Hepatitis B virus and hepatocellular carcinoma
Jordi Bruix *,Jop M.Llovet
BCLC Group,Liver Unit,Digestive Dia Institute,Hospital Clı
´nic,IDIBAPS,University of Barcelona,Villarroel 170,08036Barcelona,Spain
1.Introduction
The development of hepatocellular carcinoma (HCC)is a major global health problem [1,2].Its incidence has incread worldwide and nowadays it constitutes the 5th most frequent cancer reprenting around 5%of all cancers worldwide [2].More than 500,000new cas are diagnod per year and it reprents the third cau of cancer-related death and the first cau of death amongst cirrhotic patients [3–5].HCC incidence has striking geographical differences [1].In North America,Europe and Australia the age-adjusted incidence is less than 2/100,000per year.This increas in the South European and Mediterranean countries and reaches the highest scores (.50/100,000)in Sub-Sahara Africa and South East Asia [1].This geographic pattern overlaps with the distribution of risk factors and probably also reflects genetic characteristics inherited or acquired through oncogenic agents.While in s
ome high risk areas the incidence has decread as a result of a better health care of the population,the HCC incidence in veral areas such as US and South Europe has incread.This may be due both to an incread dia awareness with a higher diagnostic capability,but also to the emergence of risk factors which up to now would have had a minor dismi-nation and thus,no impact.This is probably the ca for hepatitis C virus (HCV)-related HCC that should reprent the final step of the HCV epidemic that spread in some Western countries after the cond world war,while in Asia it could have appeared decades before.
The most relevant oncogenic agent for HCC development is chronic hepatitis B virus (HBV)or HCV infection [1,6].Their prevalence is known to vary largely among the world population.Not unexpectedly,areas with high prevalence of viral infection have high HCC rates [1].This association allowed the establishment of the relationship between HBV
infection and HCC.Other risk factors such as alcoholism or hemochromatosis are also relevant,but have not yet reached the epidemiological and clinical importance of viral infec-tion.In most cas,HCC development complicates an underlying chronic liver dia,but in all etiological categories the tumor may occur in the abnce of cirrhosis or when minimal histologic changes have taken place.
The prent chapter reviews the evidence that links HCC development with HBV infection,summarizes the mech-anisms that may be involved in the HBV carcinogenic pathway and finally provides the guidelines to achieve early diagnosis and effective therapy.
2.Epidemiological relationship between HBV infection and HCC
Data linking HBV with HCC were raid soon after the discovery of the ‘Australia Antigen’(corresponding to the hepatitis B surface antigen –HBsAg).Several epidemio-logical surveys demonstrate the geographic coincidence between HBV endemicity and HCC prevalence [7].In South-East Asia the prevalence of HBV infection reaches 10–15%and the HCC rate ranges between 10and 25/100,000population [1,8,9].In all areas the prevalence of rological markers of past or prent HBV infection is higher in HCC patients than in tho without this tumor.Figures may vary according to the background rate in the global population and may depend on the techniques employed to asss the existence of viral infection.HBV DNA might be found in the tumor and non-tumor liver tissue of HCC patients in whom rological markers may be negative or only positive for anti-HBc [10].Accordingly,both in low and high incidence areas,the role of HBV in HCC development is very relevant.In addition to cross-ctional studies,cohort studies have established that chronic HBV infection carries an incread HCC risk and thus,of cancer related death [6,8].Again,big differen
ces
0168-8278/03/$30.00q 2003European Association for the Study of the Liver.Published by Elvier B.V.All rights rerved.
doi:10.1016/S0168-8278(03)00140-5
Journal of Hepatology 39(2003)S59–S63
/locate/jhep
*Corresponding author.Fax:þ34-4-227-9803.E-mail address:jbruix@clinic.ub.es (J.Bruix).
