ContentslistsavailableatScienceDirect
Vaccine
j o ur na l h o me pa g e : w w w . e l s e v i e r . c o m / l o c a te / v a c c i n e
The
epidemiology
of
hepatitis
B
virus
infection
in
HIV-infected
and
HIV-uninfected
pregnant
women
in
the
Western
Cape,
South
Africa
夽
M.I.
Andersson
a,∗,
T.G.
Maponga
a,
S.
Ijaz
b,
J.
Barnes
c,
G.B.
Theron
d,
S.A.
Meredith
e,
W.
Preiser
a,
R.S.
Tedder
b,faDivisionofMedicalVirology,UniversityofStellenbosch/NationalHealthLaboratoryService,Tygerberg,SouthAfrica
bBloodBorneVirusUnit,VirusReferenceDepartment,PublicHealthEngland,Colindale,London,UK
cDepartmentofCommunityHealth,UniversityofStellenbosch,SouthAfrica
dDepartmentofObstetricsandGynaecology,UniversityofStellenboschandTygerbergHospital,SouthAfrica
eDepartmentofPharmacology,UniversityofCapeTown,SouthAfrica
fDivisionofInfectionandImmunity,UniversityCollegeLondon,UK
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received13April2013
Receivedinrevisedform31July2013
Accepted9August2013
Available online 21 August 2013 Keywords: HBV HIV Antenatal Virusescape Sub-SaharanAfrica
a
b
s
t
r
a
c
t
Objectives:PersistenthepatitisBvirus(HBV)infectionisamajorcauseofmorbidityandmortalityin sub-SaharanAfrica.TheHIVepidemichasthepotentialtoaffectitsbiology.Immunisationprotocols establishedinthepre-HIVeraarebasedupondatashowingpredominantlyhorizontalinfant transmis-sion.ThisstudyaimedtodeterminewhetherHIVco-infectionwillchangetheepidemiologyofHBVboth byincreasinginfectivityandbyfavouringtheescapeofvirusesbearingphenotypicallyalteredHBsAg. Methods:Thisretrospectivecross-sectionalstudyusedantenatalsamplesfromthe2008Antenatal Sen-tinelHIVandSyphilisPrevalenceSurveyintheWesternCape,SouthAfrica.AllHIV-infectedwomen wereageandrace-matchedtoHIV-uninfectedwomen.Samplesweretestedforserologicalmarkersof HBVandHDVinfection.HBVviralload,consensussequencingandgenotypingwereperformed.Luminex technologywasusedtodetermineHBsAgphenotype.AllsamplesfromHIV-infectedwomenweretested fortracesofantiretroviraldrugsbymassspectrometry.
Results:ThisstudyshowedatrendtowardlossofimmunecontrolofHBVinHIV-infectedwomenwith3.4% ofsamplescontainingHBsAg,18.9%containedHBeAg.Incontrast,2.9%ofsamplesfromHIV-uninfected womencontainedHBsAgand17.1%oftheseHBeAg.ThemedianHBVloadintheHIV-infectedgroupwas 9.72×107IU/mlandintheHIV-uninfectedgroup1.19×106IU/ml.Genotypingshowed63/68samples
belongedtogenotypeAandtheremaindergenotypeD.Mutationsintheprecoreregionwerefoundin 35%and33%ofsamplesfromHIV-infectedandHIV-uninfectedrespectively.Althoughnomajorepitope ablationwasfound,markedvariationinHBsAgprofilesinHIV-infectedgroupwasdemonstrated.NoHDV infectionwasdetected.
Conclusion:HIV-HBVco-infectedwomenexhibitadegreeofimmuneescape.OneinsixHBV-infected pregnantwomen,irrespectiveofHIVstatusisHBeAgseropositive.HBVimmunizationofnewbornsin sub-SaharanAfricashouldbeimplemented.
© 2013 The Authors. Published by Elsevier Ltd. All rights reserved.
1. Introduction
Chronichepatitis Bvirus(HBV)infectionisa majorcauseof morbidityand mortalityin sub-SaharanAfrica(SSA)[1]despite
夽 Thisisanopen-accessarticledistributedunderthetermsoftheCreative
Com-monsAttribution-NonCommercial-NoDerivativeWorksLicense,whichpermits
non-commercialuse,distribution,andreproductioninanymedium,providedthe
originalauthorandsourcearecredited.
∗ Correspondingauthorat:DivisionofMedicalVirology,Departmentof
Pathol-ogy,StellenboschUniversity,FacultyofMedicineandHealthSciences,Francievan
ZijlAvenue,TygerbergCampus,WesternCapeProvince,SouthAfrica7505.
