Anti-Group B Streptococcus antibody in infants born to mothers with human immunodeficiency virus (HIV) infection

(1)

ContentslistsavailableatScienceDirect

Vaccine

jo u rn al h om ep a g e :w w w . e l s e v i e r . c o m / l o c a t e / v a c c i n e

Anti-Group

B

Streptococcus

antibody

in

infants

born

to

mothers

with

human

immunodeﬁciency

virus

(HIV)

infection

夽

Kirsty

Le

Doare

a,b,∗

_,

_Lauren

_Allen

c

_,

_Beate

_Kampmann

a,d

_,

_Paul

_Trafford

_Heath

b

_,

Stephen

Taylor

c

_,

_Anneke

_C.

_Hesseling

e

_,

_Andrew

_Gorringe

c

_,

_Christine

_Elizabeth

_Jones

a,b,e

a_Wellcome_{Trust/Imperial}_Centre_for_Global_Health_{Research/Department}_of_Academic_Paediatrics,_Imperial_College_London,_Norfolk_Place,_London_W2_1NY,

UK

b_Paediatric_Infectious_Diseases_Research_Group,_St_George’s,_University_of_London,_Cranmer_Terrace,_London_SW17_0RE,_UK c_Public_Health_England,_Porton_Down,_Salisbury_SP4_0JG,_UK

d_Medical_Research_Council,_Atlantic_Road,_Fajara,_Gambia

e_Desmond_Tutu_TB_Center,_Department_of_Pediatrics_and_Child_Health,_Stellenbosch_University,_Cape_Town,_South_Africa

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received26August2014 Receivedinrevisedform 10December2014 Accepted12December2014 Availableonline24December2014 Keywords: Antibody GroupBStreptococcus HIV-exposed-uninfectedinfants HIV Immunity

a

b

s

t

r

a

c

t

Background:HIV-exposeduninfectedinfantshaveincreasedinfectionriskandmortalitycomparedto HIV-unexposedinfants.HIV-exposedinfantsmaybeatincreasedriskofinvasiveGBSdiseasedueto reducedmaternalantibodyagainstGBS.

Methods:WequantifiedantibodiesthatbindtothesurfaceofwholeGroupBStreptococcus(GBS)of serotypesIa,Ib,II,IIIandVusingnovelflowcytometryassaysinSouthAfricanHIV-infectedand non-infectedmothersandtheiruninfectedinfants.Antibody-mediatedcomplementC3b/iC3bdepositiononto GBSoftheseserotypeswasalsoquantifiedbyanovelflowcytometryassay.

Results:Geometricmeanconcentration(GMC)ofbothsurface-bindinganti-GBSantibodyand antibody-mediatedcomplementdepositionontoGBSwerereducedinHIV-infectedwomen(n=46)comparedto HIV-uninfectedwomen(n=58)forST1a(surface-binding:19.3vs29.3;p=0.003;complement depo-sition:2.9vs5.3SU/mL;p=0.003),STIb(24.9vs47.6;p=0.003;2.6vs4.9SU/mL;p=0.003),STII(19.8 vs50.0;p=0.001;3.1vs6.2SU/mL;p=0.001),STIII(27.8vs60.1;p=0.001;2.8vs5.3SU/mL;p=0.001) andSTV(121.9vs185.6SU/mL;p<0.001)andintheirinfantsforSTIa(complementdeposition9.4vs 27.0SU/mL;p=0.02),STIb(13.4vs24.5SU/mL;p=0.02),STII(14.6vs42.7SU/mL;p=0.03),STIII(26.6vs 62.7SU/mL;p=0.03)andSTV(90.4vs165.8SU/mL;p=0.04).Mediantransplacentaltransferofantibody

fromHIV-infectedwomentotheirinfantswasreducedcomparedtoHIV-uninfectedwomenforGBS

serotypesII(0.42[IQR0.22–0.59]vs1.0SU/mL[0.42–1.66];p<0.001),III(0.54[0.31–1.03]vs0.95SU/mL [0.42–3.05],p=0.05)andV(0.51[0.28–0.79]vs0.75SU/mL[0.26–2.9],p=0.04).Thedifferencesbetween infantsremainedsigniﬁcantat16weeksofage.

Conclusions:MaternalHIVinfectionwasassociatedwithloweranti-GBSsurfacebindingantibody con-centrationandantibody-mediatedC3b/iC3bdepositionontoGBSbacteriaofserotypesIa,Ib,II,IIIandV. ThismayrendertheseinfantsmoresusceptibletoearlyandlateonsetGBSdisease.

1. Introduction

Theincreasingnumbers ofHIV-exposed infantswhoremain

uninfectedistestamenttothesuccessofpreventionof

mother-to-childtransmissionprogramsinresource-poorsettingsintheface

夽 PreliminarydatawaspresentedattheEuropeanSocietyofInfectiousDiseases meetingMay2014,Dublin,Ireland.

∗ Correspondingauthorat:WellcomeTrust/ImperialCentreforGlobalHealth Research,DepartmentofPaediatrics,ImperialCollegeLondon,NorfolkPlace,London W21PG,UK.Tel.:+442075895111;fax:+442075895111.

E-mailaddress:k.mehring-le-doare@imperial.ac.uk(K.LeDoare).

ofa highdiseaseburden[1].Nevertheless,thisgroupofinfants

appearstosufferfromincreasedratesoflowerrespiratory tract

infectionandmeningitiscomparedtoHIV-unexposedinfantsand

upto4-foldhighermortalityintheﬁrstyearoflife[2–4].

