Effectiveness of single-dose rifampicin after BCG vaccination to prevent leprosy in close contacts of patients with newly diagnosed leprosy: A cluster randomized controlled trial

(1)

Effectiveness

of

single-dose

rifampicin

after

BCG

vaccination

to

prevent

leprosy

in

close

contacts

of

patients

with

newly

diagnosed

leprosy:

A

cluster

randomized

controlled

trial

Renate

Richardus

a,b

,

Khorshed

Alam

c

,

Kallyan

Kundu

c

,

Johan

Chandra

Roy

c

,

Tasnuva

Zafar

c

,

Abu

Su

ﬁan

Chowdhury

c

,

Daan

Nieboer

a

,

Roel

Faber

a

,

C. Ruth

Butlin

c

,

Annemieke

Geluk

b,1

,

Jan

Hendrik

Richardus

a,1,

*

a

DepartmentofPublicHealth,ErasmusMC,UniversityMedicalCenterRotterdam,Rotterdam,TheNetherlands

b

DepartmentofInfectiousDiseases,LeidenUniversityMedicalCentre,Leiden,TheNetherlands

c_Rural_Health_Program,_The_Leprosy_Mission_{International}_Bangladesh,_Nilphamari,_Bangladesh

ARTICLE INFO Articlehistory: Received7July2019

Receivedinrevisedform28August2019 Accepted31August2019

CorrespondingEditor:EskildPetersen, Aar-hus,Denmark

Keywords: Leprosy

BacillusCalmette–Guérinvaccine Rifampicin

Householdcontacts Chemoprophylaxis Immunoprophylaxis

ABSTRACT

Objective:Toassesstheeffectivenessofsingle-doserifampicin(SDR)afterbacillusCalmette–Guérin (BCG)vaccinationinpreventingleprosyincontacts.

Methods:Thiswasasingle-centre,cluster-randomizedcontrolledtrialataleprosycontrolprogrammein northwestBangladesh.Participantswerethe14988contactsof1552newleprosypatientswhowere randomizedintotheSDR arm(n=7379)andtheSDR+arm(n=7609).Intheinterventiongroup,BCG vaccinationwasfollowedbySDR8–12weekslater.Inthecontrolgroup,BCGvaccinationonlywasgiven. Follow-up was performedat1year and2 years afterintake. Themainoutcome measurewas the occurrenceofleprosy.

Results:Theincidence rateper10000person-yearsatriskwas44 intheSDR armand 31inthe SDR+armat1year;theincidenceratewas34intheSDR armand41intheSDR+armat2years.There wasastatisticallynon-signiﬁcant(p=0.148;42%)reductionforpaucibacillary(PB)leprosyintheSDR+ armat1year.Ofallnewcases,33.6%appearedwithin8–12weeksafterBCGvaccination.

Conclusions:Inthefirstyear,SDRafterBCGvaccinationreducedtheincidenceofPBleprosyamong contactsby42%.Thiswasastatisticallynon-significantreductionduetothelimitednumberofcasesafter SDRwasadministered.TowhatextentSDRsuppressesexcessleprosycasesafterBCGvaccinationis difficulttoestablishbecausemanycasesappearedbeforetheSDRintervention.

Trialregistration:NetherlandsTrialRegister:NTR3087.

Introduction

Theglobalnumberofnewleprosycaseshasremainedstable over the last decade (Anonymous, 2017a), indicating that transmission of Mycobacterium leprae, the causative agent of leprosy, is ongoing in many endemic countries. The basic interventioninleprosycontrol ismultidrugtherapy(MDT),but this appears insufﬁcient to decrease new cases numbers and

achievetheWorldHealthOrganization(WHO)targetofreducing theburdenofleprosy(Anonymous,2018).

Close contacts of untreated leprosy cases are exposed considerably to M.leprae. Age of thecontact, bacterial load of the index patient, and close physical and genetic distance are independent risk factorsfor the development of leprosy(Moet etal.,2006).Householdcontactsofnewlydiagnosedpatientshave a 10-fold higher risk of developing leprosy compared withthe generalpopulation(Moetetal.,2008a);fordifferentcategoriesof neighbours andsocialcontacts,thisis three-to_ﬁve-fold higher (Moetetal.,2006;Moetetal.,2008a).

Many studieson immunoprophylaxis(vaccination) and che-moprophylaxisaimedatpreventingleprosyhavefocused primar-ilyoncontactsofleprosypatients.BacillusCalmette–Guérin(BCG) vaccinationisknownasavaccineagainsttuberculosis(TB)andis

* Correspondingauthorat:DepartmentofPublicHealth,ErasmusMC,University MedicalCenterRotterdam,POBox2040,3000CARotterdam,TheNetherlands.

E-mailaddress:j.richardus@erasmusmc.nl(J.H.Richardus).

