Predictors of treatment requirement in HBeAg-negative chronic hepatitis B patients with persistently normal alanine aminotransferase and high serum HBV DNA levels

(1)

Predictors

of

treatment

requirement

in

HBeAg-negative

chronic

hepatitis

B

patients

with

persistently

normal

alanine

aminotransferase

and

high

serum

HBV

DNA

levels

Aslı

Ormeci

a

,

Yucel

Aydın

a

,

Abdullah

Sumnu

b

,

Bulent

Baran

c

,

Ozlem

Mutluay

Soyer

a

,

Binnur

Pınarbasi

a

,

Suut

Gokturk

a

,

Mine

Gulluoglu

d

,

Derya

Onel

e

,

Selim

Badur

e

,

Filiz

Akyuz

a

,

Cetin

Karaca

a

,

Kadir

Demir

a

,

Fatih

Besisik

a

,

Sabahattin

Kaymakoglu

a,

*

a

DepartmentofGastroenterohepatology,IstanbulFacultyofMedicine,IstanbulUniversity,Capa,34390,Istanbul,Turkey

b

MedipolUniversity,DepartmentofNephrology,Istanbul,Turkey

c

DepartmentofGastroenterology,Koc¸UniversityHospital,Istanbul,Turkey

d

DepartmentofPathology,IstanbulFacultyofMedicine,IstanbulUniversity,Istanbul,Turkey

e_Department_of_{Microbiology,}_Istanbul_Faculty_of_Medicine,_Istanbul_University,_Istanbul,_Turkey

1. Introduction

HepatitisBvirus(HBV)infectionremainsanimportantpublic health concern; approximately 400 million people are infected

worldwide.1 _Chronic _hepatitis _B _(CHB) _has _a _wide _clinical

spectrum,rangingfromasymptomaticcarrierstatustocirrhosis and hepatocellular carcinoma.2,3 _Proper _management _of _the

diseaseisimportanttopreventmortalitybyreducingHBV-related complications.

ThepredominanttypeofCHBinfectionishepatitisBeantigen (HBeAg)-negative.ThisdevelopsafterthelossofHBeAgandcan subsequently remain in a low/non-replicative phase (inactive chronicHBVcarrierstatus)orprogresstoanactivephase.4,5The

ARTICLE INFO Articlehistory: Received7March2016

Receivedinrevisedform3September2016 Accepted5September2016

CorrespondingEditor:EskildPetersen, Aarhus,Denmark

Keywords: ChronichepatitisB HBeAg-negative PersistentlynormalALT

SUMMARY

Objectives:Serumalanineaminotransferase(ALT)isacontroversialmarkerfordiseasemonitoringin hepatitisBeantigen(HBeAg)-negativechronichepatitisB(CHB)patients.Theaimofthisstudywasto determinetheﬁbrosisstageandhistologicalactivityindex(HAI)inHBeAg-negativeCHBpatientswith persistentlynormalALT(PNALT)andhighserumHBVDNA(2000IU/ml)andtoinvestigateclinicalrisk factorsfortherequirementoftreatmentthroughtheexaminationofliverbiopsyspecimens.

Methods:HBeAg-negativeCHBpatientswithPNALT(40IU/l)andhighserumHBVDNA(2000IU/ml) wereincluded.HBVfibrosisstageandHAIwerescoredaccordingtotheIshaksystem.Multivariatelogistic regressionanalysiswasusedtoestimatetheindependentriskfactorsforfibrosisstage2and/orHAI6. Receiveroperatingcharacteristiccurveanalysiswasusedtodetermineanoptimalagecut-offforliverbiopsy. Results:A total 120 patients were enrolled. These patients had a mean HBV DNA level of 123680494500 IU/ml; the HBV DNA load was 2000–20000 IU/ml in 68 patients (56.6%) and 20000IU/mlin52(43.4%).Eighteenpatients(15%)hadmoderate-to-severehistologicalactivity(HAI6). Forty-threepatients(35.9%)hadafibrosisstage2.Forty-eightpatients(40%)hadafibrosisstage2and/or HAI6.Onmultivariatelogisticregressionanalysis,independentvariablesassociatedwithfibrosisstage2 and/orHAI6includedageandHBVDNAviralload.PatientswithHBVDNA2000–20000IU/mlweremore likelytorequiretreatmentcomparedtothosewithaviralload20000IU/ml.Theoptimalagecut-offto predictfibrosisstage2and/orHAI6was46years.

Conclusions: Signiﬁcantliverdamagewasdetectedin40%ofCHBpatientswithPNALTandhighHBV DNAuponbiopsy.AgeandHBVDNAviralloadwereindependentpredictorsofsigniﬁcantliverdamage. AbiopsytodeterminethedegreeofliverdamageisadvisableforCHBpatientsolderthan46years. ß2016TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases. ThisisanopenaccessarticleundertheCCBY-NC-NDlicense( http://creativecommons.org/licenses/by-nc-nd/4.0/).