Strepto-coccusagalactiae(groupBstreptococcus,GBS)isaleadingcauseof

neonatalpneumonia,sepsisandmeningitisinmanycountriesand

ﬁveserotypes(Ia,Ib,II,IIIandV)accountforthemajorityofdisease

[5].PublisheddatashowthatnotonlyisGBScarriageincreased

inHIV-infectedpregnantwomencomparedwithHIV-uninfected

mothers[6],butthatHIV-exposed,uninfectedinfantsalsoappear

tohaveanincreasedriskoflate-onsetGBSdiseasecomparedto

HIV-unexposedinfants[7].

http://dx.doi.org/10.1016/j.vaccine.2014.12.025

(2)

PreventionofGBSinfectionininfantsandadultsthrough

immu-nization is a theoretically attainable goal. Maternal transfer of

antibodiesis thought toprevent newbornGBS disease [8] and

GBSvaccinesofferthepotentialtopreventdiseaseinlow-income

settingswhereprenatalscreeningandintrapartumantibioticsare

generallynot feasible. Anumber of studies have demonstrated

anassociationbetweentheriskofdeveloping invasiveGBS

dis-easein thenewbornand maternalanti-capsularantibodylevels

[9–12]with opsonophagocytosis the likely effector mechanism

[13–15]. However, maternal HIV status can inﬂuence the

efﬁ-ciencyof transplacentalantibody transfer, resulting in reduced

maternally-derivedspeciﬁcantibodyconcentrationintheinfant,

even where theinfantis subsequently HIV uninfected [16–19].

Dataregardingthepersistenceofanti-GBSantibodiesin

popula-tionswithhighHIV-prevalenceandtheirfunctioninvitrowould

provideinsightintotheirprotectivepotentialindifferent

popula-tions.Thusourobjectivewastodeterminetheassociationbetween

maternalHIVinfectionstatusandtotalIgGbindingtothesurfaceof

GBSbacteriaofserotypesIa,Ib,II,IIandVandantibody-mediated

depositionofC3b/iC3bontothesurfaceoftheseGBSstrainsasa

potentialsurrogateforopsonophagocytosisinseraobtainedfrom

mothersandinfants.

2. Methods

2.1. Studysetting,eligibilityandstudymeasures

Sampleswerecollectedfrommothersandinfantsenrolledin

a cohortstudy investigating theinﬂuenceof maternal HIV and

mycobacterialsensitizationoninfantimmuneresponsestoBCG

vaccinationcarriedoutbetween2009and2011[20].Inbrief,109

pregnant women presenting to a Community Health Center in

Khayelitsha,WesternCapeProvince,SouthAfricawhodelivered

ahealthy infantover 37weeks gestation,weighingover 2.5kg,

knewthe result of their HIV test and were willing to provide

writteninformedconsentforthemselvesandtheirinfantswere

recruitedtotheoriginalstudy.Thesubsetrepresentedall

moth-ers/infantswithavailable samples.Avenous blood samplewas

collected from the mother and infant within 24h of delivery.

Allinfants had a furthervenous blood sample collectedat 16

weeks.HIV-exposedinfantshadanHIVpolymerasechainreaction

(AmplicorHIV-aDNAkit,version1.5;RocheMolecularSystems

Inc.,Branchburg,NJ)performedatages4and16weeks.Thestudy

wasapprovedbytheUniversitiesofCapeTown(382/2008)and

Stellenbosch(N08/10/278),SouthAfrica,andtheNationalHealth

ServiceResearchEthicsCommittee,England(07/H0720/178).

In2009,theHIVprevalenceamongwomenattendingantenatal

clinicsintheareawas31%[21].Duringthestudyperiod,the

Pre-ventionofMothertoChildTransmissionprogramconsistedofdual

therapyformothersandinfants,startingwiththeadministration

ofzidovudineat28ormoreweeks’gestation,thenzidovudinefor1

monthtotheinfantandasingledoseofnevirapinetobothmother

andinfant.Motherswereeligibleforhighlyactiveantiretroviral

treatmentiftheirCD4countwaslessthan200cells/␮L.

Exclu-siveinfant feedingoptionswere encouragedand motherswere

providedwithfreeformulafor6monthsiftheychoseexclusive

formulafeeding.

2.2. Laboratoryassays

Thereproducibilityoftheassayswasassessedbetweenthree

differentdaysandbetweenthreedifferentoperatorswithapanel

of18sera.Theassays hadacoefﬁcientofvariationof lessthan

30%.Controlsusedwererabbitpolyclonalsera(SSI)raisedagainst

serotypesIa,Ib,II,III&V,andnormalhumandonorsera.

2.2.1. GBSisolatesandgrowthconditions

Group B Streptococcus isolates used in this study were

H092040676 (Serotype Ia), H090820125 (Serotype Ib),

H090320548 (Serotype II), H092120162 (Serotype III) and

H091780506 (Serotype V), which were kindly provided by

ProfessorAndroullaEfstratiou,PublicHealthEngland, Colindale.

Strainsaresubsequentlyreferredtobytheirserotype.

GBSisolatesusedintheantibodysurfacebindingand

comple-mentdepositionassaysweregrowninToddHewittbrothat37◦C

withshaking(200rpm).OnceOD600nm 1.0wasreached,cultures

werecentrifugedat3060×gfor5mintopelletthebacteria.The

pelletwasre-suspendedinthesamevolumeofphosphatebuffered

saline(PBS,pH7.4,SevernBiotech,UK)containing2%

formalde-hydeand incubatedatroomtemperaturefor1h. Bacteriawere

thenwashedinPBSbythreeroundsofcentrifugationat3060×g

for5minandresuspended.Theﬁnalcellpelletwasresuspendedin

1mLPBS,whichwasstoredat4◦Cbeforeuse.

2.2.2. Samplepreparationforantibodysurfacebindingassay

Deposition of anti-GBS antibody onto the surface of whole

GBSbacteriawasmeasuredonformaldehyde-ﬁxedGBSusinga

ﬂowcytometricassayperformedin96-wellmicrotitrationplates.