E-mailaddress:anderssonm@sun.ac.za(M.I.Andersson).
theavailabilityforthelast threedecadesofa safeandeffective vaccine.Variousparameters ofHBVinfectionarealtered bythe immunosuppression caused by human immunodeficiency virus (HIV)infection.Thosepersonsco-infectedhaveagreater preva-lenceofHBVeantigen(HBeAg)whichisamarkerforinfectivity[2], higherHBVDNAlevels[3],morefrequentHBVreactivation[4]anda higherprevalenceofoccultHBVinfection[5].Tworesultingaspects ofparticularconcernarehigherinfectivityfacilitatingHBVonward transmissionandthepotentialforgenerationofimmuneescape variantsduringreactivation.ThenatureofHBVinfectionin preg-nantwomenisimportantasitdeterminestheriskoftransmission ofHBVtotheirinfantsinutero[6]andduringparturition[7].Despite
0264-410X/$–seefrontmatter © 2013 The Authors. Published by Elsevier Ltd. All rights reserved.
thisimportantrelationshiptherearefewdatadescribingtheviral parametersofHBVinfectioninHIVco-infectedpregnantwomen.
HBVimmunisationprotocolsforthepreventionofinfectionin earlylifein manyAfricancountriesarebasedondatafromthe pre-HIVerawhichindicatedthatHBVinfectionoccurred predom-inantlyhorizontallybetweensiblingsandplay-matesratherthan verticallyfrommothertochild[8,9].Theseprotocolsare there-foreaimedatpreventinghorizontaltransmissionduringchildhood ratherthanpreventingmaternal–infanttransmissionatbirth.Asa consequencethefirstdoseofvaccineisgivenlaterinthepostnatal periodratherthanasclosetoparturitionaspossible,leavingthe neonatesusceptibletoinfectionfromexposureatbirth. Further-morethereisnoantenatalscreeningprogrammeforHBVinSouth Africa,norinmuchofSSA.
TheescapeofHBVfromapartiallyincapacitatedhostimmune systemisoftenassociated withthegenerationof viralvariants bearingchangesintheHBVsurfaceantigen(HBsAg).Transmission ofthesevariantstoinfantshasthepotentialtoestablishtheminthe generalpopulation.PhenotypicchangesinHBsAgmayreducethe efficacyofthecurrentrecombinanthepatitisBvaccinesbasedon theshortHBsAgtranscriptandpromotethepropagationofimmune escapevariants.
Thisstudyaimedtodeterminewhetherthere isevidenceto supportthehypothesisthatHIVco-infectionwillchangethe epi-demiologyofHBVbothbyincreasinginfectivityandbyfavouring theescapeofvirusesbearingphenotypicallyalteredHBsAg. 2. Materialsandmethods
Thisretrospective,cross-sectionalstudyusedantenatalsamples collectedforthe2008AntenatalSentinelHIVandSyphilis Preva-lenceSurvey[10]intheWesternCapeProvinceofSouthAfrica.All HIV-infectedwomenfromwhomsampleswereavailablewere ran-domlyage-andrace-matchedtoHIV-uninfectedwomen.Approval wasobtainedfromtheUniversityofStellenboschHealthResearch EthicsCommitteeandfromtheProvincialandNational Depart-mentsofHealth.
2.1. Recruitmentofpatients
Ofthe9355pregnant womenrecruited from antenatal clin-ics around the Western Cape for the 2008 survey, 1549 were seropositiveforanti-HIVandwerematchedforageandraceto 1550HIV-uninfectedwomen. Ofthe3099serum samplesfrom theseindividuals,tenwerehaemolysedanddiscardedas unsuit-ablefor testing. The remaining 3089samples comprising 1546 HIV-uninfectedand1543HIV-infectedsurveyparticipantswere includedin thestudy. Thedate of HIV infection,thedegree of immunosuppressionandCD4countwerenotavailableforthe HIV-infectedwomen.
2.2. HBVserologicalandmoleculartesting
Samples were tested for HBsAg using the Abbott AxSYM (AbbottDiagnostics,Chicago,IL).Allsamplesinitiallyreactivefor HBsAgwereconfirmedbyneutralisationusinghightitrehuman antibodytoHBsAg(anti-HBs)intheAbbottMurexHBsAgGE34/36 immunoassaykit(Murex Biotech,Kent,England).Samples con-firmedtocontainHBsAgweretestedforHBeAgandantibodiesto HBeAg(anti-HBe)using DiaSorinETI-EBK PLUSand ETI-AB-EBK PLUS immunoassay kits (DiaSorin, Saluggia, Italy). Testing for antibodytothehepatitis DeltaVirus(anti-HDV)wasperformed onallavailable HBsAg-positivesamplesusing ETI-AB-DELTAK-2 (DiaSorin).
HBVDNAquantification,consensussequencingand genotyp-ingwereperformedaspreviouslydescribed[11–13],onsamples
containingHBsAg.Sequenceanalysiswasperformedusing DNAS-TAR(version9).