* Correspondingauthor.Tel.:+902124142000(32671–31140); fax:+902126319743.

E-mailaddress:kaymakoglus@hotmail.com(S.Kaymakoglu).

ContentslistsavailableatScienceDirect

International

Journal

of

Infectious

Diseases

j o urn a l hom e pa ge : ww w. e l s e v i e r. c om/ l o ca t e / i j i d

http://dx.doi.org/10.1016/j.ijid.2016.09.007

1201-9712/ß2016TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

(2)

distinctionbetweenthesetwoconditionsiscrucial.Observation withouttreatment can besufﬁcient for HBVcarriers; however, HBeAg-negativeCHBinfectionsrequiretreatment.

HBeAg-negative CHB is generally differentiated from the inactive carrier state by serial measurement of serum alanine aminotransferase (ALT) and HBV DNA levels. Positivity for hepatitisBsurfaceantigen(HBsAg),negativeHBeAg,andelevated ALTandserumHBVDNAlevelsarediagnosticofHBeAg-negative CHB.6 _Serum_levels _of_ALT, _an_enzyme _released_by_hepatocytes

during liver injury, usually reflect thedegree of liver damage; however,noteveryHBeAg-negativeCHB-infectedpatientexhibits elevatedALT.HBVDNAviralloadandALTlevelscanfluctuatein casesofHBeAg-negativeCHBinfection.Asinglemeasurementof ALTorHBVDNAviralloadisinsufficienttodeterminethecurrent phaseofthedisease.AproportionofHBeAg-negativeCHBpatients canhavepersistentlynormalALT(PNALT)levelsforanextended period.7

The European Association for the Study of the Liver (EASL) deﬁnePNALTasanALT below40IU/lwhencheckedevery3–4 monthsinasingleyear.8

AccordingtothelatestEASLguidelines,HBeAg-negativeCHB patients with PNALT and HBV DNA levels between 2000 and 20000IU/ml,andwhohavenoevidenceofliverdisease,donot require immediate liver biopsy or treatment. However, it is recommended that they receive careful follow-up, with ALT assessmentsevery3months,aswellasHBVDNAload measure-mentsevery6–12months,foratleast3years.8

However, there have been reports of histological injury in patients with PNALT. Furthermore, more recent studies have shownliverdamageinCHBpatientswithPNALTwhohaveviral loadsabove2000IU/ml.9,10Liverbiopsycanbeusedtoassessthe severityofnecrosisandinflammationinadditiontofibrosis,and canruleoutothercausesofliverdisease;hence,biopsyisregarded asthebestmethodtoassesstheseverityofinflammatoryactivity andfibrosis.7

Itwashypothesized that a significantproportionof HBeAg-negativeHBV-infected patients withhighHBV DNAlevels may have significant histological liver abnormalities despite PNALT. Thisprospectivestudywasthereforeperformedtodeterminethe fibrotic stage and histological activity index (HAI) in HBeAg-negative CHB patients with PNALT and high serum HBV DNA (2000IU/ml)viralloads.Theclinicalriskfactorsassociatedwith significant histological abnormalities in liver biopsy specimens werealsoinvestigated.

2. Methods

2.1. Studydesignandsetting

Thestudywasdesignedasasingle-center,prospectivestudyin the Gastroenterohepatology Department of Istanbul Faculty of Medicine,IstanbulUniversity.Thestudyprotocolsabidedbythe ethicalguidelinesasstatedinthe1975DeclarationofHelsinkiand wereapprovedby thelocalinstitutionalreview board.Written informedconsentforparticipationinthestudywasobtainedfrom eachpatient.

2.2. Patients

A total 120 patients with CHB admitted to the university hospital gastroenterology department between April 2009 and December2012wereincludedinthisstudy.Patientspresenting directlytothegastroenterologydepartmentorreferredfromother clinicalcenterstothedepartmentclinicasatertiaryhealthcare institutionwererecruited;theywereincludedaccordingtothe date at ﬁrst presentation to the department clinic. Inclusion

criteriawereage18years,diagnosiswithCHBinfection(deﬁned aspositiveHBsAgformorethan6months),HBVDNAload2000 IU/mlatleasttwicewithinthe6–12-monthintervalchecksinthe pastyear,PNALT(accordingtoatleastfourvaluesobtainedat 3-monthintervalsinthepastyear),andnopreviousorconcomitant anti-HBV therapy. Patients with liver comorbidities including hepatitisdeltavirus(HDV)infection,hepatitisCvirus(HCV) co-infection,chronicalcoholconsumption(>30gofpurealcoholper day), Wilson’s disease, HIV co-infection,autoimmune hepatitis, presentorpastevidenceofanysymptomsrelatedtochronicliver disease,andimagingorlaboratoryresultsthatindicatedcirrhosis wereexcluded fromthestudy,as werethose withevidence of immunesuppression.

