2005;10:58–60.

184.Finn RS, Kang Y-K, Mulcahy M, et al. Phase II, open-label study of brivanib as second-line therapy in pa-tients with advanced hepatocellular carcinoma. ClinCancer Res 2012;18:2090–2098.

185.Reig M, Bruix J. Pattern of tumor progression in livercancer: The missing partner in trial design. Hepatology2015;62:674–676.

186.Lencioni R, Llovet J. Modified RECIST (mRECIST)assessment for hepatocellular carcinoma. Semin LiverDis 2010;30:52–60.

187.Alizadeh AA, Aranda V, Bardelli A, et al. Toward under-standing and exploiting tumor heterogeneity. Nat Med2015;21:846–853.

188.Sangro B, Salem R. Transarterial chemoembolization andradioembolization. Semin Liver Dis 2014;34:435–443.

189.Memon K, Kulik L, Lewandowski RJ, et al. Radiographicresponse to locoregional therapy in hepatocellular car-cinoma predicts patient survival times. Gastroenterology2011;141:526–535; 535.e1–2.

190.Salem R, Lewandowski RJ, Mulcahy MF, et al. Radio-embolization for hepatocellular carcinoma using Yttrium-90 microspheres: a comprehensive report of long-termoutcomes. Gastroenterology 2010;138:52–64.

191.Hilgard P, Hamami M, Fouly AE, et al. Radioembolizationwith yttrium-90 glass microspheres in hepatocellularcarcinoma: European experience on safety and long-term survival. Hepatology 2010;52:1741–1749.

192.Sangro B, Carpanese L, Cianni R, et al. Survival afteryttrium-90 resin microsphere radioembolization of he-patocellular carcinoma across Barcelona clinic livercancer stages: a European evaluation. Hepatology 2011;54:868–878.

193.Salem R, Mazzaferro V, Sangro B. Yttrium 90 radio-embolization for the treatment of hepatocellular carci-noma: biological lessons, current challenges, and clinicalperspectives. Hepatology 2013;58:2188–2197.

194.Sangro B, Gil-Alzugaray B, Rodriguez J, et al. Liver dis-ease induced by radioembolization of liver tumors:

April 2016 Diagnosis, Staging, and Treatment of HCC 853

REVIEW

SAN

DPE

RSPE

CTIVES

description and possible risk factors. Cancer 2008;112:1538–1546.

195.Salem R, Lewandowski RJ, Kulik L, et al. Radio-embolization results in longer time-to-progression andreduced toxicity compared with chemoembolization inpatients with hepatocellular carcinoma. Gastroenter-ology 2011;140:497–507.e2.

196.Mazzaferro V, Sposito C, Bhoori S, et al. Yttrium-90radioembolization for intermediate-advanced hepatocel-lular carcinoma: a phase 2 study. Hepatology 2013;57:1826–1837.

197.El-Khoueiry AB, Melero I, Crocenzi TS, et al. Phase I/IIsafety and antitumor activity of nivolumab in patientswith advanced hepatocellular carcinoma (HCC): CA209-040. J Clin Oncol 2015;33.

198.Sprinzl MF, Galle PR. Immune control in hepatocellularcarcinoma development and progression: role of stromalcells. Semin Liver Dis 2014;34:376–388.

199.Yeo W, Chung HC, Chan SL, et al. Epigenetic therapyusing belinostat for patients with unresectable hepato-cellular carcinoma: a multicenter phase I/II study withbiomarker and pharmacokinetic analysis of tumors frompatients in the Mayo Phase II Consortium and the CancerTherapeutics Res. J Clin Oncol 2012;30:3361–3367.

200.Bruix J, Han K-H, Gores G, et al. Liver cancer:approaching a personalized care. J Hepatol 2015;62:S144–S156.

201.Zhu A, Ryoo R, Yen C, et al. Ramucirumab (RAM) assecond-line treatment in patients (pts) with advancedhepatocellular carcinoma (HCC): analysis of patients with

elevated a-fetoprotein (AFP) from the randomized phaseIII REACH study. J Clin Oncol 2015;33.

202.Santoro A, Rimassa L, Borbath I, et al. Tivantinib forsecond-line treatment of advanced hepatocellular carci-noma: a randomised, placebo-controlled phase 2 study.Lancet Oncol 2013;14:55–63.

203.Heo J, Reid T, Ruo L, et al. Randomized dose-findingclinical trial of oncolytic immunotherapeutic vacciniaJX-594 in liver cancer. Nat Med 2013;19:329–336.

204.Jennerex presents positive clinical data from phase 2 trialof JX-594 in sorafenib-refractory liver cancer patients.September 17, 2012. http://www.prnewswire.com/news-releases/jennerex-presents-positive-clinical-data-from-phase-2-trial-of-jx-594-in-sorafenib-refractory-liver-cancer-patients-170006966.html.

Received May 22, 2015. Accepted December 16, 2015.

Reprint requestsAddress requests for reprints to: Jordi Bruix Hepatic Oncology (BCLC). LiverUnit. Hospital Clínic, CIBERehd. IDIBAPS, University of Barcelona, Villarroel170. 08036 Barcelona. e-mail: [email protected]; fax: þ93 227 5792.

Conflicts of interestThe authors disclose the following: Morris Sherman has received consultinghonoraria from Bayer, Daichii Sankyo, Merck, and Celsion. Jordi Bruix hasconsulted for for Gilead, Abbvie, Kowa, Bayer, BTG, Arqule, Terumo, BMS,Boehringer Ingelheim, Kowa, Novartis, OSI, Roche and Onxeo. Maria Reighas received consulting honorari from Bayer.

FundingCIBEREHD is supported by the Instituto de Salud Carlos III. J.B. and M.R. aresupported by a grant from the Instituto de Salud Carlos III (PI14/00962 andPI15/00145). J.B. from AECC PI044031.