DigestiveandLiverDisease50(2018)1105–1114
ContentslistsavailableatScienceDirect
Digestive
and
Liver
Disease
j o ur n a l ho me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / d l d
Review
Article
Recurrence
of
hepatocellular
carcinoma
after
direct
acting
antiviral
treatment
for
hepatitis
C
virus
infection:
Literature
review
and
risk
analysis
Maria
Guarino
a,1,
Luca
Viganò
b,1,
Francesca
Romana
Ponziani
c,∗,
Edoardo
Giovanni
Giannini
d,∗∗,
Quirino
Lai
e,
Filomena
Morisco
a,
On
behalf
of
the
Special
Interest
Group
on
Hepatocellular
carcinoma
and
new
anti-HCV
therapies”
of
the
Italian
Association
for
the
Study
of
the
Liver
2aDept.ofClinicalMedicineandSurgery,GastroenterologyUnit,UniversityofNaples“FedericoII”,Naples,Italy
bDept.ofSurgery,DivisionofHepatobiliaryandGeneralSurgery,HumanitasClinicalandResearchHospital,HumanitasUniversity,Rozzano,Milan,Italy cDivisionofInternalMedicine,GastroenterologyandHepatologyUnit,PolyclinicFoundation“AgostinoGemelli”,IRCCS,CatholicUniversityofRome,Rome,
Italy
dGastroenterologyUnit,DepartmentofInternalMedicine,UniversityofGenoa,IRCCSPolyclinicHospital“SanMartino”,Genoa,Italy eHepato-bilio-pancreaticandLiverTransplantUnit,Dept.ofSurgery,SapienzaUniversityofRome,Rome,Italy
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:Received26May2018
Receivedinrevisedform26July2018 Accepted1August2018
Availableonline11August2018 Keywords: DAA HCC HCV Recurrence
a
b
s
t
r
a
c
t
Althoughstudiessuggestdecreasedincidenthepatocellularcarcinoma(HCC)aftertreatmentwith
direct-actingantivirals(DAAs)forhepatitisCvirus(HCV)infection,dataareconflictingregardingriskand
aggressivenessofrecurrenceinpatientswhohaveahistoryoftreatedHCC.Thisreviewanalysesdata
availableinliteratureinordertoelucidatetheimpactofDAAsontheriskofHCCrecurrenceafter
success-fultreatmentofthetumor.Overall24paperswereidentified.Theavailabledatacannotbeconsidered
definitive,buttheinitialalarmistdataindicatinganincreasedriskofrecurrencehavenotbeenconfirmed
bymostsubsequentstudies.Thesuggestedaggressivepattern(rapidgrowthandvascularinvasion)of
tumorrecurrenceafterDAAsstillremainstobeconfirmed.Severallimitationsoftheavailablestudies
werehighlighted,andshoulddrivefutureresearches.Thetime-to-recurrenceshouldbecomputedsince
thelastHCCtreatmentandresultsstratifiedforcirrhosisandsustainedviralresponse.Anycomparison
withhistoricalseriesisoflimitedinterestbecauseofanumberofbiasesaffectingthesestudiesand
differencesbetweenenrolledpatients.Prospectiveintention-to-treatanalyseswillbeprobablythebest
contributiontodriveclinicalpractice,providedthatarandomizedtrialcanbedifficulttodesign.
©2018EditriceGastroenterologicaItalianaS.r.l.PublishedbyElsevierLtd.Allrightsreserved.
1. Introduction
Hepatocellularcarcinoma(HCC)isthefifthmostcommon can-cerandthesecondleadingcauseofcancer-relateddeathworldwide [1].Inthelastdecades,agrowingincidenceofHCCwasobserved inmostdevelopedcountries,beinghepatitisCvirus(HCV)-related hepatitisoneofthemaindeterminantofthisphenomenon[2].The
∗ Correspondingauthor.
∗∗ Corresponding authorat: Gastroenterology Unit, Department of Internal Medicine,UniversityofGenoa,VialeBenedettoXV,No.6,16132,Genoa,Italy.
E-mailaddresses:francesca.ponziani@gmail.com(F.R.Ponziani), egiannini@unige.it(E.G.Giannini).
1 Boththeauthorsequallycontributedtothemanuscript.
introductionofsecond-generationdirectly-actingantivirals(DAAs) hasdramaticallychangedthescenarioofHCVinfection.DAAscan achievesustainedvirologicalresponse(SVR)ratesinapercentage
2StudyGroupMembers:AlessandroVitale,FrancescoPaoloRusso,UmbertoCillo,
PatriziaBurra,ClaudiaMescoli,MartinaGambato,AnnaSessa,GiuseppeCabibbo, MauroViganò,GiovanniGalati,EricaVilla,MassimoIavarone,Giuseppina Brancac-cio,MariaRendina,LuigiG.Lupo,FrancescoLosito,FabioFucilli,MarcelloPersico, RobertaD’Ambrosio,AngeloSangiovanni,AlessandroCucchetti,FrancoTrevisani eMatteoRenzulli,LucaMiele,AntonioGrieco,GianLodovicoRapaccini,Maurizio Pompili,AntonioGasbarrini,GiovanniBattisaLeviSandri,FabioMelandro,Massimo Rossi,IlariaLenci,TommasoMariaManzia,RaffaellaTortora,GiovanGiuseppeDi Costanzo,RodolfoSacco,DavideGhinolfi,ErionRreka,PaolaCarrai,Natalia Simon-etti,CarloSposito,SherrieBhoori,StefanodiSandro,FrancescoGiuseppeFoschi, AndreaCasadeiGardini,DanieleNicolini,SusannaMazzocato,AlbaKostandini,Paola Violi,UmbertoBaccarani,RiccardoPravisani,ValterVincenzi.
https://doi.org/10.1016/j.dld.2018.08.001
1106 M.Guarinoetal./DigestiveandLiverDisease50(2018)1105–1114 ofcasesashighas95–98%[3].Theirextensiveadoptionisexpected
toleadtoaprogressivedecreaseinHCV-relatedHCC,evenifan initialcumulativeincrease of HCC incidencecould beobserved becauseofthelowerriskofliverdecompensationandthehigher survivalexpectancyofcirrhoticpatientswithSVR[4,5].
Twocontemporary retrospectivestudies have advanced the hypothesisthatsuccessfultreatmentwiththenewDAAsmaybe associatedwithhigherincidenceandrecurrenceratesofHCC[6,7]. Moreover, an unusual increase of infiltrating tumors has been reportedbytheseauthors.Subsequently,controversialdatahave beenpublishedandseveralbiomolecular hypotheseshavebeen suggested[8].Thesefindingshavebeenobjectofdebateandcould havemajorimplicationsinthemanagementofHCCpatients.An urgentclarificationisneededtodriveclinicalpractice.
Thisreviewanalysesthedataavailableintheliteratureinorder toclarifytheimpactofDAAsontheriskofHCCrecurrenceafter successfultreatmentofthetumor.