have been obrved depending on the geographic area and the prence or abnce of an underlying liver dia.In the minal study by Beasley et al.including more than22,000 Taiwane men the risk in chronic HBV carriers was100 times higher than in non-infected individuals[11].A Japane study performed also in males found afifty-fold incread risk in HBV infected individuals[12].Studies performed in Alaska[13],Europe[3,14]and Canada[15] have established that the incread risk applies world-wide. However,while in Asian patients the tumor appears more frequently in an otherwi normal liver,in the Caucasian population HCCs mostly develop in patients with chronic liver dia,namely cirrhosis.Patients with HCC in a normal liver are usually younger and this may support the role of age at infection.The coexistence of additional oncogenic agents in tho areas where the infection is acquired perinatally or during childhood may also be rele-vant.This might be the ca for aflatoxin,a carcinogen that contaminates food stored in humid conditions[16].It induces a unique misn mutation of codon249of the p53 gene[17].While this is o
brved at early HCC stages in Asian patients,this genetic signature is rarely obrved in Western tumors.The annual HCC incidence in chronic HBV carriers in Asia ranges between0.4and0.6%.This figure is lower in Alaskan natives(0.26%/year)and even lower in Caucasian HBV carriers.The incidence in cirrhotics exceeds2%in all areas.Male x,advanced age (a potential surrogate for the dia duration)and incread alphafetoprotein(AFP)are associated with higher risk[6].
Thefinal demonstration of the role of HBV in the development of HCC comes from the preventive effect of vaccination.In Taiwan this has prompted a reduction of the prevalence of HBV infection in children from15to1% and also a60%reduction of HCC during childhood[18]. Considering that more than70%of HCC in developing countries is attributable to HBV,the current challenge is to ensure the availability of the vaccine in all the countries. Other preventive measures should be addresd at public health campaigns to reduce food contamination by fungi, as well as to modulate the metabolism of aflatoxins once ingested,as has been propod by the u of oltipraz.Dosage and schedule parameters for its administration have been described in a pha II study in Qidong,China[19,20],where HCC is the leading cau of death and exposure to dietary aflatoxins is widespread.However,long-term outcome of randomized controlled trials(RCTs)are not yet available.
3.Oncogenic mechanisms
The oncogenic mechanisms of HBV are not fully understood[9,21,22].HBV DNA may integrate into the genome of the hepatocytes[23]and this may cau veral hits leading to transformation.However,HBV DNA integration is random and involves always different parts of the viral genome.The viral X gene may have an important role as it is a powerful transactivator for transcription of oncogenes such as c-myc and c-jun.Additional events that may be triggered by HBV DNA integration and X gene expression deregulation include DNA repair impairment, incread cytokine production and growth factors,inhibition of apoptosis and incread expression of angiogenic agents such as vascular endothelial growth factor and nitric oxide [9,21].This last agent may act as a potent mutagen[24]and can be induced both by HBV infection or the inflammation associated to chronic hepatitis.
Recent human studies suggest that HBV positive HCC are characterized by higher chromosomal instability leading to loss of heterozygosity as compared to tumors related to HCV[25].However,there is no clear-cut HBV related genetic profile leading to HCC[26].
As mentioned above,in most cas HCCs appear in a cirrhotic liver after many years of inflammation,cellular necrosis and regeneration that define a microenvironment that is characterized by the relea of veral cytokines, growth factors and other mediators.This may affect the hepatocy
tes,promote oxidative stress and prompt the appearance of a critical genetic damage irrespective of the virus itlf.
4.Early detection and diagnostic confirmation
Mortality from HCC may decrea due to effective implementation of preventive strategies and,less impor-tantly,early detection and effective treatment.Prevention will be achieved by vaccination and by drugs that eradicate the virus prior to irreversible hepatic and/or genetic damage. In patients with chronic hepatitis B,interferon-a lamivu-dine,adefovir and other antiviral agents will reduce the incidence of cirrhosis,and also of HCC.If cirrhosis is established there is no agent that has been proven to diminish the HCC risk[6,27,28].Therefore,in this tting the only option to decrea cancer related deaths will be early detection.
Early detection might be achieved by surveillance of the population at risk and its cost-effectiveness will depend on successful therapy.Thus,careful lection of patients for surveillance is mandatory.HCC meets most of the accepted criteria to establish surveillance[29].It is a frequent event in the population at risk.This can be identified and the dia caus significant morbidity or mortality.The population should accept the surveillance strategy and this has to have a high detectio
n and diagnostic accuracy.Abnormal results have to be clearly defined and the investigations necessary must be well established.Finally,effective therapy should be available.
There are no RCTs demonstrating a survival benefit of surveillance.The available evidence is derived from cohort studies showing that the proportion of patients diagnod at an early stage,and thus amenable for effective treatment, is incread as compared to patients not included into
J.Bruix,J.M.Llovet/Journal of Hepatology39(2003)S59–S63 S60
surveillance[5,6].Cost-effectiveness has been assd from cohort studies and multistage transition modelling [5,30,31].An adequate relationship is achieved if surveil-lance is performed in patients with significant risk and if all treatments with positive impact are available and properly indicated to achieve the best possible outcome.