Brieﬂy,2␮Lofeachtestserumwasaddedto198␮LofserotypeIa,

Ib,II,IIIorVGBSbacteriaat5.14×107_CFU/mL_in_blocking_buffer

(1%BSAinPBS).Thiswasincubatedat25◦Cfor30minwithshaking

(900rpm),thenpelleted.Supernatantwasremovedandthe

pel-letwashedtwicewith200␮Lofblockingbuffer.AlexaFluor®488

GoatAnti-HumanIgG(H&L)(LifeTechnologies)(1:500)inblocking

bufferwasaddedandsamplesincubatedfor20minat4◦C,before

beingwashedtwicemorewithblockingbuffer.

2.2.3. Samplepreparationforantibody-mediatedcomplement

C3b/iC3bdepositionassay

Antibody-mediatedC3b/iC3bdepositiononthesurfaceofwhole

GBSbacteriawasmeasuredonthesurfaceofformaldehyde-ﬁxed

GBSusingaﬂowcytometricassayperformedin96-well

microtitra-tionplates.Brieﬂy,35␮LserotypeIa,Ib,II,IIIandVGBSbacteriaat

5.14×107_CFU/mL_in_blocking_buffer_(1%_BSA_in_PBS)_were_added

to10␮LIgG-depletedhumancomplement[22]and5␮Lofeach

testserum.Plateswereincubatedfor7.5minat37◦Cwith

shak-ing(900rpm)then pelleted. Supernatantwasremoved andthe

pelletwashed oncewith 200␮Lblocking buffer. Sampleswere

resuspendedin200␮Lblockingbuffercontaining1:500sheep

anti-humanC3cFITC(Abcam)andincubatedasforthesurfacebinding

assay.

2.2.4. Sampleacquisition

Assays were analyzed using a Beckman Coulter Cyan ﬂow

cytometerequippedwithaCytek96-wellmicrotitreplatereader.

Protocolswereinitiallyset-uptoanalyzeproﬁlesofbacterialevents

usingbacteria-onlycontrolsamplesandthebacteriaidentiﬁedon

thecytometerbytheforwardscatter(FS),measuringthesizeofthe

cell,andsidescatter(SS),thegranularityandinternalstructural

complexity.Ananalysis‘gate’wasdrawnaroundthepopulation

ofinterest(singleanddiplococci)andarelevanthistogramplot

createdtoanalyzetheﬂuorescence ofthe bacterialpopulation.

Foreachsample,10,000individualeventswereanalyzedfor

ﬂu-orescenceandahorizontalgatewasdrawntoinclude10%ofthe

‘noantibody’control population. A ﬂuorescence index(FI) was

calculatedforeachsample,whichinvolvedthemultiplicationof

thepercentageofbacteriainthehorizontalgate(%-gated),bythe

meanﬂuorescenceofthatpopulation(X-mean).Theﬁnalresult

for each test wasexpressedas theaverageFI of duplicate test

samplesminustheaverageFIofthebacteriaandconjugate-only

control.Astandardunit(SU)measurementforeachserum

(3)

obtainedwiththepositivecontrolserumforeachserotypewhich

wasgiven an arbitraryvalue of 1000 (kindgift fromProfessor

CarolBaker,BaylorCollegeofMedicine,Texas)togivearesultin

SU/mL.

2.3. Datamanagementandstatisticalanalysis

Samplesizewasdeterminedforthecohortstudy;thissub-study

had80%power(p=0.05)toinvestigatedifferencesbetween

anti-bodyconcentrationsinHIV-exposedandHIV-unexposedinfants

of atleast 30%with thepre-speciﬁedhypothesis that the

con-centrationswouldbelowerinHIV-exposedinfants.Thisassayis

currentlybeingformallycomparedwithELISAand

opsonophago-cytosisuptakeassaysinanotherstudy.

Statistical analyzeswere completed usingSTATA version 12

(StataCorp 2013, La Jolla, CA) and GraphPad Prism version 6.0

(GraphPadSoftwareInc.,LaJolla,CA).Duetotheskewed

distribu-tionoftheanti-GBSantibodyconcentrations,logtransformations

wererequiredtoconformtoregressionassumptions.Apairedt-test

wasusedtodetermineanydifferenceinanti-GBS-antibody

con-centrationstoeachGBSserotypebetweenbirthandat16weeks.

Comparisonofinfantanti-GBSantibodyconcentrationsby

mater-nalHIVstatuswascalculatedusingtheWilcoxonranksumtest.

Correlationofinfantanti-GBSantibodyconcentrationtoeachGBS

serotypeatbirthandat16weekswasassessedusingSpearman

correlation.Multiplelinearregressionwasusedtoascertainthe

relationshipbetweeninfant anti-GBSantibodyconcentrationto

eachGBSserotypeandthecovariatesofmaternalage,infantsex

andmaternalHIV.Inaddition,multiplelinearregressionwasused

toexploretherelationshipbetweenmaternalanti-GBSantibody

concentrationtoeach serotypeandthecovariatesmaternalage,

maternalHIVstatus. Placentaltransferwasdeﬁnedastheratio

ofinfant-to-motheranti-GBSIgGconcentrationatbirth.Missing

datawereexcludedfromanalysis.Declineinantibody

concentra-tionbetweenbirth and 16 weekswascalculated using ratioof

means.

3. Results

3.1. Participantcharacteristics

Sampleswereanalyzedfrom104women(46(44%)HIV-infected

and58(56%)HIV-uninfectedwomen)andfrom100infantsbornto

thisgroupofwomen(46(46%) HIV-exposedand54(54%)

HIV-unexposedinfants). Onemother–infantpairwasexcludedfrom

theﬁnalanalysisbecauseHIV-infectionintheinfantwasdetected

at4weeksofageandtheinfantwasreferredforrapidinitiation

ofantiretroviraltreatment.AllotherinfantsborntoHIV-infected

motherswereHIV-uninfected.Allinfantswerebornafter37weeks

gestation.Follow-upsamplesat16weekswereavailablefor93

infants(93%;38HIV-exposedand55HIV-unexposed)atamean

postnatalageof16.4weeks(SD=1.7).