2. Methods 2.1. Literaturesearch
A systematic search of PubMed, ScienceCitation Index, and Embasedatabaseswasperformedforarticlespublishedbetween January2014and January2018(cut-offdate February1, 2018) relevanttorecurrenceofHCCafterDAAs.Englishlanguage arti-cleswereselectedusingthekeywords‘hepatocellularcarcinoma’; ‘HCC’; ‘HCV’; ‘DAA/directly-actingantivirals’and ‘recurrence’to identifyallreports that maypertain to thereview issue. Man-ualcross-referencingwasperformed;andrelevantreferencesfrom selectedpaperswerereviewed.Casereportswereexcluded. 3. Results
3.1. Literatureselection
Overall, 24 papers (with 25 cohorts of patients) [6,7,9–30] reportingonHCCrecurrenceafterDAAstreatmentinpatientswith previouslyandsuccessfullytreatedHCC(liverresection,ablation ortrans-catheterarterialchemoembolization—TACE)were iden-tified.TwopapersonlyconsideredpatientswithHCCrecurrence [25,27],twoincludedalsopatientswhounderwentliver transplan-tationbeforestartingDAAs[16,29],andonefocusedonradiological featuresofrecurrences[28].ThepaperbyPettaetal.[12] com-paredtheriskofHCCrecurrencein untreatedpatientswiththe riskinpatientsreceivingDAAs,butthepaperwasexcludedfrom thepresentanalysisbecausethedataforthelattergroup(DAAs) wereobtainedfromthestudybyReigetal.[6].In2016Minami etal.reportedatime-to-eventanalysisonasmallcohortofpatients (n=27)[14],buttheypresentedanupdatedseries(n=163)[31]at theAmericanAssociationfortheStudyofLiverDisease(AASLD) meetingin2017.Boththesedatawereanalyzed.Callejaetal.[9] andBielenetal.[16]hadbeencontactedbytheauthorstoobtain additionaldataabouttheirseries.Finally, apapersummarizing theabstractspresentedattheAASLDLiverMeeting2017about HCCrecurrenceinDAAs-treatedpatientswasalsoincludedinthe presentreview[32].
3.2. Recurrencerisk
Thenumberofpatientsincludedinthestudiesrangesbetween 3and189.Mostpatientshadlimiteddiseaseburdenatbaseline, i.e. single HCC (67–94% of cases) [10,13–15,23,26,27,29]<5cm insize(rangeofmediansize1.8–4.7cm)[10,13–16,23,26,27,29] andBarcelona-ClinicLiverCancer(BCLC)stage0/A(61–100% of cases) [6,7,10,11,16,20,23,24,26,29]. The majority of patients underwent hepatic resection or percutaneous ablation.
Only four studies exclusively considered patients undergo-ing radical treatment [14,15,18,26], whereas 14 papers also included patients undergoing TACE or radiation treatments [6,7,10,11,13,16,17,19–21,23,24,29,31]and4papersdidnotdetail theinitialHCCtreatment[9,15,22,30].
Follow-up management after HCC treatment was rather homogeneous among studies (when detailed), being per-formed every 3–6months using Computed Tomography (CT), Magnetic Resonance Imaging (MRI) or ultrasonography [6,9–13,16,18,23,26,27,29].
Theshortest time betweenlast curative HCC treatmentand DAAsinitiationwasof0.5months[14].However,themean/median time was highly variable, ranging from 2 to 21.5months [6,7,9–11,13,15,16,19,21–23,25–27,29].Threestudiesreportedthe mediantime betweenHCCdiagnosisand DAAstreatment:46.8 [31], 22.8 [15] and 9.6 [30] months, respectively. Five studies reportedthemediantimebetweenthelastnegativeradiological assessmentandDAAstherapy:1.7monthsinthreestudies, maxi-mum3monthsinoneand14.4monthsinone[6,11,15,19,26].
Themean/medianfollow-upafterDAAsrangedbetween3and 21.6months[6,7,11,14,15,17–22,26,29,31].TheproportionofHCC patientstreatedwithDAAsthathadarecurrence(i.e.,numberof events/numberofpatientsatrisk)washighlyvariable,rangingfrom 0%to47.9%[6,7,9–11,13–17,19,21–24,29–31].Thesevaluesarenot ratesbecausetheydonotderivefromtime-to-eventanalyses.
ThedataoftheanalyzedstudiesaresummarizedinTable1and inSupplementaryTable1.
3.3. Time-to-eventanalyses
Sixteen papers from 15 study groups (Minami et al. published two series), including 17 cohorts of patients, performed a time-to-event analysis of HCC recurrence [6,9–11,13–15,17,18,20,21,23,24,26,29,31]. In two papers, this wasnotexplicitlyreportedinthemanuscript,butthedatawere extrapolatedfromthefigures[24]or thetables[17].Regarding thepaperbyReigetal.[6],thetime-to-eventanalysisperformed byCammàetal.wasconsidered[33].Finally,sinceTorresetal. collectedlessthan10patientsanddidnotreportrecurrences,their paperwasnotconsideredinthereviewoftime-to-eventdata[21]. Theresultsoftime-to-eventanalysesaresummarizedinTable2.
Twopapers(including3patientcohorts)[15,23]reportedthe rateofHCCrecurrenceper100person-monthsrangingfrom0.73 and1.73.ElKassasetal.reportedanadjustedrate(fortimesince HCCcomplete radiologicalresponse,sex,age,Child–Pughscore, andhistoryofgastroesophagealvarices)ofHCCrecurrenceper100 person-monthsof3.82afterDAAsexposure[26].
Twelve study groups performed a Kaplan–Meyer analysis [6,9–11,13–15,17,18,20,23,24,31], and one paper performed a Poisson regression [26].The time to recurrencewas calculated in different ways: from the last HCC treatment in 8 papers [6,9,14,17,18,20,23,29],fromtheinitiationofDAAstherapy in9 papers[9–11,13,15,17,20,26,31]andfromtheendofDAAstherapy inonestudy[24].Inthelasttwogroups,themean/mediantime betweenthelastHCCtreatmentandDAAsstartrangedfrom8to 21months.
Considering the HCC recurrence rate from the initiation of DAAstreatment,thefollowingfigureswerereported:9.6–23.0%at 6months,23.1–45.7%at12months,and38.9–54.5%at24months. Ofnote,4studiesreportedthenumberofpatientswhodeveloped HCCrecurrenceduringtheDAAstreatment[6,11,16,17].
ConsideringthetimesincethelastHCCtreatment,the recur-renceratewas0–7%at6months,8.7–30%atoneyearand27.9–42% attwoyears.Weplottedtheavailablecurvesofrecurrencesincethe lastHCCtreatmentinasinglegraph(Fig.1).Arecentmeta-analysis byCabibboetal.[34]theoreticallyprovidedthebenchmarkforHCC
M. Guarino et al. / Digestive and Liver Disease 50 (2018) 1105–1114 1107 Table1
Characteristicsoftheanalyzedstudies.