Bad on the incidence of HCC in patients with HBV cirrhosis,surveillance is indicated irrespective of the racial background.In non-cirrhotics the evidence is controversial. The0.4–0.6%yearly risk of Asian HBV carriers justifies their surveillance.The very low risk in Caucasians, however,results in a low number of detected HCCs and a reduced cost-effectiveness[6].
It is important to stress that only tho individuals who could be treated should enter the program while tho who are not candidates for curative therapy should be excluded.
Therefore,the ideal target population are patients with normal liver function or with Child-Pugh A cirrhosis.Child-Pugh C cirrhotics should be evaluated for liver transplan-tation in the abnce of an HCC.If this option is not feasible,there will be no benefit from surveillance.Child-Pugh B cirrhotics are a controversial group.Without liver transplantation it is very unlikely that any therapy will offer a significant survival benefit and hence,they should probably not become a target group for surveillance.
4.1.Surveillance for HCC
The growth of small HCC from undetectable to2cm takes about4–12months[6].Aiming to detect tumors smaller than3cm in diameter,a6-month interval has been t for surveillance.Patients with incread risk do not require shorter intervals,becau incread risk does not mean faster growth.The surveillance tools are AFP and ultrasonography(US).AFP is not a very good screening test.Its nsitivity is39–64%,its specificity 76–91%and its positive predictive value9–32%[6,15, 32,33].AFP as the only tool for surveillance has not been properly investigated and this has maintain
ed it in most surveillance strategies.US is a more effective tool.As a screening test in HBsAg carriers,it has a nsitivity of 71%and a specificity of93%,with a reduced positive predictive value of14%[15].US should be conducted by an experienced operator.If US or AFP are abnormal,the patients should be referred to a comprehensive cancer unit offering the most updated human and technical resources[6].
4.2.Recall procedures and diagnostic confirmation
This is an area of clinical rearch.There are no pros-pective studies asssing the efficacy of different strategies. In the recent EASL Monothematic Conference on HCC a panel of experts propod a t of guidelines depicted in Fig.1[6].The US discovery of a hypo-or hyperechoic nodule during follow-up should rai a HCC suspicion,depending on size.Pathological studies have shown that 50%of nodules,1cm are not malignant[34]and the available diagnostic techniques do not allow proper classification.Therefore,US control of the nodules every 3months is recommended until disappearance or progres-sive growth to.1cm.Even a true HCC may remain stable for more than1year and thus,abnce of growth during the follow-up period does not exclude malignancy.For lesions.1cm HCCs are more likely and diagnostic confirmation is required.A positivefine needle biopsy (tissue sample for cytology and histology)was felt mandatory for nodules,2cm becau the im
aging techniques are not sufficiently accurate.By contrast,in cirrhotic patients,the incidentalfinding of a hepatic nodule .2cm showing arterial hypervascularization in at least two imaging techniques(US,CT,angiography and magnetic resonance imaging)may suffice to establish the diagnosis(Table1)[34].In fact,with the characteristics, a negative biopsy will not rule out an HCC and thus,this result will not change the clinical diagnosis.AFP.400ng/ ml may also establish the diagnosis,if coinciding with a highly suspicious hepatic nodule in a cirrhotic liver[6,33]. However,transient AFP increas might be obrved coinciding with inflammatoryflares[35]and thus,care should be taken to properly characterize the hepatic nodules through imaging techniques.In any ca,the request for a diagnostic biopsy should take into account the impact of the result on the clinical decision making and the balance between the potential risks of biopsy
and Fig.1.Surveillance and recall strategy[6].(Reproduced from Bruix et al.[6],with permission.)
Table1
Diagnostic criteria for HCC[6]
X Cyto-histological criteria
X Non-invasive criteria(cirrhotic patients)
1.Radiological criteria:two coincidental imaging techniques a
–Focal lesion.2cm with arterial hypervascularization
2.Combined criteria:one imaging technique associated to AFP
–Focal lesion.2cm with arterial hypervascularization
–AFP levels.400ng/ml
a Four techniques considered:US,Spiral CT,MRI and angiography.