AllHIV-infectedwomenchoseexclusiveformulareplacement

feeding. The median (IQR) CD4 count among the HIV-infected

women was411.5(293.9–604.0)cells/␮Land the median (IQR)

viralloadwas800(357–6000)copies/mLSevenwomenhadCD4

counts of less than 200 cells/␮L; three of these were taking

highlyactiveantiretroviraltreatmentatenrollmentandfourwere

referredtocommencehighlyactiveantiretroviraltreatment

fol-lowingdelivery.Furthercharacteristicsofthestudycohorthave

beendescribed previously [20]. Atthe time of the study,local

GBSdiseaseincidencein CapeTownwasreportedat 0.67/1000

livebirthsina populationwith24%oftheseinfantsbeing

HIV-uninfected,borntoHIV-infectedmothers[23].

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 5 50 500 5000 Percentile SU/mL

STIa STIb STII STIII STV

Fig.1.Reversecumulativedistributioncurveofmaternalanti-GBSIgGsurface bind-ingbyserotypeconcentrations(SU/mL).Curvedemonstratesthepercentageofthe totalpopulationwitheachantibodyconcentrationforeachoftheﬁveGBSserotypes. ST,serotype;SU/mL,standardunits/mL

3.2. Maternalanti-GBS-antibodytoGBSserotypesIa,Ib,II,IIIand

V

Allmaternalserahadmeasurablegeometricmean

concentra-tions(GMC)ofantibodytoatleastoneserotype.Thedistribution

ofmothers’antibodiesthatboundtothesurfaceofGBSserotypes

Ia,Ib,II,IIIandVisshowninFig.1.Thepredominantsurface

bind-inganti-GBS-antibodywasagainstserotypeV,followedbyII,III,Ib

andIa(Fig.1).

WomenwithHIVinfectionhadsigniﬁcantlylowerGMC

anti-bodyconcentrationstoallﬁveserotypesatdeliverycomparedto

HIV-uninfectedwomen(Fig.2andTable1).Therewereno

asso-ciationsbetweenCD4countandviralloadonanti-GBSantibody

concentrationbyserotypeinHIV-infectedwomen.

The concentrations of antibody-mediated complement

C3b/iC3b deposition by GBS serotype are shown in Table 2.

TheGMCofantibody-mediatedC3b/iC3bdepositionwas

signiﬁ-cantlygreaterinHIV-uninfectedthanHIV-infectedwomenforall

serotypes. The GMC of antibody-mediated C3b/iC3b deposition

rangedfrom4.9SU/mL(serotype III)to6.17SU/mL(serotype V)

in HIV-uninfected women and from2.6SU/mL(serotype Ib) to

3.4SU/mL(serotypeV)inHIV-infectedwomen.

3.3. AssociationofmaternalHIVinfectionwithplacentaltransfer

ofanti-GBSantibody

Therewasapositivecorrelationbetweenpairedmaternaland

infantsurfacebindingantibodyconcentrationsforallserotypes(Ia,

r2₌_0.63,_p_<_0.001;_Ib,_r2₌_0.71,_p_<_0.001;_II,_r2₌_0.17,_p_<_0.001;_III,

r2₌_0.35,_p_<_0.001;_V,_r2₌_0.16,_p_<_0.001)._HIV-infected_women_had

signiﬁcantreductionsinplacentaltransferofanti-GBSserotypesII

andVcomparedwithHIV-uninfectedwomen(Table3).Therewas

noassociationbetweenCD4countorviralloadontransplacental

transferratios.

3.4. Infantanti-GBS-antibodybyserotypeatbirthtoGBS

serotypesIa,Ib,II,IIIandV

HIV-exposeduninfectedinfantshadsigniﬁcantlylowerGMCs

ofanti-GBSsurfacebindingantibodiestoallserotypesthan

unex-posedinfants(Fig.3andTable1).Inamultipleregressionmodel

maternalHIV-statusremainedsigniﬁcant(p<0.05).

However,therewerenosigniﬁcantassociationsfoundbetween

infant anti-GBS-antibody concentrations to each serotype and

(4)

HIV+ HI V-0 50 100 150 200 SU /m L STIb p=0.03 HIV+ HI V-0 100 200 300 400 SU /m L STII p=0.03 HIV+ HI V-0 100 200 300 400 SU /m L STIII p=0.04 HIV+ HI 0 500 1000 SU /m L STV p=0.04 HIV+ HIV -0 50 100 150 200 SU /m L STIa p=0.02

Fig.2. Maternalanti-GBSsurfacebindingIgGtoGBSserotypesIa,Ib,II,IIIandVconcentrationsatdelivery(SU/mL).STindicatesGBSserotype.HIVindicateshuman immunodeﬁciencyvirus.HorizontallinesindicateGMCresponse.

Table1

Geometricmeansurfacebindingantibodyconcentrations(95%conﬁdenceintervals)inHIV-infectedandHIV-uninfectedmothersandtheirinfantsatbirthandat16weeks ofage.