Author Design # CirrhosisN(%) SVRN(%) FUProtocol (months) Intervallast assessmentof complete response— DAAs(median, months) Intervallast HCCtreatment —DAAs (median, months) Interval DAAs— recurrence (median, months)
PreviousHCCtreatment Proportionof recurrenceseN
(%)
Type Single/multiple
N(%)
Reigetal.[6] Retrospective 58 55(94.8) 39/40(97.5)a 6 1.7 11.2 3.5 Resection,ablation,
TACE
NA 16(27.6)
Callejaetal.[9] Retrospective 70 55(78.5) 66(94.3) 6–9 NA 20d 6.7 NA NA 21(30)
Ikedaetal.[10] Retrospective 177 NA 155(89.6) 3–4 NA 10.7 20.7 Resection,ablation,
TACE,PRT
89(50.3)/88 (49.7)
61(34.5)
Cabibboetal.[11] Prospective 143 143(100) 138(96) 3–6 1.7 NA NA Resection,ablation,
TACE
101(70.6)/42 (29.4)
29(20.3)
Ogawaetal.[13] Prospective 152 90(59.2) 152(100) 3–6 NA 14.4 NA Resection,ablation,
TACE,PRT
NA 26(17.1)
Minamietal.[31] Retrospective 163 NA 150(92) NA NA NA NA Resection,ablation,RT,
TACE
65(39.8)/98 (60.1)
78(47.9)
Contietal.[7] Retrospective 59 59(100) 53(89.8) NA NA 12 NA Resection,ablation,PEI,
TACE
NA 17(28.8)
Poletal.ANRS− CO22Hepater
[15]
Prospective 189 152(80) 148/161(91.9)a NA 14.4 NA NA NA NA 24(12.7)
Poletal.ANRS− CO12CirVir[15]
Prospective 13 13(100) 8/8(100)a NA NA atleast3 37.1 Resection,ablation 13(100) 1(7.7)
Bielenetal.[16] Retrospective 19 13(81.2) 15(78.9) 6 NA 10difresection 21 Resection,ablation,
TACE
NA 6(31.6)
3ptsnodata 14difablation
Shimizuetal.[17] Retrospective 23 15(65.2) 23(100) NA NA 14 7.5 Resection,ablation,
TACE,SBRT
NA 10(43)
Nagataetal.[18] Retrospective 83 NA 77(92.7) 3–12 NA NA NA Resection,ablation 83(100) 22(29)
Zavagliaetal.[19] NA 31 31(100) 26(83.9) NA 1.7 19.3 8 Resection,ablation,
TACE
NA 1(3.2)
Kollyetal.[20] Prospective 47 40(85.1) NA NA NA 21.5d 9.6d Resection,ablation,
TACE
47(100) 20(42.6)
Torresetal.[21] Prospective 8 7(87.5) 6(75) NA NA 7.5 – Resection,ablation,
Protontherapy
NA 0
Rinaldietal.[22] Prospective 15 15(100) NA NA NA 11.3d 3 NA NA 1(6.7)
Virlogeuxetal.[23] Retrospective 23 23(100) 22(95.6) 3–6 NA 7.2 13 Resection,ablation,
TACE,RT
23(100) 11(47.8)
Cheungetal.[30] Prospective 29 29(100) 18(62.1) NA NA NA NA NA NA 2(6.9)
Idaetal.[24] Retrospective 26 NA 26(100) NA NA 4.2if
recurrence21.2 ifnorec. NA Resection,ablation, TACE 11(42.3)/15 (57.7) 12(46.2)
Elkassasetal.[26] Prospective 53 53(100) 41(77.4) 3–6 3 8 16 Resection,ablation NA 20(37.7)
Adhouteetal.[29] Retrospective 22b 22(100) 19(86.4) 3–6 NA 12 NA Resection,ablation,
TACE,LT(n=3)
NA 9(40.9)8/19
(42.1)c
PRT:particleradiationtherapy,TACE:trans-arterialchemoembolization;PEI:percutaneousethanolinjection;LT:livertransplant. aConsideringpatientsthathadalreadyreachedthe12-weekfollow-upperiod.
b Threepatientshadlivertransplantation.
c Excludingthepatientsundergoinglivertransplantation. d Meanvalue.
1108 M. Guarino et al. / Digestive and Liver Disease 50 (2018) 1105–1114 Table2
Resultsoftime-to-eventanalyses.
Author N
recurrences/100 p-months
RecurrenceratesfromlastHCCtreatment RecurrenceratesfromDAAsstart Recurrencedetails
6months 12months 18months 24months 6months 12months 18months 24months
Reigetal.[6] NA 7% 12.9% NA 13pts(81.3%)new
intrahepaticlesion:5pts singleHCC,4pts≤3HCC ≤3cm,1ptmultifocalHCC, 3ptsinfiltrativepattern and/orextra-hepatic lesions3pts(18.7%): intrahepaticgrowth
Callejaetal.[9] NA 30% 12.9% 30% 5/21pts(23.8%)aggressive,
untreatablerecurrences Ikeda etal. [10] NA NA 9.6% 30.1% 38.9% 13pts(76.5%)singleHCC 4(23.5%)2–3HCC 0%>3HCCorwith macroscopicvascular invasion Cabibbo etal. [11] NA NA 12% 26.6% 29.1% 62%BCLCA;21%BCLCB; 7%BCLCC;10%BCLCD 28ptsintrahepaticgrowth, 24ptsnodularpattern,5 ptsinfiltrativepattern,1pt macrovascularinvasion
Ogawaetal.[13] NA NA 1-yrinnon-cirrhotic:6.5%1-yrcirrhotic:23.1% NA
Minamietal.
[14,31]
NA 21.1% 29.8% 38% 54.5% 1/78pt(1.3%)extrahepatic
recurrence Poletal.ANRS—
CO22Hepater
[15]
0.73 NA 1-yr:≈10%b NA
Poletal.ANRS— CO12CirVir[15]
1.11 NA NA NA
Shimizuetal.[17] NA 0% 8.7% 23.1% 27.9% 17.4% 45.7% 52.4% 52.4% NA
Nagataetal.[18] NA 5-yr:45.1%3-yrSVR:22.9%;3-yrno-SVR:40% NA NA
Kollyetal.[20] NA 4% 19% 42% 23% 42% NA
Virlogeuxetal.[23] 1.7 ≈6%b ≈13%b ≈26%b ≈31%b NA NA
Idaetal.[24] NA NA ≈20b,c ≈35%b,c 5pts(42%)BCLC0;6pts
(50%)BCLCA;1pt(8%) BCLCC(vascularinvasion)
Elkassasetal.[26] 4.06a NA NA 95%ofptsnewsiteof
recurrence30%ofpts>3 HCC
Adhouteetal.[29] NA Timeto
progression:12 months
NA BCLCA:44%;B22%;C33%
NA:notavailable. aAdjustedforcovariates.
b Dataderivedfromthearticle’sfigure.
M.Guarinoetal./DigestiveandLiverDisease50(2018)1105–1114 1109
Fig.1.ThecurvesofHCCrecurrenceratessincethelastHCCtreatmentofsix differ-entcohortsofpatientstreatedwithDAAs[6,14,17,18,20,23]havebeenplottedina singlegraph.Allthecurvesavailableintheliteraturehavebeenincluded.
recurrencerateaftercurativetreatmentsinpatientswithactive HCVinfection.However,anycomparisonbetweenhistoricaldata ofHCCrecurrenceinuntreatedHCVpatientsandthepresentdata ofHCCrecurrenceafterDAAs-treatedpatientsisimpossible.First, theseriesafterDAAsareaffectedbyan“immortaltime”bias[35], i.e.intheDAAsgrouponlypatientswithoutrecurrenceatthetime ofthistreatmentwereconsidered.Further,patientsreceivingDAAs havehighheterogeneityintermsofintervalHCCtreatment-DAAs, andtypeandnumberoftreatmentsofHCCbeforeDAAs.
Nevertheless,somedataaboutrecurrenceafterDAAsdeserve consideration.Interestingly,theone-yearrecurrenceratewas sig-nificantly and inversely associated with the proportion of SVR achieved(p=0.008,Fig.2a):theoneyearaftertheHCCtreatment. Anassociationwasalsoobservedforcirrhosis:thehigherthe
pro-portionofcirrhoticpatients,thehighertheone-yearrecurrence rate(p=0.285,Fig.2b).
3.4. Comparativeanalyses
Ninestudies(10cohortsofpatients)comparedtheriskforHCC recurrenceofDAAstreatedpatientswiththeriskin other pop-ulations[10–12,14,15,23,26,29].Beforeconsideringtheresultsof thesestudies,thesamecautionarynotereportedforsurvival anal-ysesismandatory.AllcomparisonsofDAAspatientswithhistorical serieshavepoorreliabilitybecausetheDAAsgroupisaffectedby an“immortaltime”bias[35]andincludeheterogeneouspatients (wide range ofinterval HCC treatment-DAA, mixed radicaland non-radicaltreatmentsforHCC,andcombinationoffirstand iter-ativerecurrences).InsixstudiestheDAAs-treatedpatientswere compared withuntreated patients[10,11,15,23,26,29].Of these, onlyonepaper[25]reportedahigherrecurrencerateintheDAAs group,while3papers(4cohorts)reportedsimilarfigures[11,15,29] and 2 reporteda low recurrence riskin DAAs patients[10,23]. Intriguingly,onepaperreportedasimilarrecurrenceratebetween patientsreceivinginterferon(IFN)-basedandIFN-freetreatments [18].Twostudiescomparedthe3arms(DAAs,IFN-basedtreatment anduntreatedpatients):oneshowedsimilarrecurrencerates[14]; thesecondobservedasignificantreductionoftumorrecurrences insuccessfullytreatedpatients,withoutdifferencesbetween IFN-basedorIFN-freeregimensgroups[12].