J.Bruix,J.M.Llovet/Journal of Hepatology39(2003)S59–S63S61
the risk of invasive ansplantation)to be applied thereafter.
There are no robust data to define the recall policy for patients who prent an incread AFP value.In the abnce of a coincidental nodule on US a triple pha CT scan may be needed to rule out an HCC.If negative,a persistently incread AFP classifies the patient as a high risk individual for HCC development[6].
Applying this surveillance strategy40–80%of the detected HCC are solitary,but only half of them are suitable for curative treatment after careful evaluation[5,6].
5.Treatment of HCC
Patients diagnod with early stage HCC benefit from radical options(surgery,percutaneous ablation)that provide a high rate of complete respons according to oncologic criteria[6,36,37].An untreated control arm to test the impact of the options is not ethical and the survival benefits from treatment have been derived from cohort studies.By contrast,the abnce of effective therapies for advanced HCC has allowed the conduction of veral RCTs comparing some of the available alternatives versus no treatment[6,36].
5.1.Treatment of early HCC
There are no studies comparing the available curative treatment options–rection,liver transplantation and percutaneous ablation–against no treatment or amongst themlves.Thus,there are no data to establish thefirst-line treatment in patients with HCC in compensated cirrhosis. Each group,center and country should establish the pre-ferred schedule taking into account the results of cohort studies and the available resources[6,36].However,in patients with early HCC in decompensated liver dia,the outcome after transplantation is clearly superior to other alternatives.
Surgical rection is considered for single asymptomatic HCC in patients with prerved liver function.The Child-Pugh classification is not adequate to lect surgical candi-dates,since even Child-Pugh A may include relevant functional impairment and minor ascites.Indocyanine green retention rate or the lection of patients with normal bilirubin and no portal hypertension identify the candidates that will achieve a70%5-year survival[36,38].Vascular invasion,satellites and poor differentiation are associated with an incread risk of recurrence of up to more than70% within5years and no effective prevention of recurrence [36–38].
The optimal candidates for transplantation are patients with early HCC(single lesion,5cm or up to three lesions ,3cm)with a70%5-year survival and a recurrence rate below15%[36,37].Transplantation
is not available world-wide and there is a vere shortage of donors.Antitumoral treatments upon listing may benefit some patients,but there are no RCTs to prove their efficacy.One of the major risks after transplantation is the re-infection of the graft.Hyper-immune globulin and new antiviral agents discusd el-where in this issue markedly reduce this risk and allow optimal graft and patient survival.
Percutaneous ethanol injection(PEI)under US guidance achieves a complete respon in70–80%of solitary tumors ,3cm.Child-Pugh A cirrhotics may achieve a50%5-year survival.By contrast,survival benefits of PEI in Child-Pugh B class patients are controversial.PEI is inexpensive and carries almost no treatment related mortality.Hence,it constitutes the standard technique against which new options have to be compared[36–38].
5.2.Treatment of intermediate/advanced HCC
The survival benefits of palliative options are highly controversial due to the low number of RCTs.Chemo-embolization has been extensively evaluated[36,37]and recent Western[39]and Eastern[40]investigations have provided positivefindings that support a beneficial effect. Tamoxifen administration has no efficacy[36]and other options such as systemic chemotherapy,octreotide,intern
al radiation,proton beam radiation,antiandrogens and immunotherapy have been assd in single investigations including a limited number of patients and/or lacking an untreated control arm.Unfortunately,some of them lack any antitumoral effect and/or are associated to relevant toxicity [36,37].
In summary,HCC is a major complication of chronic HBV infection reprenting one of the most frequent caus of death.Universal vaccination will eventually eradicate this infection and thus,diminish the incidence of liver cancer.Effective antiviral therapy may prevent cirrhosis and ultimately cancer.However,if cirrhosis is already estab-lished,the only chance for long-term HCC cure will come from early detection and effective treatment. Acknowledgements
Dr.Llovet is a recipient of a contract from Programa “Ramon y Cajal”(Ministerio de Ciencia y Tecnologı´a). References
[1]Bosch FX,Ribes J,Borras J.Epidemiology of primary liver cancer.
Semin Liver Dis1999;19:271–85.
[2]Parkin DM,Bray F,Ferlay J,Pisani P.Estimating the world cancer
burden:Globocan2000.Int J Cancer2001;94:153–6.