1

AnnemiekeGelukandJanHendrikRichardusequallysharelastauthorship.

https://doi.org/10.1016/j.ijid.2019.08.035

1201-9712/©2019TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

ContentslistsavailableatScienceDirect

International

Journal

of

Infectious

Diseases

(2)

routinelygiventoinfantsaspartoftheneonatalimmunization scheme in many parts of the world. Moreover, BCG is also recognizedasprotectingagainstleprosy(Merleetal.,2010;Setia etal.,2006).SeveralvaccinetrialsusingBCGhaveestablishedits protectiveeffect against leprosy, oftenin combination withM. lepraeorrelatedMycobacteriumvaccines(Merleetal.,2010;Fine andPonnighaus,1988;Coppolaetal.,2018;Kumar,2017;Sharma et al., 2005; Richardus and Oskam, 2015). Brazil has ofﬁcially recommendedBCGsincetheearly1970sforhouseholdcontactsof leprosycases,asa boostertoroutineneonatal BCGvaccination againstTB.Since1991,theBrazilianMinistryofHealthhasadvised twodosesofBCGtobeadministeredtohouseholdcontacts.This policywasassessedinacohortstudyinBrazil(Duppreetal.,2008), andshowed56%protectionbyaboosterBCGvaccination.Therisk oftuberculoidleprosyduringtheinitialmonthswashighamong BCG-vaccinated contacts. Due to incomplete follow-up, the increasedrisk ofpaucibacillary (PB)leprosy intheﬁrstmonths afterBCGrequiresfurthersubstantiation.

Regardingchemoprophylaxis,astudyinBangladesh(theCOLEP trial)showedthatasingledoseofrifampicin(SDR)incontactsof newlydiagnosedleprosypatientsreducedtheoverallincidenceof leprosy in the ﬁrst 2 years by 57% (Moet et al., 2008b). Furthermore,thisstudyshowedthattheeffectofSDRdepended ontheBCGstatus of thecontact(Schuringet al.,2009):ifthe contact had received BCG vaccination as part of a childhood vaccinationprogramme, the protective effect of SDR was 80%. ContactswhoreceivedSDRwithoutpriorBCGvaccinationhada protectiveeffectof58%.Recently,theWHOhasincludedSDRasa recommendationinitsguidelinesforthediagnosis,treatmentand preventionofleprosy(Anonymous,2017b).

BasedonearlierstudiesinvestigatingBCGvaccinationandSDR chemoprophylaxisinthepreventionofleprosyamongcontacts,a trialwasinitiatedtoassesstheefficacyofacombinedstrategy(the MALTALEPtrial).Themainobjectiveofthistrialwastoassessthe effectivenessof SDR given after BCGvaccination in preventing leprosyinclosecontactsofpatientswithnewlydiagnosedleprosy. More specifically, the aim was to determinewhether possible excesscasesinthefirstyearafterimmunoprophylaxis,asobserved previouslyinBrazil(Duppreetal.,2008),couldbepreventedby chemoprophylaxis.

Materialsandmethods Trialdesign

Theinterventionwasaclusterrandomizedcontrolledtrialwith two treatment arms, to study the effectiveness of single-dose rifampicin(SDR+arm)givenafterBCGvaccinationinthe preven-tion of leprosy among contacts of newly diagnosed leprosy patients,versusBCGvaccinealone(SDR arm)(Figure1).During theinitialcontactsurvey,BCGwasgiventoalleligiblecontacts, followedbychemoprophylaxiswithSDR8–12weekslaterinthose contact groups randomized to receive this (FU1). Follow-up examinationswereperformedat1year(FU2)and 2years(FU3) afterreceivingBCG.Thethreefollow-uptimepointswereusedto investigate whether contacts had developed leprosy (primary outcome measure). Also, contacts were examined for adverse events at the different follow-up points. Due to operational difﬁcultiescausedbypoliticalinstabilityinthecountry,itwasnot alwayspossibletoprovideSDRexactly8weeksafterBCG,sothe rangewasbroadenedto8–12weeksafterBCG.

Eligibilitycriteriaforparticipants

Newlydiagnosedleprosypatientswereincludedwhohadbeen diagnosed withleprosy according totheRural Health Program

(RHP) guidelines, which follow those of the National Leprosy ControlProgramme(Anonymous,1998;Anonymous, 2006).The diagnosisofleprosywasmadewhenatleastoneofthecardinal signswaspresent:oneormoreskinlesionsconsistentwithleprosy andwithdefinitesensoryloss;thickenedperipheralnerve(s);and a positive skin smear result for acid-fast bacilli. Patients with negativesmearresultsandfiveorfewerskinlesionsweregrouped asPBleprosy,andthosewithpositivesmearresultsormorethan five skin lesions weregrouped as multibacillary (MB) leprosy, according to the WHO treatment criteria. MDT was started accordingtothenationalguidelines.Within2weeksafternewly diagnosedleprosypatientshadreceivedtheseconddoseofMDT(4 weeks afterthefirstdose),a householdsurveywas performed. Contactgroups wereformedof around10–15persons foreach patient.