Alignmentsof1800sequencesrepresentingalldescribed geno-types/subgenotypeswereobtainedfromeitherin-housegenerated sequencesor fromGenBank.Thesequences werecheckedboth visuallyandbyPositionSpecificScoringMatrixtoidentify intra-genotypicmotifs.Aminoacidalignmentsforwild-typeconsensus genotypespecificsequenceswerecreatedforthebasalcore pro-moter,precore,polymeraseandHBsAgregionsandusedtoidentify mutations.Genotypedeterminationwasbasedonclusteringwithin phylogenetictreesgeneratedusingthe1800representativeHBsAg nucleotidesequences.
2.3. HBsAgphenotyping
The impact of the observed amino acid changes on HBsAg antigenicitywasinvestigated viaexvivoHBsAg phenotypingas previouslydescribed[14].Briefly,Luminextechnologywasusedto measuretheinteractionofplasmaHBsAgwiththreemonoclonal antibodies(mAbs)directedagainst differentepitopesofthe“a” determinantofHBsAg.ThealterationofHBsAgantigenicitywas definedbychangesinthepatternofreactivityontheindividual solid-phasemonoclonalantibodiesexceeding2SD ofthemean whencompared toexpectedwildtype reactivity.Onlysamples fromwomeninfectedwithHBVgenotypeAwhoseplasmaHBsAg wasreactiveinthephenotypingassaysatalevelequivalenttoor greaterthan50IU/mlwereincluded.
2.4. Antiretroviraldrugresiduetesting
Atthe time of this study first linetherapy for HIV infected pregnantwomenwithaCD4countlessthan200cells/mm3 was
zidovudinepluslamivudineplusefavirenzornevirapine.Forthose womenwithaCD4countgreaterthan200cells/mm3,zidovudine
alonewasadministered.Forthose whofailedfirstlinetherapy, acombinationusinglopinavirwasprescribed.Tenofovirtherapy wasavailablefor thosepatientswho wereknowntobeHBsAg positive.Lamivudinewasusedasasurrogatefortenofovirasall womenontenofovirwerealsoadministeredlamivudine.Inorderto identifythosewomenoncombinationantiretroviraltherapy,and inparticularthoseonHBV-activetherapy,allsamplesfrom HIV-infectedwomen weretested fortracesoflamivudine,lopinavir, efavirenz and nevirapine by mass spectrometry, as described previously[15].
2.5. Statisticalanalysis
Categoricalvariablesweredescribedusingnumberand percent-ages.Quantitativevariableswereexpressedasameanandstandard deviationifnormallydistributedormedianandinterquartilerange ifnotnormallydistributed.Pearson’schi squaredtest wasused todetermineassociationbetweenindependentvariables,however wherenumbersinacellwerelessthanfive,aFisher’sexacttest wasusedinstead.Student’st-testwasusedtoexaminethe asso-ciationbetweenHIVstatusandHBVviralloadandageandHBeAg status.ANOVAwasusedtodetermineiftherewasany associa-tionbetweenHBsAgstatusandage,educationlevel,orparity.All reportedpvaluesarefortwo-tailedtests.Datawasanalysedusing Statistica,version11(StatSoftInc.,OK,USA).
3. Results 3.1. Demographics
Thebasicdemographicdataofthestudycohortareshownin Table1.
Table1
Basicdemographicdatafortheage-andsex-matchedcohortsofHIV-infected
(n=1543)andHIV-uninfected(n=1546)antenatalclinicattendeesfromthe
West-ernCape.
HIV-uninfected(n=1546) HIV-infected(n=1543)
Age Median26yrs Median26.8yrs
(IQR23,31) (IQR23,31) (Range12–44) (Range12–44) Black 1297(83.9%) 1323(85.7%) Coloured 203(13.1%) 179(11.6%) Other/unknown 46(3.0%) 41(2.7%) Education≤gd10a 998(64.6%) 886(57.4%)
Parity Median1 Median1
(IQR0,2) (IQR0,2)
(Range0–7) (Range0–7)
aSecondaryschool.