2.3. PNALTdeﬁnition

PNALTwasdeﬁnedaccordingtotheEASLguidelines;i.e.,ALT remainingbelow40IU/lwhencheckedevery3–4monthswithina singleyear.8_Serum_levels_of_ALT_were_measured_every₃_months

foratleast1yearbeforeliverbiopsy. 2.4. Serummarkers

All of the patients underwent serum biochemistry tests, including for ALT, aspartate aminotransferase (AST), alkaline phosphatase(ALP),gamma-glutamyltransferase(GGT),bilirubin, and alpha-fetoprotein, in addition to complete blood counts. ELISAswereusedtomeasureHBsAg,HBeAg,antibodiestoHBeAg (anti-HBe), antibodies to HCV (anti-HCV), HDV IgG antibodies (anti-HDV),andHIV.QuantitativeHBVDNAtestingwasperformed using the COBASAmpliPrep/COBAS Taqman 96 system (Roche, Branchburg,NJ,USA).Thedynamicmeasurementlevelofthekit rangesbetween20and170000000IU/ml.

2.5. Liverbiopsy

Allpatientsunderwentapercutaneousliverbiopsyguidedby ultrasonography. Liverbiopsieswereperformedusing18-gauge biopsy needles. The specimens obtained were fixed, paraffin-embedded, and stained with hematoxylin–eosin. Appropriate diagnosis of a biopsy specimenincluded the observation of at leastsixportalareasinthesample.Toavoiddifferencesandthe biasthatcanoccurbetweenexaminers,alldatawereexamined and evaluated by a single experienced pathologist who was blindedtotheclinicaldata.FibrosisandtheHAIwerescoredusing theIshakscoringsystem.11_Stages_of_fibrosis_ranged_from₀_(no

ﬁbrosis)to6(cirrhosis;probableordeﬁnite). 2.6. Treatmentindication

Treatment indication (significant histological abnormalities) wasdefinedasthepresenceofHAI6and/orthepresenceofstage 2fibrosisinliverbiopsyspecimens.Theoptimalagecut-offto detect treatment indication was determined through receiver operatingcharacteristic(ROC)curveanalysis.Theoptimalcut-off point was calculated using the ROC curve coordinates for treatmentindication(pointnearesttothetopleftcorner,yielding thebestrelationshipbetweensensitivityandspecificity). 2.7. Statisticalanalysis

NCSS (NumberCruncher Statistical System, 2007) and PASS (PowerAnalysisandSampleSize,2008)statisticalsoftwarewere used for the statistical analysis(NCSS LLC,Kaysville, UT, USA). Descriptive statistics such as the mean, standard deviation, frequency, and ratewereused. Furthermore, theStudentt-test

(3)

wasusedtocompareparameterswithanormalspreadbetween thegroups. Diagnostic screeningtests and area under theROC curve(AUC)analysiswereusedtodefinethecut-offpointforage. The95%confidenceintervals(CIs)weredeterminedforallresults; p<0.05wasconsideredsignificant.

3. Results

Atotal120patientswereincludedinthestudy,ofwhom58 (48.3%) were male. The mean age of the patients was 42.8811.32years(range20–72years).

Allpatients wereanti-HBe-positivewithPNALT.Noneof the patientswerepositive for anti-HCVoranti-HDVantibodies. All patientsunderwentliverbiopsiesafter1yearoffollow-up.The meanHAIscorewas2.892.30(median3,range0–8),whilethe meanstagewas1.10.95(median3,range0–18)accordingtothe Ishak scoring system. The baseline demographic and laboratory characteristicsofthe120patientsaresummarizedinTable1.

The histopathological ﬁndings on liverbiopsy are shown in

Table2.Nopatienthadliverdiseaseabovestage4.Therewerenine patients(7.5%)withaHAIscoreof6,seven(5.8%)withaHAIscore of7,andtwo(1.7%)withaHAIscoreof8;40%ofthepatientshad HAI6orstage2disease,orboth,whichindicatedtreatment.

TheHBVDNAloadofallpatientswasabove2000IU/mlatﬁrst examinationandduringthefollow-upperiod.ThemeanHBVDNA loadwas123680494500IU/ml (5.095.69log10 IU/ml). The

medianHBVDNAvaluewas18000IU/ml(4.2,range3.3–6.65log10

IU/ml).Withregardtoviralloaddistribution,68patients(56.6%)had anHBVDNAloadintherange2000–20000IU/mland43(35.9%)had anHBVDNAloadintherange20000–200000IU/ml.Ninepatients (7.5%)hadanHBVDNAloadhigherthan200000IU/ml.