Exclusioncriteriaforpatientsandcontactsaresummarizedin the previouslypublished methodologyarticle (Richardus et al., 2013).Onlyclosecontactswereincluded,i.e.householdcontacts andnext-doorneighbours.Contactswerecategorizedaccordingto their physical and genetic distance to the index patient. For physicaldistance,fourcategoriesweredeﬁnedbasedonthelocal housingsituation:sharesahouseand kitchen,sharesa kitchen only,sharesahousebutnotakitchen(togethercalledhousehold contacts), and next-door neighbours. For genetic distance,two groupsweredeﬁned:blood-related(parent,child,orsibling)and not blood-related or unclear (all others). Written informed consent was obtainedfrom all patientsand their contacts.For thosewhowereilliterate,athumbprintwasobtained.Forminors under 16years of age, the guardian’s additional consent was obtained.

Studysetting

The study was performed in the districts of Nilphamari, Rangpur, Thakurgaon, and Panchagarh in the northwest of Bangladesh. Patients entered the trialthrough the RHP of The Leprosy Mission International, Bangladesh (TLMI,B), based at theDBLMHospitalinNilphamari,areferralhospitalspecialized inthedetectionandtreatmentofleprosy.Thepopulationofthe fourdistrictsatthestartofintakewasaround7000000,and800– 900new leprosypatientsweredetected per year(Anonymous, 2011).TheprevalencerateofHIVinadultsaged15to49yearsin Bangladeshin2018was<0.1(UNAIDS).

Interventions

TheBCGvaccinewasappliedbytrainedresearchassistantsto all included contacts (0.1ml of BCG vaccine by intradermal injection).TwodifferentBCGstrainswereusedinthetrial(and inroutineneonatalvaccinationinBangladesh).TheIndianvaccine was used between 2011 and 2015 (Moscow strain 361) and theJapanesevaccinein2016and2017(Tokyostrain172).These are freeze-dried glutamate BCG vaccines composed of 0.5mg/ ampoulelivebacteriaofCalmette_–Guérin(asapproximately70% moist bacteria) and 2.0mg/ampoule sodium glutamate (as a stabilizer). The BCG vaccine was stored at the Government ImmunisationProgrammefacilities.

Rifampicincomesincapsulesof150mg,andthedosageused was the same as that recommended in the guidelines of the NationalLeprosyControlProgrammeofBangladeshandtheRHP (Table1).

Outcomes

Theprimaryoutcomemeasurewasthenumberofnewleprosy patientsemergingfromthecontactgroups.Theproportionsofnew

(3)

leprosypatientswerecomparedbetweenthetwoarmsofthetrial after1and2years.

Samplesize

IntheearlierCOLEPtrial(Moetetal.,2008b),theincidencerate (IR)ofleprosyamonghouseholdcontactsanddirectneighbours wasfoundtobe40per10000peryearintheuntreatedgroupover theﬁrst2years.ItwashypothesizedthatincontactsreceivingBCG only,this numberwould besimilarin theﬁrst yearorpossibly slightlyincreased.Alsobasedontheprevioustrial,a50%reduction

through theSDRintervention wasexpected(IR of2 per 1000). Based ontheseﬁgures(withtwo-sidedα=0.05,power=0.80),a totalofabout10000contactswouldbenecessaryineachgroupto detectreliablytheexpectedprotectiveeffectoftheBCGplusSDR combinationof50%,consideringanexpected10%losstofollow-up ofcontacts.

Intake tookplace between July2012 and January 2017. The intake took longer than originally planned, since the required numberofcontactsaccordingtothepowercalculationhadnotyet beenreached.Nevertheless,itwasnecessarytoendrecruitmentin 2017forbudgetaryreasons.Follow-upafter2yearswascompleted inJanuary2019.

Interimanalysesandstoppingguidelines

As the trial was not blinded, it was possible to assess the outcomes during thestudy. Thiswas done annually. The main stopping criterion was the occurrence of more serious adverse reactionstoBCGvaccinationamongcontactsthandescribedinthe literature.

Table1

Dosageofrifampicinchemoprophylaxisaccordingtoageandbodyweight. Age/weight Doseofrifampicin Adult>35kg 600mg

Adult<35kg 450mg Child10–14years 450mg Child5–9years 300mg

(4)

In thefirstyearof thetrial,wefoundanunexpectedly high proportionofhealthycontactsofpatients(0.4%)presentingwith PBleprosywithin12weeksafterreceivingtheBCGvaccination (i.e.,in thetimeframe beforeSDR was given)(Richarduset al., 2015).Sinceitwastooearlyinthetrialtodrawdefiniteconclusions about this finding, the study was continued according to the protocol.

