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
Liver,
Pancreas
and
Biliary
Tract
Hepatitis
C
virus
RNA
levels
at
week-2
of
telaprevir/boceprevir
administration
are
predictive
of
virological
outcome
q
Valeria
Cento
a,
Daniele
Di
Paolo
b,
Domenico
Di
Carlo
a,
Valeria
Micheli
c,
Monica
Tontodonati
d,e,
Francesco
De
Leonardis
b,
Marianna
Aragri
a,
Francesco
Paolo
Antonucci
a,
Velia
Chiara
Di
Maio
a,
Alessandro
Mancon
c,
Ilaria
Lenci
b,
Alessandra
Manunta
f,
Gloria
Taliani
g,
Antonio
Di
Biagio
h,
Laura
Ambra
Nicolini
h,
Lorenzo
Nosotti
i,
Cesare
Sarrecchia
j,
Massimo
Siciliano
k,
Simona
Landonio
l,
Adriano
Pellicelli
m,
Adriano
Gasbarrini
k,
Jacopo
Vecchiet
d,
Carlo
Federico
Magni
l,
Sergio
Babudieri
f,
Maria
Stella
Mura
f,
Massimo
Andreoni
j,
Giustino
Parruti
e,
Giuliano
Rizzardini
l,
Mario
Angelico
b,
Carlo
Federico
Perno
a,
Francesca
Ceccherini-Silberstein
a,∗aDepartmentofExperimentalMedicineandSurgery,UniversityofRome“TorVergata”,Rome,Italy bHepatologyUnit,UniversityHospitalofRome“TorVergata”,Rome,Italy
cUnitofMicrobiology,HospitalSaccoofMilan,Milan,Italy dInfectiousDiseaseClinic,Chieti,Italy
eInfectiousDiseaseUnit,PescaraGeneralHospital,Pescara,Italy
fInfectiousDiseasesUnit,DepartmentofClinicalandExperimentalMedicine,UniversityofSassari,Italy g“LaSapienza”University,Rome,Italy
hS.MartinoHospital,Genova,Italy
iHepatologyUnit,NationalInstituteofHealth,MigrationandPoverty,Rome,Italy jInfectiousDisease,UniversityHospitalofRome“TorVergata”,Rome,Italy kGastroenterology,CatholicUniversityofRome,Rome,Italy
lDivisionofInfectiousDisease,HospitalSaccoofMilan,Milan,Italy mHepatologyUnit,SanCamilloForlaniniHospital,Rome,Italy
a
r
t
i
c
l
e
i
n
f
o
Articlehistory: Received8July2014 Accepted18November2014 Availableonlinexxx Keywords: Earlyresponse NS3proteaseinhibitors Viralkinetics Virologicalfailurea
b
s
t
r
a
c
t
Background: Triple therapy with telaprevir/boceprevir+pegylated-interferon+ribavirin can achieve excellentantiviralefficacy,butitcanbeburdenedwithresistancedevelopmentatfailure.
Aims:ToevaluatekineticsofhepatitisCvirus(HCV)RNAdecayandearlyresistancedevelopment,in ordertopromptlyidentifypatientsathighestriskoffailuretofirstgenerationproteaseinhibitors.
Methods: HCV-RNA was prospectively quantified in 158 patients receiving pegylated-interferon+ribavirin+telaprevir (N=114) or+boceprevir (N=44), at early time-points and during perprotocolfollow-up.Drugresistancewascontextuallyevaluatedbypopulationsequencing.
Results:HCV-RNAatweek-2wassignificantlyhigher inpatientsexperiencingvirologicalfailureto triple-therapythaninpatientswithsustainedviralresponse(2.3[1.9–2.8]versus1.2[0.3–1.7]logIU/mL,
p<0.001).A100IU/mLcut-offvalueforweek-2HCV-RNAhadthehighestsensitivity(86%)inpredicting virologicalsuccess.Indeed,23/23(100%)patientswithundetectableHCV-RNAreachedsuccess,versus
26/34(76.5%)patientswithHCV-RNA<100IU/mL,andonly11/31(35.5%)withHCV-RNA>100IU/mL (p<0.001).Furthermore,differentlyfromfailingpatients,noneofthepatientwithundetectableHCV-RNA atweek-2hadbaseline/earlyresistance.
q Thisworkhasbeenpresentedinpartatthe48thAnnualMeetingoftheEuropeanAssociationfortheStudyoftheLiver(EASL),heldinAmsterdam,Netherlands,April24–28, 2013andatTheLiverMeeting®2014AmericanAssociationfortheStudyoftheLiver,November1–5,2013,WashingtonDC,USA.
∗ Correspondingauthorat:DepartmentofExperimentalMedicineandSurgery,UniversityofRomeTorVergata,ViaMontpellier1,Rome00133,Italy.Tel.:+390672596553; fax:+390672596039.
E-mailaddress:[email protected](F.Ceccherini-Silberstein).
http://dx.doi.org/10.1016/j.dld.2014.11.010
Conclusions:Withtripletherapybasedonfirstgenerationproteaseinhibitors,suboptimalHCV-RNAdecay atweek-2combinedwithearlydetectionofresistancecanhelpidentifyingpatientswithhigherriskof virologicalfailure,thusrequiringaclosermonitoringduringtherapy.
©2014EditriceGastroenterologicaItalianaS.r.l.PublishedbyElsevierLtd.Allrightsreserved.
1. Introduction
AccordingtotheHCV-kineticsmodel,initiallybasedon inter-feron (IFN) monotherapy [1], antiviral treatment of chronic hepatitis C leads toa biphasic decay of plasma HCV-RNA. Ini-tially, treatment actsby blockingviralproduction, determining a veryfastfirstphaseofHCV-RNA declinecharacterizedbythe clearance offreecirculatingvirions.Afterwards,theprogressive clearanceofinfectedcellsdeterminesamuchslowersecondphase ofviraemiadecline.Thismodelwaslaterconfirmedalsoin pegy-lated IFN (pegIFN), pegIFN+ribavirin (RBV) and in treatments includingdirectactingantiviralagents(DAAs),suchastelaprevir (TVR)[2–6].
WithIFNtreatment,thedividinglinebetweenfirstandsecond phasewassetatday2[1],butinthecontextofTVR-treatment,viral dynamicsaremuchmorerapidandtheabovementionedlinemay bemovedbackwards[3].