3.2. HBVandHDVserology
Ninety-sevensamplesfromthe3089studiedwereconfirmed to contain HBsAg; the prevalence in HIV-uninfected pregnant womenwas2.9%(44/1546)comparedwith3.4%(53/1543)in HIV-infectedpregnantwomen(Table2).Asignificantassociationwas foundbetweenHBsAgseropositivityandalowereducationalgrade (p=0.03),butnotbetweenHBsAgstatusandHIVstatus(p=0.404), age(p=0.52)orparity(p=0.27).AllHBsAg-positivesamplesof suf-ficientvolume(87)weretestedforanti-HDVandallwerenegative. 3.3. HBeAgprevalence
Only94sampleswereofsufficientvolumeforHBeAgtestingof which17containedHBeAgbutnoanti-HBe,7/41(17.1%)inthe HIV-uninfectedgroupcomparedwith10/53(18.9%)intheHIV-infected group(Table2).IntheHIV-uninfectedgroup,onesamplecontained neither HBeAg nor anti-HBe and one contained both antibody andantigen.IntheHIV-uninfectedgroup,theHBeAgseropositive mothersweresignificantlyyoungerthantheHBeAgseronegative mothers(21.7yrsSD=5.59vs.27.6yrsSD=4.74,p=0.016).In con-trast,therewasnosignificantdifferencein agebetweenHBeAg seropositiveandseronegativewomenintheHIV-infectedgroup. 3.4. HBVDNAlevels
In those samples from mothers who were seropositive for HBeAg, the HBV DNA was of similar range, but the median was different, in those seven not infected with HIV (120–260,000,000IU/ml,median1.19×106IU/ml)andthoseten
HIV-infected (190–560,000,000IU/ml, median 9.72×107IU/ml)
(Fig.1andTable2).
TheserumHBVDNAwasbelowthelevelofdetectionin14/75 whoseserumcontainedanti-HBe.TheproportioninwhomDNA
Table2
HBVserologyandHBVDNAlevelsinsamplesfromHIV-uninfectedandHIV-infected
antenatalclinicattendees.
HIV-uninfected HIV-infected p-Value HBsAgpositivea(%) 44/1546(2.9) 53/1543(3.4) 0.404
HBeAgpositive,AntiHBe
negative(%)
7c/41b(15.9) 10/53(18.9) 0.872
HBeAgnegative,Anti-HBe
positive(%)
32c/41b(72.7) 43/53(81.1) 0.912
HBVDNA>10IU/ml(%) 34/41(77.2) 44/53(83.0) 1.000
aConfirmedbyneutralisation.
bThreeHBsAg-positivesampleshadinsufficientvolumesleftforfurthertesting.
c OnesamplewasnegativeforbothHBeAgandanti-HBeandanotheronepositive
forbothHBeAgandanti-HBe.
couldbe quantifiedwassimilarin the HIVuninfected mothers (27/32) and the co-infectedmothers (34/43). Serum HBV DNA loadswasofsimilarmagnitudeinthe27HIV-uninfectedwomen (10–140,000IU/ml, median 108IU/ml) and the 34 HIV-infected women (10–1,600,000IU/ml, median 300IU/ml) (Fig. 1 and Table2).
3.5. HBVgenotyping,mutationalanalysisandphenotyping
HBV DNA sequence and genotype analysis was successfully undertakenon25and43samplesfromHIV-uninfectedand HIV-infected women, respectively. Phylogenetic analysis of HBsAg indicated63personstobeinfectedwithvirusbelongingto geno-typeA.TheremainingfivepersonsharbouredgenotypeDviruses; fourwereHIV-uninfectedwomenandoneHIV-infected.OneHBV sequencefromanHIV-uninfectedpatientborertV173Lalone,part oftheantiviralresistanceprofileforlamivudine;noneofthe42HBV sequencesfromtheHIV-infectedcohortcarrieddetectabledrug resistancemutations.
Mappingof amino acidchanges fromthewildtype consen-sussequencewasundertakenacrosstheHBsAgregioninorder toidentify potentialmutation ‘hotspots’ (Fig.2).Two samples fromHIV-uninfectedandthreefromHIV-infectedindividuals car-riedviruseswithanumberofcodingchangesbetweencodons120 and150.OnesamplefromanHIV-uninfectedandtwofrom HIV-infectedmotherscontainedvirusescarryingaprematurestopat codon182.Thesignificanceofchangesoutsidecodons120–150on HBsAgantigenicityremainunknown.
Epitopemappinganalysiswasundertakenin20HIV-uninfected and 33HIV-infectedmotherswithHBVgenotypeA virus infec-tionandHBsAglevelsequaltoorgreaterthan50IU/ml.Noantigen exhibitedmajorepitopeablation.However,thevariationsdefined bySDexceedancewereclearlymoremarkedinthesurface anti-genprofilesinsamplesfromHIVinfectedmothers(Fig.3).Three reactionsfallingoutside2SDofthemeanwereseeninsamples fromHIV-infectedmothers.Oneviruscarryingextensivelymutated HBsAgsequences(M1T,R79H,L98V,M133T,P135L,D144G,D194V) andtwolessscarredviruses(G44E,P203G/EandS204N)had reac-tivitiesononeormoresolidphasesoutside2SDofthemean.