STIa STIb STII STIII STV

HIV+mother 19.3(12.7–29.3) 24.9(19.6–31.6) 21.8(13.4–35.7) 27.8(15.9–48.6) 121.9(88.0–169.0)

HIV−mother 37.0(24.8–55.2) 47.6(33.3–68.1) 50.0(28.7–87.0) 60.1(36.0–100.5) 185.6(126.4–272.4)

HIV-exposedinfants(birth) 9.4(6.1–14.5) 13.4(9.6–18.7) 14.6(8.1–26.3) 26.6(17.8–39.8) 90.4(61.16–133.6)

HIV-unexposedinfants(birth) 27.0(16.5–44.3) 24.5(16.4–36.5) 42.7(23.4–77.8) 62.7(38.1–103.1) 165.8(112.0–245.4)

HIV-exposedinfants(16w) 7.0(3.8–13.0) 8.7(6.3–12.1) 17.6(8.5–36.3) 48.0(28.3–81.3) 68.54(38.0–123.6)

HIV-unexposedinfants(16w) 21.6(13.6–34.2) 14.9(10.6–21.1) 68.3(38.8–120.3) 97.4(63.0–150.7) 180.2(106.3–305.4)

STIa,GBSserotypeIa;STIb,GBSserotypeIb;STII,GBSserotypeSTII;STIII,GBSserotypeIII;STV,GBSserotypeV;HIV+,HIVinfected;HIV−,HIVuninfected;16w,16weeksof age.

Table2

Geometricmeanconcentrationofantibody-mediatedC3b/iC3bdeposition(95%conﬁdenceintervals)onGBSserotypesIa,Ib,II,IIIandVinHIV-infectedandHIV-uninfected mothersandtheirinfantsatbirth.

STIa pvalue STIb pvalue STII pvalue STIII pvalue STV pvalue

HIV+mother 2.9(2.3–3.6) 2.6(1.9–3.2) 3.1(2.4–3.9) 2.8(2.2–3.6) 3.4(2.8–4.2)

HIV−mother 5.3(4.6–6.2) 0.003 4.9(4.2–5.4) 0.003 6.2(5.6–6.8) 0.001 5.3(4.5–6.2) 0.001 6.2(4.9–6.9) <0.0001

HIV-unexposedinfants 6.1(5.9–6.3) 5.9(4.6–6.2) 6.9(5.6–7.2) 6.6(6.4–6.7) 6.3(5.8–6.9)

HIV-exposedinfants 3.8(3.0–4.8) 0.02 3.1(3.0–4.1) 0.02 4.2(3.9–4.9) 0.02 3.9(3.1–5.0) 0.03 4.2(3.3–5.4) 0.04

(5)

Table3

InﬂuenceofmaternalHIV-infectionontransplacentaltransferofanti-GBSsurfacebindingantibodytoGBSserotypesIa,Ib,II,IIIandV.

GBSserotype HIV-infectedmothers:infants(IQR) HIV-uninfectedmothers:infants(IQR) %ReductioninHIV-infected pvalue

STIa 0.66(0.34–0.99) 0.6(0.39–0.96) 0% 0.86

STIb 0.48(0.17–0.88) 0.52(0.32–0.78) 8% 0.48

STII 0.42(0.22–0.59) 1.0(0.42–1.66) 58% <0.001

STIII 0.54(0.31–1.03) 0.95(0.4–3.05) 43% 0.05

STV 0.51(0.28–0.79) 0.75(0.26–2.9) 32% 0.04

STIa,GBSserotypeIa;STIb,GBSserotypeIb;STII,GBSserotypeSTII;STIII,GBSserotypeIII;STV,GBSserotypeV;IQR,interquartilerange;HIV+,HIVinfected,HIV−,HIV uninfected;16w,16weeksofage.

SigniﬁcantlylowerGMCofantibody-mediatedC3b/iC3b

depo-sition was noted amongst HIV-exposed infants compared to

HIV-unexposedinfantsforallserotypes(Table2).Notably,theGMC

ofantibody-mediatedC3b/iC3bdepositionwashigherininfants

thanintheirmothersforallserotypes.

3.5. Infantanti-GBSantibodyat16weeksofage

Anti-GBSsurfacebindingantibodyGMCsremainedsigniﬁcantly

lowerat16weeksinHIV-exposeduninfectedinfantscomparedto

unexposedinfantsforallserotypes(Table1).HIV-exposedinfants

had signiﬁcantly slowermedian rateof antibody decline

com-paredtoHIV-unexposedinfantsat16weekscomparedtobirthto

STII(ratioofdecline1.03[IQR−0.35to4.6]vs0.21SU/mL[−0.62

to0.54],p=0.04)and STIII(1.53[−0.21to5.36]vs−0.09SU/mL

(−0.60to0.93), p=0.03) but notto ST1a (−0.12 [IQR−0.64 to

0.13] vs −0.04 [−0.39 to 0.45]), ST1b (−0.52 [−0.65 to 0.46] vs −0.36[−0.18to0.74]) orSTV (−0.40 [−0.61to 2.9]vs 0.14 [−0.52to2.7]). HIV+ HI V-0 50 100 150 200 SU /m L STIa p=0.03 HIV+ _HIV- 0 100 200 300 SU /m L p=0.03 STII HIV+ _HI V-0 50 100 150 200 SU /m L STIb p=0.02 HIV+ HIV -0 100 200 300 SU /m L p=0.005 STIII HIV+ _HI V-0 100 200 300 400 500 SU /m L p=0.00STV3

Fig.3.Infantanti-GBSIgGserotypespeciﬁcantibodyconcentrations(SU/mL).STindicatesGBSserotype.HIVindicateshumanimmunodeﬁciencyvirus.Horizontallines indicateGMC.

(6)

4. Discussion

Ourﬁndingsdemonstratethattotalsurfacebindingand

func-tional maternal antibody concentration to each of the GBS

serotypes(Ia,Ib,II,IIIandV)isloweratdeliveryinHIV-infected

womencomparedtotheirHIV-uninfectedpeers.Further,forthree

ofthe ﬁveserotypes (II, III, V) thetransplacentaltransfer ratio

wassigniﬁcantlylowerfromHIV-infectedwomentotheir

unin-fectedinfants.ItfollowsthereforethatinfantsborntoHIV-infected

womenhadloweranti-GBSsurfacebindingantibody

concentra-tionsatbirthforallserotypes;thisdifferencewasstillevidentat

16weeksofage.TheimplicationofourﬁndingsisthatHIV-exposed

infantsmaybemorevulnerabletoearlyandlateonsetGBSdisease

thaninfantsborntoHIV-uninfectedmothers.