TVR and boceprevir (BOC), approved by the U.S. Food and DrugAdministration(FDA)andEuropeanMedicinesAgency(EMA) in 2011, are the first-generation protease inhibitors (PIs) cur-rently availablein clinicalpractice.Bothareadministeredusing a response-guided protocol, in which viral decline determines treatment-duration [7–11]. Allguidelines setthe first viraemia check-point at week-4. Nevertheless, given the rapid HCV-dynamicsduringPI-basedtripletherapy,earliertime-pointsmay beadditionallyinformativeonexpectedtreatmentoutcome,and thereforebecomeusefulinclinicalpractice.
Moreover,atypicalfeatureofHCVistheabilitytodevelop/select resistanceassociatedvariants(RAVs)duringtreatment,asa conse-quenceofpotentialnaturalresistanceandlowgeneticbarrierof first-generationPIs[12–14].Virological-failuretoTVRandBOCis indeedassociatedwithRAVsdevelopmentinthevastmajorityof cases[15–18].
WhenRAVsarepresentatbaseline,eitherasmajorviral popula-tionorasminorityvariants,theycouldgreatlyaffectviralresponse totreatment,particularlyinmonotherapy,determininga subop-timal viraldecay andthusfurther increasinginresistance level [14,19–26].Thispointshouldbetakenintoaccounttofully deter-minethekineticsofHCV-RNAdecay.
In the present study, a large heterogeneous population of patients infected with HCV genotype 1 treated with TVR- or BOC-based triple-therapy wasanalyzed, in order toinvestigate HCV-kineticsaccordingtopatients’complexityinreal-lifesettings. ThekineticsofviralresponsewasassessedshortlyafterPI’sstart, andwascorrelatedwithbothclinicaloutcomeandviralgenetic background,focusingonbaseline/earlydetectionofRAVs.Several cut-offscategorizingearlyHCV-RNAdecaywerethenevaluated, inordertoprovideausefultoolforthemonitoringofvirological responsetofirst-generationPIsinclinicalpractice.
2. Methods 2.1. Patients
Chronically HCV genotype 1 infected patients, consecutively seenatseveralItalianclinicalcentresbetweenJanuary2011and August 2013and startingatriple-therapybasedonPegIFN/RBV plusBOCorTVR,wereconsideredforinclusion.Onlypatientswith
available treatment outcome were considered for the analysis. Exclusion criteria were age under 18 years and other chronic liver diseases. Patients who stopped triple-therapy early for any other reasons than virological breakthrough or stopping rules were also excluded. Treatment schedules and stopping rulesfollowedTVR/BOCprescribinginformation[8,9].Thechoice betweenaBOC-orTVR-basedregimenwasattheinvestigator’s discretion.
This study was conducted in accordance with the princi-ples of the Declaration of Helsinki and approved by the local EthicsCommittees.Allenrolledpatientsprovidedwritteninformed consent.
2.2. Patientsmonitoring
FibrosisstagingwasdeterminedusingeitherFibroscan® (Echo-sens,Paris,France),Fibrotest®(Biopredictive,Paris,France)orliver biopsy,andinterpretedbyanexpertpathologist.
HCV-RNA viral load quantification was performed using theCOBASAmpliprep/COBASTaqMan HCV quantitativetest v2.0 (RocheDiagnostics)orAbbottRealTimeHCVassay(Abbott Labo-ratories,AbbottPark,IL,USA)withlowerlimitofdetection(LLOD) of15and12IU/mL,respectively.Inadditiontostandardviraemia check-points[7],HCV-RNAwasalsodeterminedat48h,week-1 andweek-2afterPIstart(TVRorBOC).
Plasmasampleswerecollectedandstoredat−80◦Caftereach
visit.
2.3. NS3-proteasesequencing
Genotypicresistancetest(GRT)onNS3-proteasesequences(aa 1–181)wasperformedbyanhome-madepopulation-sequencing protocol as elsewhere described [12]. Baseline-GRT was per-formedfor110patientsincludedintheanalysis,onthebasisof samples’availability.For39patientswasalsoavailablean addi-tional GRT at early time points, between 48h and week-4 of triple-therapy.
ThefollowingPIRAVswereconsideredintheanalyses:36AGLM, 41R,43ISV,54ASV,55IA,80K,155IKMQT,156GSTV,168AEGNTVY and170AT.
2.4. Statisticalanalysis
Resultsareexpressedasmedianvaluesandinterquartilerange (IQR).ValueswerecomparedusingtheMann–WhitneyU-test.
Sensitivity,specificityandpositivepredictivevalueswere cal-culatedtoevaluatethepredictionofvirological-successinrelation toHCV-RNAvaluesafter48h,2weeksand4weekssincePI-start. CorrelationcoefficientbetweenbaselineHCV-RNAandHCV-RNA at 48h and at week-2 of triple therapy was determinedusing Spearman rankcorrelationtest.AROC curveanalysiswasused todeterminetheoptimalHCV-RNAcut-offfortreatmentoutcome prediction.
Linear logistic regression analysis was used toestimate the associationbetweensustainedviralresponse(SVR)andHCV-RNA valuesatweek-2and week-4since PI-start,stratifiedaccording tothepredictioncut-off.HCV-genotype,gender,age,diagnosisof cirrhosis,null-responsetopreviouspegIFN+RBVadministrations,
lowbaselineviraemia(≤600,000IU/mL)andpresenceofatleast onebaseline/earlyRAVwereusedaspotentialconfounders.
All the analyses were performed using the statistical open sourceenvironmentR(version3.1.1)andSPSSsoftwarepackage (version19.0)forWindows(SPSSInc.,Chicago,IL).
3. Results
3.1. Studypopulation
Atotalof287patientswereenrolledforaPI-basedtreatment (BOC=79,TVR=208)(Supp.Fig.S1).Outofthese:84(29.3%) pre-maturelyinterruptedtreatment,duetovirological-failure(48/84, 57.1%),adverse events (31/84, 36.9%)or poor adherence (5/84, 6.0%);93/287(32.4%)arestillon-treatment,while110/287patients (38.3%)reachedEndOfTreatment(EOT)withundetectable HCV-RNA.Among theEOT patients, 104(94.5%) reached SVR, while 6 patients showed a relapse after treatment discontinuation (BOC=3,TVR=3).OnlypatientsexperiencingeitheratleastSVR12
(N=104) or virological failure (N=54) were consideredfor the analysis.