Randomization

Eachcontactgroupwasallocatedrandomlytooneofthetwo studyarms(arm1:BCGonly;arm2:BCGplusSDR)bymeansofa computer-generatedrandomsequencewitha 1:1ratiofor each arm. A block size of 10 was used. A randomization table was createdwith2000sequentialstudynumbers(oneforeachcontact group).Eachstudynumberreceivedarandomnumbergenerated inMSExcelandthiswasfixed.Thetablewasthensortedbyblock number and random number. Within each block of 10 study numbers,thehighestfiverandomnumberswereassignedSDR,the lowestfivewereassignednoSDR.Theallocationwasgeneratedby thedatabasemanager (RF);participants wereenrolled by field staff. On inclusion of a new index patient, the local database manager(KK)enteredtheindexintothedatabase.Randomization intoanarmofthetrialwasachievedbyautomaticallyassigning eachnextstudynumbertothecontactgroup,thusassigningthe pre-allocatedrandomizationgroupofthestudynumber. Blinding

Blinding was not possible because there were no placebo capsulesofrifampicinavailableandwewerenotabletolocateany companythatcouldproducetheseespeciallyforthistrial. Statisticalmethods

Forthecalculationoftheprimaryoutcomemeasure,westarted atFU1,thetimewhenSDRwasprovidedinthetreatmentarm(SDR +) of the trial. Contacts who developed leprosy after BCG vaccination,butbeforeFU1,werenotincludedinthecalculation oftheprimaryoutcomemeasure.IRsper10000person-yearsat risk(PYAR)werecalculatedforyear1(FU2)andyear2(FU3)of follow-up. The numbers at risk were calculated by adding the numberofnewcasesofleprosytothenumberofcontactswithout leprosy at the same follow-up moment. The probability of developingleprosyat2yearswasconvertedtotheIRassuming aconstanthazardduringtheperiod(rate= log(1 leprosy/total)/ 2).Toobtainconﬁdenceintervals(CI),we appliedthestandard errorsoftheprobabilityofdevelopingleprosy(sqrt(1/leprosy+1/ noleprosy))aroundthelog(rate).Additionally,thenumberneeded totreatforBCG+SDRwasestimated.Asigniﬁcancelevelof5%was usedinalltests.

ThestatisticalanalysesweredoneusingSAS9.4.Techniquesfor theanalysisofsurveysampleswereusedtoaccountforclustering at the level of the index patient in the sample. Bivariate associations were investigated using proc surveyfreq and the Rao–ScottChi-squaretestinsteadofthePearsonChi-squaretest. Additionalanalyses

The effectiveness of BCG alone and BCG with SDR was investigated in different subgroups and odds ratios(OR) were reported,whicharecomparabletorelativerisksduetothelow prevalenceofleprosy.Additionally,thenumberneededtotreat (NNT) per subgroup of contacts was reported. Clustering was accountedforthroughtheprocsurveylogisticinsteadofordinary logisticregression.

Results

Atotalof1552indexpatientswereincluded,ofwhom1077 (70%)werePBleprosypatientsand475(30%)wereMBpatients. TheintakeofPBleprosyindexpatientswasintentionallyended whenaround1000hadbeenincluded,toensureanintakeofat least300MBpatients.Thenumberofparticipantsineacharmof thetrialisshowninFigure1.Atotalof20947eligiblehousehold contactswereidentiﬁed.Reasonsforexclusionweresteroiduse (n=9), pregnancy (n=241), liver disease or jaundice (n=70), malignancies(n=7),historyoforundertreatmentforTB(n=122), historyof leprosy(n=462),leprosypatientorsuspect atintake (n=228),refusalofinformedconsent(n=1136),under5yearsold (n=1900),residingtemporarilyinthearea(n=1314),orsuffering from another serious illness (n=673). Some contacts were excluded becausethey had more than one exclusion criterion. HIVwasnottestedwithinthetrial,butwhenreported,wasusedas anexclusioncriterion.Afterexclusion,14988contactsenteredthe trial.

The contacts in both arms of the trial were well-balanced (Table2).Ofthe14988contactsincluded,7245contactsinthe SDR armwerecheckedatFU1,7033at FU2,and 6898atFU3 (Figure1).Atotalof7322contactsintheSDR+armreceivedSDRat FU1;7042werechecked atFU2 and 6906at FU3.Of the7322 contacts randomizedtoreceive SDR, 283didnot receiveit for variousreasons. These contactswerenotincludedin theeffect calculations.

Amongtheincludedcontacts, 27newleprosypatientswere foundinthefirstyear(atFU2)intheSDR arm,and19inthe SDR+arm. Subsequently, 24 new patients were found in the secondyear(atFU3)intheSDR arm,and 29intheSDR+arm (Tables3–5).TheIRof leprosyper10 000PYAR was44in the SDR arm and 31 in the SDR+arm at 1year, and 34 in the SDR armand41intheSDR+armat 2years.Thereductionin incidenceof leprosyin theSDR+groupcompared totheSDR groupwas42%(95%CI 13%to70%);Rao–ScottChi-square=2.1 (df=1),p= 0.148.TheoverallNNTwas714(95%CI 2000to313) forPBleprosyinthefirstyear.ThereductioninnewPBleprosy cases in the BCG + SDR group occurredin the first year after treatment;inyear2,nostatisticallysignificantdifferenceinthe numberofnewPBleprosycaseswasfoundbetweenthegroups. SupplementaryMaterialTablesS1andS2showtheeffectof BCGonlyandBCGwithSDRprophylaxisbyvariablecategory1and 2yearsafterBCGvaccination.Nosignificantdifferencesofinterest werefound.AnegativeNNTindicatesastatisticallynon-significant difference.