Considering the studies presented at the AASLD meeting in 2017,14abstractsanalyzedthedataof5346patients[32].Therisk forHCCrecurrenceinDAAstreatedcaseswasunchangedin8 stud-ies(incomparisonwithuntreatedpatientsin7studiesandwith patientstreatedwithIFNinonestudy)andwasreducedin3(in comparisonwithuntreatedpatientsin2studiesandwithpatients treatedwithIFNinonestudy).
Insummary,onlyonepaperreportedanincreasedriskofHCC recurrenceinpatientstreatedwithDAAscomparedtountreated ones[26],andnonereportedanincreasedriskinpatientstreated withDAAsincomparisonwiththosetreatedwithIFN.Theresults ofcomparativestudiesaresummarizedinFig.3.
Fig.2.TheHCCrecurrenceratesoneyearaftertheDAAsadministrationreportedintheliteraturehavebeenconsidered.TheirassociationwiththeproportionofSVRachieved afterDAAsandtheproportionofcirrhoticpatientsincludedinthecohortshavebeenanalyzed.(a)Aninverseassociationbetweentheone-yearrecurrencerateandthe proportionofSVRachievedafterDAAs.(b)Adirectassociationbetweentheone-yearrecurrencerateandtheproportionofcirrhoticpatientsincludedinthecohorts.
1110 M.Guarinoetal./DigestiveandLiverDisease50(2018)1105–1114
Fig.3. TheresultsofstudiesanalysingtheriskofHCCrecurrenceinpatientstreated withDAAsincomparisonwiththeriskofHCCrecurrenceinuntreatedpatientsorin patientsreceivingINFaresummarized.Thenumberofpapersshowinganincreased, unchangedordecreasedrecurrenceriskintheDAAsgroupvs.theno-treatment group(leftpartofthefigure)orvs.theINFgroup(rightpart)isshown.
3.5. Predictorsofrecurrence
Twelve studies analyzed the predictors of HCC recurrence in DAAs treated patients [6,7,10,11,13,17,18,20,23,26,29,31]. One additional paper analyzing predictors of recurrence was not considered because it included patients undergoing liver transplantation[16].
TheSVRratewasoneofthemostcommonlyanalyzed param-eter.Itsprevalencerangedbetween62%and100%.Nagataetal. [18]reportedthatthecumulativeincidenceofHCCrecurrencein patientswhoachievedSVRwasmuchlowerthaninthose with-outSVR(at3years,22.9%vs.40.0%;p=0.022),butthisdifference wasnotconfirmedbyotherauthors[10,11,26].Nevertheless,as reportedabove,inthepresentreviewweobservedasignificantand inverseassociationbetweentheSVRrateandone-yearrecurrence risk(Fig.2a),supportingthedataofNagataetal.[18].
ThetimebetweenthelastHCCtreatmentandDAAsstarthas beenidentifiedasapredictorofrecurrenceby5authors,i.e.the shorterthetime,thehighertheriskofrecurrence[6,13,20,29,31].
Ogawaetal.[13]reportedahigherrecurrencerisk(HazardRatio —HR2.3195%CI1.04–5.15,p=0.041)whennon-curative treat-ments (trans-arterial therapy/radiotherapy) were adopted; the highestriskwasobservedwhenthetimebetweenHCCtreatment andDAAsinitiationwas<1year.Althoughitmaybequestioned whetherTACEcanbeconsideredpotentiallycurative,other stud-iesdidnotconfirmitsassociationwithHCCrecurrenceinpatients treatedwithDAAs[6,7,10,20,23].
Interestingly,4 authors [10,11,13,31] observed that patients whounderwentmorethanoneHCCtreatmentbeforeDAAs initia-tionhadahigherrecurrenceratethanthosetreatedonlyonce(in thestudybyIkedaetal.50%vs19.1%).Inagreement,Cabibboetal. [11]reportedahazardratioof2.22(95%CI1.02–4.83,p=0.043)for post-DAAsrecurrenceinpatientswithclinicalhistoryofpriorHCC recurrenceascomparedtopatientswithoutpriorHCCrecurrence). Thefollowing additional predictors of HCC recurrence were reported:tumorsize11,livercirrhosis13,liverstiffness7,patient’s age7,20,serumalpha-fetoprotein-L3(AFP-L3)31and Des-gamma-CarboxyProthrombin(DCP)[31],anti-HBc[17],andWisteria flori-bundaagglutininpositiveMac-2bindingprotein(WFA+M2BP)[18]. 3.6. Aggressivenessofrecurrences
Someauthorsspeculatedaboutamajoraggressivenessofthe recurrenttumorinDAAstreatedpatients[6,10,11,26–28].
How-ever,itshouldbenotedthatadefinitionofHCCaggressivenessdoes notexistinliterature,oratleastanagreementonitsdefinitionhas notbeenreachedasyet.
AlthoughmostrecurringtumorwerediagnosedinBCLCstage0 orA[6,11,16,20,22–24,26,29],theauthorsidentifiedassignsofHCC aggressivenessthefollowingfactors:fasttumorgrowthwithalow responseratetoablation[27],percentageofinfiltrative[6,11]or multifocal[26]tumorpattern,andradiologicalsignsof microvas-cularinvasion(mVI)[28].Specifically,Renzullietal.[28]compared theprevalenceofradiologicalfeaturespredictiveofmVIinHCC recurringafterDAAstherapyandintumorsdiagnosedbeforeDAA treatmentinthesamepatients.TheyshowedthatmVIwaspresent in70.7%ofnodulesrecurredafterDAAstreatment,asignificantly higherproportionwithrespecttoHCCdetectedbeforeDAAs(33.3%, p=0.0007).Thisdifferencewasconfirmedeveninthecaseofsmall noduleswithadiameterbetween10and20mm.ElKassasetal.[26] observedthat30%ofpatientswitharecurrentHCCafterDAAs expo-surehadmorethanthreenodulesvs.18%intheDAAsnon-exposed populationwithrecurrence,butthelimitedsamplesizedidnot allowtoclarifywhetherornotthisdifferencewassignificant.
Considering the studies presented at the AASLD meeting in 2017,threereportedanaggressivepatternofrecurrenceafterDAAs treatmentandthreenoticedarapidHCCrecurrence[32].However, acase-controlstudydidnotconfirmthesedatabecausethepattern ofrecurrencewassimilarinpatientswithandwithoutexposureto DAAs[29].
Recently,Critelliet al.[36] studied,although ina non-DAAs related contest, molecular characteristics of the HCC aggres-siveness, showingthat growthspeed and outcome of HCC are associated tospecificmolecularsignatures ofneo-angiogenesis, toprominentfeaturesofepithelial-mesenchymaltransitionwith extremelypoorcelldifferentiation,toclear-cutactivationof trans-forming growth factor-beta 1 (TGF-1) signaling and to the findingofprogrammedcelldeathprotein1(PD-1)/programmed death-ligand1(PD-L1)overexpressionassignofsevere microen-vironmentimmunosuppression[36].Moreover,ina preliminary previousand prospectivestudy[37],thesameauthorsreported that25%ofnewHCCsdiagnosedundersurveillancehaveavery rapidgrowthwithadoublingvolumetime<2months,regardless ofinitialBCLCclassification.Thisfast-growingsubgroupofpatients canbebiologicallycharacterizedveryaccuratelybya5-gene tran-scriptomichepaticsignatureincludingangiopoietin-2(ANGPT2), delta-like ligand4 (DLL4),neuropilin (NRP)/tolloid (TLL)-like 2 (NETO2),endothelialcell-specificmolecule-1(ESM1),andnuclear receptorsubfamily4,groupA,member1(NR4A1).Thisgene sig-naturewasassociatedwithaworseprognosis(mediansurvivalof 11vs41months).