TwelveHIV-HCVcoinfectedpatientswereincludedinthestudy population(BOC=2andTVR=10),and9/12reachedSVR(2 TVR-treated coinfected patients resulted partial respondersand the thirdexperiencedarelapseafterTVRdiscontinuation).
Baselinedemographicandvirologicalcharacteristicsofthe158 patientsincludedintheanalysisarereportedinTable1.HCV GT-1bwasthemostrepresentedsubtype(88/158patients [55.7%]), but no statistically significant differences were highlighted betweenGT-1aandGT-1bfortheprevalenceofcirrhosis(56.5% versus 42.0%, p=0.072), baseline HCV-RNA (median [IQR]=6.0 [5.4–6.5]logIU/mL versus 5.6 [5.0–6.4]logIU/mL, p=0.055) and baselineALTvalues(median[IQR]=96[65–140]IU/mLversus88 [64–116]IU/mL,p=0.322).
Overall,18.4%ofpatients werenaïvetoanti-HCV treatment. Amongtreatment-experiencedpatients, themajority were pre-viousnon-responders(83/126,65.9%),and43/126 (34.1%)were previousrelapsers. Baseline lead-in (N=51) 48h (N=67) Baseline PI (N=158) Week 1 (N=44) Week 2 (N=88) Week 4 (N=154) L o gH C V -R N A ( IU /m L ) 1.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0
Fig.1. Box-plotsrepresentingtheoveralldistributionofhepatitisCvirusRNAvalues atbaselineandatearlytimepoints,uptoweek4,oftriple-therapyadministration. IU,internationalunits;PI,proteaseinhibitor.
3.2. EarlyHCV-RNAdecaykineticanditsimpactonthe achievementofarapidviralresponse
Arapidviralresponse(RVR)wasobservedin73/154(47.4%) patientswithavailableHCV-RNAatweek-4(34.9%withBOCand 52.3% with TVR, p=0.053). Out of 73 RVR-patients, 63 (86.3%) reachedthenSVRversus38/81(46.9%)non-RVRpatients(p<0.001). Basedontheavailabilityofsamples,earlytime-pointsafter PI-startwereanalyzedinasubsetofpatients:48hin67patients,and week-2in88patients.TheoveralldistributionofHCV-RNAvalues atearlytime-points,uptoweek-4oftriple-therapyadministration, isreportedinFig.1.
After 48h of PI-administration, the median (IQR) HCV-RNA decay was of −3.1 (−3.4; −2.5)logIU/mL, corresponding to a median (IQR) weekly viraemia decay-slope of −10.8 (−11.8; −8.9)logIU/mLperweek.
After48h,2patientshad alreadyHCV-RNAbelowthelower limitofdetection(TND).BothofthemreachedRVRandthenSVR. Moreover,RVRwasachievedin22/31(71.0%)patientswith48h HCV-RNAdetectablebut<1000IU/mLand12/34(35.3%)patients Table1
Baselineclinicalcharacteristicsofthe158proteaseinhibitortreatedpatientsincludedintheanalysis.
Patientsreceivingboceprevir Patientsreceivingtelaprevir Overall
Patients,N 44 114 158 Males,N(%) 32(72.7) 82(71.9) 114(72.2) HCV-1subtype,N(%) 1a 17(38.6) 52(45.6) 69(43.7) 1b 26(59.1) 62(54.4) 88(55.7) 1g 1(2.3) 0(0.0) 1(0.6)
Age(years),Median(IQR) 54(45–61) 53(48–61) 53(47–61)
PatientswithunfavourableIL-28Bgenotype(CT/TT),N(%)a 26(83.9) 67(81.7) 93(82.3)
TimesinceHCVdiagnosis(years),Median(IQR) 14(8–18) 16(7–20) 15(7–20)
Stageofliverdisease,N(%)
F0–F2 10(22.7) 28(24.6) 38(24.1)
F3 21(47.7) 23(20.2) 44(27.8)
Cirrhosis(F4) 13(29.5) 63(55.3) 76(48.1)
Naivepatients,N(%) 3(6.8) 26(22.8) 29(18.4)
VirologicaloutcometopreviousSOC,N(%)
Non-responder 9(20.5) 8(7.0) 17(10.8)
Null-responder 14(31.8) 25(21.9) 39(24.7)
Partial-responder 8(18.2) 19(16.7) 27(17.1)
Relapse 10(22.7) 33(28.9) 43(27.2)
Unknown 0(0.0) 3(2.6) 3(1.9)
BaselineLead-inHCV-RNA(logIU/mL),Median(IQR) 6(5.5–6.6) 6.1(5.9–6.2)b 6.1(5.6–6.5)
BaselinePIHCV-RNA(logIU/mL),Median(IQR)c 5.1(3.7–5.7) 6(5.5–6.6) 5.8(5.1–6.4)
BaselineALT,Median(IQR) 86(50–117) 92(65–122) 90(64–122)
HCV,hepatitisCvirus;IQR,interquartilerange;SOC,standardofcare;IU,internationalunits;PI,proteaseinhibitor;ALT,alaninetransaminase.
aIL-28Bgenotypewasavailablefor31boceprevir-treatedpatientsandfor82telaprevir-treated.
b A4-weekssensitivitytestwithpeg-interferon+ribavirinwasperformedineightpatientsbeforetelaprevirstart. c HepatitisCvirusRNAvaluebeforeproteaseinhibitorstart.
Table2
Triple-therapyoutcomeinrelationto48handweek2hepatitisCvirusRNAvalues.