MutationsintheprecorewhichimpactonHBeAgexpression werefoundin33%and35%ofsamplesfromHIV-uninfectedand HIV-infectedindividualsrespectively,withthelossofmethionine atcodon1beingthemostcommonlyobservedpathwayfor achiev-ing e-null virusesconstitutively unable to express HBeAg. The A1762T/G1764Abasal corepromoter variantwasobserved ata similarlevel(38%)insamplestestedfrombothHIV-infectedand uninfectedwomen.NosignificantassociationbetweenHBVDNA levelandthepresenceofthesebasalcorepromotermutationsor pre-coremutationswasnotedineitherpatientgroup.
3.6. Antiretroviraldrugresiduetesting
Antiretroviraldrugresiduesweredetectedinthreeof50 HIV-HBVco-infectedsamples(1.5%).Lamivudineandnevirapineintwo andlamivudineandlopinavirinathirdsample.Threesampleswere ofinsufficientvolumefortesting.
4. Discussion
Unlikedatafromwell-resourcedcountrieswhichshowa sig-nificantdifferenceinHBsAgprevalenceinHIV-infectedcompared touninfectedpersons[16],thisstudyconfirmsprevious antena-taldatafromAfricashowinglittledifferenceintheprevalenceof HBsAginHIV-infectedcomparedwithHIV-uninfectedwomen,as foundelsewhereinSouthAfrica,CôteD’Ivoire,Malawiand Tanza-nia[17–20].WehavefoundhoweversomeevidenceoflossofHBV
HB V V ira l lo ad (l o g 10 HB V DNA IU/m l) Median 25%-75% Range HBeAg: positive
HIV uninfected HIV infected
0 1 2 3 4 5 6 7 8 9 10 HBeAg: negative
HIV uninfected HIV infected
Fig.1. MedianplasmaHBVviralloads(expressedinIU/ml)accordingtoHIVstatusinHBeAgpositive(left)andinHBeAgnegative,anti-HBepositive(right)samplesdisplayed
asbox(interquartilerange(IQR))andwhisker(range)plots.
immunecontrolinHIVco-infectedwomenwhodemonstrateda trendtowardshigherHBVviralloads.
InthisantenatalpopulationfromtheWesternCape,asmanyas oneinsix(18.1%)ofHBV-infectedmothersareHBeAg-seropositive and anti-HBe negative, irrespective of HIV status. This is a higherproportionthanexpectedandcontrastswithOshitaniand colleagueswho founda differenceinHBeAg prevalencein HIV-infectedcompared withHIV-uninfected women(25%vs. 12.3%) [21].Ourfigureislowerthanthehighprevalence(5/12)ofHBeAg positivitythat wefoundpreviouslyin a cohortof HIV-HBV co-infectedpregnantwomenwithlowCD4countsattendingatertiary referralhospital[22].Thismayreflectdifferencesinimmune com-petencebetweenthetwostudycohorts.CD4datawasnotavailable forthecurrentstudytoconfirmthissuggestion.
AlthoughanassociationbetweenhighHBsAgprevalenceand loweducationwasfound,thereasonsforthisarenotclear.Low educationalattainment maybe a surrogatefor rural childhood whereeducationispoorerthaninurbanareas,andHBsAg preva-lencehigher.
OurdataalsoindicatethatamongHIV-uninfectedwomen,those whoseserumcontainsHBeAgaresignificantlyyoungerthanthose whoareHBeAgseronegative.Thisraisesthequestionofwhether HIVishavingapopulationbasedeffectonHBVepidemiologyalso intheHIV-uninfectedpopulation.
WHOguidelinesrecommendHBVvaccinetobeadministered within24hofdelivery[23],howeverthefirstHBVvaccinedosein sub-SaharanAfricaisgivenanytimebetweenfourandsixweeks postnatally,atimeappropriatetopreventhorizontaltransmission,
Fig.2.‘Hotspot’analysisofaminoacidchangesacrosstheHBsAg.Solidbarsaresequencesfrom25HIV-uninfectedwomenandthehatchedbarsfromHIV-infectedwomen.
Fig.3.EpitopevarianceofplasmaHBsAgexpressedasbox(middlequartiles)and
whiskers(range)in20 HIV-uninfectedand33HIV-infectedwomen.Reactivity
againsteachmonoclonalantibodysolidphaseisshownindependently(MAbP2D3
isagainstalinearepitopeandMAbsHB07andHB05areagainstdifferent
conforma-tionaldeterminantsinthesecondloopofHBsAg)andisexpressedasthepercentage
reactivityofeachmAbaspartofthetotalreactivityforthatsample.䊉denotesthose
sampleswhosereactivityfell2standarddeviationsoutsidetheexpectedrangefor
thatspecificMAb.