3.1. Analysisoffactorsassociatedwithsignificantnecroinflammatory activityandfibrosis

Theeffect ofeach variableon significantnecroinflammatory activitywasassessed.UnivariateanalysisindicatedthatALT,AST, age,andmalesexwereassociatedwithsignificant necroinflam-matoryactivity(HAI6)inHBeAg-negativepatientswithPNALT (Table3).Moreover,agewasassociatedwithsignificantfibrosisin thesepatients(Table4).

Table1

Baselinedemographicandlaboratorycharacteristicsofthe120study patients

Patients(N=120) Age,years,meanSD 42.811.32 Sex,n(%) Male Female 58(48.3) 62(51.7) AST(IU/l) MeanSD Median(range) 236.06 22(12–40) ALT(IU/l) MeanSD Median(range) 258.25 24(12–40) ALP(IU/l) MeanSD Median(range) 126.3255.23 120(34–332) GGT(IU/l) MeanSD Median(range) 22.6414.84 20(7–131) Totalbilirubin,mg/dl MeanSD Median(range) 0.640.33 0.6(0.2–1.2) Alpha-fetoprotein,ng/ml MeanSD Median(range) 3.022.95 2(0.10–15.0) HBVDNA(log10IU/ml)

MeanSD Median(range) 5.095.69 4.2(3.3–6.65) Stage MeanSD Median(range) 1.10.95 1(0–6) HAI MeanSD Median(range) 2.892.30 3(0–18)

ALP,alkalinephosphatase;ALT,alanineaminotransferase;AST,aspartate aminotransferase;GGT,gamma-glutamyltransferase;HAI,histological activityindex;HBV,hepatitisBvirus;SD,standarddeviation.

Table2

Histopathologicalﬁndingsinbiopsyspecimensofchronic hepatitis B patients with persistently normal alanine aminotransferase,accordingtotheIshakscoringsystem

Liverbiopsyﬁndings n(%) Stage 0 38(31.7%) 1 39(32.5%) 2 38(31.7%) 3 2(1.7%) 4 3(2.5%) HAI <6 102(85%) 6 18(15%) HAI6orstage2 Negative 72(60.0%) Positive 48(40.0%) HAI,histologicalactivityindex.

Table3

Univariate analysis of factors associated with signiﬁcant necroinﬂammatory activityinHBeAg-negativepatientswithpersistentlynormalalanine aminotrans-ferase

Necroinﬂammatoryactivity p-Value HAI<6 HAI6 Age,years n=102 n=18 MeanSD 41.8611.29 48.679.98 0.018c_* Median(range) 41.5(20–72) 49(30–68) Sex 0.003b ** Male 43 15 Female 59 3 AST,IU/l n=102 n=18 MeanSD 22.645.86 25.836.63 0.038c_* Median(range) 22(12–40) 25.5(14–37) ALT,IU/l n=102 n=18 MeanSD 24.248.05 29.338.24 0.015c * Median(range) 23(11–40) 29.5(12–40) ALP,IU/l n=102 n=18 MeanSD 123.5355.48 145.6251.4 0.076a Median(range) 120(46–332) 150(34–210) GGT,IU/l n=102 n=18 MeanSD 22.5115.57 23.59.14 0.261a Median(range) 20(7–131) 23(8–48) Totalbilirubin,mg/dl n=102 n=18 MeanSD 0.640.32 0.70.4 0.766a Median(range) 0.6(0–2.19) 0.67(0.31–1.78) Alpha-fetoproteinng/ml n=102 n=18 MeanSD 3.183.12 2.231.76 0.434a Median(range) 2.2(0.1–15) 1.95(0.15–5) Plateletcount,109 /l n=102 n=18 MeanSD 243.564.4 234.249.0 0.560c Median(range) 233.5(95–421) 226(166–350) HBVDNA,log10IU/ml n=102 n=18

MeanSD 5.135.72 4.725.03 0.763a

Median(range) 4.2(3.3–6.6) 4.06(3.34–5.64)

ALP,alkalinephosphatase;ALT,alanineaminotransferase;AST,aspartate amino-transferase;GGT,gamma-glutamyltransferase;HAI,histologicalactivityindex; HBeAg,hepatitisBeantigen;HBV,hepatitisBvirus;SD,standarddeviation.

a

Mann–WhitneyU-test;b

Yatescontinuitycorrectiontest;c

Studentt-test;*p<0.05, **p<0.01.

(4)

3.2. Univariateandmultivariateanalysesoffactorsassociatedwith significantfibrosisand/ornecroinflammation

Theeffectofeachvariableontreatmentindicationwasassessed in HBeAg-negative patients with PNALT. When patients were divided into two groups according to their viral load (2000– 20000IU/mlvs.20000IU/ml),univariateanalysisindicatedthat ageandHBVDNAviralload(2000–20000IU/ml)wereassociated with significant fibrosis and/or necroinflammatory activity (Table5).

