Vous êtes sur la page 1sur 23

A MERICAN A SSOCIATION FOR

T HE STUDY OF LIVER D I S E ASES

PRACTICE GUIDELINE | HEPATOLOGY, VOL. 67, NO. 1, 2018

AASLD Guidelines for the Treatment
of Hepatocellular Carcinoma
Julie K. Heimbach,1 Laura M. Kulik,2 Richard S. Finn,3 Claude B. Sirlin,4 Michael M. Abecassis,5 Lewis R. Roberts,6
Andrew X. Zhu,7 M. Hassan Murad,8 and Jorge A. Marrero9

aspects of HCC that lacked sufficient evidence to war-
Guiding Principles and rant a systematic review.
Objectives The guideline focuses on a broad spectrum of clinical
practice, including surveillance of patients with cirrhosis
GUIDING PRINCIPLES for HCC, establishing the diagnosis of HCC, and vari-
ous therapeutic options for the treatment of HCC. To
This document presents official recommendations address other issues on HCC such as epidemiology, stag-
of the American Association for the Study of Liver ing, and additional aspects of diagnosis and treatment,
Diseases (AASLD) on the surveillance, diagnosis, and the authors have created a new guidance document that
treatment of hepatocellular carcinoma (HCC) occur- will be published soon and is based upon the previous
ring in the setting of adults with cirrhosis. Unlike pre- HCC AASLD guidelines by Bruix and Sherman.(2)
vious AASLD practice guidelines, the current
guideline was developed in compliance with the Insti-
tute of Medicine standards for trustworthy practice KEY QUESTIONS
guidelines and uses the Grading of Recommendation
Assessment, Development and Evaluation (GRADE) The guideline developers from the AASLD identi-
approach.(1) Multiple systematic reviews of the litera- fied key questions that health care providers are faced
ture were conducted to support the recommendations with frequently in the evaluation and management of
in this practice guideline. An enhanced understanding patients with HCC. These questions were:
of the guideline can be obtained by reading the appli- 1. Should adults with cirrhosis undergo surveillance
cable portions of the systematic reviews. In addition, for HCC? If so, which surveillance test is best?
more detailed information may be found in the associ- 2. Should adults with cirrhosis and suspected HCC
ated guidance document related to clinically important undergo diagnostic evaluation with multiphasic

Abbreviations: AASLD, American Association for the Study of Liver Diseases; AFP, alpha-fetoprotein; CI, confidence interval; CT, computed tomog-
raphy; DEB-TACE, drug-eluting beads TACE; GRADE, Grading of Recommendation Assessment, Development and Evaluation; HAIC, hepatic
arterial infusion chemotherapy; HCC, hepatocellular carcinoma; HBV, hepatitis B virus; HCV, hepatitis C virus; HR, hazard ratio; LRT, local-
regional therapy; MELD, Model for End-Stage Liver Disease; mRECIST, modified Response Evaluation Criteria in Solid Tumors; MRI, magnetic
resonance imaging; NAFLD, nonalcoholic fatty liver disease; OPTN, Organ Procurement and Transplantation Network; OR, odds ratio; OS, overall
survival; PEI, percutaneous ethanol injection; PVT, portal vein thrombosis; RCT, randomized controlled trial; RFA, radiofrequency ablation; RR, rel-
ative risk; TACE, transarterial chemoembolization; TACI, transarterial chemoinfusion; TAE, transarterial embolization; TARE, transarterial radio-
embolization; US, ultrasound; Y90, yttrium-90.
The funding for the development of this Practice Guideline was provided by the American Association for the Study of Liver Diseases.
Additional Supporting Information may be found at onlinelibrary.wiley.com/doi/10.1002/hep.29086/suppinfo.
Received January 10, 2017; accepted January 10, 2017.
Copyright VC 2017 by the American Association for the Study of Liver Diseases.

View this article online at wileyonlinelibrary.com.
DOI 10.1002/hep.29086
Potential conflict of interest: Laura M. Kulik is on the advisory board for Gilead, Bayer, Eisai, Salix, and Bristol-Myers Squibb. Richard Finn
consults for Pfizer, Bayer, Novartis, Merck, and Bristol-Myers Squibb. Claude B. Sirlin consults for and has received grants from Virtualscopics. Lewis
R. Roberts consults for Wako, Medscape, and Axis; advises Tavec and Bayer; is on the speakers’ bureau for Onlive; and has received grants from Ariad,
BTG, and Gilead. Andrew Zhu consults for Bristol-Myers Squibb, Eisai, Merck, Novartis, Sanofi, and Bayer.

358

HEPATOLOGY, Vol. 67, No. 1, 2018 HEIMBACH ET AL.

computed tomography (CT) or multiphasic mag- TARGET AUDIENCE
netic resonance imaging (MRI)?
3. Should adults with cirrhosis and an indetermi- This guideline is intended primarily for health care
nate hepatic nodule undergo a biopsy, repeated providers who care for patients with cirrhosis. Addi-
imaging, or alternative imaging for the diagnostic tionally, the guideline may inform policy decisions
evaluation? regarding patients with HCC.
4. Should adults with Child-Pugh class A cirrhosis
and early-stage HCC (T1 or T2) be treated with
resection or local-regional (LRT) therapy?
Background
5. Should adults with cirrhosis and HCC that has BURDEN OF DISEASE
been resected or ablated successfully undergo
adjuvant therapy? According to the World Health Organization,
6. Should adults with cirrhosis awaiting liver trans- HCC is the fifth most common tumor worldwide and
plantation and HCC (T1) be treated or undergo the second most common cause of cancer-related death
observation? (http://globocan.iarc.fr/old/FactSheets/cancers/liver-new.
7. Should adults with cirrhosis and HCC (Organ asp). Male-to-female predominance is greater than 2:1
Procurement and Transplantation Network with liver cancer, and approximately 83% of the estimat-
[OPTN] T2) awaiting liver transplantation ed 782,000 new HCC cases in 2012 occurred in less
undergo transplantation alone or transplantation developed regions of the world, with East and South
with bridging therapy while waiting? Asia plus sub-Saharan Africa being the regions of high-
8. Should adults with cirrhosis awaiting liver trans- est incidence, Southern Europe and North America
plantation and HCC beyond Milan criteria (T3) being the regions of intermediate incidence, and North-
undergo transplantation after being down-staged ern Europe and South Central Asia being the regions of
to within Milan criteria? lowest incidence.(3)
9. Should adults with cirrhosis and HCC (T2 or The incidence of HCC has been rising rapidly in the
T3, no vascular involvement) who are not can- United States over the last 20 years.(4) According to esti-
didates for resection or transplantation be mates from the Surveillance Epidemiology End Results
treated with transarterial chemoembolization, (SEER) program of the National Cancer Institute, the
transarterial radioembolization, or external United States will have witnessed an estimated 39,230
radiation? cases of HCC and 27,170 HCC deaths in 2016 (https://
10. Should adults with Child-Pugh class A/B cir- seer.cancer.gov). In addition, a recent study using the
rhosis and advanced HCC with macrovascular SEER registry projects that the incidence of HCC will
invasion and/or metastatic disease be treated continue to rise until 2030,(5) with the highest increase in
with systemic or locoregional therapies or no Hispanics, followed by African Americans and then Cau-
therapy? casians, with a decrease noted among Asian Americans.

ARTICLE INFORMATION:
From the 1Division of Transplant Surgery, William J. von Liebig Transplant Center, Mayo Clinic, Rochester, MN; 2Department of Medi-
cine, Division of Gastroenterology and Hepatology, Northwestern University, Chicago, IL; 3Department of Medicine, Division of Hema-
tology and Oncology, David Geffen School of Medicine at the University of California, Los Angeles, Santa Monica Geffen School of
Medicine at UCLA, Los Angeles, California; 4Liver Imaging Group, Department of Radiology, University of California, San Diego;
5
Northwestern University Feinberg School of Medicine, Chicago, IL; 6Division of Gastroenterology and Hepatology, Mayo Clinic, Roches-
ter, MN; 7Massachusetts General Hospital Cancer Center, Harvard Medical School, Boston, MA; 8Mayo Clinic Evidence-based Practice
Center, Mayo Clinic, Rochester, MN; 9Digestive and Liver Diseases Division, Department of Internal Medicine, UT Southwestern Medi-
cal Center, Dallas, TX.

ADDRESS CORRESPONDENCE AND REPRINT REQUESTS TO:
Julie K. Heimbach, M.D. 200 First Street SW
Mayo Clinic Rochester, MN 55905-0001
William J. von Liebig Transplant Center E-mail: heimbach.julie@mayo.edu
Charlton 10 Tel.: 507-266-6640

359

Policy makers: There is a need for substantial debate and involvement of stakeholders. Liver Disease (MELD) classification. prompting Petrick et al. HIGH-RISK GROUP The presence of cirrhosis represents a key risk factor Methods of Guideline for the development of HCC. the recommendation is prefaced by “The AASLD recommends. Health care workers: Be prepared to help patients make a decision that is consistent with their values using decision aids and shared decision making. For conditional recommendations. Determinants of the strength of a recommendation Quality of evidence Balance of benefit and harms Patient values and preferences Resources and costs 3. have also shown that metabolic disorders—defined as The key questions posed above reflect common sce- nonalcoholic fatty liver disease (NAFLD) and the meta. and burdens is uncertain.. For strong recommendations. January 2018 TABLE 1. For policy makers. Recent data and therapeutic options substantially. this recommen- dation type could imply a need for substantial debate and involvement of all stakeholders and is likely insufficient to be used as a qual- ity measure. patients with cirrhosis attributed primarily to the hepatitis C virus (HCV) epi.. the balance of benefits. We used the Child- den of HCC than any other risk factor including HCV Pugh classification to define the underlying degree of infection.” In contrast. the recommendation is prefaced by “The AASLD suggests. For clinicians. Implication of the strength of a recommendation Strong Population: Most people in this situation would want the recommended course of action and only a small proportion would not. harms.8) and the HCC incidence rate An experienced methodologist moderated and facil- among patients with cirrhosis has been shown to be itated the process of selecting the aforementioned key 360 . this would imply that patients should receive the recommended course of action.. the recommendation could be adopted as a policy in most situations and potentially could be used as a quality measure. work for this practice guideline. narios in this patient population and provide the frame- bolic syndrome—contribute numerically more to the bur.(9) Therefore. but many would not. RR 5 0.g. For patients. For policy makers. but many would not. Policy makers: The recommendation can be adapted as policy in most situations. The prevalence of cir- rhosis among patients with HCC has been estimated Development to be 85%-95%. HEPATOLOGY. mainly because it is more commonly used in this context.(7.. a strong recommendation implies that most patients in this situation would want the recommended course of action and only a small proportion would not. with consis- tent benefits and few side effects. a conditional recommendation (also sometimes termed a “weak” recommendation) for patients would imply that the majority of patients in this situation would want the recommended course of action. The fact that patients with HCC tive efforts should target the birth cohort with the highest have underlying liver disease impacts the management prevalence of HCV infection (1945-1965). constitute a high-risk group for efforts at prevention demic.(4) to suggest that preven.” The increase in incidence of HCC in the United States is 2%-4% per year.(6) which is due primarily to the high prevalence liver dysfunction instead of the Model for End-Stage of NAFLD in the population overall.g. Conditional Population: The majority of people in this situation would want the recommended course of action.HEIMBACH ET AL.. Health care workers: Most people should receive the recommended course of action. and early detection.. Rating the quality of evidence Study design Initial rating of quality of evidence Rate down when: Rate up when: RCT High Risk of bias Large effect (e. For clinicians making a conditional recommendation. RR 5 0. and they should be prepared to help patients make a decision that is consistent with their own values using a shared decision-making approach.5) Moderate Inconsistency Very large effect (e. The GRADE Approach 1.2) Imprecision Dose response gradient Observational Low Indirectness All plausible confounding would increase the association Very low Publication bias 2.