Onthisbasis,somehypotheseshavebeensuggestedtoexplain the potential aggressiveness in the DAAs contest. It has been assumedthatamajorroleisplayedbyadownregulationofthe anti-tumorresponsebytheimmunesystemcausedbythe bru-talDAAs-inducedHCVclearance,whichcouldboostthegrowthof invisiblefociofmalignantcells[6,27].Itiswellknownthat recur-renceafteracompleteresponsetoablativetreatmentmaybedue todisseminationofcellsbeforethetreatmentandtothe appear-anceofnewoncogeniccloneswithintheunderlyingcirrhoticliver fromalreadycommittedcellswithgeneticdamage[38].Infact, theviralclearancecausesrapiddownregulationofIFNstimulating genes(ISGs) inthe liverand blood. Theobserved immunologi-calchangesincludeadown-regulationofantigen-specificTcells, witharestorationofproliferativeHCV-specificCD8+Tcellsafter 12weeksofDAAtherapyinmostSVRpatients[39],andadeclineof IFN-inducibleprotein-10(IP-10),monocytechemoattractant pro-tein1,macrophageinflammatoryprotein1betaandIL-18levels [40–42].Incontrasttotherecoveryofvirus-specificCD8+Tcells andnaturalkiller(NK)cells,DAAstherapyinHCVpatientsisnot
M.Guarinoetal./DigestiveandLiverDisease50(2018)1105–1114 1111 followedby arestoration of themucosal-associatedinvariantT
(MAIT)cells[43].Moreover,duringDAAstherapyNKcells,which mediatetarget-cellapoptosisthroughsurfaceexpressionof TNF-relatedapoptosis-inducingligand(TRAIL),showadecreaseinTRAIL expression[44,45].
Accordingtothis“immuneescapehypothesis”,itcanbeinferred that,onceHCVinfectioniscleared,smalltumorsalreadypresent but not radiologically evident might accelerate their growth. Nonetheless,itisverydifficulttoprovewhetherthismechanism isa majorcauseof thesupposed greateraggressivenessofHCC recurrence.
Finally,CarrandGuerra[46]developedanHCCaggressiveness index,basedontumordiameter,multifocality,portalveininvasion andalpha-fetoproteinlevels,whichwasabletostratifypatientsin3 groupswithdifferentsurvival.Todate,nopaperreportinga poten-tialmajoraggressivenessofHCCrecurrenceafterDAAsadopted eithertheCarrandcoll.indexorhomogeneousandvalidated crite-riatodefinethetumoraggressiveness.Aprospectivestudyaimed toevaluatetherecurrencepatternandtheoutcomeofpatientswith recurrentHCCexposedornottoDAAstreatmentseemstobethe onlywaytosolvethisdebateandreachanagreementaboutthe reliabledefinitionofHCCaggressiveness.
3.7. Managementofrecurrences
Treatment of recurring HCCs included different approaches acrossstudies.Idaetal.[24]usedcurativetreatments(liver resec-tionorradiofrequencyablation)inalmostallcasesofrecurrence, whileinthecohortfromCabibboetal.[11]TACEwasthemost commontreatment(45%). In thestudy byAbdelazizet al. [27] chemoembolizationwasadoptedin51.1%ofcases,whereaslocal ablationandresectionwereusedinaminorityofcases(8.9%and 4.4%,respectively).Diseasecontrolwasachievedinabouthalfof cases(51.1%),whileinonethirdofpatients(33.3%)only support-ivetreatmentwaspossible.Similarly,inthestudiesbyReigetal. [6]andCallejaetal.[9]18.7%(3/16)and23.8%(5/21)ofpatients, respectively,werequalifiedforsupportivecare.Converselyahigher percentageofuntreatablerecurrencewasreportedbyBielenetal. [16](50%).
3.8. Limitationsoftheavailablestudies
Severallimitationsoftheanalyzedstudiesshouldbe consid-ered.First,only8studies(including9cohortsofpatients)hada prospectivedesign[11,13,15,20–22,26,30],6ofthemincludinga time-to-eventanalysis.Inaddition,fewpaperscollectedan ade-quatenumberofpatients(>50patientsin10papersand>100in5) toprovidereliableinformation.
Second,theinterval betweenHCC treatmentandDAAs ther-apy,whichhasacrucialimportanceinsettingtheriskoftumor recurrence,wasextremelyvariable.Infact,therecurrencecurve obtainedbyameta-analysisoftheuntreatedharmsofstudiesabout HCV-patientsundergoingcurative treatmentfor HCCshowed a parabolicshape,withthemaximalriskrateoverthefirstyears[34]. Consequently,consideringthetimingofDAAsinitiation,atleast3 differenttemporalscenarioswithadifferent“intrinsic”riskcanbe identified(Fig.4):
• HCC recurrence in patientsexposed to DAAs within thefirst 3years aftersuccessfullytreated HCC(highrisk ofrecurrence witharateof22/100py);
• HCC recurrence in patients exposed to DAAs between 3 and 5yearsaftersuccessfullytreatedHCC(intermediateriskof recur-rencewitharateof8/100py);
Fig.4. TheHCCcumulativerecurrencerateaftercurativetreatmentforHCCin HCV-patientswithoutDAAstreatmentisextractedfromarecentmeta-analysis (redcurve)[34].Ithasaparabolicshape,withthemaximalrecurrenceriskover thefirstyears.Consequently,atleast3differenttemporalscenarioswitha dif-ferent“intrinsic”riskofHCCrecurrencecanbeidentifiedaccordingtothetiming ofDAAsintroduction:(1)HCCrecurrenceriskinpatientswiththeintroduction ofDAAsduringthefirst3yearsishigh(22/100p/y,yellowarea);(2)HCC recur-renceriskinpatientswiththeintroductionofDAAsbetween3and5yearsis intermediate(8/100p/y,greenarea);(3)HCCrecurrenceinpatientswiththe intro-ductionofDAAsafter5yearsislow(2/100p/y,bluearea).Theoverallnumberof recurrence/100person/yearofuntreatedHCVinfectionhavebeenderivedfromthe meta-analysisbyCabibboetal.[34].
• HCCrecurrenceinpatientsexposedtoDAAs5yearsafter success-fullytreatedHCC(lowriskofrecurrencewitharateof2/100py). Therefore,thetimetoHCCrecurrenceshouldbecalculated con-sideringthelastHCCcurativetreatmentasthestartingpoint.The alarmistdataabouthighrecurrenceratesinDAAstreatedpatients couldbesimplyrelatedtothefactthatDAAswereinitiatedinmost patientsduringtheperiodofhighestrecurrencerisk.
Third, 14 papers included both patients receiving poten-tially curative treatments (surgery and ablation) and palliative treatments(chemoembolizationandradiationtreatments)before DAAstherapy[6,7,10,11,13,16,17,19–21,23,24,29,31].Fourpapers included patients with multiple HCC treatments [10,11,24,31], while 13 papers didnot detail the number of HCC treatments receivedbeforeDAAs[6,7,9,13,15–17,19,21,22,26,29,30],making itdifficulttoprovidesolidevidenceontheeffectofDAAstherapy ontheindividualrisk.