Histopathologicalﬁndingsandreferralfortreatment(HAI6 and/orstage2)accordingtotheHBVDNAviralloadgroupare reportedinTable6.TheresultsshowedthatanHBVDNAviralload of 2000–20000 IU/ml (odds ratio 3.098, 95% CI 1.228–7.814, p=0.017)and olderage(odds ratio1.055,95% CI 1.010–1.103, p=0.017)wereriskfactorsforrequiringtreatment(HAI6and/or ﬁbrosis2)onmultivariateanalysis(Table7).

3.3. AUCfortheassociationofagewithtreatmentindication Agewasusedtopredicttheprobabilityofbeingdiagnosedwith signiﬁcanthistologicalabnormalitiesinHBeAg-negativepatients with PNALT. After assessment of the frequency of treatment indicationin all patients, thecut-off point for patientage was determinedtobe46yearsandabove.TheAUCoftheageassociated withrequiringatreatmentinterventioninthesepatientswas0.71 (95% CI 0.580–0.847; sensitivity = 84.62%,speciﬁcity = 59.81%, p=0.012)(Figure1).

4. Discussion

Inthepresentstudy,liverhistologycharacteristicsin HBeAg-negative patients with PNALT and HBV DNA 2000 IU/ml, in additiontoriskfactorsforadvancedﬁbrosisand necroinﬂamma-toryactivity,wereassessed.

HBV DNAviralload, ALT,and liverhistologyplay important rolesindeﬁningtheseverityofliverdiseaseinCHBpatients.8_The

Table4

UnivariateanalysisoffactorsassociatedwithsigniﬁcantﬁbrosisinHBeAg-negative patientswithpersistentlynormalalanineaminotransferase

Stage p-Value Stage<2 Stage2 Age,years n=77 n=43 MeanSD 41.2510.68 45.8111.98 0.034c * Median(range) 41(20–68) 47(23–72) Sex Male 36(46.8%) 22(51.2%) 0.785b Female 41(53.2%) 21(48.8%) AST,IU/l n=77 n=43 MeanSD 22.916.25 23.495.75 0.617c Median(range) 22(14–40) 23(12–37) ALT,IU/l n=77 n=43 MeanSD 24.788.25 25.48.32 0.696c Median(range) 24(11–40) 24(14–40) ALP,IU/l n=77 n=43 MeanSD 133.5956.49 111.5650.19 0.081a Median(range) 120(48–332) 110(34–218) GGT,IU/l n=77 n=43 MeanSD 23.4416.81 21.039.81 0.749a Median(range) 20(7–131) 20(7–58) Totalbilirubin,mg/dl n=77 n=43 MeanSD 0.660.34 0.610.31 0.384a Median(range) 0.6(0–2.19) 0.6(0.2–1.78) Alpha-fetoprotein,ng/ml n=77 n=43 MeanSD 3.213.12 2.752.69 0.564a Median(range) 2.2(0.1–15) 2.05(0.13–14) Plateletcount,109 /l n=77 n=43 MeanSD 248.763.6 230.258.4 0.119c Median(range) 239(138–421) 230(95–369) HBVDNA,log10IU/ml n=77 n=43

MeanSD 5.167.75 4.925.53 0.164a

Median(range) 4.3(3.3–6.6) 4.04(3.3–6.3)

a

Studentt-test;*p<0.05.

Table5

Univariateanalysisoffactorsassociatedwithsignificantfibrosisand/or necroin-flammationinHBeAg-negativepatientswithpersistentlynormalalanine amino-transferase

Treatmentindication(HAI6and/or stage2) p-Value Negative Positive Age,years n=72 n=48 MeanSD 40.8210.39 45.9812.06 0.014c * Median(range) 41(20–64) 46.5(23–72) Sex Male 34(45.8) 24(52.1) 0.628b Female 39(54.2) 23(47.9) AST,IU/l n=72 n=48 MeanSD 23.086.16 23.175.96 0.942c Median(range) 22(14–40) 22(12–37) ALT,IU/l n=72 n=48 MeanSD 24.888.20 25.198.40 0.840c Median(range) 24(11–40) 24(12–40) ALP,IU/l n=72 n=48 MeanSD 132.7756.88 115.7451.39 0.130c Median(range) 120(48–332) 115(34–218) GGT,IU/l n=72 n=48 MeanSD 23.2917.06 21.5810.38 0.971a Median(range) 20(7–131) 20(7–58) Totalbilirubin,mg/dl n=72 n=48 MeanSD 0.660.35 0.620.31 0.503a Median(range) 0.6(0–2.19) 0.6(0.2–1.78) Alpha-fetoprotein,ng/ml n=72 n=48 MeanSD 3.283.22 2.722.60 0.509a Median(range) 2.3(0.1–15) 2(0.13–14) Plateletcount,109_/l _n₌₇₂ _n₌₄₈ MeanSD 250.865.2 229.0855.5 0.060c Median(range) 239.5(138–421) 226(95–369) HBVDNA,log10IU/ml n=72 n=48

MeanSD 5.185.76 4.885.51 0.085a

Median(range) 4.30(3.30–6.65) 4.01(3.30–6.35) HBVDNA,3.30–4.30

vs.4.30log10IU/ml

34vs.38 34vs.14 0.008a**

a

Studentt-test;*p<0.05, **p<0.01.