the questions with sparse. No. recurrence. AND IF benefits and harms. TABLE 2. Based on this assess- ment.(1) In this approach. Technical remarks are added to every 6 months. A group of AASLD content experts worked recommendations to help reconcile the level of the rec- collaboratively with an independent research group ommendation with the quality of the evidence and to specializing in conducting systematic reviews to syn. and risk of bias. The AASLD suggests surveillance using ultra- play a role in determining the strength of recommen. moderate. sound (US). For the key GRADE approach (Table 1). the guideline writing group based its recom. no vascular involvement) who chemoembolization or external radiation are not candidates for resection or transplantation 10 Adults with Child-Pugh class A/B cir. Evidence profiles for the thesize the available evidence. SHOULD ADULTS WITH review. Systemic therapy Local-regional therapy Survival rhosis and advanced HCC with or no therapy macrovascular invasion and/or metastatic disease questions. Vol. The research group pro. Following a comprehensive analysis of each systematic 1. SO. Strength of Recommendation: Strong ences. morbidity rhosis and stage T1 or T2 HCC 5 Adults with cirrhosis and HCC suc. with or without alpha-fetoprotein (AFP). progression to transplantation and T2 HCC T3/waitlist dropout 8 Adults with cirrhosis awaiting liver Down-staging and No transplant Posttransplant survival. Other factors—including the bal. CIRRHOSIS UNDERGO mendations on the quality of the evidence. progression to transplantation and T1 HCC T3/waitlist dropout 7 Adults with cirrhosis awaiting liver Bridging therapy Observation Survival. consistency. WHICH SURVEILLANCE TEST and other clinical considerations. or very low based on the domains of precision. Clinical Questions Evaluated Question Population Intervention Comparison Outcome 1 Adults with cirrhosis Surveillance for HCC No surveillance Survival 2 Adults with cirrhosis and suspected Diagnostic evaluation Diagnostic evaluation Sensitivity and specificity HCC with multiphasic CT with multiphasic MRI 3 Adults with cirrhosis and an indeter. Adjuvant therapy No adjuvant therapy Survival cessfully resected or ablated 6 Adults with cirrhosis awaiting liver Local-regional therapy Observation Survival. Biopsy Repeated or alternative imaging Sensitivity and specificity minate hepatic nodule 4 Adults with Child-Pugh class A cir. directness. patients’ values and prefer. high. indirect evidence. facilitate implementation. balance of SURVEILLANCE FOR HCC. The strength 1A. transplantation and T3 HCC transplant recurrence 9 Adults with cirrhosis and HCC (T2 or Transarterial Transarterial radioembolization Survival T3. and feasibility of the recommended action—all 1B. low. 2018 HEIMBACH ET AL.HEPATOLOGY. 1. relevant stud- quality of evidence in each systematic review is rated as ies are summarized after each recommendation. patients’ values and preferences. dations. 361 . Resection Local-regional therapy Survival. the guideline writing group generated AASLD IS BEST? recommendations that are graded as either strong Recommendations (apply to most patients with minimal variation) or con- ditional (apply to a majority of patients). 67. quality of evidence. corresponding systematic review for each of the key vided curated evidence summaries following the questions are presented in the Appendix. Quality/Certainty of Evidence: Moderate ance of benefits and harms. The AASLD recommends surveillance of adults of recommendation is not only determined by the with cirrhosis because it improves overall survival.

1. ment of a surveillance program(11) given that patients surveillance led to 163 per 1000 more patients detected with cirrhosis are a high-risk group(7) and can be readi.33) compared with no mortality. There were six studies that controlled for lead- to 8 months.8% among the 1. 4735 patients who underwent HCC surveillance. Of the 23 stud- cannot be recommended at this time. what type of plus AFP had a pooled risk ratio of 1. riers. January 2018 Quality/Certainty of Evidence: Low EVIDENCE AND RATIONALE Strength of Recommendation: Conditional 1C. P < to have chronic viral hepatitis B and cirrhosis due to 0. 1.001). 1. P < 0. with The goal of surveillance and screening is to reduce an OR of 2. 1. ies evaluated. surveillance led to more curative treatments com- HCC(2) summarize the populations at the highest risk pared with no surveillance (61.71] be part of surveillance. It is not possible to determine which type of sur. 1. There were 38 Strength of Recommendation: Conditional observational cohort studies that evaluated surveillance in patients with cirrhosis. and if so. hepatitis B virus (HBV) car. com- veillance test—US alone or the combination of pared with only 27.6% versus 28. which uses the data unless they are on the transplant waiting list. 1. The AASLD suggests not performing surveil.67-2. HEPATOLOGY. improvement in survival seen with sur- HCV. 10 were considered high-quality studies in which the 3-year survival with surveillance was greater than no surveillance (45.8% versus 38. P < 2.55-2.60).8%. er US alone or US plus AFP improved survival. The optimal interval of surveillance ranges from 4 0. it is unknown whether the addi. tainties.86 (95% CI. surveillance is best.(10) HCC meets the criteria for the develop. Of the studies identified. alone.(12) However.17. US lance in adults with cirrhosis. time bias.001. making the overall quality of the evidence moderate (Supporting Table 1). There tion of biomarkers such as AFP allows for improved were no studies that directly compared US alone versus survival. only four used US randomized trials in Western populations with cirrho.001). and 2. Both source of controversy is which surveillance tests should US alone and US plus AFP led to similar rates of cura- be used. with an odds ratio (OR) of survival.(13) low anticipated survival for these patients. detection of early-stage HCC compared with no surveil- er surveillance truly leads to a reduction in mortality in lance. The pooled 3-year survival rate was 50. there are no and AFP. There were no randomized controlled trials (RCTs) of Quality/Certainty of the Evidence: Low surveillance in patients with cirrhosis. A randomized surveillance study performed in veillance appears to be due to higher early-stage detec- another high-risk group.53]).(2) Because of these uncer. whereas US alone 362 . 1.16 (95% CI.2%.68).76-1. 2. Although it is well established that US should tive treatment (OR. surveillance.89-2.HEIMBACH ET AL. surveillance also led to an increase in the detection of early-stage HCC.97) for improving survival.88-2. with an OR of 2.1% without surveillance. P < 0. The previous AASLD guidelines on tion.9% among the 6115 patients with- US plus AFP—leads to a greater improvement in out previous surveillance. and the improvement in survival persisted (3- 3.04 (95% CI.19 for US plus AFP [95% CI.80-2. month intervals. at early stages compared with no surveillance. In terms of anticipated absolute effects. of early-stage detection or curative therapy.83-2. Modification in surveillance strategy based on eti.90 (95% confidence interval [CI]. whereas the rest relied on US and AFP at 6- sis secondary to chronic HCV or fatty liver disease. given the from a recent systematic review on surveillance. tion and higher curative treatment rates.001). The Technical Remarks majority of the data was reported with 3-year survival.23 for US [95% CI. whereas this population of patients with cirrhosis. year survival rates of 39. 1. The use of US plus AFP improves and thus there is some controversy surrounding wheth. Thus. The previous AASLD guidelines recommend US plus AFP to determine which was superior in terms US as the primary modality. In addi- ly identified. showed a 37% reduction in mortality for those The surveillance tests used most commonly were US who underwent surveillance. The evidence profile of surveillance for HCC is lance of patients with Child-Pugh class C cirrhosis included in Supporting Table 1. Another US alone had an OR of 2.) BACKGROUND In addition to improved survival. the aim of this question was to determine The studies were also evaluated to determine wheth- whether current data are in support of HCC surveil.7% for surveillance versus ology of liver diseases or risk stratification models 29.11 (95% CI.

and multiphasic MRI with gadoxetate Recommendation disodium (a specific gadolinium-based compound that accumulates in hepatocytes and permits characteriza- 2. There was no statis. All studies were performed at academic centers. 2. Commonly used HCC UNDERGO DIAGNOSTIC methods in clinical practice include multiphasic CT EVALUATION WITH with extracellular agents.56-1. (1) no description of the trigger to perform a diagnos. including its operator dependency and reliabili- biopsy. There were no randomized comparative studies of CT versus 1. 2018 HEIMBACH ET AL. there are serious issues when comparing these surveil. considerations. and treatment may be initi- needed. tical difference between the two strategies.(14) confidently to be HCC. the pretest probability of HCC is sufficiently high. diag- nostic imaging is used to noninvasively verify the pres- FUTURE RESEARCH ence of HCC (diagnosis) and determine its extent (radiological staging). through. There were 19 observational studies put. scan time. 67. compared the per-lesion diagnostic accuracy of CT 363 . No. guide management. which uses the data from a de novo systematic review on imaging Technical Remarks in HCC performed to address this question. multiphasic MRI with extra- MULTIPHASIC CT OR cellular agents (gadolinium-based compounds that stay in the extracellular space and permit characterization of MULTIPHASIC MRI? blood flow). These MRI with gadoxetate disodium. examination costs and charges. The selection of the optimal modality and con. MRI because of similar diagnostic performance characteristics. the diagnosis of HCC can be US with US plus AFP as surveillance strategies are established noninvasively. better patients for possible liver transplantation. radiologist expertise. had a slightly lower pooled risk ratio of 1. Because of the greater technical complexity of lance tests for their impact on survival. patient preference. Unlike most studies including appropriate study design comparing other malignancies.75 (95% CI. MRI. patient preference. generalizability to practices without liver MRI tic test. no studies identified that compared multiphasic trast agent for a particular patient depends on MRI with an extracellular agent versus multiphasic multiple factors beyond diagnostic accuracy.98) for improving survival. such as des-gamma carboxy pro- HCC imaging criteria can be assumed reliably and thrombin. The AASLD recommends diagnostic evaluation tion of hepatocellular “function” in addition to blood for HCC with either multiphasic CT or multiphasic flow). and In addition. The goals are to measure tumor Given the current burden of HCC and the projected burden. EVIDENCE AND RATIONALE Quality/Certainty of Evidence: Low for CT versus MRI The evidence profile of diagnostic accuracy for Strength of Recommendation: Strong HCC is included in Supporting Table 2. scheduling backlog. which include: multiphasic MRI compared with multiphasic CT. and other novel serum tests. The rationale is that in patients with cirrhosis. However.HEPATOLOGY. (2) some studies appear to evaluate AFP or US expertise is not yet established. Although there is strong con- sensus that the imaging diagnosis of HCC requires 2. SHOULD ADULTS WITH multiphasic imaging. Such studies should evaluate the characteristics ated based on imaging alone. and safety 1. AFP L3. rather than the combination. ty as a surveillance test in specific patient populations. and (4) most importantly. 1. and no data on include modality availability. institutional technical in patients with cirrhosis and suspected HCC that capability. and help prioritize continued increase in incidence of this tumor. it would be important to determine the pretest probability of lesions that may mimic HCC whether other serum biomarkers in addition to AFP at imaging is sufficiently low such that a lesion meeting complement US. BACKGROUND mine an improvement in survival. there is not agreement about CIRRHOSIS AND SUSPECTED which diagnostic imaging test to use. without confirmatory of US. In patients with cirrhosis and suspected HCC. (3) no mention of the performance characteristics of these tests. Vol. the studies were not powered to deter.

87 versus 0. Advan.89].08). difficulty holding still. 0.63 [95% CI. 0.64-0.94 [95% CI. or biopsy.63-0. including beyond diagnostic accuracy inform the selection of follow-up imaging. reporting true positive. whereas the specificity was.78 [95% CI. a multicenter trial of US transplan- extracellular agent versus multiphasic CT and showed tation patients with HCC underwent both MRI and a similar pooled sensitivity.49. three studies compared multiphasic MRI with an collected. lower throughput.46 versus 0.96 [95% CI. 0.) and this also showed similar sensitivity and specificity. Quality/Certainty of Evidence: Very Low tages include greater soft tissue contrast.76 [95% CI. false comprehensive assessment of nodule and background negative.57.02).69 versus 0. January 2018 and MRI. An additional 14 liver tissue properties.97] versus 0. REPEATED IMAGING. CT is faster. HEPATOLOGY. particular patient population (NCT01082224. SHOULD ADULTS WITH agent. but pared multiphasic MRI using an extracellular agent it exposes patients to radiation. and true negative values.049). ing backlogs.001) with similar specificity (0. CT at multiple fixed time points while awaiting trans- cificity of 0. P 5 0. as sensitivity cannot be plexity. and further studies are [95% CI.82 [95% CI.81]. and absence of ionizing radia- studies reported only detection rate (sensitivity).73 nose HCC noninvasively. 0. but tion. The sensitivity of MRI for <1 cm was signifi. From a patient perspective. or high- limitations of the included studies. inconsistency across volume ascites). 0. United States—lower availability and longer schedul- all and for lesions of different sizes are reviewed below. Examining accu. interpreted in the absence of data on specificity and/or increased susceptibility to artifact.7 (P 5 0. MRI with an extracellular require IV access and contrast agents. tiphasic CT. P INDETERMINATE HEPATIC 5 0. 0.HEIMBACH ET AL. and—especially outside the mance characteristics of these imaging modalities over.90. and recognition that multiple factors with cirrhosis and an indeterminate nodule. NODULE UNDERGO A BIOPSY. ney injury or chronic renal failure. P < 0. multiphasic MRI with an extracellular 0. P 5 0. less consistent positive predictive value. and possible publication bias.87 [95% CI. P < 0. imaging with an alternative optimal imaging modalities in individual patients. lower (0.72-0. concerns about generalizability to nonaca. dications. The AASLD suggests several options in patients demic settings. more Strength of Recommendation: Conditional 364 .93] versus 0.90-0. higher charges. Eight studies compared multiphasic MRI with FUTURE RESEARCH gadoxetate disodium versus multiphasic CT. Both modalities versus multiphasic CT. agent. 3A. With regard to overall accuracy.(15. P 5 0. For HCC <1 cm. larger number of potential contrain- studies. modality or alternative contrast agent. false positive. Of note.02) with similar specificity needed.47).98].16) 0.79-0. The perfor. has important advantages and disadvantages. multi- with gadoxetate disodium provided higher pooled sen. more spacious. two studies compared multiphasic CT versus multiphasic MRI with an extracellular 3. Mitigating factors include the low quality of the evidence. Quality of evidence was low image quality (largely because of patient factors such and was downgraded because of the methodological as breath holding. 0. the differences in pooled diagnostic THE DIAGNOSTIC EVALUATION? performance are insufficient to recommend MRI over Recommendations CT. 0.62). CIRRHOSIS AND AN cantly higher compared with CT (0. OR Although multiphasic MRI may be marginally ALTERNATIVE IMAGING FOR more sensitive than CT in a pooled analysis of com- parative studies. there were six further elucidate which technique is optimal in this studies that compared multiphasic MRI versus CT. but Compared with multiphasic CT. and multiphasic MRI with gadoxetate disodium When looking specifically at lesions larger than 2 8. Disadvantages include greater technical com- these are not further discussed. with a higher pooled spe. the use of agent provided higher pooled sensitivity than CT which may be problematic in patients with acute kid- (0. phasic contrast-enhanced US also can be used to diag- sitivity than CT (0.69. MRI Although not used widely in North America. at a trend level. eight studies com.88] versus 0.69]. plantation has recently completed enrollment and may racy in HCC between 1 and 2 cm. and provokes less claustrophobia.(17-24) Prospective studies should include mul- (0. longer scan times. and data on costs and patient preference should be cm. multiphasic MRI cannot recommend one option over the other.81] versus 0.71-0.