HCV-RNAat48h(IU/mL) p-valuea HCV-RNAatweek2(IU/mL) p-valuea
TND <1000 >1000 TND <100 >100
PatientswithavailableHCV-RNAvalue,N 2 31 34 23 34 31
PatientsexperiencingRVR,N(%)a 2(100) 22(71.0) 12(35.3) 0.004 23(100) 19(55.9) 3(9.7) <0.001
PatientsexperiencingSVR,N(%) 2(100) 25(80.6) 20(58.8) 0.102 23(100) 26(76.5) 11(35.5) <0.001 Patientsexperiencingvirologicalfailure,N(%)b – 6(19.4) 14(41.2) 0.102 – 8(23.5) 20(64.5) <0.001
Virologicalbreakthrough,N(%) – 6(100) 13(92.9) – 6(75.0) 20(100)
Relapse,N(%) – – 1(7.1) – 2(25.0) –
HCV,hepatitisCvirus;TND,targetnotdetected(belowthelowerlimitofdetection);IU,internationalunits;RVR,rapidvirologicresponse;SVR,sustainedvirologicalresponse.
aP-valuewascalculatedbyChiSquaretestfortrend.
with48hHCV-RNA>1000IU/mL(p=0.004)(Table2).The abso-lutevalueofHCV-RNAat48hwasindependentbyHCV-genotype, previousresponsetopegIFN+RBVdual-therapyanddiagnosisof cirrhosis(Fig.2,panelA),aswellasbythetypeofPIused(p=0.905 byMann–Whitneytest,datanotshown).
Atweek-2,viraemia slopecontinuedtodiminishas patients approached undetectability,following aclassicalbiphasic kinet-ics.Indeed,themedian(IQR)HCV-RNAdecaywasof−4.1(−4.9; −3.6)logIU/mLand theweeklyviraemia decay-slope reached a median(IQR)valueof−2.1(−2.4;−1.8)logIU/mLperweek,thus muchslowercomparedtothatestimatedat48h.Atweek-2,31/88 (35.2%) patients still had viraemia >100IU/mL (Table 2): 87.1% (27/31) of them were previousnon-responders,supportingthe significantly slowersecond phase kinetic observed in this cat-egory of subjects versus previousrelapsers or naïve (p=0.002; Fig.2,panelA).Interestingly,only3/31(9.7%)patientswith HCV-RNA>100IU/mLatweek-2experiencedanRVRafterwards,versus 23/23TNDpatients(p<0.001byFisherexacttest)(Table2).
Therefore,earlydeterminationofHCV-RNAat48hand week-2 already allows the identificationof: (a)patients with higher chancesofreachingRVRand(b)patientswithsuboptimal viral-response.NodifferenceswerehighlightedinearlyHCV-RNAlevels followingTVRorBOCadministration.
3.3. ImpactofearlyHCV-RNAdecaykineticontreatment outcome
HCV-RNA valueat week-2 wassignificantly associated with virological-outcome, being lower in patients reaching SVR (median[IQR]HCV-RNA2w=1.2[0.3–1.7])incomparisontothose
experiencing virological-failure (median [IQR] HCV-RNA2w=2.3
[1.9–2.8];p<0.001byMann–WhitneyUtest)(Fig.2,panelB).On thecontrary,HCV-RNAvalueat48hdidnotreachstatistical signif-icance(p=0.139)(Fig.2,panelB).
Inparticular,virological-failurewasneverobservedin23TND patientsatweek-2(4BOCand19TVR),whereasitoccurredin8/34 patients(23.5%)with≤100IU/mL(1BOCand7TVR)andin20/31 patients(64.5%)(7BOCand13TVR)with>100IU/mLatweek-2 (p<0.001byChi2TestforTrend)(Table2).Notably,inour
popula-tion,only1RVRpatientexperiencedvirological-failure,andhehad HCV-RNA=430IU/mLatweek-2.
3.4. DeterminantsofearlyHCV-RNAdecayduringPI administration
TherapidHCV-RNAdecaysobservedatboth48hand week-2 of triple-therapy administration were highly homogenous in
H CV -RN A 4 8 h ( lo g IU /m L ) 0 1.0 2.0 3.0 4.0 5.0 Naïve patients (N=8) GT-1a (N=27) GT-1b (N=40) Previous relapsers (N=22) Previous non-responders (N=37) Cirrhotic patients (N=31) Non-cirrhotic patients (N=36) SVR to triple therapy (N=47)
Virological failure
to triple therapy (N=20) Naïve patients (N=18) GT-1a (N=36) GT-1b (N=52) Previous relapsers (N=24) Previous non-responders (N=46) Cirrhotic patients (N=45) Non-cirrhotic patients (N=43) SVR to triple therapy (N=62)
Virological failure
to triple therapy (N=26)
A
B
p=0.222a p=0.638b p=0.421a p=0.139a H C V -R N A 4 8 h ( lo g IU /m L ) 0 1.0 2.0 3.0 4.0 5.0 p=0.078a p=0.002b p=0.104a P<0.001a H C V -R N A 2 w ee k s (l o g IU /m L ) 0 1.0 2.0 3.0 4.0 5.0 6.0 H C V -R N A 2 w ee k s (l o g IU /m L ) 0 1.0 2.0 3.0 4.0 5.0 6.0Fig.2. Box-plotsreportingthedistributionofhepatitisCvirusRNAvaluesat48handweek-2oftriple-therapyadministration.Patientswerestratifiedaccordingto baselinecharacteristics(A),suchasinfectinghepatitisCvirussubtype,previoustreatmentexperienceanddiagnosisofcirrhosis;andaccordingtofinaloutcomeof telaprevir/boceprevir-basedtripletherapy(B).Thenumberofpatientsincludedineachcategoryarereported.ap-valueswerecalculatedthroughMann–WhitneyUtest.b
<100 IU/mL (N=57) >100 IU/mL (N=31) Week 2 HCV-RNA level L o g H C V -R N A 4 8 h -2 w dec a y ( IU /m L) 2.0 0 -2.0 -4.0 2.2 P<0.001
Median (IQR) = -1.7 (-2.3; -0.9) Median (IQR) = -0.9 (-1.2; -0.2)
Fig.3.Box-plotrepresentingthelogarithmichepatitisCvirusRNAdecaybetween 48handweek-2oftriple-therapyadministrationinpatientswithhepatitisCvirus RNAatweek-2lowerorhigherthan100IU/mL.Thenumberofpatientsincludedin eachcategoryarereported.P-valuewascalculatedthroughMann–WhitneyUtest. IU,internationalunits;IQR,interquartilerange.
allpatientsanalyzed,irrespectiveofHCV-genotype(p=0.848and p=0.849byMann–WhitneyUtest,respectively),previous treat-ment outcome (p=0.367 and p=0.831 by Kruskal–Wallis test, respectively),and diagnosisof cirrhosis (p=0.505 and p=0.753 byMann–WhitneyUtest, respectively)(Supp.Fig. S2,panels A andB).