butprobably toolatetopreventreliablyperinataltransmission andtheevolutionofpersistentHBVinfection.Inlightofthis,our observationshaveimportantimplicationsforcurrenthepatitisB immunisationschedulesinsub-SaharanAfricawherethetiming ofneonatalvaccineispredicatedupondatageneratedbeforethe HIVpandemic.ThiswasatatimewhenHBeAgseropositiveand,by inference,highinfectivitymotherswererarelyencounteredinthe labourwards[24].Mostmotherswerethereforeconsideredtobe oflowinfectivity.HBVtransmissiontoinfantswasdeemedtooccur duringearlychildhood[8] ratherthan throughmaternal–infant transmissioninspiteofonebaselineHBVvaccinestudyreporting a10.3%HBeAgprevalenceinHBV-infectedfemales[25].Ourdata clearlyindicatethatcontrarytopreviousbeliefs,highHBV infec-tivitymothers,whetherdefinedbyHBeAgseropositivityorbyHBV DNA,arenotuncommonintheantenatalpopulationinSouthAfrica andindicatethatcurrentnationalvaccineschedulesmaybe inade-quatetopreventmother-to-childtransmissionofHBV.Klinglerand colleagueshaveproducedmodellingdatawhichsuggestaddinga birthdoseofHBVvaccinetothecurrentregimemaybecost effec-tive[26].Thismaynotbeasimpracticalasitfirstseemsgiventhe majorityofwomeninSSAgivebirthinhealthcarefacilities[27].
AlthoughHBV DNA load is themostimportant predictor of HBVtransmission frommother to child [7],HBeAg remains an importantmarkerforpotentialinfectivityinpregnantwomen, par-ticularlyinthosesituationswhereviralloadestimatesmaynotbe
available.Notonlyisitassociatedwithahighviralload,butitis alsothoughttocrosstheplacentatotolerizetheneonateleading toanincreasedriskofchronicHBVinfection[28].HBeAgtesting issimpletoimplement,notexpensiveandmaystillbeusefulin resourcepoorsettingstoidentifythosewomenatincreasedriskof transmittingHBVleadingtoviralpersistenceintheiroffspring.
Sequence analysis indicated the presence of codon changes acrosstheHBsAg,precoreand basalcorepromoterregions.The highprevalenceofe-nullvirusesbearingpre-core/coreandbasal corepromotermutationsmayindicateapropensityforearly pro-gressiontoseriousliverdiseaseinthefuture[29]andthosemaking healthcarepredictionsshouldbearthisinmind.NoHBVprimary resistancemutationswereseenprobablyreflectingthefactthat veryfewwomenwereonantiretroviraltherapy,withantiviraldrug residuesdetectedinonlythreeoutof50samples.
VariationintheHBsAgsequenceinpersistentHBVinfections wasnotableandseemsphenotypicallymorecommonintheHIV co-infectedhost.Therewasnoinstanceofmajorepitopelossascanbe foundintheclassicvaccineescapeG145Rvariant[14,30].However, thepossiblerescuebyparturitionofvirusesbearingalteredcoding andexpressionofHBsAgderivedfromHIV-infectedmotherswhose virusesarebeginningtoescapehostimmunesurveillanceclearly exists.Whetherdrugregimeswherelamivudineisthesole HBV-activeagentwhichresultsinasignificantriskofdevelopingHBV resistance[31]couldalsopotentiatetheemergenceofthevaccine escape-likeviruses[32,33]remainstobeseen.Inthisstudy sur-prisinglyfewwomenwereonantiretroviraltherapy,inparticular fewwereonagentsactiveagainstHBV,probablyreflectingpoor roll-outofantiretroviraltherapyatthetimeofthestudy. Never-thelessthepossibleinfantrescueofvirusesbearingalteredHBsAg lendsurgencytotheneedtoalterthetimingofHBVimmunisation andtheespousalofthemorepotentanti-HBVdrugsliketenofovir andentecavirforthetreatmentofthosewomenwhoareHBeAg seropositiveinthefuture.
Thereis one importantlimitation of this study which must beconsideredwhenextrapolatingfromthesedata.Thewomen recruitedtothisstudywereonlythosewhopresentedtoantenatal clinicsforbookingandasignificantpartofthepregnantpopulation wouldhavebeenexcludedfromthisstudythroughnotattending clinic.Thereasonsforpooraccesstohealthcarearemultipleand rangefromfinancialrestraints,lackofunderstandingofthe ben-efitsofearlybookingtopoorprovisionofhealthcareservicesin someareas.Thosepregnantwomenwhodonotaccessantenatal clinicsarethereforemorelikely tohavepooreraccesstohealth careandpotentiallytohavemoreadvancedHIVdisease.Inthese womenthelossofcontrolofHBVislikelytobemoreprofoundand theymayrepresentareservoirofmutatedHBVvariants.However, noimmunestatusdataintheHIV-infectedwomenwascollected norwasthereanywaytoestablishthetimingoftheirHIVinfection. ThisstudyindicatesthatHIV-infectedpregnantwomen,who mayremainrelativelywellandimmune-competent,nevertheless haveevidenceofHBVescapewhencomparedtoHIV-uninfected women.OneinsixHBV-infectedpregnantwomen,irrespectiveof HIVstatus,isHBeAgseropositiveandlikelytotransmitHBVtotheir offspring.ThesedatasupportthecallforashiftofHBV immunisa-tionclosertothetimeofbirthor,wherethisisdeemedimpractical considerationshouldbegiven totheadditionofa birthdoseof HBVvaccinetothecurrentschedule.Thedatahighlighttheneed forfurtherstudiestodeterminetheriskandvirologicaloutcomeof verticaltransmissionfrommothersinthispopulation.