however. temporaneous biopsy as the reference standard. gressed to definite HCC during follow-up. Khalili et al. using con- data are able to elucidate an optimal strategy. but its tified that directly address this question. Lesions that do detected at surveillance ultrasound are malignant. Biopsy has the single-center. Tanabe et al.(25) the OPTN. Biopsy had a false negative rate of 30%. biopsy Because many if not most indeterminate small hepat- may not be an optimal strategy in all cases. the ous LI-RADS categories are associated with different aim of this question was to determine whether current likelihood of HCC in patients with cirrhosis.23. intermediate probability of HCC.7% and 93. of lesions initially categorized as intermediate probabili- given that biopsy confirmation of <20 mm HCC would ty progressed to HCC. and repeated biopsies may be nec. Biopsy is ic nodules are nonmalignant. The indeterminate lesions were categorized as is an additional drawback. Stringent imaging criteria with high specificity for with the standard. Imaging Reporting and Data System (LI- as patients with suspicious imaging findings or growth RADS).(26) and previous AASLD were rebiopsied up to three times. and 38% of lesions initially cate- not change management or contribute to liver trans. whereas 7% plantation with a single small. and exclude malignancy. Given not meet these guidelines or are smaller than 1 the low likelihood of malignancy. although two routine use is not suggested. biopsy has a risk of hepatic nodules detected during surveillance ultrasound bleeding. MRI. the sensitivity was only 33%. whereas benign nodule.to 2-cm indeterminate nodules initially ance and/or capsule appearance. probably HCC based only on imaging features. In its previous HCC clinical practice guidelines. and has a risk of cation are needed. with American College of Radiology through its Liver 100% specificity. plantation priority. Darnell et al.(27) Sampling error. history of indeterminate lesions detected at CT or ing. Because of its many limitations. 1. Technical Remarks Based on an extensive search strategy detailed in the sys- tematic review.(38) evaluated the natural complications. which a diagnosis is required to affect therapeutic Forner et al. Follow-up imaging may lesions initially categorized as probably benign pro- be especially relevant in patients awaiting liver trans. They found a sensitivity and sue representative of the nodule rather than a truly specificity of MRI to be 61. which was biopsy.(2) with biopsy reserved for 1-2 cm nodules with arterial the AASLD recommended biopsy for all indeterminate phase hyperenhancement or in the presence of a synchro- lesions initially detected by surveillance ultrasound. 3B. tumor seeding. Vol. Biopsy may be required in selected cases. there were no comparative studies iden- 1.29-37) tion.(19. A negative biopsy may not probably benign. 2018 HEIMBACH ET AL. for all indeterminate hepatic nodules may be impractical and suggested an alternative strategy of close follow-up BACKGROUND imaging with sequential contrast imaging using an alter- nate technique for most indeterminate 2 cm nodules. nodules. 365 . taneous biopsy of 2 cm nodules in addition to MRI and contrast-enhanced US. gorized as probably HCC progressed to definite HCC.6%. strategies for risk stratifi- expensive.(2) and include arterial phase hyperen.(25) No essary to establish a diagnosis. they argued that biopsy cm are considered indeterminate.(17) in 2008 reported outcomes for 2 cm decision making. may cause anxiety or pain.HEPATOLOGY. contrast-enhanced US was 51. thereby permitting earlier interven. as characterized by CT or MRI. noncomparative studies were identified potential to establish a timely diagnosis in cases in that examined the role of biopsy. which uses the data from a de novo systematic review on imaging in HCC performed to address this question. (28) reported that in patients with cirrhosis.1% compared 2. Because there is controversy regard. Numerous other studies also reported low the presumed rationale being that biopsy can establish a likelihoods of malignancy among 2 cm indeterminate definitive diagnosis. The AASLD suggests against routine biopsy of EVIDENCE AND RATIONALE every indeterminate nodule.7% and 96. 67. guidelines. and the possibility that a in patients with cirrhosis.(39) in 2015 showed that the vari- ing optimal workup for an indeterminate nodule. indeterminate nodule. especially for very small lesions. with nous HCC. No. When both tests 10 mm HCC have been developed by the were in concordance. however. The authors performed percu- negative biopsy is due to the failure to obtain tis. Quality/Certainty of Evidence: Very Low The evidence profile is included in Supporting Table Strength of Recommendation: Conditional 2. Similarly. including tumor track seeding and bleed. only 14%- hancement in combination with washout appear- 23% of 1. In 2011.

are favorably located may be equally well treated tion. (TACE) or other forms of ablative therapy. whereas therapeutic options 4. and variability Serste et al. number. CHILD-PUGH CLASS A CIRRHOSIS AND EARLY-STAGE BACKGROUND HCC (T1 OR T2) BE TREATED Because cirrhosis is one of the primary risk factors WITH RESECTION OR for HCC. 2. An individualized diagnostic 3. and location of the tumors.to 2-cm indeterminate nodules are nonmalignant histologically and unlikely to progress to 1. In on the size. multifocal. such as patient preference or HCC) and a lower prevalence of other liver dis- drop-off from the transplant waiting list. Stage T1 and T2 HCC include a wide range of workup based on clinical context and imaging findings tumor sizes from <1 cm to 5 cm. and referral to a specialty center.5-3 cm. feasibility of biopsy. imaging with an alternative modality or con. or near major vascular or biliary structures may have limited ablative options. and tiveness of available therapies depend in large part institutional expertise may be the optimal approach. This tests was 98% and 81%. tional nonimaging features to more precisely predict in whom there is a higher etiologic prevalence of lesion progression. The impact of these demo- graphic differences on oncologic outcomes of dif- 4. A study by across studies or in clinical practice. respectively. and biopsy for a is seen not only in what is defined as resectable series of 74 patients with nodules identified by surveil. from a purely technical standpoint but also in lance ultrasound. on either CT or MRI. a strategy of types of LRT—such as transarterial radioemboliza- obtaining a biopsy of all indeterminate nodules would tion (TARE) and transarterial chemoembolization result in a considerable number of unnecessary biopsies. These include ablative strategies such as 366 . January 2018 Taken together. Direct comparative studies of resection versus other HCC during imaging follow-up. imaging. eases such as NAFLD or HCV compared with Western patients.5 cm) that elect to treat a probable HCC without biopsy confirma.39) potentially including end. are limited for patients who present with advanced liv- Pugh class A cirrhosis and resectable T1 or T2 HCC er disease and/or advanced tumor stages. Future research is needed to standardize the defini. Randomized trials performed to date comparing tion of and independently verify the prognostic value radiofrequency ablation (RFA) to resection have of different nodule characteristics and to identify addi. 2. Other diagnostic options include follow-up though indirect evidence favors resection. selected circumstances.(38. though practitioners and patients need to be aware by either resection or ablation. The authors concluded that sensitivity patient-related factors such as acceptable degree of and specificity of the combination of the two diagnostic portal hypertension and performance status. and the effec- such as nodule characteristics. Thus. Therefore. the selection of treatment modality depends LOCOREGIONAL THERAPY? as much on the underlying liver function and the degree of portal hypertension as on the oncologic stage Recommendation of the tumor. potentially resect- Strength of Recommendation: Conditional able tumors. these studies suggest that a substan.(40) performed CT. indeterminate nodules do require further eval. 4. options exist for those presenting with well- Quality/Certainty of Evidence: Moderate compensated cirrhosis and smaller. Technical Remarks tial proportion of 1. Multiple tumors that are bilobar or cen- FUTURE RESEARCH trally located may not be resectable. multiple undergo resection over radiofrequency ablation. and that biopsy variability leads to challenges in comparing study could be reserved for those without definitive findings findings. uation. MRI. The AASLD suggests that adults with Child. tumors larger than that such treatment may affect transplant priority. single tumors (<2. HBV (including noncirrhotic HBV–associated points other than survival. a multidisciplinary group may Whereas smaller. such However. SHOULD ADULTS WITH ferent therapies is unknown. as radiation and microwave—are not available. HEPATOLOGY.HEIMBACH ET AL. been performed primarily in East Asian patients. The definition of resectability is not uniform trast agent.

3).78) as (some trials accept two lesions up to 4 cm). The additional endpoints of 2-year event- from analyses of studies comparing resection to abla. 57. 95% CI. radiographic evidence of extrahepatic disease or macro.71. (2) without well as 2-year survival (HR. free survival and local progression favored resection tion of “resectable” tumors and may lead to biased regardless of inclusion of the potentially biased trial. 0. the complication rate was higher for In addressing this particular question. plication rates (4.4% for patients was too low to reach a firm conclusion. noted that the existing evidence was reviewed to com.(47) ablation—and found moderate quality evidence that This report included 544 Child-Pugh class A patients RFA prolonged survival. When a third trial with a high risk of bias is added to vascular invasion. 95% CI. 0.(44) The results of the two patients with resectable HCC as those (1) with one to low-risk-of-bias trials demonstrate that hepatic resec- three unilobar lesions.765). radiofrequency. with the largest being <5 cm) and found resection to EVIDENCE AND RATIONALE be superior in 1 and 3 years of follow-up (HR. A subgroup Importantly.(45. Not unexpectedly. No. we did not include a review of the resection group (P 5 0. addressed in an RCT. 95% CI.(41) but examining the three patients.56.HEPATOLOGY. of bias and moderate evidence quality.1% for review concluded that the total number of included RFA. and cryoabla.38. This systematic review did not cover survival for patients with single HCC lesions <3 cm the use of TACE or TARE. 3. resection. 8. a number of clinical and increased risk of bias in the third study is related to an laboratory variables and circumstances. the difference in survival between resec- minimal or no portal hypertension and in the absence tion and RFA became insignificant (overall survival: of synthetic dysfunction (Barcelona Clinic Liver Can.46) One single-center RCT compared resection setting of compensated cirrhosis (minimal or no portal with RFA combined with TACE (TACE was per- hypertension and preserved synthetic function).17-0. 0. However. it should be resection compared with RFA (OR. HR. and (3) occurring in the setting of the analysis. 0.0% for RFA. including the unusually high number of patients (n 5 19) who availability of alternative therapies.44-1. and 5 years for unresectable HCC. analyses and conclusions. Given formed first. 66.. with an upper size limit of 5 cm tion is more effective than RFA regarding overall sur- for single lesions and 3 cm for more than one lesion vival (hazard ratio [HR].5% for resection.101).84). including a total of 578 performed by Weis et al.(41) on treatment ing given that it is known that RFA is more effective for early-stage HCC in patients with Child-Pugh class in lesions <3 cm. to and exceeding Milan criteria (up to five tumors. potentially resectable HCC occurring in the tion.(45) Another RCT tri- studies comparing transplantation to either resection al compared resection with TACE alone for lesions up or ablative therapies. In addition to the trials included in the systematic pare resection with ablative therapy (also comparing review by Weis et al. intention to treat. 2018 HEIMBACH ET AL. the specific question of A or B cirrhosis. followed by RFA within 4 weeks) and that liver transplantation is reserved for patients with demonstrated improved survival at 1. chemical. which utilizes the data from a recent system. thus potentially overstating the ben- ity constitutes a limitation of the interpretation of data efit of RFA. 67. two recently published RCTs confirm the find- optimal therapeutic option for patients with early-stage ings of improved survival for patients after resec- (T1-T2). both arms of the systematic review.2% for RFA. and the authors observed similar com- resection comparison and the RFA versus other tech. recurrence rates (56% for resection.(42. This is not surpris- atic review performed by Weis et al. In both the RFA versus from 15 centers. as well as surgical resection. there were three RCTs that compared analysis for outcomes of smaller single lesions was not RFA with resection. The reason for an cer stage 0 or A). 0.(42-44) Two of these three trials had a low risk individual RCT trials included in the systematic 367 . 0. Vol. P 5 0. However. P 5 niques comparison. can influence the switched from the RFA arm to the resection arm yet individual clinician’s decision to proceed with resec.40-0. The absence of a standard definition of resectabil. P 5 0. 0.001). the authors of the systematic 0. treated with resection versus RFA has not been ple other comparative groups—including RCTs com. Most studies define trial had a high risk of bias. 1.(41) comparing resection with different ablative options) specifically to determine the RFA.(46) The evidence profile is included in Supporting Lesion size was a risk factor for worse outcome in Table 3. though it covered multi. and 4-year survival rates (74.353).4. P < 0.43) and one tion.007).15). A recent multicenter retrospec- paring RFA with percutaneous ethanol or acetic acid tive report from Italy did examine this question. 2. were still counted within the RFA group because of tion. microwave.