Since the observed viraemia decays from baseline-to-48h and from baseline-to-week-2 were homogeneous, the absolute values of 48hand week-2 HCV-RNAwere expectedtodepend uponbaseline viraemia.Indeed, Spearman correlationtest con-firmeda strongpositive associationof HCV-RNA atthese early time-pointswithbaselinevalues.Thisindicatesthat,byincreasing baselineviraemia,alsoviraemiaatearlytimepointsconsensually increase,withastrongerconcordanceat48h(linearcorrelation coefficient=0.61,p<0.001)incomparisontoweek-2(linear cor-relationcoefficient=0.51,p<0.001).Thereasonforthisdifference couldlieintheobservationthat HCV-RNAvalueatweek-2was notonlydependentuponbaselineviraemia,butalsoontheslope ofsecond-phaseHCV-RNAdecay.Indeed,themedian(IQR) HCV-RNA48h-week2decaywas−0.9(−1.2;−0.2)logIU/mLinpatients with>100IU/mLat week-2 and −1.7 (−2.3;−0.9)logIU/mL in patients with <100IU/mL (p<0.001 by Mann–Whitney U test) (Fig.3).
3.5. PredictivevalueofHCV-RNAdeterminationafter48hand2 weeksoftriple-therapy
ToidentifytheoptimalHCV-RNAvalue(atboth48hand week-2time-points)abletodiscriminatevirological-outcome,bothin termsofsensitivityandspecificity,ROCcurveanalyseshavebeen performed.ROCcurveanalysisidentifiedaweek-2HCV-RNAvalue of 100IU/mL as an optimal cut-off in predicting SVR (sensitiv-ity=81%,specificity=71%).Overall,thepositivepredictivevaluefor HCV-RNA<100IU/mLatweek-2oftriple-therapywasof86%.
At48h, ROC curveanalysisidentified as optimal cut-off for sensitivity in predictingSVR an HCV-RNA valueof 1000IU/mL. Nevertheless,thisoptimizedcut-offofHCV-RNAat48hoftriple therapyhadlowersensitivityandspecificityinpredictingSVR (sen-sitivity=57%,specificity=70%)incomparisontoHCV-RNAcut-off of100IU/mLatweek-2,reachingapositivepredictivevalueof82%. TheachievementofRVRhadthebestspecificityinpredicting SVR(81%),butsensitivitywaslowerrespectingtoweek-2cut-offof 100IU/mL(63%versus81%,respectively).Positivepredictivevalue wasidenticaltothatobtainedwithweek-2cut-off(86%).
Therefore,inourstudypopulation,forSVRprediction,theuse of100IU/mLcut-offatweek-2seemstobethemostsuitablein comparisonto1000IU/mLcut-offat48handdetectableHCV-RNA valuesatweek-4.
By univariable logistic regression analysis, HCV-RNA≤ 100IU/mLat week-2wasalsoa strong predictorof SVR (Odds Ratio,OR[CI]=13.0[4.1–41.4],p<0.001)(Table3),evenstronger thantheachievementofanRVR(OR[CI]=8.5[2.5–28.8],p<0.001). Notably,theassociationbetweenHCV-RNAvalue≤100IU/mLand virologicalsuccesswasalsoconfirmedbymultivariableanalysis (OR[CI]=20.3[2.7–152.7],p=0.003)(Table3),aftercorrectionby baseline/earlydetectionofresistance,gender,age,HCV-genotype, cirrhosis,previousnull-responsetopegIFN+RBVandlowbaseline viraemia(≤600,000IU/mL).AlsotheRVRcorrelationwithSVRwas confirmedaftercorrectionbyconfoundingvariables(OR[CI]=18.7 [2.2–157.5],p=0.007).Amongconfoundervariables,alsofemale genderwassignificantlyassociatedwithsuccessbymultivariable analysis analysis (OR [CI]=11.0 [1.2–99.3], p=0.032) (Table 3). Probablyduetothehighnumberofmissingvalues,unfavourable IL-28Bgenotype(CTorTT)wasnotsignificantlyassociatedwith success by univariable analysis (OR [CI]=0.2 [0.0–1.9], p=0.64, datanotshown),anditwasnotconsideredasconfoundervariable inmultivariableanalysis.
3.6. Baseline/earlyresistanceandearlyvirologicalresponse NoneofthepatientswithundetectableHCV-RNAatweek-2 pre-sentedevidencesofbaseline/earlyRAVsbypopulationsequencing (0/17and0/8withavailableGRTatbaselineandwithinthefirst 4weeksofPIadministration,respectively).Allthesepatientsare SVR12.Onthecontrary,atleastonebaseline/earlyRAVwasfound
Table3
Univariableandmultivariablelogisticregressionmodelsforassociationwithvirologicalsuccesstotripletherapy.
Characteristic CrudeOR 95%C.I. p-value AdjustedOR 95%C.I. p-value
Lower Upper Lower Upper
HCV-RNA≤100IU/mLatweek2 13.0 4.1 41.4 <0.001 20.3 2.7 152.7 0.003
Atleastonebaseline/earlyresistancemutation 0.6 0.2 2.2 0.441 0.5 0.1 2.7 0.389
Gender(maleversusfemale) 2.7 0.9 8.3 0.083 11.0 1.2 99.3 0.032
Age(1yearhigher) 1.0 0.9 1.1 0.234 1.0 0.9 1.1 0.307
HCVgenotype(1aversus1b) 1.5 0.6 4.1 0.397 0.8 0.1 4.0 0.743
Cirrhosis(yesversusno) 1.0 0.4 2.6 0.975 0.6 0.1 2.7 0.500
PreviousnullrespondertoSOC 0.5 0.2 1.4 0.158 2.6 0.3 20.3 0.355
BaselineVL≤600,000IU/mL 2.3 0.8 6.4 0.116 1.4 0.3 7.9 0.672
in14/47patientswithdetectableHCV-RNAatweek-2(p=0.001by Fisherexacttest).