Fourth, two main determinants of HCC recurrence should be always considered, i.e. liver cirrhosis and SVR. As we foundinthis review,both theseparametersmayimpact recur-rence risk. The variability of the reportedrecurrence rate may be therefore ascribed, at least in part, to the different pro-portion of cirrhotic patients included in the studies (range 59.2%–100%),which wasdetailedin 16papers(with17 cohorts ofpatients)[6,7,9,11,13,15–17,19–23,24,26,29,30],andoftheSVR rateobserved(range62%–100%),whichwasreportedinallbuttwo papers[20,22].Unfortunately,onlyfewstudies[10,11,13,18,20,26] stratifiedtherecurrencerateaccordingtothepresenceofcirrhosis and/ortheachievementofSVR.
Finally,anycomparativeanalysisbetweenDAAsgroupand his-toricalseriessuffers from atleastone majorlimitation,i.e. the “immortaltime”bias[35].Infact,intheDAAsgrouponlypatients
1112 M.Guarinoetal./DigestiveandLiverDisease50(2018)1105–1114 withoutrecurrenceatthetimeofthistreatmentwereconsidered.
Aprospectiveintention-to-treatanalysis,startingfromthetimeof HCCcureandwithanearlyDAAsinitiationisthereforemandatory toknowtherealriskofcancerrecurrenceinpatientundergoing antiviraltreatment.
4. Discussion
ThepotentialrisksandbenefitsofDAAtherapyinpatientswith ahistoryofHCCarestillamatterofdebate.Someauthors sug-gestedareductionoftherecurrenceriskincomparisontopatients withuntreatedHCVinfection,althoughtheseconclusionsmustbe interpretedinthecontextoftheclinicalheterogeneitywithinand betweenstudiesandmethodologicallimitationsofcurrentdata. However,alltheseindirectcomparisonswithhistoricaldatasuffer frombiasesfavoring(immortalbias)ordisfavoring(analysesnot consideringthetimeelapsedbetweenHCCcureandinitiationof thetherapy)DAAstreatedpatientswithrespecttothoseuntreated ortreatedwithIFN.Moreover,thepatientsmaydifferintermof thepreviousoncologichistoryandproportionofcirrhoticcases.
PatientswithahistoryofHCCareatriskforrecurrencethat canberelatedtointrinsictumorfactorsandunderlinedliver dis-ease such as active viremia and degree of liver function [47]. It hasbeen hypothesized that the HCV eradicationinduced by DAAsresultsin reducedimmune surveillancedysregulatingthe anti-tumorresponseandboostingthegrowthofstillundetected microscopicHCCtumorclones[6,27].Ontheotherhandsuccessful antiviraltreatmentcanresultinfibrosisregressionand improve-mentsinportalhypertensionandliverfunction,showntobethe majordriverofdeathinpatientswithsuccessfullytreatedHCCand activeHCVinfection[48].Inthiscontext,DAAsmayreducerisk ofHCCrecurrencebyHCVeradicationandliverfunction improve-ment[49].
TheidentificationoffactorsevaluatingtheriskofHCC recur-renceis mandatorytodefinepatientssubgroupsin whomDAA therapyshouldbeavoided.Severalpredictorsforrecurrencewere reportedinliterature,likethehistoryofpriorHCCrecurrenceand theinterval betweenHCCsuccessful treatmentandDAA initia-tion,withlongerintervalsbeingassociatedwithlowerriskofHCC recurrencebecauseofalowerrisk thatresidualtumorcells are stillpresentattheDAAinitiation.Inthisdirection,someauthors suggestedtodelayanddeferDAAtreatmenttillthereisa well-establishedcompleteradiologicalresponse,reducinginthisway thechanceofmisclassificationbias[50,51].
An aggressivepattern (rapid growth and vascular invasion) oftumor recurrenceafter DAAshas beenpointed out, but still remainstobeconfirmed.MostpatientswithHCCrecurrenceacross studies were found at an early stage and underwent curative treatments,suggestingrecurrencefollowingDAA therapyis not aggressive;however,few data regardingtreatmentresponse or post-recurrencelong-termprognosisareavailable.Consideringthe poorevidencesagainstDAAsandtheclearbenefitsintermsofliver functionincaseofSVR,thechoiceoftreatingthesepatientswith antiviraldrugsseemstobejustified.
The limitations of the available studies should drive future researchesandavoidthepitfallsofseveralalreadypublished stud-ies. First, time-to-event analyses are mandatory. Based on the “immuneescape”hypothesis,theimpactofDAAstherapyontumor recurrencejustifiestherelationshipbetweentheonsetof recur-rence and the time elapsed since the last curative treatment. Therefore,itisadvisabletocalculatethetimetorecurrence con-sideringthelastHCCcurativetreatmentasstartingpoint.Further, itsoundslogicaltoconsideranintervalperiodbetweenthe radio-logicalevidenceofHCCeradicationandthestartofDAAstoexclude thepresenceof“prevalent”tumors,evenifanagreementaboutthe adequatedurationofthistimeintervalislacking.
Second,theproportionofcirrhoticpatientsandthepercentage ofSVRshouldbeconsideredtoproperlyinterpretthedata.The crucialcontributionofthesetwofactorstotheriskofrecurrence, evenafterDAAs,strengthensthehypothesisthatDAAstreatmentis irrelevant(orevenprotectivefromrecurrence,accordingtosome studies)inthenaturalhistoryofHCC.
Finally,anycomparisonwithhistoricalseriesisoflimited inter-est because of a number of biases affecting these studies and differencesbetweenenrolledpatients. Prospective intention-to-treatanalyseswillbeprobablythebestcontributiontodriveclinical practice,providedthatarandomizedtrialcanbedifficulttodesign. 5. Conclusions
TheavailabledataabouttheriskofHCCrecurrenceafterDAAs treatmentinHCVpatientscannotbeconsidereddefinitive,butthe initialalarmistdataindicatinganincreasedriskofrecurrencehave notbeenconfirmedbymostsubsequentstudies.Nonetheless,since therearenoclearevidencesofadetrimentaleffectofDAAsonHCC recurrence,webelievethatDAAsshouldnotbedeniedtothese patients,aswellas,itismandatorytomaintainanactive surveil-lanceforHCC.
Conflictofinterest Nonedeclared. Acknowledgements
WethanktheothermembersofSpecialInterestGroupon “Hep-atocellular carcinoma andnew anti-HCV therapies” of the Italian Association for the Study of the Liver (AISF): Alessandro Vitale, Francesco Paolo Russo, Umberto Cillo, Patrizia Burra, Claudia Mescoli,MartinaGambatofromPaduaUniversityHospital.Anna Sessa from Dept. of Clinical Medicine and Surgery, Gastroen-terology Unit, University of Naples “Federico II”, Naples, Italy. Giuseppe Cabibbo from Section of Gastroenterology, DIBIMIS, Universityof Palermo.MauroViganòfromHepatologyDivision, SanGiuseppeHospital,UniversityofMilan.GiovanniGalatifrom InternalMedicineandHepatologyUnit,CampusBioMedico Uni-versity of Rome. Erica Villa from Division of Gastroenterology, ModenaHospitalUniversityofModenaandReggioEmilia, Mod-ena. Massimo Iavarone from Gastroenterology and Hepatology Unit,Fondazione IRCCS Ca GrandaOspedale Maggiore Policlin-ico. Giuseppina Brancaccio from Infectious Diseases and Viral Hepatitis, Department of Mental and Physical Health and Pre-ventiveMedicine,UniversityofCampaniaLuigiVanvitelli,Naples. MariaRendinaandLuigiG.LupofromUniversityHospitalofBari. FrancescoLosito,FabioFucillifromIRCCSfromCastellanaGrotte, Bari.MarcelloPersicofromUniversityofSalerno,Baronissi,Salerno. RobertaD’Ambrosio,AngeloSangiovannifromIRCCSCa’Granda Maggiore Hospital, University of Milan. Alessandro Cucchetti, FrancoTrevisanieMatteoRenzullifromS.Orsola-Malpighi Hospi-tal,Bologna.LucaMiele,AntonioGrieco,GianLodovicoRapaccini, MaurizioPompiliandAntonioGasbarrinifromUniversitàCattolica delS.Cuore.Roma.GiovanniBattisaLeviSandrifromSanCamillo Forlanini Hospital, Roma.Fabio Melandro, Massimo Rossifrom SapienzaUniversityofRome,Roma.IlariaLenci,TommasoMaria ManziafromPoliclinicoTorVergata,Roma.RaffaellaTortora, Gio-vanGiuseppeDiCostanzofromCardarelliHospital,Napoli.Rodolfo Sacco,DavideGhinolfi,ErionRreka,PaolaCarrai,NataliaSimonetti fromPisaUniversityHospital,Pisa.CarloSposito,SherrieBhoori fromFondazioneIRCCSIstitutoNazionaleTumori,Milano.Stefano diSandrofromASSTNiguardaHospital,Milan.FrancescoGiuseppe FoschifromFaenzaHospital,AreaVastaRomagna,Ravenna.Andrea
M.Guarinoetal./DigestiveandLiverDisease50(2018)1105–1114 1113 CasadeiGardinifrom IRCCS,Meldola. Daniele Nicolini,Susanna
Mazzocato,AlbaKostandinifromPolytechnicUniversityofMarche, Ancona.PaolaViolifromVeronaUniversityHospital,Italy.Umberto Baccarani,RiccardoPravisanifromUniversityofUdine,Italy.Valter VincenzifromSanMartinoHospital,Belluno.