the ideal adjuvant ther. The systematic review Quality/Certainty of Evidence: Low by Wang et al. To date. In addition. January 2018 review. HCC is uncertain and requires further study.(48) and has never been shown to improve survival in 2. such as stereotactic body agents in advanced disease has hampered the develop- radiation and microwave ablation.. at the time of surgery. The distinction RFA (P 5 0. and the presence or absence of lym- phovascular invasion. Of the agents evaluated in the adjuvant setting. The efficacy of several cytotoxic with LRT. FUTURE RESEARCH There is a clear need for the development of new. the effectiveness of embolization strategies most of the adjuvant agents studied did not have clinical such as transarterial approaches (TACE and TARE) evidence that they improve survival in any stage of have not been systematically compared with either HCC. addition. though other classification systems are chemotherapy regimens has also been tested in RCTs also used. such as size. with the RCTs. Technical Remarks however.(52) yet it ultimately did not show any improvement in outcomes for the adjuvant treatment of 5. the role of statin therapy in the adjuvant setting is 368 .(57) Final- apy would have an antineoplastic component aimed at ly.(50) with a decrease in the development of secondary tumors.(53) Resection of HCC CIRRHOSIS AND HCC THAT HAS with curative intent or ablation is associated with rates BEEN RESECTED OR ABLATED of recurrence at 5 years as high as 75%.(54) on adjuvant treat- cessful resection or ablation. The risk of recurrence after surgical resection or advanced HCC. advanced disease. the original tumor and a chemopreventive effect aimed lowing resection remained favorable compared with at the development of a de novo tumor.(47) Therefore. SHOULD ADULTS WITH HCC in randomized studies.(55) RCTs of adju- 1. The AASLD suggests against the routine use of Table 5. review performed by Wang et al. This of these two scenarios is difficult and often based on the subgroup analysis was not performed in the other two time of the recurrence (e. ADJUVANT THERAPY? EVIDENCE AND RATIONALE Recommendation The evidence profile is included in Supporting 5. latter believed to be related to the development of a de novo tumor).(51) The lack of proven active other than RFA techniques. identified that adjuvant interferon ther- Strength of Recommendation: Conditional apy can improve both recurrence-free and overall sur- vival in patients with virus-associated liver disease.HEIMBACH ET AL.(56) which limits their use in the adju- ablation is related to characteristics of the tumor vant setting. degree of dif- ferentiation. but larger studies The comparative effectiveness of ablative strategies did not confirm a benefit. ment for HCC after treatment.03) in patients with smaller tumors. Huang et al.g. the impact of HCV eradication by includes both recurrence of the primary tumor and the direct-acting antiviral therapies on the future risk of development of de novo tumors. internal radiation. and heparanase Solid Tumors (mRECIST) may be the most inhibitor PI-88 therapy were included in the systemat- common criteria used to evaluate radiological ic review and failed to improve recurrence-free or over- response in patients affected by HCC and treated all survival. effective chemotherapy agents for treatment of HCC BACKGROUND in both the advanced setting and in the adjuvant set- Given the unique biology of HCC in which risk ting. remain unclear. SUCCESSFULLY UNDERGO there is a clear need for adjuvant systemic therapies.(49) Early studies with the adjuvant use of FUTURE RESEARCH acyclic retinoids were promising.(42) demonstrated that survival fol. the side effects of interferon are significant. resection or ablative strategies in Child-Pugh class A only sorafenib has been shown to improve survival in patients with T1 or T2 HCC. limiting its use in clinical practice. which uses the data from a recent systematic adjuvant therapy for patients with HCC following suc. The modified Response Evaluation Criteria in vant chemotherapy. In ment of agents targeting early-stage disease. HEPATOLOGY. early versus late.

The median age was 60 years. SHOULD ADULTS WITH remain at risk for HCC recurrence and will require continued monitoring. 7. anticipated waiting patients with T1 HCC who were or were not treated time. This is the outcome assessed CIRRHOSIS AWAITING LIVER by the study by Huo et al. the anticipated wait time. decompensation. may grow Quality/Certainty of Evidence: Very Low to beyond T2 criteria and/or metastasize during the Strength of Recommendation: Conditional observation period. Importantly. the recent reports of an associated reduction in HCC LRT of a T1 HCC may be of significant benefit to risk for patients with HBV who are on statin patients who are well compensated and have no other therapy. and two allocation policy prioritizes patients with OPTN T2 additional patients died of non-HCC causes. and the rate of tumor growth was estimated gov/governance/policies). There were no RCTs. 100 patients dependent in large part on an assessment of the (87%) progressed from T1 to T2 at a median of 6. rate of growth of the lesion. and only two of these versus treatment depends on several factors trials address the question of waitlist outcomes for including patient preference. This is the question addressed in the observational study by Mehta et al. No. if a T1 lesion is to be 0. Eighty-seven noncom- 2.5% 2 or 3 lesions each between 1-3 cm) but not for those within 6 months. six undergo LRT. Therefore. the possibility that the tumor. BE TREATED OR UNDERGO If the patient has other indications for transplant OBSERVATION? other than the presence of HCC. especially if these complications are not captured by the current MELD- Recommendation Na score.(59) which is discussed TRANSPLANTATION AND T1 HCC below. and organ other patients (5. current liver T2 at a median of 5.3%) progressed from T1 to beyond allocation policy. imaging over treatment for patients with cirrhosis If observation is contemplated.3%) remained within T1.14 cm per month. unknown. This recommendation is intended for patients EVIDENCE AND RATIONALE who are already on the liver transplantation wait- The data are summarized in Supporting Table 6.4 2 cm (T1) and are listed for liver transplantation is years. a key consideration is awaiting liver transplantation who develop T1 HCC. 67.hrsa. though it may warrant investigation given the patient increased priority for transplantation. and AFP. The choice of observation with follow-up imaging parative trials were identified. and 15. Risks for wait list dropout treated with LRT.(60) which is Technical Remarks discussed below.HEPATOLOGY. The median follow-up was 2. 1... The stage HCC (either a single lesion between 2-5 cm. In the United States. proportions in Child-Pugh class A (48%) and Child- rhosis who have a single HCC nodule between 1 and Pugh class B/C (52%). degree of liver with LRT. Six patients (5. The AASLD suggests observation with follow-up sion to treat with LRT requires careful consideration. denying included AFP >500 and rapid growth. it may not reach stage T2.1% within 1 year. the patient will 6. Future allocation policy revisions may bridging or downstaging therapies before transplan- impact this recommendation.transplant. if untreated. The study by Mehta et al. in 2 years. 1. 2018 HEIMBACH ET AL. The authors 369 .(60) is a retrospective observational study of 114 patients with T1 HCC BACKGROUND listed for liver transplantation at a single United States institution between 2004-2012 who were not treated The decision to offer LRT consisting of either local with LRT. such as encephalopathy or ascites.1 months from listing. or cumulative probability of waitlist dropout was 4.(58) indication for transplantation. with equal ablation or transarterial treatment to patients with cir.6% with- with OPTN stage T1 (https://optn. as they may be able to avoid transplantation. and during the observation period.9 patient’s underlying liver function and ability to safely months. tation. the deci- 6. Vol. list—and thus presumably with an indication for including the findings of a de novo systematic review transplantation in addition to HCC—and is based of all studies that enrolled adults with cirrhosis on current organ allocation policies in the United awaiting liver transplantation and treated with States.

bridging therapy. 5. eralizable to areas with shorter wait times. Quality/Certainty of Evidence: Very Low The study by Huo et al. ablative therapy. Notably. the prac- addition. they recommended LRT rather Strength of Recommendation: Conditional than observation for patients with T1 HCC with high 7B. TRANSPLANT WITH BRIDGING synonymous with the Milan criteria. It is important to liver-directed therapy over another for the purposes of note that this study was performed in an area with pro. though based on their observations of the patients Quality/Certainty of Evidence: Very Low who dropped out. was to validate a potential transplantation by way of MELD exception point allocation score proposal called the HCC-MELD allocation since February 2002. TACE and ablation—to induce tumor death and patients with T1 HCC had a 6-month waitlist dropout deter tumor progression beyond the Milan rate of 5. January 2018 concluded that observation for patients with T1 HCC T2 (Milan) criteria to decrease progression of disease waiting for liver transplantation is an acceptable strate. In 4. Bridging is defined as the use of LRT—such as RFA. predictive markers of poor biologic behavior tices for liver transplantation for HCC may differ such as rapid progression would also better inform based on geographic location and access to living decisions about nontreatment of T1 HCC with regard and deceased donor organs. deceased donor liver transplantation. The AASLD does not recommend one form of AFP >500 or with rapid growth. which may limit the gener.HEIMBACH ET AL. and TACE. Given that organ availability is variable. have been THERAPY WHILE WAITING? granted additional HCC MELD exception points since 2002 because of an excellent overall survival Recommendations with a low risk for HCC recurrence posttransplanta- 7A. percutaneous ethanol injection (PEI) or acetic TACE.(59) reported on outcomes Strength of Recommendation: Conditional for 390 patients in Taiwan with T1 (n 5 94) and T2 (n 5 296) HCC who were eligible for transplantation Technical Remarks but who were treated instead with LRT. gy. HEPATOLOGY. The AASLD suggests bridging to transplant in tion (10%-15%). the primary aim Milan criteria have been granted access to liver of the study by Huo et al. The risk of hepatic decompensation due to LRT with RFA. Patients in the United States with HCC within alizability of the findings. In addition. Patients treated with RFA combination of different types of LRT such as had the lowest rate of waitlist dropout. yttrium-90 (Y90).3% for tumor progression beyond T2 criteria. The MELD allocation system with additional pri- oritization for HCC is not practiced worldwide. 7. to gain a better understanding of its natural history. multiple changes to the policy to reduce access combined FUTURE RESEARCH with ever-increasing waiting times have impacted the interpretation of studies before and in the ear- Additional longitudinal data from multicenter ly days following adoption of MELD allocation cohorts of patients with T1 HCC would be beneficial compared with current practice. Overall. though this represented only 2% of patients treated 2. Patients with T2 tumors. Although patients score rather than to observe the impact of LRT on with T2 HCC have continued to have access to waitlisted patients with T1 or T2 HCC. and the findings may not be gen. bridging to liver transplantation for patients within longed waiting time. SHOULD ADULTS WITH CIRRHOSIS AND OPTN T2 HCC BACKGROUND AWAITING LIVER TRANSPLANTATION UNDERGO The primary aim of bridging therapy is to mini- TRANSPLANT ALONE OR mize the risk of HCC progression while awaiting liv- er transplantation. and subsequent dropout from the waiting list. to a risk/benefit analysis. or a acid injection. OPTN T2 (Milan) criteria.(61) Progression beyond the Milan patients listed for liver transplantation within OPTN criteria while awaiting transplantation eliminates 370 . criteria. Patients were treated with a number of different methods including 1. 3. cal transplantation patient. a majority of patients in the study must be considered when selecting patients for had HBV and were of slightly older age than the typi.