Table6showsthenumberofnewcasesatthedifferent follow-uppointsincludingFU1at8–12weeksafterBCG.Thistableshows that50outofatotalof149newcases(33.6%)occurredwithin3 monthsafter BCGvaccination. Allexcept oneof thesewerePB leprosycases.Laterinthetrial,moreMBcasesarose(eightMB casesafter1yearandsixafter2years).

TherateofdocumentedadverseeventsafterBCGinthetrial waslow(0.34%)andcomparabletoratesreportedinstudiesfrom othercountries(Krysztopa-Grzybowskaetal.,2012;Douradoetal., 2003;Turnbulletal., 2002;Grange,1998).Thesecomplications consistedprimarily(80%) ofskinulcerations, whichareknown, commonand benign adverse events afterBCG vaccination; we havedescribedthesepreviously(Richardusetal.,2018).Exceptfor theorangeurinediscolourationcausedbyrifampicin,noadverse eventswerereportedafterSDR.

Discussion

IntheﬁrstyearaftertheprovisionofSDRtocontactswhohad ﬁrstreceivedBCGvaccination,thenumberofPBleprosypatients

(5)

Table3

Analysisofallcasesofleprosyincontactsofpatientswithnewlydiagnosedleprosy.a

Treatment Leprosy Noleprosy Totalnumberatrisk Incidencerateper10000PYAR 95%CI BCG 1yearFU 27 7250 7277 44 30–64 2yearFU 24 7118 7142 34 23–50 1–2yearsFU 51 BCGandSDR 1yearFU 19 7228 7247 31 20–48 2yearFU 29 7087 7116 41 28–59 1–2yearsFU 48

PYAR,person-yearsatrisk;CI,conﬁdenceinterval;BCG,bacillusCalmette–Guérin;SDR,singledoseofrifampicin;FU,follow-up.

a_Numbers_are_provided_by_treatment_arm_at₁_and₂_years_of_follow-up,_with_incidence_rates_per₁₀₀₀₀_PYAR_(95%_CI).

Table2

Baselinecharacteristicsatintakeofcontactsofnewlydiagnosedleprosypatients(N=14988)bytreatmentallocation.a

Variable BCG(n=7379) % BCGandSDR(n=7609) % p-Value Ageatintake(inyears)

5–14 2203 29.85 2302 30.25 0.68 15–29 2051 27.80 2113 27.77 30–44 1586 21.49 1610 21.16 45 1539 20.86 1584 20.82 Sex Male 3358 45.51 3407 44.78 0.27 Female 4021 54.49 4202 55.22

Geneticdistancetoindexpatient

Blood-related 1662 22.52 1647 21.65 0.34

Notblood-related(orunclear) 5717 77.48 5962 78.35 Typeofleprosyinindexpatient

Paucibacillary 5009 67.88 5367 70.53 0.31 Multibacillary 2370 32.12 2242 29.47 BCGscar Present 4201 56.93 4369 57.42 0.67 Absent 3172 42.99 3236 42.53 Unknown 6 0.08 4 0.05

Physicaldistancetoindexpatient

Householdcontact 2192 29.71 2117 27.82 0.09

Neighbour 5187 70.29 5492 72.18

BCG,bacillusCalmette–Guérin;SDR,singledoseofrifampicin.

a

Valuesarenumbersandpercentagesoftotalnumbersofcontacts.

Table5

AnalysisofMBleprosyincontactsofpatientswithnewlydiagnosedleprosy.a

Treatment MB Noleprosy Totalnumberatrisk Incidencerateper10000PYAR 95%CI BCG 1yearFU 3 7274 7277 4.9 1.6–15 2yearFU 0 7142 7142 0.00 – 1–2yearsFU 3 BCGandSDR 1yearFU 5 7242 7247 8.1 3.4–20 2yearFU 6 7110 7116 8.4 3.8–19 1–2yearsFU 11

MB,multibacillary;PYAR,person-yearsatrisk;CI,conﬁdenceinterval;BCG,bacillusCalmette–Guérin;SDR,singledoseofrifampicin;FU,follow-up.

a

Numbersareprovidedbytreatmentarmat1and2yearsoffollow-up,withincidenceratesper10000person-yearsatrisk(95%conﬁdenceinterval). Table4

AnalysisofPBleprosyincontactsofpatientswithnewlydiagnosedleprosy.a

Treatment PBleprosy Noleprosy Totalnumberatrisk Incidencerateper10000PYAR 95%CI BCG 1yearFU 24 7253 7277 39b _26–58 2yearFU 24 7118 7142 34 23–50 1–2yearsFU 48 BCGandSDR 1yearFU 14 7233 7247 23b _13–38 2yearFU 23 7093 7116 32 22–49 1–2yearsFU 37

PB,paucibacillary;PYAR,person-yearsatrisk;CI,conﬁdenceinterval;BCG,bacillusCalmette–Guérin;SDR,singledoseofrifampicin;FU,follow-up.

a_Numbers_are_provided_by_treatment_arm_at₁_and₂_years_of_follow-up,_with_incidence_rates_per₁₀₀₀₀_person-years_at_risk_(95%_conﬁdence_interval). b

OverallreductioninincidenceofPBleprosyintheSDR+groupinyear1:42%(95%CI 13%to70%);Rao–ScottChi-square=2.1(df=1),p=0.148;overallnumberneededto treat=714(95%CI 2000to313).