Table6

HistopathologicalﬁndingsandreferralfortreatmentindifferentHBVDNAviral loadgroups HBVDNA p-Value 3.30–4.30log10IU/ml (n=68)a 4.30log10IU/ml (n=52)b Stage<2 Stage2 38(55.9%) 30(44.1%) 39(75%) 13(25%) 0.024* HAI<6 HAI6 56(82.4%) 12(17.6%) 46(88.5%) 6(11.5%) 0.253 Treatmentindicated Treatmentnot indicated 34(50%) 34(50%) 14(26.9%) 38(73.1%) 0.008**

HBV,hepatitisBvirus;HAI,histologicalactivityindex.

a

HBVDNA2000–20000IU/ml.

b _HBV_DNA

20000IU/ml. *p<0.05;**p<0.01.

(5)

ALTleveliscommonlyusedtoassesstheactivityofliverdisease andtoidentifypatientswhorequiretreatment.However,ALTmay beinﬂuencedbyvariousfactors,makingitanimperfectsurrogate marker.

EASLguidelines recommendnon-invasivefollow-upforliver ﬁbrosisinPNALTpatients withHBV DNA2000IU/ml with 3-month ALT assessment intervals.8 _Liver _biopsy _is _not

recom-mendedurgently, especially in PNALTpatients with HBV DNA levelsbetween2000and 20000 IU/ml. However, manystudies have reported signiﬁcant histological damage to the livers of HBeAg-negativeCHBpatientswithPNALTbuthighHBVDNAviral loads.9,10,12,13_Kumar_et_al._detected_stage₂_and_above_advanced

ﬁbrosisinover40%ofHBeAg-negativeCHBpatientswithPNALT.9

In thepresent study, thetop ALT limit of 40 IU/l wasused in accordancewiththeEASLguidelines,yetstage2andabovefibrosis wasdetectedin35.9%(n=43)ofpatients;moreover,significant necroinflammatory(HAI6)activitywasdetectedin15%(n=18) ofpatients.Thesedataindicatethatasizableproportionofpatients withPNALThaddeveloped significantliverdamage,asseen on liverbiopsy.

AccordingtotheEASLguidelines,highHBVDNAlevelsarearisk factorforliverdamage.8_HBV_DNA_is_also_a_predictive_factor_for

hepatocellularcarcinomaandcirrhosis.14–16_{Papatheodoridis}_et_al.

performedasystematicreviewandconcludedthatPNALTpatients withHBVDNAloadsof2000–20000IU/mlareindicatedforALT assessmentevery6months,HBVDNA loadmeasurementevery year,andfollow-upwithtransientelastographyandliverbiopsyin patientswithelevatedALTlevels.17Theysuggestclosefollow-up of patients who have HBV DNA loads of 2000–20000 IU/ml. Conversely, when patients were separated into two groups accordingtotheirviralload(2000–20000IU/mlvs.20000IU/ ml)inthepresentstudy,referralfortreatmentwassigniﬁcantly higher in the former group. An HBV DNA viral load of 2000– 20000IU/mlwasdeterminedtobeapredictivefactorforrequiring treatment intervention on univariate and multivariate analysis (oddsratio3.098,95%CI1.228–7.814,p=0.017).Itwasalsofound thathistologically signiﬁcantliverdiseaseisnotrarein HBeAg-negativepatientswithPNALTandwithserumHBVDNAviralloads of2000–20000IU/ml.Inpatientsinfectedlaterinlife,CHBmaybe indolent owing to immunoclearance; this results in HBeAg seroconversion and a relatively low HBV DNA viral load.18,19 ThesearethepatientsinwhomlowHBVDNAloadswithnormal ALTlevelsarelikely (inthosewho areHBeAg-negative).Asthe presentdatashow,whilethissituationiscomplex,histologically activeliverdiseasecanbesuspectedin patientswithHBVDNA viralloadsof2000–20000IU/ml.Inordertoassessthephaseof infection in CHB, serial HBV DNA and ALT measurements are required,asthenatureofCHBinfectionnecessitatesthesetests. PatientswithHBVDNAlevelsbetween2000and20000IU/mland normalALTareconsideredtobewithinthe‘grayzone’.The follow-upperiodinthisgroupofpatientsisveryimportantinorderto assessthephaseofinfection.

Inthisstudy,patientswithHBVDNAvaluesbetween2000and 20000 IU/ml and with persistently normal aminotransferase profileswerefollowedfor1year.Theshortfollow-upperiodmight a limitation of our study with regard to the definitI˙on of persistently normal ALT. A longer follow-up period would be beneficial for understanding the aforementioned ‘gray zone’ patients.