(67) reported outcomes included dropout because of HCC progression and because of all causes. bridging therapy is conditionally Strength of Recommendation: Conditional 371 . one study enrolled only patients meeting Milan receiving bridging LRT for waitlisted patients with criteria. The RR of recur. bers in each respective therapy. criteria. including geographically variable define criteria. downstaging therapies before transplantation. using the Scientific Registry Transplant Recipients including the findings of a de novo systematic review data: 76% versus 71% (P 5 0.(62-64) This question assesses the benefit associated with lower rates of waitlist dropout of of the addition of bridging therapy for patients with 8. but the difference did not reach statisti.(65) Furthermore. Studies have demonstrated that the present time and the relatively low risk of harm without liver-directed therapy. or multitherapies. and two did not specifically logistical reasons.HEPATOLOGY. CIRRHOSIS AND HCC BEYOND cal significance. This stratification did not The addition of biomarker data may also help stratify reveal any significant difference among the various HCC with regard to its biologic behavior and response outcomes. TACE 1 RFA. Despite this limited Quality/Certainty of Evidence: Very Low evidence. No. access to exception points. 2018 HEIMBACH ET AL.9% at and 12 months.32 and 0. respectively. transarterial embolization (TAE). The quality of the evidence overall wait time within the United States for deceased donor was very low because of studies with significant risk transplants in patients with HCC. An RCT comparing bridging LRT versus not ies. Among the comparative stud. maintaining recommended because of selection bias for the tumor burden within or below T2 while waiting for patients selected to receive LRT as well as shorter transplantation is the only way to continue earning waiting time during the study period compared with exception points. The AASLD suggests that patients beyond the rence was <1 in patients treated with TACE 1 RFA Milan criteria (T3) should be considered for liver trans- and RFA alone with a markedly wide CI and was plantation after successful downstaging into the Milan limited to single studies with relatively small num. 0. 3-year overall survival (OS) after liver transplantation has been reported to EVIDENCE AND RATIONALE be significantly improved in patients with HCC who received LRT compared with those who did not The data are summarized in Supporting Table 6. 1.7% at 6 months and 22. The hepatic dysfunction. and overall recurrence-free survival after FUTURE RESEARCH liver transplantation. RFA.38. MILAN CRITERIA? Outcomes were noted to be similar when examined Recommendation by TACE. 67. recur- rence rate. six enrolled patients both within and exceed- HCC is unlikely to be performed due primarily to ing the Milan criteria. despite the lack of random- ization and potential for selection bias regarding DOWNSTAGING TO WITHIN which patients were selected to receive bridging. Posttransplantation recurrence and MILAN CRITERIA (T3) BE survival rates were not significantly different between TRANSPLANTED FOLLOWING the two reported cohorts. Noncomparative studies of LRT have been respectively. 8. Importantly. Vol. There with those exceeding 6 months being considered for were 18 comparative studies that reported the out- LRT if deemed appropriate based on the degree of come of interest. SHOULD ADULTS WITH respectively). The data were analyzed stratifying outcomes based on pretransplantation using all included studies and among the subset of radiographic response using mRECIST may help to those performed in the United States to control for delineate the true potential benefit derived from LRT. lower dropout because of progression and lower dropout from all causes in patients who received bridging LRT (relative risk [RR]. T2 HCC awaiting LT. the dropout rate is as for the intervention compared with the potential high as 25% and 38% at 6 months and 12 months. 8. Greater attention to of bias and imprecision. though there were no RCTs. and thus. the MELD era effect. there was a trend toward to LRT.(66) The decision to of all studies that enrolled adults with cirrhosis await- bridge patients with HCC to transplantation is large- ing liver transplantation and treated with bridging or ly dependent upon their anticipated waiting time.03). benefit.

Only three of these studies compared following down-staging to determine efficacy of down-staging of T3 tumors versus T2 tumors with no down-staging and subsequent optimal timing for down-staging before liver transplantation.and post-MELD era tive. most commonly the nonrandomized United States study.HEIMBACH ET AL. or they may appeal with a T3 including the findings of a de novo systematic review of tumor. reported to be beyond T2 treated with LRT. Down-staging of T3 which is not useful in patient selection and makes patients compared with no therapy (in T2 patients) direct comparison of results challenging. through the use of LRT. There were no compara- in tumor burden to within Milan criteria based on tive studies for transplantation of T3 with and without radiographic findings. The optimal form of liver-directed therapy for the to determine whether patients with HCC burden purposes of downstaging cannot be determined beyond Milan criteria should undergo liver transplan- based on the available data.03-1. free survival (1 and 5 years). the remaining patients in the LRT group. remaining studies were noncomparative. with those within T2 who were not treated with LRT. 3.32). whereas the liver transplantation. mRE. definition of successful down. The third compara- response to therapy is gauged. Y90.17 (95% CI. though some studies define down-staging. Currently. Hołowko et al. agreed-upon waiting period were no RCTs. within these criteria. Other studies use explant plantation overall (1. fully down-staged to within the Milan criteria may There were 12 of 50 patients who were successfully become eligible for HCC MELD exception points down-staged from T3 to within T2 at the time of trans- after undergoing review by their respective regional plantation. The outcomes reported in the three down-staging as a complete absence of tumor by comparative studies were limited to post–liver trans- radiographic findings. The 5-year BACKGROUND observed survival with down-staging had an RR of 1. and 5 years) and recurrence- pathology to define successful down-staging. 1. may be granted by appeal to the regional review board system for patients initially presenting with EVIDENCE AND RATIONALE T3 HCC after successful down-staging to within The data are summarized in Supporting Table 6. There is no standard.(69) provided the only comparative yet burden to predefined criteria. ciated with down-staging and transplantation. as summa- 4. patients had a very short wait list futility support limiting access to organs to patients time: 28 days in the 50 patients receiving LRT (TACE. T2/Milan criteria.(68) This variability is large. most of ly because of differences in tumor burden before LRT. days in those without LRT before transplantation. RFA. the severe organ shortage and concerns about patients. though this is not a practice that is widely all studies that enrolled adults with cirrhosis awaiting accepted. HEPATOLOGY. CIST) and lack of a standardized time period at which noting no difference in 5-year OS. MELD exception criteria. tation after successful down-staging to within Milan 2. Many studies define down-staging as a reduction rized in Supporting Table 6. Although 123 patients undergoing transplantation between 2000 some may consider the Milan criteria to be too restric. type of LRT used. Reported success with down-staging is between the 12 that were down-staged compared with highly variable (24%-90%). HCC organ allocation policy may be liver transplantation and treated them with bridging or revised in the future to allow access to standard- down-staging therapies before transplantation. before liver transplantation was associated with similar overall and recurrence-free survival. relative to no down-staging. January 2018 Technical Remarks addition to tumor size and number to meet criteria for successful down-staging. No significant difference in OS was noted review board. there 3. EASL. Down-staging is defined as a reduction in tumor Heckman et al. whom were stage T2 at the time of transplantation. spanning both pre. Within the United States. or resection) before transplantation and 24 patients who exceed these criteria who can be success. some have tive study was from Asia and consisted predominately of proposed the incorporation of tumor markers in living donor liver transplantations. There ized MELD exception for down-staged patients were a total of 24 studies examined for outcomes asso- rather than requiring appeal. and 2006. which includes Milan criteria. with down-staging 372 . Furthermore. This key question attempts 1. in the United States. as well as differing methods to assess radio.(70) present outcomes for patients staging. RECIST. compared graphic response (WHO. In this series.

RFS 90% versus 86%). Overall.7). and Parikh et al. The AASLD does not recommend one form of TAE) were small. to the Milan criteria and were compared with 110 NO VASCULAR INVOLVEMENT) patients who presented within Milan criteria and thus underwent LT without LRT. and thus the results of these studies may with Child-Pugh class C or poor performance sta- be affected by the inclusion of nontransplant candidates tus. Among the 9A. 2018 HEIMBACH ET AL. TACE. employed multitherapies.15-27. No. More 373 . drug-eluting beads TACE [DEB-TACE].g. and num- gets for future studies. Effectiveness of down-staging ber of lesions. and vival in the setting of patients with HCC not posttransplant details such as recurrence and survival. amenable to either resection or transplantation. consisting mostly of TACE. LRT over another. tumor characteristics such as Bar. 67. the risk of harm. These outcomes are comparable to what TACE: Moderate has been reported posttransplantataion among patients Transarterial Bland Embolization: Very Low with HCC within Milan criteria.. though these differences did not reach significance EXTERNAL RADIATION? (OS 83. transarterial chemoinfusion [TACI]. with unresectable HCC. including TACE and bland TAE are widely used in patients patient demographics. before transplantation can only be determined if the many variables that can confound these analyses are stan- BACKGROUND dardized. The available evidence is for Child-Pugh class A patients who are not deemed liver transplant candidates and highly selected Child-Pugh class B. HCC. the results to date are encouraging but inadequate to FUTURE RESEARCH make a recommendation. Strength of Recommendation: Strong RFA.HEPATOLOGY.1%). 1. As discussed below. There are for other reasons (e. location. PEI. the 5-year post LT Quality/Certainty of Evidence: OS was 77. The number of studies TARE: Very Low that examined various individual modalities (including External Radiation: Very Low Y90. RFA is another treatment strategy that may be after down-staging as well as the optimal waiting period used for selected patients with unresectable T2 between down-staging and transplantation are key tar. Noncomparative studies examining the success of down-staging may include 1. The majority of the remaining studies that examined Recommendations down-staging were noncomparative studies. Determining the variables which predict outcomes 3. OR stage. SHOULD ADULTS WITH patients were successfully down-staged radiographically CIRRHOSIS AND HCC (T2 OR T3. advanced age or significant no data to support the use of LRT for patients comorbidities).(72) have proposed criteria that should be included in all down-staging studies. either as bridge to transplan- ume and waiting time. and 9B. no posttransplantation that averaged 20. despite TRANSPLANTATION BE the down-staged patients being at a more advanced TREATED WITH TACE.4%). A small number of T3 WHO ARE NOT CANDIDATES patients underwent resection for down-staging. vascular involvement) who are not candidates for resec- with the lowest recurrence rate noted to be in studies that tion or transplantation. The highest 5-year OS was reported in those treated Quality/Certainty of Evidence: Very low with multitherapies (84. center characteristics such as vol. with a range of 1-4 for each modality. tation or as a recommended treatment to extend sur- celona Clinic Liver Cancer stage. and use of LRT should be weighed against in whom LRT is palliative in its intent.9%. 14 reported recurrence rates ment in adults with cirrhosis and HCC (T2 or T3.(71) A total of 51 T3 9.6%. The AASLD recommends LRT over no treat- 21 noncomparative studies. depending on the size. and the lowest 5-year OS Strength of Recommendation: Conditional was seen in those that were treated with TACI (54. treatment details. The data for the use of TARE and external beam radiotherapy is emerging. Vol.4% (CI 0. There FOR RESECTION OR was a trend favoring LRT for both OS and RFS. 2.7% versus 78. TARE. A lack of a comparative group beyond historical controls Technical Remarks severely limits interpretation.