(6)

wasreducedby42%comparedtothegroupthatdidnotreceive SDR.NoadditionaleffectofSDRwasseeninthesecondyear.A large proportion (33.6%) appeared within 8–12 weeks after vaccination, the window period between vaccination and the provisionofSDR.

By providing rifampicin (a bactericidaldrug) at 8–12 weeks afterBCGvaccination,weenvisagedthepreventionofnewleprosy casesamongcontactsinthefirstyearafterBCGvaccination.This wasdescribedinBrazilbyDuppreetal.(2008),whoshowedthat theriskofPBleprosywashighduringtheinitialmonthsamong those contacts vaccinated with BCG: among the 58 newcases detected during 18 years of contact follow-up, leprosy was diagnosed in 21 of these contacts (36%) relatively soon after vaccination(2–10months); 18out ofthese 21contactshad PB leprosy.WealsofoundanunexpectedlyhighproportionofnewPB leprosycasesfollowingBCGvaccination;however,this phenome-nonhadalreadyoccurredintheperiodbetweenBCGvaccination and SDR provision.This time interval was selected in order to ensurethatrifampicinwouldnotaffecttheefficacyofBCG,which isalivevaccine.Atthetimeofconceptualizationofthetrial,we hadnoindicationtoexpectthatthiswouldoccurthisearlyafter BCG.Mosttrials haveonlyincludedlong-term follow-up,often starting1yearaftervaccination.TheBraziliantrial(Duppreetal. (2008))diagnosedthenewleprosycases2–10monthsafterBCG vaccination,whichwasalsolaterthanwhatwasfoundinourtrial. Inpreviousstudies,eitherthe numberof caseswas toolow to confirmearly‘induction’ofleprosyafterBCG(KarongaPrevention TrialGroup,1996; HW W,1960), or it was not specified when exactly leprosy occurred after vaccination (Lwin et al., 1985; Bagshaweetal.,1989).So,atthetimeSDRwasprovidedinthe currentstudy, mostexcesscases had probablyalready become manifest.

WhatwouldhavebeentheresultofthetrialifSDRhadbeen givenbeforeBCGvaccination?Therewasnopublishedevidenceto supportourdecision ontheorderof BCGand SDR. We simply followed the logic of the primary research question regarding whether SDR would suppress the excess cases after BCG vaccination and designed the study in that order. Also, the interventionstrategyconsideredthebactericidaleffectofSDRon livebacteriasuchasBCG.Inhindsight,itmayhavebeenpreferable toﬁrstprovideSDR,andthisshouldbeexploredinafuturestudy. ThelevelofprotectionofferedbySDRinthepresentstudywas 42%,whichislessthanthelevelreportedintheCOLEPstudy(57%) conducted10yearspreviouslyinthesamepopulation(Moetetal., 2008b).However,ourcontactpopulationonlyincludedhousehold andﬁrstneighbourcontacts,whiletheCOLEPstudyalsoincluded secondneighboursand socialcontacts.Thefurthercontactsare physicallyremovedfromtheindexcase,themorepronouncedis theeffectofSDRinprotectingagainstleprosy.Thisisprobablydue toalowerexposurerateandhencealowerbacterialloadofthese

furtherdistancedcontacts,renderingasingledoseofrifampicin more effective (Moet et al., 2008b; Feenstra et al., 2012). ImmunologicalscreeningoftheeffectofSDRonM.lepraeinfection incontactscouldprovideinsightintotheextent,howfast,andhow durable M. leprae infection is reduced by this single dose of antibiotic.

The observations from this trial give rise to interesting hypothesesregardingtheimmunologicalmechanismsunderlying theeffectofBCGvaccinationgiventocontactsofnewlydiagnosed leprosycases.ItispossiblethatBCGacceleratespro-inflammatory T-helper1(Th1)immunitytoM.lepraeantigens,therebyrevealing incipient forms of PB leprosy. Alternatively, BCG vaccination is also known to induce trained immunity and thereby non-specifically activates protective innate responses (Arts et al., 2018;Kleinnijenhuisetal.,2012).Inapreviousstudy(Richardus etal.,2018),weshowedthatBCGvaccinationinducedsignificant Th1-typeimmunity(higherlevelsofinterferongamma)inthose whopresentedwithhighlocalinflammationresponses, implicat-ing that efficient protection against M. leprae is dependent on an adequate Th1 response (Ottenhoff, 2012), although the concomitantinflammationmayresultincollateraltissuedamage (Geluk,2018).