Acknowledgements
WesternCapeGovernmentandNationalDepartmentofHealth forallowingaccesstoantenatalsurveysamples,ProfessorMartin Kiddforassistancewiththestatisticalanalysis.
Funding: Wellcome Trust (087859), NHLS K-funding (KNC110).
References
[1]HoffmannCJ,ThioCL.ClinicalimplicationsofHIVandhepatitisBco-infection inAsiaandAfrica.LancetInfectDis2007;7(6):402–9.
[2]GilsonRJ,HawkinsAE,BeechamMR,RossE,WaiteJ,BriggsM,etal.Interactions betweenHIVandhepatitisBvirusinhomosexualmen:effectsonthenatural historyofinfection.AIDS1997;11(5):597–606.
[3]ColinJF,Cazals-HatemD,LoriotMA,Martinot-PeignouxM,PharmBN,Auperin A,etal.Influenceofhumanimmunodeficiencyvirusinfectiononchronic hep-atitisBinhomosexualmen.Hepatology1999;29(4):1306–10.
[4]DiMartinoV,ThevenotT,ColinJF,BoyerN,MartinotM,DegosF,etal. Influ-enceofHIVinfectionontheresponsetointerferontherapyandthelong-term outcomeofchronichepatitisB.Gastroenterology2002;123(6):1812–22.
[5]PuotiM,TortiC,BrunoR,FiliceG,CarosiG.Naturalhistoryofchronichepatitis Binco-infectedpatients.JHepatol2006;44(1Suppl):S65–70.
[6]XuDZ,YanYP,ChoiBC,XuJQ,MenK,ZhangJX,etal.Riskfactorsandmechanism oftransplacentaltransmissionofhepatitisBvirus:acase-controlstudy.JMed Virol2002;67(1):20–6.
[7]BurkRD,HwangLY,HoGY,ShafritzDA,BeasleyRP.Outcomeofperinatal hepatitisBvirusexposureisdependentonmaternalvirusload.JInfectDis 1994;170(6):1418–23.
[8]BothaJF,RitchieMJ,DusheikoGM,MoutonHW,KewMC.HepatitisBvirus carrierstateinblackchildreninOvamboland:roleofperinatalandhorizontal infection.Lancet1984;1(8388):1210–2.
[9]WhittleHC,BradleyAK,McLauchlanK,AjdukiewiczAB,HowardCR, Zuc-kermanAJ,etal.HepatitisBvirusinfectionintwoGambianvillages.Lancet 1983;1(8335):1203–6.
[10]DepartmentofHealth.2008NationalAntenatalSentinelHIVandSyphilis
Prevalence Survey, South Africa; 2009 http://www.info.gov.za/view/
DownloadFileAction?id=1090[accessed08.07.13].
[11]GarsonJA,GrantPR,AyliffeU,FernsRB,TedderRS.Real-timePCR quantita-tionofhepatitisBvirusDNAusingautomatedsamplepreparationandmurine cytomegalovirusinternalcontrol.JVirolMethods2005;126(1–2):207–13.
[12]HealthProtectionAgency.RecommendedprotocolforHBVgenotypingand
for antiviral resistance analysis. Available from: http://www.hpa.org.uk/
ProductsServices/InfectiousDiseases/LaboratoriesAndReferenceFacilities/ VirusReferenceDepartment/BloodBorneVirusesUnit/[accessed29.11.12]. [13]TedderRS,RodgerAJ,FriesL,IjazS,ThurszM,RosenbergW,etal.Thediversity
andmanagementofchronichepatitisBvirusinfectionsintheUnitedKingdom: Awake-upcall.ClinInfectDis2013;56(7):951–60.
[14]IjazS,SzypulskaR,AndrewsN,TedderRS.Investigatingtheimpactofhepatitis Bsurfacegenepolymorphismonantigenicityusingex-vivophenotyping.JGen Virol2012;93(Pt.II):2473–9.
[15]CohenK,vanCutsemG,BoulleA,McIlleronH,GoemaereE,SmithPJ,etal.Effect ofrifampicin-basedantituberculartherapyonnevirapineplasma concentra-tionsinSouthAfricanadultswithHIV-associatedtuberculosis.JAntimicrob Chemother2008;61(2):389–93.