(88) This trial was not ade- als were analyzed together for overall mortality.53.81. which is not the population TAE were compared with placebo for T2 or T3 HCC addressed in this question. P 5 0. Nonsignificant trends were noted The data used for this question are based on recent in 1-. In this analysis. Neither trial reported on mor- not amenable to resection or transplantation.79. There were two patients in each arm with the meta-analysis by Llovet and Bruix. establishing TACE as an effective Abdel-Rahman and Elsayed(86) performed a systematic strategy for unresectable multinodular HCC occurring in review to determine the potential benefit of TARE using patients with compensated cirrhosis. The analysis dem. In this report. suggesting that refinements in techniques may have macrovascular invasion who are not eligible for resec. 0. icant intervention effect (HR. P 5 0. able tumors and identified 19 RCTs that met this lated the number of subjects required to be included in a inclusion criterion. gression for waitlisted patients with HCC receiving 0. TARE compared with TACE.11). there were who were successfully down-staged to either undergo liver two additional RCTs included.53. 0.77) Of the nine includ. the over- attempt to determine the optimal therapy for those all survival for the TACE-only groups in these studies is patients with larger (>2.005). Looking specifically at the Kolligs et al.07) was noted. and quality of demonstrated a similar adverse event frequency in each life also included.71-1. there were a total of 28 patients (three RCTs). Looking at the overall outcomes for the been used as a treatment strategy for HCC. there were nine RCTs arm.(75. so additional showed no significant effect of TACE or TAE com. adverse events. 95% CI. 2-. 2 were noted to have a high risk of bias. One of treated with TACE versus placebo (41% versus 27%.HEIMBACH ET AL.88. published between 1990 and 2005.(89) This analysis included 1948 meta-analysis to accept or reject an intervention effect at patients and found a benefit to the combination of 1028.83. The authors calcu. had an impact on outcomes following TACE alone. 0. The prima.06. though the data from TACE formed at a United States transplantation center has just were pooled with the data from TAE. An additional meta-analysis of trials performed pri- P 5 0. TACE or for advanced stage cancer. tion or liver transplantation. 0. In this analysis. a pared with placebo on survival (overall HR.34-0. compared with TACE alone for T2 and T3 unresect- 0. whereas the other was a planned interim formed a more recent systematic review that questioned analysis of TARE plus sorafenib versus sorafenib alone the beneficial effect of TACE. external beam radiotherapy and TARE have also the analysis.5 cm) or multinodular T2 or superior to the TACE-only groups from the earlier stud- T3 tumors with no evidence of distant metastasis or ies.(85) identified four RCTs and eight observational trials. included. and this still failed to dem.02. two onstrated improvement in 2-year survival for patients RCTs were identified with a total of 64 patients. P 5 0. total of 545 patients. a subgroup analysis of marily in China published in 2015 assessed the avail- TACE only was performed. with advances in technology to improve preci. Both trials were classified as ry outcome was all-cause mortality. P 5 0. The intent more recent trials of TACE only versus TACE plus an of this question was to review the existing evidence to ablative strategy without a placebo control arm. 95% CI. Importantly. 95% CI. 0.58-1. stage HCC. and 3-year survival in favor of DEB-TACE existing systematic reviews. the trials compared TARE with TACE for intermediate- OR.) However. HEPATOLOGY. 0. no signif.(74) per. Notably. rent data are too sparse to make an assessment of efficacy.(87) trial for identified on the use of TACE (six RCTs) or TAE TARE versus TACE. having a high risk of bias and low quality.(75-84) Compared treated with TACE. Though the cur- ed trials.64-1. transplantation (n 5 3) or RFA (n 5 1). Y90 microsphere radioembolization. tality or disease progression. January 2018 recently. The two trials been reported and has demonstrated longer time to pro- with high risk of bias showed a significant effect (HR. with 13 patients treated with TARE and 15 reporting on a total of 645 patients. Oliveri et al. data is anticipated in the near future. Both trials comes of tumor response. When all nine tri. A meta-analysis performed compared with conventional TACE. number of study subjects were available to be included in sion. Pooled analysis by Llovet and Bruix(73) comparing TACE with placebo of objective response and of complications showed no identified seven RCTs on TACE versus placebo with a difference between the two therapies. with secondary out. quately powered to detect a survival advantage. 0. A meta-analysis specifically comparing DEB- TACE with conventional TACE treatment performed EVIDENCE AND RATIONALE by Facciorusso et al. 95% single-center RCT comparing TARE with TACE per- CI.10. sis of the HRs from seven trials with low risk of bias the authors identified five ongoing trials. and therefore only about two-thirds of the required TACE and PEI in both survival at 1 and 2 years as 374 .27). Analy. able data for the combination of TACE plus PEI onstrate a statistically significant benefit to TACE (HR.017.

3 and 5 years posttreatment as advanced cirrhosis. and the performance status of the strategy and which ablative strategy should be used. and the eligibility 3. although studies were mixed and included some patients with Child- Pugh class B cirrhosis. generally dependent on the extent of the vascular inva- tors may also determine which patients may benefit from sion and/or metastatic disease. The prognosis and treatment decision is important in determining efficacy of therapy. size. inadequate evidence to inform the balance of ben- bination therapy appears to be beneficial compared efit versus harm. 1. There was heterogeneity in the included systemic therapy over LRT. patient. 2018 HEIMBACH ET AL. Even for patients with metastatic disease. Another meta-analysis of trials of TACE selection of treatment type may vary depending alone versus TACE plus external beam radiotherapy on the extent of macrovascular invasion and/or has also been performed by Huo and Eslick that metastatic disease. Most patients involved in the studies had Child- beam radiotherapy in combination with TACE. no therapy may be the best well as for local tumor control. further multicenter RCTs are clear. the degree of underlying cir- included 11 RCT and 25 trials overall. Pugh class A cirrhosis. well as in local tumor response rates and decreased Technical Remarks AFP values. No. SHOULD ADULTS WITH often leads to rapid tumor progression with disease- CHILD-PUGH CLASS A/B related symptoms. The intent of this question was to review the existing Recommendation evidence to determine the optimal treatment recommen- dation for those patients with advanced HCC (macro- 10. the evidence supporting the routine use of DISEASE BE TREATED WITH many of these approaches has not been established. Quality/Certainty of Evidence: Moderate The evidence of a de novo systematic review includ- Strength of Recommendation: Strong ing all studies that enrolled adults with advanced HCC 375 . Advanced HCC is a heterogeneous group. par- ticularly those with limited extrahepatic tumor burden. regardless of the treatment strategy used. the severity of underly- combination therapy of TACE or TARE plus an ablative ing cirrhosis. The ly needed. While various LRTs are provided in this INVASION AND/OR METASTATIC setting. Therefore. 67. The AASLD recommends the use of systemic vascular invasion and/or metastatic disease) in the setting therapy over no therapy for patients with Child-Pugh of underlying Child-Pugh class A/B cirrhosis. underlying sion and/or metastatic disease) represent a unique clin- liver disease. FUTURE RESEARCH Additional data on the efficacy of TARE and external BACKGROUND beam radiotherapy is anticipated in the near future. with TACE alone. many patients with lim- ited extrahepatic metastatic disease burden and concur- CIRRHOSIS AND ADVANCED HCC rent macrovascular vascular invasion have been treated WITH MACROVASCULAR with LRT. The RCTs analyzed in option. Vol. because there was studies. location. The patients who benefited the most from the combination therapy were those with preserved liv. It was not possible to make a recommendation for er function. It is unknown LRT based on the available evidence. and degree of liver dysfunction) are most ical challenge. Patient fac. whether RFA or PEI would be equivalent to external 4. SYSTEMIC THERAPY OR LRT OR and thus far. the presence of concurrent macrovascular invasion 10. 2.HEPATOLOGY. 2. and when demonstrated significant benefit to the combination patients have very poor performance status and/or therapy for OS at 1. 1. class A cirrhosis or well-selected patients with Child- Pugh class B cirrhosis plus advanced HCC with macro- EVIDENCE AND RATIONALE vascular invasion and/or metastatic disease. and although the authors concluded that com. Efforts will likely need to focus on defining which patient Patients with advanced HCC (macrovascular inva- characteristics (tumor number. It is not possible to identify a preferred type of criteria were variable among the studies. NO THERAPY? the prognosis remains poor.(90) This analysis rhosis. this trial were all from Asian centers. and patient’s performance status.

one comparison with placebo.87) and demonstrated a trend for improve.115). The median 226 patients randomized. HR.39-1. 95% afenib group (4.011).(98) compared the efficacy of (sorafenib. median OS in comparison with placebo in the whole study population (sorafenib. Sorafenib significantly improved the median OS in the 40 received sorafenib and 57 received radiation.7 an survival did not differ significantly between the sor- months versus placebo.94. In an observational placebo arm.55-0. 8. hepatic arterial infusion chemotherapy (HAIC)— HR. No RCTs have been pub- treated with either sorafenib or LRT. 10. ing sorafenib versus placebo in patients with advanced and HAIC) combined with sorafenib versus sorafenib HCC and underlying Child-Pugh B cirrhosis alone is being studied in ongoing phase 3 clinical trials (NCT01405573).3 months. 0.68. tion. OS was significantly longer in the HAIC group than rovascular invasion and 155 (69%) patients had extra. 95% CI. 8. and with metastatic disease to either lungs or lymph clinical trials with combined strategies using sorafenib nodes (HR.68. There have been four published (NCT01730937.(97) compared the hepatic disease. 0. There were no comparative patients with macrovascular invasion and 1220 patients trials and only a few noncomparative studies that with metastatic disease. Of the 15 stud. Two phase 3 trials are compar- The definitive benefits of sorafenib in advanced ing the survival benefits of sorafenib versus radioembo- HCC with underlying Child-Pugh class B cirrhosis lization in advanced HCC with macrovascular has not been clearly established. Of note. four were RCTs. Medi- entire population included in the study (sorafenib.69.HEIMBACH ET AL. and the extent of metastatic patients with advanced HCC with PVT in the main disease was only provided for lungs and lymph nodes.82. NCT01829035. retrospective study. 95% CI. Nakazawa et al. macrovascular invasion or metastatic disease. 7. 0. 0. Sorafenib significantly improved the 0. For patients with metastatic 376 . Song et al.(92.9 CI. progression and OS than the chemoembolization alone 159 patients (53%) had extrahepatic disease versus the and sorafenib groups (P < 0. there were an additional designed to compare the outcome of sorafenib with 2001 patients enrolled in the sorafenib arm. In another retrospective observa- ment both for patients with macrovascular invasion tional study. January 2018 is summarized in Supporting Table 7. HEPATOLOGY.03) with advanced HCC with macrovascular invasion.(96) The placebo arm. ease) is a randomized phase 3 trial with sorafenib in Specific to patients with macrovascular disease.50-0. there were 108 196) with sorafenib (n 5 66) in patients with advanced patients (35%) with macrovascular invasion versus the HCC with portal vein thrombosis (PVT).57-1. Of the 97 patients included. linifa- ies identified. in the sorafenib arm.49-0.5 months.95) Collectively. the extent of macrovascular survival benefits of sorafenib versus radiation in invasion was not detailed.9 months. 0. reinforcing the benefits of sor- addressed the question of whether patients should be afenib in advanced HCC. TACE alone (n 5 295) or TACE with radiation (n 5 patic metastasis.64-1. 0.1 months versus placebo. TACE/radiation group had longer median time to rovascular invasion. HR.15). 6. 0.3 months) and the radiation group (5. of the sorafenib in advanced HCC with PVT. 80 (35%) patients had mac.5 months versus place. lished to critically assess the relative benefits of one evidence that exists in patients with advanced sorafenib versus LRT in advanced HCC with either HCC (macrovascular invasion and/or metastatic dis. 4. the added benefits of LRT (radiation.93) and demonstrated a positive trend in both patients with Given the recognized poor prognosis for patients macrovascular invasion (HR.91) mixed with chemotherapy released in the artery—with Similarly. 0.93) and for patients with which involves an actual infusion catheter directly in metastatic disease (sorafenib. NCT01482442). trunk or its first branch. single-center retrospective observational study (N 5 of the total 602 patients enrolled. 95% CI.(92. months. In the sorafenib arm. Additionally.001). NCT02774187.63.2 months. TACE. which had 123 patients (41%) with mac. 0.(52. P 5 hepatic disease. which had 150 patients (50%) with extra. In addi- randomized phase 3 trial conducted in Italy is evaluat. in the Asia-Pacific phase 3 trial. the hepatic artery as opposed to embolized particles bo. FUTURE RESEARCH bo. HR. 95% CI.93) and LRT are ongoing. The only level. phase 3 randomized trials comparing sorafenib versus and NCT01214343). though an ongoing invasion (NCT01135056. in the sorafenib group (7. In the pivotal SHARP trial. The four RCTs were not nib. 95% CI. 4.18).9 months versus place. and the other 11 were nib) or the combination of sorafenib with erloti- observational studies. 231 patients had 557) has attempted to compare the relative benefits of macrovascular invasion and 309 patients had extrahe.9 months. P 5 0. 8. 0. 0. either other targeted agents (sunitinib. 0. with 688 LRT in advanced HCC. brivanib.85.1 versus 5.