Atbaseline,asexpected,RAVsweredetectedmorefrequently inpatientsinfectedwithGT-1a(15/50,30.0%)thaninGT-1b(3/56, 5.4%). The mostcommondetectedbaseline RAVwastheQ80K, foundin11/110(10.0%,9/11GT-1a)patientsanalyzed,followed byT54Sin4/110(3.6%,2/4GT-1a).Inaddition,oneGT-1apatient showedtheco-presenceofV36L+Q80KRAVs.Virologicalfailure wasobservedin5/11patientswithbaselineQ80K(allGT-1a),3/4 patientswithT54SandinthepatientwithV36L+Q80K.
DenovoRAVsdevelopmentatearlytime-pointswasobserved in7/39(17.9%)patients.AllhadHCV-RNAdetectableatweek-2and 5/7experiencedvirologicalfailure.
Overall,RAVseitheratbaselineordenovodevelopedduring PI-administrationhavebeenexclusivelyobservedinpatientsshowing detectableHCV-RNAvaluesatweek-2,whiletheywereabsentinall patientsreachingundetectableHCV-RNAvaluesatthistime-point. Thecombinationofresistanceandslowervirologicalresponsecan thusplayasynergicroleindeterminingtreatmentfailure.
4. Discussion
Thepresentstudyanalyzestheclinicalusefulnessofearlyviral responsetoTVRorBOCtreatmentinalargeandheterogeneous population ofpatients infectedwithHCVgenotype1.Wefound that,independentlyfromthePIemployed,a100IU/mLcut-offof HCV-RNA at week-2of triple therapy wasable todiscriminate patients withincreasedrisk of virological-failure from patients who, despitehepaticimpairmentorprevioustreatment experi-ence,aremostlikelytoreachtherapeuticsuccess.
ItiswellknownthatHCV-RNAdecayduringantiviraltreatment followsabiphasicprofile,asaconsequenceoftheadministration ofdrugswhichactbyblockingviralproduction[1,3].Theblockof denovovirionproductionfrominfectedhepatocytesdetermines arapidfirstphaseofHCV-RNAdeclineinserum.Aslowersecond phasemainlydependsupontheclearanceofinfectedcells, medi-atedthroughdeathorlossofreplicativeintermediates[3].PI-based tripletherapywasassociatedwithadeeperandfasterfirst-phase, aswellaswithafastersecond-phase,comparedtoregimens not-includingaDAA[3,27,28].
Inthisstudy,wefoundthatthefirst-phasedecline,classically comprisedwithinthefirst48hoftriple-therapyadministration, wasindeedveryintenseand, furthermore,highlyhomogeneous amongallpatientsanalyzed.Baselineclinicalcharacteristicshadno impactonthisfirst-phasedecline,andHCV-RNAvaluesat48hwere correlatedonlywithbaselineHCV-RNAvalues.ThisearlyHCV-RNA kineticswasalsocomparableamongthetwoPIs,supporting previ-ousresultsindicatingthattherapywithalead-inphasefollowedby additionofasingleDAAachievesasimilarearlyHCV-RNAreduction asincludingtheDAAfromthebeginning[29].
Recently,weshowedthattelapreviradministrationin difficult-to-treatpatients(i.e.previousnonrespondersorcirrhotic)leads to a slower second-phase decline in respect to naïve, non-cirrhoticpatients[6].Interestingly,thesecondphasedeclinewas particularly compromisedinthosewhoexperienced virological-breakthrough,andresultedinhigherHCV-RNAvaluesatweek-2of PI-basedtripletherapy.
Similarly,alsoinourwiderpopulation,patientswithdetectable HCV-RNA at week-2 more frequently experienced virological-failure to triple-therapy, either containingBOC orTVR. Indeed, virological-failure wasobservedin64.5% ofpatients with HCV-RNA>100IU/mLatweek-2versus0%inthosewithundetectable HCV-RNAvalues.
In our study population, two main mechanisms cooperated in determiningHCV-RNAlevelsatweek-2oftriple-therapy:(a)
baseline viraemia values and (b) the slope of second-phase (48h–week2)viraemiadecline.Highbaselineviraemiaisaknown riskfactorforvirological-failure[30,31],andthusitscorrelation withlowerresponsivenessisnotsurprising.Theslopeofsecond phasedecline,ontheotherhand,directlydependsontreatment efficacy,aspreviouslydemonstrated[3].Thiscorrelationwas con-firmedinthisstudy,sincepatientswithhigherweek-2viraemia were,inthemajorityofcases,thosewithpoorsensitivitytoIFN (previousnon-responders).
Overall,the determination of HCV-RNA at week-2of triple-therapywashighlysensitiveinpredictingSVR,and allowedthe identificationofpatientswithoutstandingviral-response.Indeed, inourstudy,atweek-2,23/88(26.1%)patientshadundetectable HCV-RNA,with23/23(100%)patientsreaching anRVRandSVR afterwards,versusonly 3/31(9.7%) RVRand 11/31 (35.5%) SVR amongpatientswithHCV-RNA>100IU/mL.Moreover,6cirrhotic patients and 4patients previously non-responder todual ther-apyhadundetectableviraemiaatweek-2,andnotablyallofthem achievedSVR.
Thecut-offsetat100IU/mLforHCV-RNAatweek-2of triple-therapyshowedthehighestsensitivityinpredictingSVRamong otherearlyHCV-RNAassessments,includingtheachievementof anRVR,withapositivepredictivevalueof86%.Thestrengthofthis correlationamongHCV-RNA≤100IU/mLatweek-2andSVRwas alsoconfirmedbyregressionanalysis.Indeed,thisearlyviraemia check-pointwasfoundtobea predictorofSVR asgoodas the achievementofanRVR,evenaftercorrectionbythemost com-monclinicalandviralparametersclassicallyinvolvedinsensitivity tofirst-generationPIsadministration.
OurcurrentstudyshowsalsothattheanalysisofNS3sequence atbaselinemayfurtherstrengthentheabilitytoidentifypatients withhigherprobabilityof success.Indeed,baseline presenceof RAVs(includingtheQ80K)orearlydevelopmentwereexclusively observedinpatientswithstilldetectableHCV-RNAatweek-2,and theassociationofRAVswithsuboptimalHCV-RNAdecay uncov-eredaconditionofhigherriskofvirological-failure,suggestingthe needforaclosevirologicalfollow-up.