AppendixA. Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttps://doi.org/10.1016/j.dld.2018.08.001. References
[1]El-SeragHB.Epidemiologyofviralhepatitisandhepatocellularcarcinoma. Gas-troenterology2012;142:1264–73.
[2]JeongSW,JangJY,ChungRT.HepatitisCvirusandhepatocarcinogenesis.Clin MolHepatol2012;18:347–56.
[3]JakobsenJC,NielsenEE,FeinbergJ,KatakamKK,FobianK,HauserG,etal. Direct-actingantiviralsforchronichepatitisC.CochraneDatabaseSystRev 2017;6:CD012143.
[4]IoannouGN,GreenPK,BerryK.HCVeradicationinducedbydirect-acting antivi-ralagentsreducestheriskofhepatocellularcarcinoma.JHepatol2017,pii: S0168-8278(17)32273-0.
[5]CucchettiA,D’AmicoG,TrevisaniF,MorelliMC,VitaleA,PinnaAD,etal.Effect ofdirect-actingantiviralsonfutureoccurrenceofhepatocellularcarcinomain compensatedcirrhoticpatients.DigLiverDis2018;50:156–62.
[6]ReigM,MarinoZ,PerellòC,I ˜narrairaeguiM,RibeiroA,LensS,etal. Unex-pectedhighrateofearlytumorrecurrenceinpatientswithHCV-relatedHCC undergoinginterferon-freetherapy.JHepatol2016;16:719–26.
[7]ContiF,BuonfiglioliF,ScuteriA,CrespiC,BolondiL,CaraceniP,etal.Early occurrenceandrecurrenceofhepatocellularcarcinomainHCV-related cirrho-sistreatedwithdirect-actingantivirals.JHepatol2016;65:727–33.
[8]NaultJC,ColomboM.Hepatocellularcarcinomaanddirectactingantiviral treat-ments:controversyaftertherevolution.JHepatol2016;65:663–5.
[9]CallejaJL,CrespoJ,RinconD,Ruiz-AntoránB,FernandezI,PerellóC,etal. Effectiveness,safetyandclinicaloutcomesofdirect-actingantiviraltherapyin HCVgenotype1infection:resultsfromaSpanishreal-worldcohort.JHepatol 2017;66:1138–48.
[10]Ikeda K,KawamuraY, KobayashiM, KominamiY,Fujiyama S,SezakiH, etal.Direct-actingantiviralsdecreasedtumorrecurrenceafterinitial treat-ment of hepatitis C virus-related hepatocellular carcinoma. Dig Dis Sci 2017;62:2932–42.
[11]CabibboG,PettaS,CalvarusoV,CacciolaI,CannavòMR,MadoniaS,etal.Is earlyrecurrenceofhepatocellularcarcinomainHCVcirrhoticpatientsaffected bytreatmentwithdirect-actingantivirals?Aprospectivemulticentrestudy. AlimentPharmacolTher2017;46:688–95.
[12]PettaS,CabibboG,BarbaraM,AttardoS,BucciL,FarinatiF,etal.Hepatocellular carcinomarecurrenceinpatientswithcurativeresectionorablation:impactof HCVeradicationdoesnotdependontheuseofinterferon.AlimentPharmacol Ther2017;45:160–8.
[13]OgawaE,FurusyoN,NomuraH,DohmenK,HigashiN,TakahashiK,etal. Short-term riskofhepatocellularcarcinomaafterhepatitis Cvirus eradi-cationfollowingdirect-actinganti-viraltreatment.AlimentPharmacolTher 2018;47:104–13.
[14]MinamiT,TateishiR,NakagomiR,FujiwaraN,SatoM,EnookuK,etal.The impactofdirect-actingantiviralsonearlytumorrecurrenceafter radiofre-quencyablationinhepatitisC-relatedhepatocellularcarcinoma.JHepatol 2016;65:1272–3.
[15]ANRSCollaborativeStudyGroup.Lackofevidenceofaneffectofdirect-acting antiviralsontherecurrenceofhepatocellularcarcinoma:datafromthreeANRS cohorts.JHepatol2016;65:734–40.
[16]BielenR,MorenoC,VanVlierbergheH,BourgeoisS,MulkayJP,Vanwolleghem T,etal.Theriskofearlyoccurrenceandrecurrenceofhepatocellularcarcinoma inhepatitisC-infectedpatientstreatedwithdirect-actingantiviralswithand withoutpegylated:aBelgianexperience.JViralHepat2017;24:976–81. [17]ShimizuH,MatsuiK,IwabuchiS,FujikawaT,NagataM,TakatsukaK,etal.
RelationshipofhepatitisBvirusinfectiontotherecurrenceof hepatocellu-larcarcinomaafterdirectactingantivirals.IndianJGastroenterol2017;36: 235–8.
[18]NagataH,NakagawaM,AsahinaY,SatoA,AsanoY,TsunodaT,etal.Effect ofinterferon-basedand-freetherapyonearlyoccurrenceandrecurrenceof hepatocellularcarcinomainchronichepatitisC.JHepatol2017;67:933–9. [19]ZavagliaC,OkolicsanyiS,CesariniL,MazzarelliC,PontecorviV,CiaccioA,
etal.Istheriskofneoplasticrecurrenceincreasedafterprescribing direct-actingantiviralsforHCVpatientswhoseHCVwaspreviouslycured?JHepatol 2017;66:236–7.
[20]KollyP,WaidmannO,VermehrenJ,MorenoC,VögeliI,BergT,etal. Hepatocel-lularcarcinomarecurrenceafterdirectantiviralagenttreatment:aEuropean multicentricstudy.JHepatol2017;66:876–8.
[21]TorresHA,VautheyJN,EconomidesMP,MahaleP,KasebA.Hepatocellular car-cinomarecurrenceaftertreatmentwithdirect-actingantivirals:first,dono harmbywithdrawingtreatment.JHepatol2017;65:862–4.
[22]RinaldiL,DiFranciaR,CoppolaN,GuerreraB,ImparatoM,MonariC,etal. HepatocellularcarcinomainHCVcirrhosisafterviralclearancewithdirect actingantiviraltherapy:preliminaryevidenceandpossiblemeanings.WCRJ 2016;3:e748.
[23]VirlogeuxV,PradatP,Hartig-LavieK,BaillyF,MaynardM,OuzielG,etal. Direct-actingantiviraltherapydecreaseshepatocellularcarcinomarecurrenceratein cirrhoticpatientswithchronichepatitisC.LiverInt2017;37:1122–7. [24]IdaH,HagiwaraS,KonoM,MinamiT,ChishinaH,ArizumiT,etal.