McGlynn KA. Laversanne M. Jung DH.328:1490... Falck-Ytter Y.D.122:3430-3436. 21) Manini MA. Phase 8) Kanwal F. Barbaro of hepatocellular carcinoma in Eastern versus Western popula. Goetz MB. Contrast-induced nephropathy in CT: incidence. Update on biomarkers review. Wilson SR.130:417-422. Zeringue A. Michael W.L. Jawad 16) Tao SM. PS. Piscaglia F.. No. Todd Stra.S. M. Elia F..D. curative treatment. J Hepatol 2014. (Chair). States forecast through 2030. who of screening for hepatocellular carcinoma. Yang B-H. HEPATOLOGY 2008. Fasiha Prospective validation of the noninvasive diagnostic criteria for Kanwal. M. Choi SJ...D. Talwalkar. Int J Cancer 2016.. Hassan Murad.. Valery PC. LV. ment of the practice guideline and provided the peer 14) Chaiteerakij R. Piscaglia F. 18) Jang H-J. Simon P. McGlynn KA. Increasing prevalence of HCC and cirrhosis in nib are ongoing in an attempt to improve survival in patients with chronic hepatitis C virus infection.D. Comparison 22) Pompili M.A. M.D. risk factors Fredric Gordon.D. Cancer 2016. Fornari F. (Commit.. Merriman. Kwon OS. M. M.. Management of hepatocellular carcinoma: AASLD guidelines for the diagnosis of hepatocellular carcinoma. Ulahannan S. The impact of vascular and nonvascular findings on the M.D. Piscaglia F. Zhu AX. F.H.122:1757-1765. HEPATOLOGY 2011. 15) Thomsen HS.D. 17) Forner A. Camaggi V..A. and Michael Volk. Tan WL. Leise. Shim YS. Whitney E.. Dig Liver Dis 2015.34:280-287. Asch SM. Ayuso C. F. dations. 67. Fuller SR. Jayant A.D. Dig 4) Petrick JL.D. Orefice R. Nephrogenic systemic fibrosis: history and epide- Tram T. Christine et al. 1. Bianchi L. Population attributable fractions in cirrhosis. et al. et al.11: mittee approved the scope and directed the develop- e1001624.S.26:3310-3318. Radiol Clin North Am 2009. et al.D.D. M.A. S..47:827-831.S. 9) El-Serag HB..A. 1978-2007. 20) Leoni S. gy 2011.. (Board Liaison). Roberts LR.D.60:995-1001. cancer. M.S. J Radiol 2009. Addissie BD. Tiro J.A. R. Grading quality of evidence and strength of recommen. Eur M. Characterization of primary and recurrent nodules in liv- er cirrhosis using contrast-enhanced ultrasound: which vascular 1) Atkins D. M. et al.P. 3) Choo SP. M. Semeraro S. Graubard BI. Wichmann JL.. 365:1118-1127. Donato F. Vidili G.S. Julie Heim.A.D. Early detection. ease (NCT01761266 and NCT02576509).R.40:206-215. Bray F.13: B. 3 trials comparing lenvatinib or nivolumab with sorafe. 599-609. Golfieri R.P. F. International trends in liver cancer incidence. F. Kim YS. 19) Leoni S.H.47:97-104. cularization of small nodules arising in the cirrhotic liver. M. BMJ 2004. Manini MA. LJ.S.. M.HEPATOLOGY.D. M.D. Sole M.A.P. Iavarone M. F. Hepatocellular sorafenib with or without stereotactic body radiation in carcinoma in cirrhosis: incidence and risk factors.S..P. Altekruse SF.A.A. et al. Kramer JR. Briss PA. Kelly SP. Riccardi L. Schoepf UJ. Horslen.H. M. Small nodules (1–2ccm) in liver cirrhosis: characterization with contrast-enhanced ultrasound. Zagni I. M. et al. Tran.L.. Fraquelli M.A.L. J Natl Cancer Inst Monogr Acknowledgment: This practice guideline was pro. Rosenberg lular carcinoma in cirrhosis. Joseph Ahn. vitz. The diagnostic and economic impact of 139:1534-1545. Forzenigo overall and by histologic subtype. Goh BK. Eccles M. J Clin Oncol 2016. approach. disease. Fanigliulo L. Contrast-enhanced ultrasound assessment of arterial vas- tions.D.. J Cancer Res Clin participated in the selection of the clinical questions Oncol 2004. 1112) through a randomized phase 3 trial comparing 7) Fattovich G. contrast imaging techniques in the diagnosis of small hepatocel- 5) Petrick JL. Members of the committee included Raphael of hepatocellular carcinoma.D. 2015. J Natl Cancer Inst 1980. and strategies for prevention. hepatocellular carcinoma. and survival rates for hepatocellular carcinoma surveillance in patients with cirrhosis: a meta-analysis.140:1182-1188.53:1020-1022. Tang Z-Y.A.47:775-782.C. Helen S.D.D.D. Romeo R.I. et al.C. Marinelli REFERENCES S. Tai WM.D.A. PLoS Med 2014. M. and provided expertise regarding the GRADE 13) Singal AG. Hoang T. et al. noninvasive diagnosis of small hepatocellular carcinoma based on the EASL and AASLD criteria. Fried. Biolato M. Borghi A. 237-245. Stroffolini T..1997:15-19.. Am J Gastroenterol 2010. Screening fundamentals. duced in tandem with three de novo systematic 11) Cole P. Vilana R. Hepatocellular carcinoma. F. Eur Radiol 2016. 2018 HEIMBACH ET AL.A.D. The AASLD Practice Guidelines Com. Lu GM. Flottorp criteria should be adopted? Ultraschall Med 2013. Pini P. small atypical hepatocellular carcinomas from dysplastic nodules Duffy AG. Future of hepatocellular carcinoma incidence in the United 24) Shin SK.127(5 Suppl.. bach. there is an attempt to assess the added benefits of risk factors for hepatocellular carcinoma in the United States. Best D.L. an update. 12) Zhang B-H. M. M.e1.L.L. Clin Gastroenterol Hepatol 2015. 10) Smith R.D.A. Jacqueline G. O’Leary.S.. Contrast-enhanced ultrasound for the differentiation of 6) Makarova-Rusher OV.D. Kim TK. miology.L. Granito A. F. Zhang Ahmad. M.. Michael L. F.D. M. Basic issues in population screening for reviews that were written by the same writing group.72:418-424. McNeel TS. Gastroenterolo- gy 2004. N Engl J Med 2011. 377 . Sangiovanni A.S. Diagnosis of hepatic nodules 20 mm or smaller in cirrhosis: Hsu. Randomized controlled trial including M.D. 1):S35-S50. Braunlin M.. James R. Schilsky.A. Te. Gastroenterolo- patients who have advanced HCC with metastatic dis.64:1263-1272. Jackson. 23) Sangiovanni A. patients with advanced HCC (NCT01730937).59:638-644. tee Liaison). Liver Dis 2008. Ayuso JR. Altekruse SF.. Amit Singal. Clinical and economical impact of 2010 2) Bruix J. of stereotactic body radiation to sorafenib (RTOG Cancer 2016. Michael D. Gut 2010.34:1787-1794. (Vice-Chair). Pillai A.D. Vol. Sherman M. et al. M. M. B. Morrison AS.105: F. Spivey. Galassi M. M..

placebo-controlled trial. A ran- 28) Khalili K. Wu H. Non-invasive diagnosis of hepatocellular carcinoma 2 cm F. et al.59:89-97. Guindi domized trial comparing radiofrequency ablation and surgical M. Kyoung Kim T. Long-term effectiveness of resection and radio- 33) Yu JS. Zhou WP. Hecht EM. Cucchetti A. Bedossa P. Marrero JA. MRI in the cirrhotic liver. Cochrane Database the outcome? Seeding risk for percutaneous approach of HCC. Effect and safety of inter- MR imaging: evaluation in nodules 20 mm or smaller detected feron for hepatocellular carcinoma: a systematic review and meta- in cirrhosis at screening US. Wang J. Takasaki KT. Mamidipalli A. randomised. Li AJ. Kim DC. 2000. Liver Imaging Reporting and Data System with 55) Zhuang L.hrsa. Garcıa-Criado A. Adachi S. Kochak Yazdi L. observations detected at CT and MR imaging. 44) Chen M-S. Yang Z. Lee JH.org/quality-safety/resour. tis B virus-induced mild cirrhosis. Small resectable multiple hepatocellular carcinoma beyond Milan crite- hypervascular enhancing lesions on arterial phase images of mul. Rofsky NM. Cancer 35) Khan AS.237:938-944. Afdhal NH.55:800-806.gov/governance/policies. Outcome of small local ablative therapy and partial hepatectomy for small hepato- (10-20 mm) arterial phase-enhancing nodules seen on triphasic cellular carcinoma. Ayuso C. Indeterminate 1-2-cm nodules found on hepa. et al. 50) Muto Y. Piccirillo MC. Yao N. de Matthaeis N. Lancet Oncol 2015. Matsui O. 54) Wang J. raphy and magnetic resonance imaging for the diagnosis of small 26) Organ Procurement and Transplantation Network 2016.61:82-88. Sherman M. Forner A. polyprenoic acid. Li A-J. Yan L. rhotic liver: evaluation and comparison with whole explanted liv. D’Onofrio L.33:437-447. Hahn WY. Cheng Z. Xia F. Federico B. 2016. Mazzaferro V. some. What is the risk and interventions for hepatocellular carcinoma.56:1317-1323. Ricci S. et al. Marelli L.54:2048-2054. Radiology 2016. PLoS One 2015. January 2018 25) American College of Radiology. Chung J-J. Pelletier SJ. Shi H. HEPATOLOGY 2011. Cancer Treat Rev 2007. 52) Llovet JM. Ninomiya M. Du L.275:698-707. Barrau V. Reig M.32:453. Liu C-L. Bianchi L. Zheng Y. Zhang YC. Acta Radiolog. ing rim in magnetic resonance imaging diagnosis of small hepa. Different risk factors and prognosis for early and late intrahepatic 459. 49) Poon RT-P. Partial 1659. Haider MA.32:39-45. Radiology hepatocellular carcinoma: a randomized double-blind placebo- 2011. optn. dimeglumine-enhanced hepatic arterial-phase MRI of the hepati. 48) Vincenzi B. Jakobsen JC. J Hepatol 2014. recurrence after resection of hepatocellular carcinoma. Ann Surg. PLoS One 2013. N Engl in cirrhosis. A prospective randomized trial comparing percutaneous 30) O’Malley ME. Prevention of second primary tumors by an tocellular carcinoma in the cirrhotic liver. Lai ECH. Tremosini S. Pan Z-y. J Hepatol 2012.16:1344-1354. HEPATOLOGY. Hussain HK. phases on dynamic magnetic resonance imaging. et al. doi: http://www. Can J Gastroenterol 2013. Davies N. Vullierme M-P. Gane E.100:1523-1528. Small hypervascu. Accuracy and disagreement of computed tomog- ces/LIRADS. Kim KW.103:348-356. Wolfson T. Diagnostic accuracy assessing fat. Vilana R. Moriwaki findings for characterizing small (1–2-cm) hepatic nodules in H. Yang J. Ma K. Clin Imaging 2008. Menezes 1996. Analysis of gadobenate dimeglumine–enhanced MR 51) Okita K. Yu D. Fan S-T.243:321-328. J Hepatol 2012. Costa EAC. 43) Feng K. intensity at dynamic MRI. Kwak HS.57:794-802. J Comput AssistTomogr 2008. Accessed on December 1. et al. Lee YH. lar hepatocellular carcinoma: limited value of portal and delayed Results of a multicenter Italian survey. Li J-Q. 3. Blanc J- et al. Mazzaferro V.334:1561-1567.10:e0133488. Sherman M. controlled study. 46) Yin L. Kanki A. 53) Bruix J. The clinical quency ablation versus partial hepatectomy for hepatocellular car- outcome of small (<20 mm) arterially enhancing nodules on cinoma within the Milan criteria. Spoto C.259:730-738. Li X.102:1654. Lee KH. RJ. Li H. Kaneko S. Randomized clinical trial of chemoembolization plus radiofre- 31) Byrnes V.89:500-507. Reig M. of adjuvant therapy after potentially curative treatment 39) Darnell A. Radiology 2005. et al. analysis. Am J 45) Liu H. Kudo 38) Tanabe M. Ozenne V.359:378-390. Babb JS. et al. hepatectomy vs transcatheter arterial chemoembolization for 32) Hwang SH. et al. HEPATOLOGY 2012.  for hepatocellular carcinoma. et al. carcinoma.transplant. et al. ria: a RCT. liver CT in patients with cirrhosis or chronic liver disease. Liang H-H. 37) Rimola J. et al. Yan J. Kim KW. in patients with hepatocellular 2010. Clinical according to EASL criteria and mRECIST criteria in hepatocel- implication of small (<20 mm) enhancing hepatic nodules lular carcinoma patients treated with loco-regional therapies: a observed only during three-dimensional gadobenate literature-based meta-analysis. Zeng X. Gastroenterol 2005. et al. ica 2008. only during the hepatic arterial phase at MR imaging of the cir. tocellular carcinoma surveillance: biopsy for all. http:// hepatocellular carcinoma and dysplastic nodules: role of biopsy. Wong J. Kim JH. Ng IO-L. A random- 29) Holland AE. data system. Rimola J. Izumi N. et al. 378 .32:360-366. M. Jang H-J. Ann Surg 2006. Takayama Y. J Hepatol 2013. Fu SY.27:351-363. Farges O. J Gastroenterol 2015. Kim CS. Am J Gastroenterol 2007. et al. tiphase dynamic computed tomography in cirrhotic liver. Importance of small (20-mm) enhancing lesions seen tion in the treatment of small hepatocellular carcinoma. Franke A.49:735-743. Lau WY. Han YM. N Engl J Med 36) Kim TK.252:903-912. Jang HJ. Saviano A. A meta-analysis 281:152173. Ardito F. Imaging outcomes of Liver Imaging resection or ablation (STORM): a phase 3. Wang ZG. Lo C-M. Pan ZY. et al. Meng Z. 2016. He XD. Saito A. and 4 blind. Liang WJ. Sorafenib in advanced hepatocellular carcinoma. Patch D. 47) Pompili M. J Magn Reson Imaging acyclic retinoid. Schoppmeyer K. Syst Rev 2013:1-73. Zhang Y- er. Adjuvant sorafenib for hepatocellular carcinoma after Ferreira MPFD. Tanaka K. Kim JH.50:191-202. Burroughs 41) Weis S. Jacks LM. Forner A.acr. AK. et al. frequency ablation for single hepatocellular carcinoma 3 cm. double- Reporting and Data System version 2014 category 2. Di Maio M. Silletta M. Q. Lee ized controlled trial of radiofrequency ablation and surgical resec- VS.8:e61361. or none? 2010. Takayama T. 42) Huang J. 27) Stigliano R. Wang S. Hepatoma Prevention Study Group. Hilgard P. M€ ossner J. Chau G-Y. Weadock WJ. Yu J-S. capsule and signal J Med 2008. Br J Surg 2016. Kiryu S. Value of delayed hypointensity and delayed enhanc. Guo R-P. Moriwaki H. resection for HCC conforming to the Milan criteria. Chung JJ. Johnson TD.HEIMBACH ET AL. Prognostic relevance of objective response 34) Kim YK. Seeding following percutaneous diagnostic and therapeutic Radiofrequency (thermal) ablation versus no intervention or other approaches for hepatocellular carcinoma. Peretinoin after curative therapy of hepatitis C-related patients at high risk for hepatocellular carcinoma. Radiology 2015. Liver imaging reporting and 40) Serste T.