Pharmacoeconomicanalysesandclinicalstudiessuggestthat the“costperSVR”stillfavourstheuseofpegIFN-RBVdual-therapy, ratherthanPI-basedtriple-therapy,inpatientscomplying “easy-to-treatcriteria”,suchasthosewhoexhibitexcellentviralresponse duringthefirst4weeksoftreatmentwithpegIFNandRBValone [30,32–35]. On the other hand, for patients who do not fully comply withpredictors of viral-response(i.e. beingdrug-naïve, non-cirrhotic,lowbaselineHCV-RNA,preferentiallyIL28-Bnon-TT andachievinganRVR),additionalevaluationofHCV-RNAatearly time-pointsduringPI-basedtriple-therapycouldactuallylimit:(a) theriskoffailurewiththedevelopmentofRAVs[17];(b)theonset ofseriousadverseevents;and(c)costsrelatedtodrugsand man-agementofcomplications.
Overall,thisstudyhasitsstrengthsandlimitations.Evenifthis isthelargeststudy,toourknowledge,analyzingearlyviral-kinetics withfirst-generationPIs in clinical practice,thepredictive val-uesofearlyHCV-RNAdecaycouldchangebyanalyzingawider population.Theadinterimnatureofthisstudyalsoaccountsfor aquitehighrateofvirological-failureobserved(54/158[34.2%]). Indeed,even ifthemajority ofpatientsstarted treatmentfairly closetogether,treatment-failuresgenerallyoccurafterfewweeks oftherapy,whilealongerperiodoftimeisnecessaryfortheSVR evaluation,therefore resultingin anunderestimation ofits fre-quency.Forthis reason,theoverallSVRratesshouldnotbeyet extrapolatedinthisstudy,andindeedtheassessmentoftreatment efficacywasnotanendpoint.
Inconclusion,duetotherapidevolutionoftreatmentscenarios withdifferentefficacyandcosts,itisimportanttoimplement tail-oredtreatments.Thepresentstudyprovidesaproofofconcept,to
beconfirmedondifferentpopulationsandonalongertimespan, thatarapidearlyHCV-RNAdecay,withundetectableHCV-RNAat 2weeksafterfirst-generationPIinitiation,mayallowtopredict agoodtherapeuticoutcome,evenindifficult-to-treatpatients.On thecontrary,asuboptimalearlydecay,especiallyifassociatedwith thedevelopmentofRAVs,speaksinfavourofahighlyprobable virologicalfailure.
Conflictofinterest
A.Gasbarrini,A.DiBiagio,M.Angelico,C.F.PernoandF. Ceccherini-Silberstein have no conflict of interest with respect to this manuscript,however:A.Gasbarrinihasreceivedpriorconsultancy honorariafromMSD,gilead,jansenn,abbvie,alfawassermann.A. DiBiagiohasreceivedpriorresearchfundingand/orconsultancy honorariafromBristolMyersSquibb,Gilead,Janssen,MerckSharp &Dohme, Roche, and ViiV.M. Angelico has received funds for attending symposia,speaking, organizingeducational activities, grantresearchsupportandconsultancyfromRoche,Gilead,Merck Sharp&Dohme,JanssenCilag.C.F.Pernohasreceivedfundsfor attending symposia,speaking, organizingeducational activities, grantresearchsupport,consultancyandadvisoryboard member-ship, fromAbbott, Boehringer Ingelheim, Bristol MyersSquibb, Gilead,MerckSharp&Dohme,JanssenCilag,Pfizer,Tibotec,Roche, andViiV.F.Ceccherini-Silbersteinhasreceivedfundsfor attend-ing symposia,speaking, organizing educationalactivities, grant researchsupport,consultancyand advisoryboardfromAbbVie, MerckSharp&Dohme,Gilead,JanssenCilag,Roche,BristolMyers Squibb,andViiV.Theotherauthorsofthisstudydonothave any-thingtodisclose regardingfunding fromindustry orconflict of interestwithrespecttothismanuscript.
Acknowledgements
Thisworkwassupported by theItalian Ministryof Instruc-tion,UniversityandResearch(MIUR)(AccordidiProgramma2011: RBAP11YS7K001[HIRMA],BandieraInterOmicsProtocolloPB05 1◦),byMerckSharpe &Dohmeand byAviraliaFoundation.The
fundershadnoroleinstudydesign,datacollectionandanalysis, decisiontopublish,orpreparationofthemanuscript.
WethankRocheandAbbottforprovidinglaboratorymaterial forHCV-RNAdetermination,theClinicalHepatologyandMolecular Virologygroupsof“TorVergata”UniversityHospital,and particu-larlyFrancescaStafoggiafortechnicalsupport.
AppendixA. Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttp://dx.doi.org/10.1016/j.dld.2014.11.010. References
[1]NeumannAU,LamNP,DahariH,etal.HepatitisCviraldynamicsinvivoand theantiviralefficacyofinterferon-alphatherapy.Science1998;282:103–7.
[2]DixitNM,Layden-AlmerJE,LaydenTJ,etal.Modellinghowribavirinimproves interferonresponseratesinhepatitisCvirusinfection.Nature2004;432:922–4.
[3]Guedj J,Perelson AS.Second-phasehepatitisCvirusRNA declineduring telaprevir-basedtherapyincreaseswithdrugeffectiveness:implicationsfor treatmentduration.Hepatology2011;53:1801–8.
[4]PowersKA,DixitNM,RibeiroRM,etal.Modelingviralanddrugkinetics: hep-atitisCvirustreatmentwithpegylatedinterferonalfa-2b.SeminarsinLiver Disease2003;23(Suppl.1):13–8.
[5]ShudoE,RibeiroRM,PerelsonAS.ModellingthekineticsofhepatitisCvirus RNAdeclineover4weeksoftreatmentwithpegylatedinterferonalpha-2b. JournalofViralHepatitis2008;15:379–82.
[6]CentoV,DiMaioC,ValentiF,etal.HCVkineticsandquasispecies evolu-tionwithinthefirsthoursoftelaprevirtripletherapyindifficult-to-treat HCV-infectedpatients.TheLiverMeeting2013,AASLDAbstracts.Hepatology 2013;58(Suppl.1):939A#1509.
[7]EASL Recommendations onTreatment ofHepatitis C. InternationalLiver CongressTM2014;2014.
[8]EuropeanMedicinesAgency(EMA).Incivek(telaprevir)FilmCoatedTablets. EUPrescribingInformation.2011May1.