Thedurationofdiseaseisanotherfactorelicitingliverdamage. PatientswithPNALTwhowereinfectedperinatallyhave signifi-cantlyworseliverdamagethanthoseinfectedinadulthood,since thedamage of fibrosis increaseswithtime. Patients who were infectedprenatally,duringbirth,orintheirfirstorsecondyearof lifemayhavemoreextensiveliverdamageduetothefactthatthey havearelativelylongerimmune-tolerancephasefollowedbyan equallyprolonged immune-clearance phase.20–22 Inthe present patientgroup,thistranslatesintotheprobabilitythatthosewitha long disease duration may have significant histological liver damage.

SinceHBVtendstoexhibitavaryingcourse,asinglehighHBV DNA load at a single point in time is not sufﬁcient for an appropriateassessmentofthedisease.7_In_the_present_study,_the

HBVDNAviralloadwasmeasuredtwicebeforeliverbiopsy,which wasperformed1monthafterthelastmeasurement.Ifanyofthe HBVDNAviralloadmeasurementswerebelow2000IU/ml,that patientwasexcludedfromthestudy.

Age is also one of the most important factors that affect histological activity and the stage of disease.23 _All _guidelines

recommend that patient age should be considered in making treatmentdecisions,andpatientsover40yearsold(over30years old according to theEASL; over 40 years old accordingto the AmericanAssociationfortheStudyofLiverDiseasesandtheAsian Paciﬁc Association for the Study of the Liver) are currently considered at risk.7,8,23 _In _the _present _study, _age _was _used _to

predicttheprobabilityofrequiringtreatmentinHBeAg-negative patientswithPNALT.Followingliverhistopathologyassessment, theagecut-offwasdeterminedtobe46yearsandover.Treatment wasindicatedin50%ofpatientsolderthan46yearsand33.3%of

Table7

Multivariateanalysisfortreatmentindication

HAI6and/orstage2 Multivariateanalysis

OR 95%CI p-Value Age,years 1.055 1.010–1.103 0.017* Plateletcount,109/l 1.000 1.000–1.000 0.055 HBVDNA(3.30–4.30log10IU/ml)a 3.098 1.228–7.814 0.017*

Sex,male 1.025 0.405–2.594 0.958 CI,conﬁdenceinterval;HAI,histologicalactivityindex;HBV,hepatitisBvirus;OR, oddsratio.

a

HBVDNA=2000–20000IU/ml. *p<0.05.

Figure1.Areaunderthecurve(AUC)ofageassociatedwithtreatmentindicationin HBeAg-negativepersistentlynormalALTpatients(AUC=0.71,95%CI0.580–0.847; sensitivity=84.62%,speciﬁcity=59.81%,p=0.012).

(6)

patientswhowereyounger.TheAUCoftheageassociatedwith signiﬁcantliverhistopathologyinHBeAg-negativePNALTpatients was0.71(95%CI0.580–0.847;sensitivity=84.62%,speciﬁcity= 59.81%,p=0.012).

A limitation of the present study is that the patients’ HBV genotypes werenot investigated, since HBV genotyping is not routinelyperformedinclinicalpractice.

In this cohort, treatment necessity in the HBeAg-negative CHBinfectionpatientswithHBVDNAlevelsbetween2000and 20000 IU/mlwas signiﬁcantly greaterthanfound in previous studies in the literature. As mentioned before, longer and more detailed follow-up of these ‘gray zone’ patients, or re-evaluationof thepatientswithnon-invasivemethods(suchas theAST-to-plateletratioindex(APRI),ﬁbrosis4score,FibroTest, and vibration-controlled transient elastography), is recom-mended.

Inconclusion,signiﬁcantliverdamagerequiringtreatmentwas detectedin40%ofCHBpatientswithPNALTandahighHBVDNA load(2000IU/ml)after undergoingliverbiopsy.Furthermore, HBeAg-negativeCHBpatientswithPNALTdonothaveabenign prognosis, and patients with an HBV DNA viral loadof 2000– 20000 IU/ml who are 46 years old are more likely to be candidatesfortreatment.

Ethicalapproval:Thestudyprotocolwaspreparedinaccordance with the1975 version of the Declaration of Helsinki and was approvedbytheIstanbulMedicalFacultyEthicsCommittee.

Informed consent: Informed consent was obtained from all individualparticipantsincludedinthestudy.

Conﬂict of interest: The authors declare that they have no conﬂictsofinterest.

References

1.SorrellMF,BelongiaEA,CostaJ,GareenIF,GremJL,InadomiJM,etal.National InstitutesofHealthConsensusDevelopmentConferenceStatement: manage-mentofhepatitisB.AnnInternMed2009;150:104–10.