Hepatocel-lularcarcinomaafterachievementofsustainedviralresponsewithdaclatasvir andasunaprevirinpatientswithchronichepatitisCvirusinfection.DigDis 2017;35:565–73.
[25]KozbialK,MoserS,SchwarzerR,LaferlH,Al-ZoairyR,StauberR,etal. Unex-pectedhighincidenceofhepatocellularcarcinomaincirrhoticpatientswith sustainedvirologicresponsefollowinginterferon-freedirect-actingantiviral treatment.JHepatol2016;65:856–8.
[26]ElKassasM,FunkAL,SalaheldinM,ShimakawaY,EltabbakhM,JeanK,etal. IncreasedrecurrenceratesofhepatocellularcarcinomaafterDAAtherapyin ahepatitisC-infectedEgyptiancohort:acomparativeanalysis.JViralHepat 2018;25:623–30.
[27]AbdelazizAO,NabilMM,AbdelmaksoudAH,ShoushaHI,CordieAA, Has-sanEM,etal.De-novoversusrecurrenthepatocellularcarcinomafollowing direct-actingantiviraltherapyforhepatitisCvirus.EurJGastroenterolHepatol 2018;30:39–43.
[28]RenzulliM,BuonfiglioliF,ContiF,BrocchiS,SerioI,FoschiFG,etal. Imag-ingfeaturesofmicrovascularinvasioninhepatocellularcarcinomadeveloped after direct-acting antiviral therapyin HCV-related cirrhosis. Eur Radiol 2018;28:506–13.
[29]AdhouteX,PenarandaG,RaoulJL,SellierF,CastellaniP,OulesV,etal. Hep-atocellularcarcinomarecurrenceinhepatitisCvirus-relatedcirrhosistreated withdirect-actingantivirals:acase-controlstudy.EurJGastroenterolHepatol 2018;30:368–75.
[30]CheungMCM,WalkerAJ,HudsonBE,VermaS,McLauchlanJ,MutimerDJ, etal.Outcomesaftersuccessfuldirect-actingantiviraltherapyforpatients withchronichepatitisCanddecompensatedcirrhosis.JHepatol2016;65: 741–7.
[31]MinamiT,TateishiR,WakeT,NishibatakeM,NakagomiR,SatoM,etal. Hep-atocellularcarcinomarecurrenceaftercurativetreatmentsinpatientswith chronichepatitisCwhounderwentdirect-actingantiviraltherapy.Hepatology 2017;66:760A–1A.
[32]ColomboM,BoccaccioV.HepatitisCeradicationwithDAAandriskofliver cancerrecurrence:thedebateunrests.JViralHepat2018;25:620–2. [33]CammàC,CabibboG,CraxìA.DirectantiviralagentsandriskforHCCearly
recurrence:muchadoaboutnothing.JHepatol2016;65:861–2.
[34]CabibboG,PettaS,BarbàraM,MissaleG,VirdoneR, CaturelliE,etal. A meta-analysisofsingleHCV-untreatedarmofstudiesevaluatingoutcomes aftercurativetreatmentsofHCV-relatedhepatocellularcarcinoma.LiverInt 2017;37:1157–66.
[35]LévesqueLE,HanleyJA,KezouhA,SuissaS.Problemofimmortaltimebiasin cohortstudies:exampleusingstatinsforpreventingprogressionofdiabetes. BMJ2010;340,b5087.
[36]CritelliR,MilosaF,FaillaciF,CondelloR,TurolaE,MarziL,etal. Microen-vironment inflammatory infiltrate drives growth speed and outcome of hepatocellular carcinoma: a prospective clinical study. Cell Death Dis 2017;8:e3017.
[37]Villa E, Critelli R, Lei B, Marzocchi G, Cammà C, Giannelli G, et al. Neoangiogenesis-relatedgenesare hallmarksoffast-growing hepatocellu-larcarcinomasandworst survival.Resultsfroma prospectivestudy.Gut 2016;65:861–9.
[38]BruixJ,GoresGJ,MazzaferroV.Hepatocellularcarcinoma:clinicalfrontiersand perspectives.Gut2014;63:844–55.
[39]MartinB,HenneckeN,LohmannV,KayserA,Neumann-HaefelinC,Kukolj G,etal.RestorationofHCV-specificCD8+Tcellfunctionbyinterferon-free therapy.JHepatol2014;61:538–43.
[40]Burchill MA, Golden-Mason L, Wind-Rotolo M, Rosen HR. Memory re-differentiationandreducedlymphocyteactivationinchronicHCV-infected patientsreceivingdirect-actingantivirals.JViralHepat2015;22:983–91. [41]CallendretB,EcclestonHB,HallS,SatterfieldW,CaponeS,FolgoriA,etal.
T-cellimmunityandhepatitisCvirusreinfectionaftercureofchronic hep-atitisCwithaninterferon-freeantiviralregimeninachimpanzee.Hepatology 2014;60:1531–40.
[42]CarlinAF,AristizabalP,Song Q,WangH,PaulsonMS,StammLM,etal. Temporaldynamicsofinflammatorycytokines/chemokinesduring sofosbu-virandribavirintherapyforgenotype2and3hepatitisCinfection.Hepatology 2015;62:1047–58.
[43]BolteFJ,O’KeefeAC,WebbLM,SertiE,RiveraE,LiangTJ,etal.Intra-hepatic depletionofmucosal-associatedinvarianttcellsinhepatitisCvirus-induced liverinflammation.Gastroenterology2017;153:1392–403.
[44]SertiE,Chepa-LotreaX,KimYJ,KeaneM,FryzekN,LiangTJ,etal.Successful interferon-freetherapyofchronichepatitisCvirusinfectionnormalizesnatural killercellfunction.Gastroenterology2015;149:190–200.
[45]SpaanM,vanOordG,KreefftK,HouJ,HansenBE,JanssenHL,etal. Immuno-logicalanalysisduringinterferon-freetherapyforchronichepatitisCvirus infectionrevealsmodulationofthenaturalkillercellcompartment.JInfect Dis2016;213:216–23.
[46]CarrBI,GuerraVA.Hepatocellularcarcinomaaggressivenessindexandits rela-tionshiptoliverenzymelevels.Oncology2016;90:215–20.
1114 M.Guarinoetal./DigestiveandLiverDisease50(2018)1105–1114 [47]PoonRT,FanST,NgIO,LoCM,LiuCL,WongJ.Differentriskfactorsand
progno-sisforearlyandlateintrahepaticrecurrenceafterresectionofhepatocellular carcinoma.Cancer2000;89:500–7.
[48]CabibboG,PettaS,BarbaraM,AttardoS,BucciL,FarinatiF,etal.Hepatic decom-pensationisthemajordriverofdeathinHCV-infectedcirrhoticpatientswith successfullytreatedearlyhepatocellularcarcinoma.JHepatol2017;67:65–71. [49]FosterGR,IrvingWL,CheungMC,WalkerAJ,HudsonBE,VermaS,etal.Impact ofdirectactingantiviraltherapyinpatientswithchronichepatitisCand decompensatedcirrhosis.JHepatol2016;64:1224–31.
[50]Saraiya N, Yopp AC, Rich NE, Odewole M, Parikh ND, Singal AG. Sys-tematicreviewwithmeta-analysis:recurrenceofhepatocellularcarcinoma followingdirect-actingantiviraltherapy.AlimentPharmacolTher2018;48: 127–37.
[51]RussoFP,TessariM,ImondiA,LynchEN,FarinatiF.HCVclearancebydirect antiviraltherapyandoccurrence/recurrenceofhepatocellularcarcinoma:still anissue?HepatomaRes2018;4:25.