[9]EuropeanMedicinesAgency(EMA).Victrelis(boceprevir)Capsules.EU Pre-scribingInformation.2011May1.
[10]AssociazioneItalianaper loStudio del Fegato.Indicazioni e gestionedei trattamentianti-HCVnei pazienti con epatite cronica:linee di indirizzo dell’AssociazioneItalianaperloStudiodelFegato(AISF).31-3-0014. [11]Bruno S, Bollani S, Zignego AL, et al. Undetectable HCV-RNA at
treatment-week8 results in high-sustained virological response in HCV G1treatment-experiencedpatientswithadvancedliverdisease:the Inter-nationalItalian/Spanish Boceprevir/Peginterferon/Ribavirin Name Patients Program.JournalofViralHepatitis2014[Epubaheadofprint].
[12]CentoV,MirabelliC,SalpiniR,etal.HCVgenotypesaredifferentlypronetothe developmentofresistancetolinearandmacrocyclicproteaseinhibitors.PLoS ONE2012;7:e39652.
[13]DelangL,VliegenI,FroeyenM,etal.Comparativestudyofthegeneticbarriers andpathwaystowardsresistanceofselectiveinhibitorsofhepatitisCvirus replication.AntimicrobialAgentsandChemotherapy2011;55:4103–13.
[14]RongL,DahariH,RibeiroRM,etal.Rapidemergenceofproteaseinhibitor resistanceinhepatitisCvirus.ScienceTranslationalMedicine2010;2:30–2.
[15]JacobsonIM,McHutchisonJG,DusheikoG,etal.Telaprevirforpreviously untreatedchronichepatitisCvirusinfection.NewEnglandJournalofMedicine 2011;364:2405–16.
[16]ZeuzemS,AndreoneP,PolS,etal.TelaprevirforretreatmentofHCVinfection. NewEnglandJournalofMedicine2011;364:2417–28.
[17]PawlotskyJM.Treatmentfailureandresistancewithdirect-actingantiviral drugsagainsthepatitisCvirus.Hepatology2011;53:1742–51.
[18]HalfonP,LocarniniS.HepatitisCvirusresistancetoproteaseinhibitors.Journal ofHepatology2011;55:192–206.
[19]Bartels DJ, Zhou Y, Zhang EZ, et al. Natural prevalence of hepatitis C virus variants with decreased sensitivity to NS3.4A protease inhibitors in treatment-naive subjects. Journal of Infectious Diseases 2008;198: 800–7.
[20]CentoV,DiMaioVC,DiPaoloD,etal.NS3-resistanceinadvancedHCVpatients treatedwithBOC/TRV-basedtherapy:impactonvirologicalfailureofbaseline andearlytimepoints.TheLiverMeeting2013,AASLDAbstracts.Hepatology 2013;58(Suppl.1):1156A#1960.
[21]FrancoS,BellidoR,AparicioE,etal.NaturalprevalenceofHCVminorityvariants thatarehighlyresistanttoNS3/4Aproteaseinhibitors.JournalofViralHepatitis 2011;18:e578–82.
[22]Lopez-LabradorFX, MoyaA, Gonzalez-CandelasF.Mappingnatural poly-morphismsofhepatitisCvirusNS3/4Aproteaseandantiviralresistanceto inhibitorsinworldwideisolates.AntiviralTherapy2008;13:481–94.
[23]ValletS,VironF,HenquellC,etal.NS3proteasepolymorphismandnatural resistancetoproteaseinhibitorsinFrenchpatientsinfectedwithHCV geno-types1–5.AntiviralTherapy2011;16:1093–102.
[24]Vicenti I, Rosi A, Saladini F, et al. Naturally occurringhepatitis C virus (HCV)NS3/4Aproteaseinhibitorresistance-relatedmutationsinHCV geno-type 1-infectedsubjects in Italy.Journal of AntimicrobialChemotherapy 2012;67:984–7.
[25]OgertRA,HoweJA,VierlingJM,etal.Resistance-associatedaminoacidvariants associatedwithboceprevirpluspegylatedinterferon-alpha2bandribavirin inpatientswithchronichepatitisCintheSPRINT-1trial.AntiviralTherapy 2013;18:387–97.
[26]IzquierdoL,HelleF,FrancoisC,etal.SimeprevirforthetreatmentofhepatitisC virusinfection.PharmacogenomicsandPersonalizedMedicine2014;7:241–9.
[27]ChatterjeeA,SmithPF,PerelsonAS.HepatitisCviralkinetics:thepast,present, andfuture.ClinicsinLiverDisease2013;17:13–26.
[28]ChatterjeeA,GuedjJ,PerelsonAS.MathematicalmodellingofHCVinfection: whatcanitteachusintheeraofdirect-actingantiviralagents.AntiviralTherapy 2012;17(6PtB):1171–82.
[29]Rong L, Guedj J, Dahari H, et al. Treatment of hepatitis C with an interferon-basedlead-inphase:aperspectivefrommathematicalmodelling. AntiviralTherapy2014;19:469–77.
[30]AndriulliA,DiM,MargaglioneVM,etal.IdentificationofnaiveHCV-1patients withchronichepatitiswhomaybenefitfromdualtherapywithpeg-interferon andribavirin.JournalofHepatology2014;60:16–21.
[31]BergT,SarrazinC,HerrmannE,etal.Predictionoftreatmentoutcomein patientswithchronichepatitisC:significanceofbaselineparametersandviral dynamicsduringtherapy.Hepatology2003;37:600–9.
[32]PearlmanBL,EhlebenC.HepatitisCgenotype1viruswithlowviralload andrapidvirologicresponsetopeginterferon/ribavirinobviatesaprotease inhibitor.Hepatology2014;59:71–7.
[33]SerfatyL.IstherestillaroleforPEGIFN+RBVtherapyinpatientswithHCV genotype1?LiverInternational2014;34(Suppl.1):11–2.
[34]CammaC,PettaS,CabibboG,etal.Cost-effectivenessofboceprevirortelaprevir forpreviouslytreatedpatientswithgenotype1chronichepatitisC.Journalof Hepatology2013;59:658–66.
[35]CammaC,PettaS,EneaM,etal.Cost-effectivenessofbocepreviror telapre-virforuntreatedpatientswithgenotype1chronichepatitisC.Hepatology 2012;56:850–60.