[16]HepatitisTCL.Bandhumanimmunodeficiencyviruscoinfection.Hepatology 2009;49(5Suppl):S138–45.
[17]BurnettRJ,NgobeniJM,FrancoisG,HoosenAA,Leroux-RoelsG,MeheusA,etal. IncreasedexposuretohepatitisBvirusinfectioninHIV-positiveSouthAfrican antenatalwomen.IntJSTDAIDS2007;18(3):152–6.
[18]RouetF,ChaixML,InwoleyA,MsellatiP,VihoI,CombeP,etal.HBVandHCV prevalenceandviraemiainHIV-positiveandHIV-negativepregnantwomen in Abidjan,Coted‘Ivoire:theANRS 1236study.JMed Virol 2004;74(1): 34–40.
[19]AhmedSD,CuevasLE,BrabinBJ,KazembeP,BroadheadR,VerhoeffFH,etal. SeroprevalenceofhepatitisBandCandHIVinMalawianpregnantwomen.J Infect1998;37(3):248–51.
[20]MsuyaSE,MbizvoEM,HussainA,SamNE,Stray-PedersenB.Seroprevalenceof hepatitisBandCvirusesamongwomenofchildbearingageinMoshiUrban, Tanzania.EastAfrMedJ2006;83(2):91–4.
[21]Oshitani H,KasoloFC,MpabalwaniM,MizutaK,LuoNP,SuzukiH,etal. PrevalenceofhepatitisBantigensinhumanimmunodeficiencyvirustype1 seropositiveandseronegativepregnantwomeninZambia.TransRSocTrop MedHyg1996;90(3):235–6.
[22]AnderssonMI,MapongaTG,IjazS,TheronG,PreiserW,TedderRS.HighHBV viralloadsinHIV-infectedpregnantwomenatatertiaryhospital,SouthAfrica. JAcquirImmuneDeficSyndr2012;60(4):e111–2.
[23]WorldHealthOrganisation.HepatitisBvaccines:WHOpositionpaper.Wkly EpidemiolRec2009;84(40):405–19.
[24]GuidozziF,SchoubBD,JohnsonS,SongE.ShouldpregnanturbansouthAfrican womenbescreenedforhepatitisB?SAfrMedJ1993;83(2):103–5.
[25]ProzeskyOW,SzmunessW,StevensCE,KewMC,HarleyEJ,HoylandJA,etal. BaselineepidemiologicalstudiesforahepatitisBvaccineintrialinKangwane. SAfrMedJ1983;64(23):891–3.
[26]KlinglerC,ThoumiAI,MrithinjayamV.S.Cost-effectivenessanalysisofan addi-tionalbirthdoseofHepatitisBvaccinetopreventperinataltransmissionina medicalsettinginMozambique.Vaccine2012;31(1):252–9.
[27]Shisana O, Rehle T, Simbayi L. South African National HIV prevalence,
incidence and communication survey; 2008. http://www.hsrcpress.ac.za/
product.php?productid=2279&freedownload=1[accessed21.07.13]. [28]HsuHY,ChangMH,HsiehKH,LeeCY,LinHH,HwangLH,etal.Cellularimmune
responsetoHBcAginmother-to-infanttransmissionofhepatitisBvirus. Hep-atology1992;15(5):770–6.
[29]LiuS,ZhangH,GuC,YinJ,HeY,XieJ,etal.AssociationbetweenhepatitisB virusmutationsandtheriskofhepatocellularcarcinoma:ameta-analysis.J NatlCancerInst2009;101(15):1066–82.
[30]Carman WF, Zanetti AR, Karayiannis P, Waters J, Manzillo G, Tanzi E, et al. Vaccine-induced escape mutant of hepatitis B virus. Lancet 1990;336(8711):325–9.
[31]Benhamou Y, Bochet M,ThibaultV,Di MartinoV,Caumes E,Bricaire F, etal.Long-termincidenceofhepatitisBvirusresistancetolamivudinein human immunodeficiencyvirus-infectedpatients.Hepatology1999;30(5): 1302–6.
[32]SheldonJ,RamosB,Garcia-SamaniegoJ,RiosP,BartholomeuszA,RomeroM, etal.SelectionofhepatitisBvirus(HBV)vaccineescapemutantsin HBV-infectedandHBV/HIV-coinfected patientsfailingantiretroviraldrugswith anti-HBVactivity.JAcquirImmuneDeficSyndr2007;46(3):279–82.
[33]ClementsCJ,CoghlanB,CreatiM,LocarniniS,TedderRS,TorresiJ.Global controlofhepatitisBvirus:doestreatment-inducedantigenicchangeaffect immunization?BullWorldHealthOrgan2010;88(1):66–73.