RectovaginalcolonizationwithGBSismorefrequentin

HIV-infected women compared to HIV-uninfected women [6,24].

Maternalcolonizationisthemajorriskfactorforinfant

acquisi-tionatbirthand forearlyonsetGBSdisease.Thusinfantsborn

toHIV infectedmothersmayhavebothincreasedsusceptibility

and increasedexposure. It is generally understood that

mater-nalcolonizationstimulatesmaternalimmunity;thustheﬁnding

oflower maternalantibodyconcentrationsdespite higher

colo-nizationratessuggestsimpairmentofanti-GBS-speciﬁcantibody

production.Thisisalsoseenwithpolysaccharide-speciﬁcantibody

productioninHIV-infectedadults,suchasinpneumococcal

dis-ease[25].ThefactthatHIV-infectedwomenhavehigherratesof

GBScolonizationthanHIV-uninfectedwomenmaymeanthat

anti-bodyproductionsecondarytocolonizationwithGBSisdeﬁcientin

thecontextofmaternalHIV-infectionincreasingtheriskofinvasive

GBSdiseaseinHIV-exposedinfants.

Itisdifﬁculttocorrelateouranti-GBSantibodyresultswithwhat

isknownregardingmaternalcolonizationandinfantGBSdiseasein

thisSouthAfricanpopulation.However,thehighestconcentrations

ofantibodiesinwomenatdeliveryweretoSTVandIIIbacteria.The

latterisconsistentwiththeknownpredominanceofSTIIIamong

carriedstrainsinSouthAfricanwomen(37%)butSTViscarriedby

aminorityofwomen(10%)[24,26].Inourcohortthelowest

lev-elsofantibodyweretoSTIabacteria,whichisconsistentwithan

increasedincidenceofdiseaseduetothisserotype[26].Our

ﬁnd-ingsmaysuggestdifferencesintheimmuneresponsetodifferent

colonizingserotypes.

It islikely that lower maternalantibody concentrations and

alteredtransplacentaltransferofantibodyinHIVinfectionplaya

roleintheincreasedsusceptibilitytoinfectiousdiseasesininfancy

[19,27],butthereisalackofdataontheimpactofmaternalHIV

infectiononneonatalinfectionsingeneral,includingGBS

infec-tions[28].Infantanti-GBSsurfacebindingantibodybyserotype

remains lowerat 16 weeksin HIV-exposed,uninfected infants,

whomightthereforebemorevulnerabletolateonsetGBSdisease.

ThiscorrespondstorecentdatafromCapeTownwhich suggest

that56%ofGBSdiseaseoccurredafter7daysoflifeina

popula-tionwhereoneinfourinfantswereHIV-exposed[23].Apartfrom

thereducedpassiveimmunityprovidedbymaternalantibodyin

thecontextofmaternalHIVinfectionitmightbethatthereare

differencesintheimmuneresponsetoinfectionofHIV-exposed,

uninfectedinfantssecondarytoinuteroHIVexposure,assomedata

suggestthattheseinfantshavealteredlymphocytedifferentiation

andfunctiondespiteremainingHIV-uninfected[29,30].However,

howthiscorrespondstoclinicalinfectionintheneonatalperiodis

notknown.

Ourﬁndingsareconsistentwithstudiesthatdemonstratethe

higherriskofinfectiousmorbidityandmortalityinHIV-exposed

infantscomparedtoHIV-unexposedinfants[31–33].RecentSouth

AfricanstudiesdemonstratedahighprevalenceofneonatalGBS

diseaseinapopulationwithahighHIVburden[23,24,34].AStudy

fromMalawialsodemonstratedhighratesofearlyandlateonset

GBSinfectionin apopulationwithhighHIVprevalence[35].In

contrast,onestudyfromtheUSAdidnotﬁndevidenceofincreased

susceptibilitytoGBSinfectioninHIV-exposed,uninfectedinfants

suggesting either regional screening and treatment differences

forbothHIVinfectionandGBScarriageorpopulation-based

dif-ferences in late onset GBS disease that might be unrelated to

HIV-status[36].

Ourstudyhassomelimitations.Firstly,wedidnotassess

col-onizationinthemothersparticipatinginthestudyandthuswere

unabletocorrelateantibodyresponsestomaternalcolonization.

Whilstnoinfantdevelopedearlyorlateonsetsepsisormeningitis

duringthestudy,oursamplesizewasalsotoosmalltodetectthis

asavalidendpointandthiswasnottheaimofthestudy.

Theuseoftwonovelﬂowcytometricassaysthatquantify

anti-bodybindingtothesurfaceofGBSbacteriaandantibody-mediated

C3b/iC3bdepositiononthebacterialsurfacemayprovideamore

completeassessmentofantibodyfunctionthanthemeasurement

ofanti-polysaccharideIgGalone.However,theresponsemeasured

isnotwhollyserotype-speciﬁcaspartoftheantibodybindingis

tocellwallproteinsaswellasthepolysaccharidecapsule.Despite

this,weseeaserotype-speciﬁcresponsewithaclear

differentia-tionbetweenthedistributionofantibodybindingtothedifferent

serotypes.

ThepotentialofatrivalentmaternalGBSvaccine

encompass-ingthemajordisease-causingserotypes(Ia,Ib,III)[8]toprotect

againstneonataldiseaseremains promising.However,it isvital

thatongoingevidencetosupportitsuseinvulnerablepopulations

withhighHIV-prevalenceiscollectedinordertosupport

immu-nizationstrategiesinSub-SaharanAfrica.