2.Fattovich G.Natural history andprognosis ofhepatitis B. SeminLiver Dis 2003;23:47–58.

3.McMahonBJ.ThenaturalhistoryofchronichepatitisBvirusinfection.Semin LiverDis2004;24:17–21.

4.FunkML,RosenbergDM,LokAS.World-wideepidemiologyofHBeAg-negative chronichepatitisBandassociatedprecoreandcorepromotervariants.JViral Hepat2002;9:52–61.

5.HadziyannisSJ,PapatheodoridisGV.HepatitisBeantigennegativechronic hepatitisB:naturalhistoryandtreatment.SeminLiverDis2006;26:130–41.

6.PapatheodoridisGV,ManolakopoulosS,ArchimandritisAJ.Currenttreatment indicationsandstrategiesinchronichepatitisBvirusinfection.WorldJ Gastro-enterol2008;14:6902–10.

7.TerraultNA,BzowejNH,ChangKM,HwangJP,JonasMM,MuradMH.AASLD guidelinesfortreatmentofchronichepatitisB.Hepatology2016;63:261–83.

8.EuropeanAssociationfortheStudyoftheLiver.EASLclinicalpractice guide-lines:managementofchronichepatitisBvirusinfection.JHepatol2012;57: 167–85.

9.KumarM,SarinSK,HissarS,PandeC,SakhujaP,SharmaBC,etal.Virologicand histologicfeaturesofchronichepatitisBvirus-infectedasymptomaticpatients withpersistentlynormalALT.Gastroenterology2008;134:1376–84.

10.LaiM,HyattBJ,NasserI,CurryM,AfdhalNH.Theclinicalsigniﬁcanceof persistentlynormalALTinchronichepatitisBinfection.JHepatol2007;47: 760–7.

11.IshakK,BaptistaA,BianchiL,CalleaF,DeGrooteJ,GudatF,etal.Histological gradingandstagingofchronichepatitis.JHepatol1995;22:696–9.

12.MontazeriG,RahbanM,MohamadnejadM,ZamaniF,HooshyarA,FazlolahiA, etal.LiverhistologyandHBV-DNAlevelsinchronicallyHBVinfectedpatients withpersistentlynormalalanineaminotransferase.ArchIranMed2010;13: 193–202.

13.LinCL,LiaoLY,LiuCJ,YuMW,ChenPJ,LaiMY,etal.HepatitisBviralfactorsin HBeAg-negativecarrierswithpersistentlynormalserumalanine aminotrans-feraselevels.Hepatology2007;45:1193–8.

14.ChenCJ,YangHI,SuJ,JenCL,YouSL,LuSN,etal.,REVEAL-HBVStudyGroup. Riskofhepatocellularcarcinomaacrossabiologicalgradientofserumhepatitis BvirusDNAlevel.JAMA2006;295:65–73.

15.YimHJ,LokAS.NaturalhistoryofchronichepatitisBvirusinfection:whatwe knewin1981andwhatweknowin2005.Hepatology2006;43:173–81.

16.IloejeUH,YangHI,SuJ,JenCL,YouSL,ChenCJ,RiskEvaluationofViralLoad ElevationandAssociatedLiverDisease/Cancer-InHBV(theREVEAL-HBV)Study Group.PredictingcirrhosisriskbasedonthelevelofcirculatinghepatitisBviral load.Gastroenterology2006;130:678–86.

17.PapatheodoridisGV,ManolakopoulosS,LiawYF,LokA.Follow-upand indica-tionsforliverbiopsyinHBeAg-negativechronichepatitisBvirusinfectionwith persistentlynormalALT:asystematicreview.JHepatol2012;57:196–202.

18.SarinSK,KumarM.ShouldchronicHBVinfectedpatientswithnormalALT treated:debate.HepatolInt2008;2:179–84.

19.LaiCL,RatziuV,YuenMF,PoynardT.ViralhepatitisB.Lancet2003;362:2089–94.

20.YuenMF,YuanHJ,HuiCK,WongDK,WongWM,ChanAO,etal.Alarge populationstudyofspontaneousHBeAgseroconversionandacute exacerba-tionofchronichepatitisBinfection:implicationsforantiviraltherapy.Gut 2003;52:416–9.

21.HsuYS,ChienRN,YehCT,SheenIS,ChiouHY,ChuCM,etal.Long-termoutcome afterspontaneousHBeAgseroconversioninpatientswithchronichepatitisB. Hepatology2002;35:1522–7.

22.SarinSK,KumarM,LauGK,AbbasZ,ChanHL,ChenCJ,etal.Asian-Paciﬁc clinicalpracticeguidelinesonthemanagementofhepatitisB:a2015update. HepatolInt2016;10:1–98.

23.YapaliS,TalaatN,LokAS.ManagementofhepatitisB:ourpracticeandhowit relatestotheguidelines.ClinGastroenterolHepatol2014;12:16–26.