379 . J Hepatol 2015. Fontes P. Rodriguez M.15:3169-3177. J Hepatol 2001. Lam CM. Garcia-Valdecasas JC. et al. Ou TM. atic treatment in patients with unresectable hepatocellular carcino- 68) Toso C. et al. Coll S.HEPATOLOGY. 2018 HEIMBACH ET AL. Arterial embolisation or chemoembolisation versus symptom- Lancet Oncol 2012. Freise CE. et al. pensity score landmark analysis. et al. cellular carcinoma: an international consensus conference report.27:1578-1583. Obi S. Muscatiello N. improves survival. Gay F. 85) Facciorusso A. Perrier A. Merion RM. Castells A. et al.37:429-442. domized study with 131 cases. Bridging locoregional therapy for hepatocel. Coll S. Ink O. downstaging for hepatocellular carcinoma. 60) Mehta N. J Hepatol 2016. Hickey R.44:528-538. et al. Monta~ na X. Yoshida H. N Engl J 78) Groupe Francophone d’Etude et de Traitement du Carcinome Med 1996.334:693-700. 81) Pelletier G. Barone M. Clin Transplant 2008. Am J Transplant 2008. Elsayed Z. Clin cellular carcinoma. Liu CL. Qian BY. 83) Llovet JM. CL. I~ narrairaegui M. Semin Liver Dis 1999. Bruix J. able hepatocellular carcinoma (HCC). Multicentre randomised phase III trial comparing with the “wait and not ablate” approach until meeting T2 criteria Tamoxifen alone or with transarterial lipiodol chemoembolisation for liver transplant listing. Gores GJ. Kerlan RK. 71) Kim PTW. Bossuyt PMM. Yao FY. in patients with unresectable hepatocellular carcinoma. 1990. Juzi JT. Wojtaszek M.359:1734-1739. A multicenter randomized trial. HEPATOLOGY cation policy. World J 70) Hołowko W. HEPATOLOGY 2003. Hirose R. Singal AG.20:764-768. Roberts JP.24:625-629. Liver Trans. Recommendations for liver transplantation for hepato. et al. Sato S. Adam R. Roberts JP.22:469-475. ized trial. Bilbao JI. Mentha G. Lin H-C. Treatment of unresectable hepatocellular carci- United States. Lancet 2002. Castells A. 73) Llovet J. Monta~ na X. Jakobs T. Wong GL-H. Br u C. Statin and the risk of hepatocellular carcinoma and percutaneous tumor ablation in patients with hepatocellular carci- death in a hospital-based hepatitis B-infected population: a pro. 58) Hsiang JC. hepatocellular carcinomas in patients with cirrhosis. ment in patients with advanced hepatocellular carcinoma: results ity in patients with small hepatocellular carcinoma undergoing of a randomized. Huang Y-H. Marsh JW. Davern TJ.8:958-976. Postop- 63) Bismuth H. Transarterial embolization versus symptomatic treat- Y-Y. 1. No. Bonnetain F. deVera MB. Kallini J. Bouche O.7:2009-2019. J Hepatol E364. Evaluation of transcatheter arterial embolization prior to Chan HL-Y. controlled trial in a single institution.19:311-322. 66) Freeman RB Jr. J Hepatol 2010:52: 84) Llovet J. Amesur NB. Mouli S. Validation of the HCC-MELD for dropout probabil. Derhy S. Sun Y. 87) Kolligs FT. FY. Holloway SE. Intention to treat outcome of T1 hepatocellular carcinoma Fratte S. Wong VW-S. for unresectable hepatocellular carcinoma in cirrhotic patients 61) Mazzaferro V. Liver Int 2004. Wang J. sion compared with chemoembolization in patients with hepato- plant 2015. Real MI. Merriman R. Langer B. Lesurtel M. have prognostic power. Fuster J. Wetterslev J. Vetter D. Ngan H.35:1715-1721. Transarterial chemoemboliza. Lee CC. Anciaux ML. 74) Oliveri RS. Vilana R. Cochrane carcinoma.e2. et al. lization for unresectable hepatocellular carcinoma. Systematic review of randomized trials for et al. et al. A randomized trial of hepatic arterial chemoembolization with hepatocellular carcinoma.63:1190-1197. Fujishima T. et al. Regalia E. Cui YL. Tso WK. Majno P. Ayuso 59) Huo T-I. Kneteman NM. sation for unresectable hepatocellular carcinoma. Wr oblewski T. antivirals. Pulvirenti A. Waljee AK. The place of ma: a randomised controlled trial.18:A58. Aponte J. Chiang J-H. 56) Chen K. Eur J Bozzetti F. World J Hepatol 2015. Liver transplantation for the treatment of small Cancer 2008. Jennings LW. Hepatocellulaire. noma: a randomized controlled trial.29:129-134. Planas R. Tse Y-K. HEPATOLOGY 1998. Current systemic treatment of hepatocellular carcinoma: a unresectable hepatocellular carcinoma: chemoembolization review of the literature. Davis GL.21:1142-1152.27:311-318. Chiou MDC. Liver Transplant 2003. McKenna GJ. noma with lipiodol chemoembolization: a multicenter random- 67) Clavien P-A. Ducreux M. cellular carcinoma. Chinnakotla S. et al. Downstaging hepatocellular et al. Farmer DG. Risk factors for liver transplant waitlist dropout in patients et al. Chemoembolization improves survival in patients with unresect- 69) Heckman JT. et al. Andreola S. 2002. Luboinski M. Clin Transplant 2012. Bass NM. Nikolai B.9:684-692. Ziarkiewicz-Wroblewska B. et al. Roche A. Majno P. 88) Salem R. et al. Y90 radioembolization significantly prolongs time to progres- carcinoma: a systematic review and pooled analysis. 79) Li Q. Yip TC-F. Bolondi L. Aponte J. Liver transplantation for hepato. et al. 75) Akamatsu M. Kerlan R. Fidelman N. et al. Rougier P. OGY 1998. Real MI. Treatment of unresectable hepatocellular carci- 62) Llovet JM.65:727-733. Gluud C. 930-936. Dig Surg 2006. Abergel A.13:e11-e22. Gra˛t M. Sarkar M. Horng CT. Nagel JM. Transplant 2013.26:E359. Berg 82) Pelletier G. 72) Parikh ND. Gastroenterology 2016. out from the waiting list for liver transplantation in patients with Randomized controlled trial of transarterial lipiodol chemoembo- hepatocellular carcinoma: implications for the current organ allo. 76) Bruix J. 65) Park S-J. Di Leo A. Cochrane Data- lar carcinoma in HCV-related cirrhosis treated with direct-acting base Syst Rev 2011:1-60.11:181-184. Dodge JL. 67.35:1164-1171. 80) Lo C. Scuteri A. Gordon AC. Guidinger MK. noma (HCC) by lipiodol-targeted transcatheter arterial chemo- Grande L. Hepatol 2015. Onaca N.34:11.7:1412-1420. Kobry n K. chemotherapy for patients with hepatocellular carcinoma: a ran- 64) Yao F. Tumor biology and pre-transplant locore. Early occurrence and recurrence of hepatocellu. Bruix J. Ann Surg Oncol Transarterial chemoembolization: evidences from the literature 2008.151:1155-1163. Yao 77) Doffo€el M.27:1572-1577. Llovet JM. Pilot randomized trial of selective internal gional treatments determine outcomes in patients with T3 hepa. Planas R. A follow-up analysis of the pattern and predictors of drop. J Hepatol 1998. and applications in hepatocellular carcinoma patients. (Federation Francophone de Cancerologie Digestive 9402). Su C-W. Poon RT. erative transhepatic arterial chemoembolization and portal vein cellular carcinoma. radiation therapy vs chemoembolization in unresectable hepato- tocellular carcinoma undergoing liver transplantation. lular carcinoma prior to liver transplantation. Koike Y. carcinoma: the tumor-node-metastasis classification does not HEPATOLOGY 1993. Ann Transplant 2015. 1997-2006. Liver transplantation for small hepatocellular embolization (TACE). Hagège H.22:178-187. Database Syst Rev 2016:1-31.23:235-240. Vol. Groupe CHC. Crespi C. Yttrium-90 microsphere radio- tion prior to liver transplantation in patients with hepatocellular embolisation for unresectable hepatocellular carcinoma. Transarterial (chemo)emboli- Caraceni P. Liver Int 2015. et al. Doci R. Liver and intestine transplantation in the Thong D. Hsu CW. Steffick DE. et al. HEPATOL- locoregional therapy. Gabr A. Licinio R. 57) Conti F. Liver Transpl 2016. 86) Abdel-Rahman OM. Tsai YY. Buonfiglioli F.

BMC Pacific region with advanced hepatocellular carcinoma: a phase Gastroenterol 2014. Shim JH. Bae SH. Lancet 98) Song DS. et al.com/doi/10.33:559-566. Qin S. Jung J. Transcatheter arterial chemoembolization 96) Kim G-A. Qin S. Effica.33:172-179. Transarterial 95) Zhu AX. cinoma according to baseline status: subset analyses of the phase III Sorafenib Asia–Pacific trial. Mazzaferro V. al. Takada J. Tanaka Y. Bolondi L.HEIMBACH ET AL. Ryu M-H. et al. portal vein tumor thrombosis.50:445-454. portal vein tumor thrombosis: propensity score analysis. et al. Chung IJ. Kang Y-K. 2014. Zhu Y. A comparative study between sorafenib and hepatic arterial infu- 93) Cheng A-L. HEPATOLOGY. Song MJ.26:320-329. J Clin Oncol 2014. Tsao C-J. Raoul J-L. et al. Craxi A. sion chemotherapy for advanced hepatocellular carcinoma with cy and safety of sorafenib in patients with advanced hepatocellular car. January 2018 89) Fu Y. placebo-controlled trial. Supporting Information tocellular carcinoma: results of a randomized phase III trial. tion therapy and sorafenib for advanced hepatocellular carcinoma 91) Bruix J. et al. Huang WT. Jang JY. Ross PJ.wiley. Yoon SM. Pan H.10:25-34. Chen Z. Evans TRJ. radiotherapy in unresectable hepatocellular carcinoma with major et al. JAMA Oncol 2015. J Clin Oncol Med 2015. 97) Nakazawa T. Santoro A. Chen Z. 90) Huo YR. et al. plus radiotherapy compared with chemoembolization alone for et al. Linifanib versus sorafenib in patients with advanced hepa. 380 .29086/suppinfo. Overall survival in response to sorafenib versus 92) Cheng A-L. Shibuya A. Yun Q. randomized. Chung WJ. SEARCH: a phase III.8:10388-10400. Comparison of chemoembolization with and without radia- hepatocellular carcinoma. double- (PEI) for the treatment of unresectable hepatocellular carcinoma: blind. 94) Cainap C. chemoembolization (TACE) plus percutaneous ethanol injection Carrilho FJ. Kim JH. Zhu X. Kim JS. Guan Z. placebo-controlled trial of sorafenib plus erlotinib in a meta-analysis of randomized controlled trials. Additional Supporting Information may be found at onlinelibrary. Int J Clin Exp patients with advanced hepatocellular carcinoma.14:84. Okuwaki Y. Li Q.1:756. Kim YS.e6. with portal vein tumor thrombosis: a propensity score analysis. J Gastroenterol 2015. Tsao C-J. Oncol 2009. Cheng Y. Efficacy and safety of sorafenib in patients in the Asia. Hidaka H.1002/hep. J Hepatol 2012. III randomised. Eur J Cancer 2012. Qin S. Kim JS. et al. Eslick GD. Sherman M. double-blind.48:1452-1465.57:821-829. Rosmorduc O. Zhao X. advanced hepatocellular carcinoma: subanalyses of a phase III tri. Efficacy and safety of sorafenib in patients with J Vasc Interv Radiol 2015.