Acknowledgements

We thank themothers and infantswho participated in this

study.ProfessorAndroullaEfstratiou,PublicHealthEngland,

Col-indaleforaccesstoGBSstrainsandProfessorCarolBaker,Baylor

CollegeofMedicine,TexasforaccesstoherHumanpositiveserum

samples.Mr.MohammedAbdul-Khalidforhisstatisticaladviceand

support.

Conﬂictofintereststatement:PTHisaninvestigatorforresearch

studiesdoneonbehalfofStGeorges,UniversityofLondon

(Lon-don,UK)andfundedbyNovartisvaccines(Siena,Italy)andserves

as a consultant to Novartis vaccines on GBS vaccines. KLD is

fundedbyaWellcomeTrustGlobalHealthFellowship(London,UK),

RoyalCollegeofPhysiciansThomasWattEden Fellowship

(Lon-don,UK)andBHIVA/GileadRegistrarAward(London,UK). BKis

fundedbyanMRCProgramgrant(London,UK)andanNIHRSenior

Researchfellowship(London,UK).CJwasfundedbyESPID(Geneva,

Switzerland)andTheThrasherResearchFund(Atlanta,USA).

Fund-ing:Thisworkis supportedby aWellcomeTrustGlobalHealth

Fellowship(GrantNumberKLD2013);TheThomasWattEden

Fel-lowship(RoyalCollegeof PhysiciansGrant Number01012013);

andtheGilead/BHIVARegistrarAward.Contributionstatement:KLD

developedthemanuscriptandoriginalresearchidea.LA

partici-patedintheresearch.PTH,BK,ST,CJdevelopedtheoriginalideaand

substantiallycontributedtothedevelopmentofthemanuscript.

AppendixA. Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound,

intheonlineversion,athttp://dx.doi.org/10.1016/j.vaccine.2014.

12.025.

References

[1]JointUnitedNationsProgrammeonHIVandAIDS(UNAIDS).Globalreport: UNAIDSreportontheglobalAIDSepidemic:2010;2010.

(7)

[2]MarindaE,HumphreyJH,IliffPJ,MutasaK,NathooKJ,PiwozEG,etal.Child mortalityaccordingtomaternalandinfantHIVstatusinZimbabwe.Pediatr InfectDisJ2007;26:519–26.

[3]McNallyLM,JeenaPM,GajeeK,ThulaSA,SturmAW,CassolS,etal.Effectof age,polymicrobialdisease,andmaternalHIVstatusontreatmentresponseand causeofseverepneumoniainSouthAfricanchildren:aprospectivedescriptive study.Lancet2007;369:1440–51.

[4]KuhnL,KasondeP,SinkalaM,KankasaC,SemrauK,ScottN,etal.Doesseverity ofHIVdiseaseinHIV-infectedmothersaffectmortalityandmorbidityamong theiruninfectedinfants?ClinInfectDis2005;41:1654–61.

[5]Le DoareK,HeathPT. Anoverview ofglobalGBS epidemiology.Vaccine 2013;31:D7–12.

[6]GrayKJ,KafulafulaG,MatembaM,KamdoloziM,MembeG,FrenchN.Group BstreptococcusandHIVinfectioninpregnantwomen.Malawi,2008–2010. EmergInfectDis2011;17:1932–5.

[7]EpalzaC,GoetghebuerT,HainautM,PrayezF,BarlowP,DedisteA,etal.High incidenceofinvasivegroupBstreptococcalinfectionsinHIV-exposed unin-fectedinfants.Pediatrics2010;126:e631–8.

[8]BakerCJ,CareyVJ,RenchMA,EdwardsMS,HillierSL,KasperDL,etal.Maternal antibodyatdeliveryprotectsneonatesfromearlyonsetgroupBstreptococcal disease.JInfectDis2014;209:781–8.

[9]KasperDL,PaolettiLC,WesselsMR,GuttormsenHK,CareyVJ,JenningsHJ, etal.ImmuneresponsetotypeIIIgroupBstreptococcalpolysaccharide-tetanus toxoidconjugatevaccine.JClinInvest1996;98:2308–14.

[10]Baker CJ, Kasper DL. Correlation of maternal antibody deﬁciency with susceptibility to neonatal group Bstreptococcal infection.N Engl J Med 1976;294:753–6.

[11]SuaraRO,AdegbolaRA,MulhollandEK,GreenwoodBM,BakerCJ. Seropreva-lence ofantibodiesto groupBstreptococcalpolysaccharides inGambian mothersandtheirnewborns.JNatlMedAssoc1998;90:109–14.

[12]LinFY,WeismanLE,AzimiPH,Philips3rdJB,ClarkP,ReganJ,etal.Levelof maternalIgGanti-groupBstreptococcustypeIIIantibodycorrelatedwith pro-tectionofneonatesagainstearly-onsetdiseasecausedbythispathogen.JInfect Dis2004;190:928–34.

[13]HallMA,EdwardsMS,BakerCJ.Complementandantibodyparticipationin opsonophagocytosisoftypeIVandVgroupBstreptococci.InfectImmun 1992;60:5030–5.

[14]Edwards MS, Lane HJ, Hillier SL, Rench MA, Baker CJ. Persistence of functional antibodies to group B streptococcal capsular polysaccharides following immunization with glycoconjugate vaccines. Vaccine 2012;30: 4123–6.

[15]BakerCJ,PaolettiLC,RenchMA,GuttormsenHK,EdwardsMS,KasperDL. Immune response ofhealthywomen to 2different groupB streptococ-caltypeVcapsularpolysaccharide-proteinconjugatevaccines.JInfectDis 2004;189:1103–12.

[16]CaceresVM,StrebelPM,SutterRW.Factorsdeterminingprevalenceof mater-nalantibodytomeaslesvirusthroughoutinfancy:areview.ClinInfectDis 2000;31:110–9.