ThisstudyinvestigatedtheeffectofBCGwithorwithoutSDRin onehighlyendemicareaintheIndiansub-continentwithaspeciﬁc PB:MBratio(2:1insteadoftheusual1:1reportedworldwide)(van Hooijetal.,2016;vanHooijetal.,2018;vanHooijetal.,2019),alow socio-economic status, and speciﬁc demographic, genetic, and culturalcharacteristics.WhetherBCGwouldgivesimilar protec-tioninotherareasoftheworldisquestionable.Furthermore,in BangladeshtheMoscowstrain361andTokyostrain172areused; elsewhere,theuseofotherBCGstrainsforvaccinationcouldlead todifferentresults(Fine,1995;Zhangetal.,2013).

Thistrialwasnotdesignedtoestablishtheprotectiveeffectof BCGagainstleprosy.Itwasassumedthatthisisagivenbasedon theliterature(Merleetal.,2010;KarongaPreventionTrialGroup, 1996;Cunhaetal.,2008);thereforeanarmwithoutBCGwasnot includedinthetrial.However,wedoubtthattheprotectiveeffect ofBCGalonewaslargeinthisstudy.TheIRofleprosyat2years amongthehouseholdcontactsandnext-doorneighboursinthe non-interventionarmintheCOLEPstudywas39.35per10000 PYAR(Moetetal.,2008b).TheIRwas33.72per10000PYARinthe BCG-onlyarmat2yearsintheMALTALEPtrial.Thisimpliesa14.3% reductioninleprosyincidencebyBCGvaccinationcomparedtono intervention.ABraziliantrial(Duppreetal.,2008)showedthatthe protectionconferredbyaboosterBCGvaccinationwas56%and wasnotsubstantiallyaffectedbypreviousBCGvaccination.More speciﬁcally,thiseffectwas83–85%fortheindeterminateandMB formsofleprosy,butanon-signiﬁcanteffectof26%wasfoundfor thePBleprosyforms.ThismightexplainthelackofeffectofBCGin ourtrialwhencomparedtonointervention;inBangladesh,most patientshavethePBformofleprosy(Anonymous,2017a).

Inasubgroupanalysis(SupplementaryMaterial),wefoundno signiﬁcant difference in the development of leprosy between revaccinated(BCGscar-positive)versusprimaryvaccinated(BCG scar-naïve) contacts.In theirmeta-analysis, Merleet al. (Merle etal.(2010))alsofoundnostatisticaldifferenceinBCGprotection againstleprosybetweenstudiesinwhich individualswere vac-cinatedonceandstudiesinwhichindividualsreceivedabooster vaccinationontopoftheneonatalvaccination.

There may be better alternatives to BCG vaccination as immunoprophylaxis in leprosy, with new candidate leprosy vaccinesinthepipeline,suchasMIP(Sharmaetal.,2005;Kumar, 2017)andLepVax(Duthieetal.,2018).TheMIPvaccinehasonly been evaluatedin UttarPradesh, India, withboth patientsand contactsvaccinated.Theprotectiveefﬁcacywas68%,60%,and28% after 3, 6, and 9 years, respectively (Sharmaet al., 2005).For

Table6

Newleprosycasesamongcontactsofnewlydiagnosedleprosycasesidentiﬁed accordingtothetimepointsofdiagnosis.

8–12weeks 1year 2years Total BCG PB 23 24 24 71 MB 0 3a 0 3 BCGandSDR PB 26 14 23 63 MB 1 5 6 12 Total 50 46 53 149

BCG,bacillusCalmette–Guérin;PB,paucibacillary;MB,multibacillary;SDR,single doseofrifampicin;BI,bacterialindex;BL,borderlinelepromatous;BT,borderline tuberculous.

a

OnlyonenewMBleprosycasehadaBIof2+(BL);therestoftheMBcaseswere smear-negative(MBBT).

(7)

LepVax,post-exposureprophylaxistestedinnine-banded arma-dillosappearedsafe,andunlikeBCG,itdiminishedthe neurologi-caldisruptionscausedbyM.lepraeinfection(Duthieetal.,2018). Furthertrialsareneededtoinvestigatethesevaccinesbeforethey canbeintroducedintheﬁeld.

Astrengthofthistrialisthatitwasrandomizedandcontrolled and field-based.Anextensive numberof leprosycontacts were included (n=14988). Also, because it was based in a leprosy-endemicarea,implementationwasclosetoclinicalfieldpractice. Theloss tofollow-up was less than6%, which was lowerthan expected.Alimitationisthatitwasnotpossibletomakeit double-blind (placebo was not available), which mayhave biased the results.Evenwhenusingaharmlessdoseofadissimilarvitamin pilltopreventparticipantsfromknowingwhetherornottheyhad beengivenanintervention,thiswouldnothavepreventedbiasby thefieldstaff,sincetheywouldhaveknownthedifference.For instance,thefieldstaffmayexpectandlookmorecloselyforsigns and symptomsofleprosy in thosewho havenot receivedSDR. Furthermore, a limitation was that intake took longer than expectedandthereforewecouldnotreachthe10000 contacts perarmwesetouttoinclude,leadingtolesspowerandtherefore lessstatisticallysignificantresults.