[3]Fattovich G,Giustina G,Schalm SW,Hadziyannis S,Sanchez-Tapias
J,Almasio P,et al.Occurrence of hepatocellular carcinoma and decompensation in western European patients with cirrhosis type
B.The EUROHEP Study Group on Hepatitis B Virus and Cirrhosis.
Hepatology1995;21:77–82.
J.Bruix,J.M.Llovet/Journal of Hepatology39(2003)S59–S63 S62
[4]Degos F,Christidis C,Ganne-Carrie N,Farmachidi JP,Degott C,
Guettier C,et al.Hepatitis C virus related cirrhosis:time to occurrence of hepatocellular carcinoma and death.Gut2000;47:131–6.
[5]Bolondi L,Sofia S,Siringo S,Gaiani S,Casali A,Zironi G,et al.
Surveillance programme of cirrhotic patients for early diagnosis and treatment of hepatocellular carci
noma:a cost-effectiveness analysis.
Gut2001;48:251–9.
[6]Bruix J,Sherman M,Llovet JM,Beaugrand M,Lencioni R,
Burroughs AK,et al.Clinical Management of Hepatocellular carcinoma:conclusions of the Barcelona-2000EASL Conference.
J Hepatol2001;35:421–30.
[7]Beasley RP.Hepatitis B virus as the etiologic agent in hepatocellular
carcinoma.Hepatology1982;2:21–26s.
[8]Chang MH,Chen DS.Prospects for hepatitis B virus eradication and
control of hepatocellular carcinoma.Baillieres Best Pract Res Clin Gastroenterol1999;13:511–7.
[9]Okuda K.Hepatocellular carcinoma.J Hepatol2000;32:225–37.
[10]Brechot C,Thiers V,Kremsdorf D,Nalpas B,Pol S,Paterlini-Brechot
P.Persistent hepatitis B virus infection in subjects without hepatitis B surface antigen:clinically significant or purely‘occult’?Hepatology 2001;34:194–203.
[11]Beasley RP,Hwang LY,Lin CC,Chien CS.Hepatocellular carcinoma
and hepatitis B virus.A prospective study of22,707men in Taiwan.
Lancet1981;2:1129–33.
[12]Sakuma K,Saitoh N,Kasai M,Jitsukawa H,Yoshino I,Yamaguchi
M,et al.Relative risks of death due to liver dia among Japane male adults having various status for hepatitis B s and e antigen/ antibody in rum:a prospective study.Hepatology1988;8:1642–6.
[13]McMahon BJ,Bulkow L,Harpster A,Snowball M,Lanier A,Sacco F,
et al.Screening for hepatocellular carcinoma in Alaska natives infected with chronic hepatitis B:a16-year population-bad study.
Hepatology2000;32:842–6.
[14]Fattovich G,Brollo L,Giustina G,Noventa F,Pontisso P,Alberti A,
et al.Natural history and prognostic factors for chronic hepatitis type
B.Gut1991;32:294–8.
[15]Sherman M,Peltekian KM,Lee C.Screening for hepatocellular
carcinoma in chronic carriers of hepatitis B virus:incidence and prevalence of hepatocellular carcinoma in a North American urban population.Hepatology1995;22:432–8.
[16]Jackson PE,Groopman JD.Aflatoxin and liver cancer.Baillieres Best
Pract Res Clin Gastroenterol1999;13:545–55.
[17]Ozturk M.p53mutation in hepatocellular carcinoma after aflatoxin
exposure.Lancet1991;338:1356–9.
[18]Chang MH,Chen CJ,Lai MS,Hsu HM,Wu TC,Kong MS,et al.
Universal hepatitis B vaccination in Taiwan and the incidence of hepatocellular carcinoma in children.N Engl J Med1997;336: 1855–9.
[19]Kensler TW,He X,Otieno M,Egner PA,Jacobson LP,Chen B,et al.
Oltipraz chemoprevention trial in Qidong,People’s Republic of China:modulation of rum aflatoxin albumin adduct biomarkers.
Cancer Epidemiol Biomarkers Prev1998;7:127–34.
[20]Wang JS,Shen X,He X,Zhu YR,Zhang BC,Wang JB,et al.
Protective alterations in pha1and2metabolism of aflatoxin B1by oltipraz in residents of Qidong,People’s Republic of China.J Natl Cancer Inst1999;91:347–54.