[17]CumberlandP,ShulmanCE,MaplePA,BulmerJN,DormanEK,KawuondoK, etal.MaternalHIVinfectionandplacentalmalariareducetransplacental anti-bodytransferandtetanusantibodylevelsinnewbornsinKenya.JInfectDis 2007;196:550–7.

[18]HealyCM,MunozFM,RenchMA,HalasaNB,EdwardsKM,BakerCJ.Prevalence ofpertussisantibodiesinmaternaldelivery,cord,andinfantserum.JInfectDis 2004;190:335–40.

[19]JonesCE,NaidooS,DeBeerC,EsserM,KampmannB,HesselingAC.Maternal HIVinfectionandantibodyresponsesagainstvaccine-preventablediseasesin uninfectedinfants.JAmMedAssoc2011;305:576–84.

[20]JonesCE,HesselingAC,Tena-CokiNG,ScribaTJ,ChegouNN,KiddM,etal.The impactofHIVexposureandmaternalMycobacteriumtuberculosisinfectionon infantimmuneresponsestobacilleCalmette-Guérinvaccination.AIDS2014[in press].

[21]BarronP,PillayY,DohertyT,ShermanG,JacksonD,BhardwajS,etal. Eliminat-ingmother-to-childHIVtransmissioninSouthAfrica.BullWorldHealthOrgan 2013;91:70–4.

[22]BrookesC,KuismaE,AlexanderF,AllenL,TiptonT,RamS,etal.Development ofalargescalehumancomplementsourceforuseinbacterialimmunoassays. JImmunolMethods2013;391:39–49.

[23]FrigatiL,vanderMerweH,TheronG,RabieH,HarveyJ,CottonMF.A retrospec-tivereviewofgroupBstreptococcalinfection–2010–2011.CapeTown,South Africa:WorldSocietyforPaediatricInfectiousDiseases;2013.OP1,abstract1.

[24]CutlandCL,MaribeA,GroomeMJ,VelaphiSC,KhoosalM,WadulaJ,etal. Inva-sivegroupBstreptococcussepsisinyoungSouthAfricaninfantsintherea ofhighmaternalHIVinfectionrates(2004–2008).CapeTown,SouthAfrica: WorldSocietyforPaediatricInfectiousDiseases;2013.OP2,abstract2.

[25]GuptaA,MathadJS,YangWT,SinghHK,GupteN,MaveV,etal.Maternal pneu-mococcalcapsularIgGantibodiesandtransplacentaltransferarelowinSouth AsianHIV-infectedmother–infantpairs.Vaccine2014;32:1466–72.

[26]MadzivhandilaM,AdrianPV,CutlandCL,KuwandaL,MadhiSA,PoPSTT. DistributionofpilusislandsofgroupBstreptococcusassociatedwith mater-nal colonization and invasivedisease in South Africa. J Med Microbiol 2013;62:249–53.

[27]SimaniOE,AdrianPV,ViolariA,KuwandaL,OtwombeK,NunesMC,etal.Effect ofin-uteroHIVexposureandantiretroviraltreatmentstrategiesonmeasles susceptibilityandimmunogenicityofmeaslesvaccine.AIDS2013;27:1583–91.

[28]SealeAC,MwanikiM,NewtonCR,BerkleyJA.Maternalandearlyonsetneonatal bacterialsepsis:burdenandstrategiesforpreventioninsub-SaharanAfrica. LancetInfectDis2009;9:428–38.

[29]HyginoJ,LimaPG,FilhoRG,SilvaAA,SaramagoCS,AndradeRM,etal.Altered immunologicalreactivityinHIV-1-exposeduninfectedneonates.ClinImmunol 2008;127:340–7.

[30]ClericiM,SaresellaM,ColomboF,FossatiS,SalaN,BricalliD,etal.T-lymphocyte maturationabnormalitiesinuninfectednewbornsandchildrenwithvertical exposuretoHIV.Blood2000;96:3866–71.

[31]Mussi-PinhataMM,FreimanisL,YamamotoAY,KorelitzJ,PintoJA,CruzML, etal.InfectiousdiseasemorbidityamongyoungHIV-1-exposedbutuninfected infantsinLatinAmericanandCaribbeancountries:theNationalInstituteof ChildHealthandHumanDevelopmentInternationalSiteDevelopment Initia-tivePerinatalStudy.Pediatrics2007;119:e694–704.

[32]KoyanagiA,HumphreyJH,NtoziniR,NathooK,MoultonLH,IliffP,etal. Morbidityamonghumanimmunodeficiencyvirus-exposedbutuninfected, human immunodeficiency virus-infected, and human immunodeficiency virus-unexposedinfantsin Zimbabwebefore availabilityof highlyactive antiretroviraltherapy.PediatrInfectDisJ2011;30:45–51.

[33]NewellML,CoovadiaH,Cortina-BorjaM,RollinsN,GaillardP,DabisF,etal. MortalityofinfectedanduninfectedinfantsborntoHIV-infectedmothersin Africa:apooledanalysis.Lancet2004;364:1236–43.

[34]QuanV,CohenC,MadhiSA.PaediatricinvasivegroupBstreptococcal dis-easeinSouthAfrica,2004to2011.CapeTown,SouthAfrica:WorldSociety forPaediatricInfectiousDiseases;2013.OP3,abstract3.

[35]GrayKJ,BennettSL,FrenchN,PhiriAJ,GrahamSM.InvasivegroupB strepto-coccalinfectionininfants,Malawi.EmergInfectDis2007;13:223–9.

[36]MayerE,MendezH,GesnerM,DesaiN,HammerschlagMR.HIV-exposed unin-fectedinfantsarenotatriskofdevelopinggroupBstreptococalinfection.Cape Town,SouthAfrica:WorldSocietyforPaediatricInfectiousDiseases;2013.OP4, abstract4.