[21]Brechot C,Gozuacik D,Murakami Y,Paterlini-Brechot P.Molecular
bas for the development of hepatitis B virus(HBV)-related hepatocellular carcinoma(HCC).Semin Cancer Biol2000;10: 211–31.[22]Chen PJ,Chen DS.Hepatitis B virus infection and hepatocellular
carcinoma:molecular genetics and clinical perspectives.Semin Liver Dis1999;19:253–62.
[23]Shafritz DA,Shouval D,Sherman HI,Hadziyannis SJ,Kew MC.
Integration of hepatitis B virus DNA into the genome of liver cells in chronic liver dia and hepatocellular carcinoma.Studies in percutaneous liver biopsies and post-mortem tissue specimens.N Engl J Med1981;305:1067–73.
[24]Halliwell B.Oxygen and nitrogen are pro-carcinogens.Damage to
DNA by reactive oxygen,chlorine and nitrogen species:measure-ment,mechanism and the effects of nutrition.Mutat Res1999;443: 37–52.
[25]Laurent-Puig P,Legoix P,Bluteau O,Belghiti J,Franco D,Binot F,
et al.Genetic alterations associated with hepatocellular carcinomas define distinct pathways of hepatocarcinogenesis.Gastroenterology 2001;120:1763–73.
[26]Buendia MA.Genetics of hepatocellular carcinoma.Semin Cancer
Biol2000;10:185–200.
[27]Baffis V,Shrier I,Sherker AH,Szilagyi A.U of interferon for
prevention of hepatocellular carcinoma in cirrhotic patients with hepatitis B or hepatitis C virus infection.Ann Intern Med1999;131: 696–701.
[28]Camma C,Giunta M,Andreone P,Craxi A.Interferon and prevention
of hepatocellular carcinoma in viral cirrhosis:an evidence-bad approach.J Hepatol2001;34:593–602.
[29]Prorok PC.Epidemiologic approach for cancer screening.Problems in
design and analysis of trials.Am J Pediatr Hematol Oncol1992;14: 117–28.
[30]Sarasin FP,Giostra E,Hadengue A.Cost-effectiveness of screening
for detection of small hepatocellular carcinoma in western patients with Child-Pugh class A cirrhosis.Am J Med1996;101:422–34. [31]Collier J,Sherman M.Screening for hepatocellular carcinoma.
Hepatology1998;27:273–8.
[32]Sherman M.Alphafetoprotein:an obituary.J Hepatol2001;34:603–5.
[33]Trevisani F,D’Intino PE,Morlli-Labate AM,Mazzella G,Accogli
E,Caraceni P,et al.Serum alpha-fetoprotein for diagnosis of hepatocellular carcinoma in patients with chronic liver dia: influence of HBsAg and anti-HCV status.J Hepatol2001;34:570–5.
[34]Nakashima T,Kojiro M.Pathologic characteristics of hepatocellular
carcinoma.Semin Liver Dis1986;6:259–66.
[35]Lok AS,Lai CL.Alpha-Fetoprotein monitoring in Chine patients
with chronic hepatitis B virus infection:role in the early detection of hepatocellular carcinoma.Hepatology1989;9:110–5.
[36]Bruix J,Llovet JM.Prognostic prediction and treatment strategy in
HCC.Hepatology2002;35:519–24.
[37]Befeler AS,Di Bisceglie AM.Hepatocellular carcinoma:diagnosis
and treatment.Gastroenterology2002;122:1609–19.
[38]Arii S,Yamaoka Y,Futagawa S,Inoue K,Kobayashi K,Kojiro M,
et al.Results of surgical and non-surgical treatment for small-sized hepatocellular carcinomas:a retrospective and nationwide survey in Japan.Hepatology2000;32:1224–9.
[39]Llovet JM,Real MI,Montanya X,Planas R,Coll S,Aponte AJ,et al.
Arterial embolization,chemoembolization versus symptomatic treat-ment in patients with unrectable hepatocellular carcinoma:a randomized controlled trial.Lancet2002;359:1734–9.
[40]Lo CM,Ngan H,Tso WK,Liu CL,Lam CM,Poon RT,et al.
Randomized controlled trial of transarterial lipiodol chemoemboliz-ation for unrectable hepatocellular carcinoma.Hepatology2002;35: 1164–71.
J.Bruix,J.M.Llovet/Journal of Hepatology39(2003)S59–S63S63