Inconclusion,itisdifﬁculttoestablishtheextenttowhichSDR suppressesexcessleprosycasesamongcontactsintheyearafter BCGvaccination.Basedonthisstudy,wecannotrecommendBCG vaccinationfollowedbySDRasaroutineinterventioninleprosy control.However,wedoadvisecontactsurveysfollowedbySDR to eligible contacts of new leprosy cases. Recently, the WHO includedSDRintheguidelinesforitsleprosyeliminationstrategy (Anonymous, 2017b). Implementationstudies onthe effective-ness of SDR as leprosy post-exposure prophylaxis (LPEP) are currently ongoing (Barth-Jaeggi et al., 2016; Steinmann et al., 2018).

Authorcontributions

All authors contributed to the design of the study and manuscriptpreparation.Allauthorshavereadandapprovedthe ﬁnalmanuscript.

Conﬂictofinterest

Theauthorsdeclarethattheyhavenocompetinginterests.The BCGvaccinewasprovidedfree ofchargebytheGovernmentof Bangladesh.

Ethicalapproval

TheNationalResearchEthicsCommittee(BangladeshMedical ResearchCouncil)approvedthestudy protocol(Ref.No.BMRC/ NREC/2010-2013/1534).

Funding

This work was supportedby the Order of Malta Grants for Leprosy Research (MALTALEP), the Q.M. Gastmann-Wichers Foundation (to AG and JHR), the Netherlands Leprosy Relief Foundation (NLR; ILEP#: 702.02.73), and the LeprosyResearch Initiative (LRI;ILEP#: 703.15.07)both togetherwith theTuring Foundation.

AppendixA.Supplementarydata

Supplementarymaterialrelatedtothisarticlecanbefound,inthe onlineversion,atdoi:https://doi.org/10.1016/j.ijid.2019.08.035.

References

Anonymous.WHOexpertcommitteeonleprosy.WorldHealthOrganTechRepSer 1998;874:1–43.

Anonymous.Globalstrategyforfurtherreducingtheleprosyburdenandsustaining leprosycontrolactivities2006-2010.Operationalguidelines.LeprRev2006;77 (3)IX,X,1–50.

Anonymous.Bangladeshbureauofstatistics:Bangladeshpopulation&housing census2011.2011Availablefrom: http://www.bbs.gov.bd/site/page/47856ad0-7e1c-4aab-bd78-892733bc06eb/-.[Accessed20April2019].

Anonymous. Global leprosy update, 2016: accelerating reduction of disease burden.ReleveEpidemiolHebd2017a;92(35)501–19[publishedOnlineFirst: 2017/09/02].

Anonymous.Guidelinesforthediagnosis,treatmentandpreventionofleprosy.New Delhi:WorldHealthOrganization,RegionalOfﬁceforSouth-EastAsia;2017.

Anonymous.Globalleprosyupdate,2017:reducingthedisease burdendue to leprosy.WklyEpidemiolRec2018;35(93):445–56.

ArtsRJW,MoorlagS,NovakovicB,LiY,WangSY,OostingM,etal.BCGvaccination protectsagainstexperimentalviralinfectioninhumansthroughtheinduction ofcytokinesassociatedwithtrainedimmunity.CellHostMicrobe2018;23(1) 89–100.e5,doi:http://dx.doi.org/10.1016/j.chom.2017.12.010[publishedOnline First:2018/01/13].

BagshaweA,ScottGC,RussellDA,WigleySC,MerianosA,BerryG.BCGvaccination inleprosy:ﬁnalresultsofthetrialinKarimui,PapuaNewGuinea,1963-79.Bull WorldHealthOrgan1989;67(4)389–99[publishedOnlineFirst:1989/01/01].

Barth-JaeggiT,SteinmannP,MierasL,vanBrakelW,RichardusJH,TiwariA,etal. Leprosy post-exposureprophylaxis (LPEP)programme: study protocol for evaluatingthefeasibilityandimpactoncasedetectionratesofcontacttracing andsingledoserifampicin.BMJOpen2016;6(11)e013633,doi:http://dx.doi.org/ 10.1136/bmjopen-2016-013633[publishedOnlineFirst:2016/11/20].

CoppolaM,vandenEedenSJF,RobbinsN,WilsonL,FrankenK,AdamsLB,etal. Vaccines for leprosy and tuberculosis: opportunities for shared research, development,andapplication.FrontImmunol2018;9:308,doi:http://dx.doi. org/10.3389/ﬁmmu.2018.00308[publishedOnlineFirst:2018/03/15].

CunhaSS,AlexanderN,BarretoML,PereiraES,DouradoI,MarojaMdeF,etal.BCG revaccinationdoesnot protectagainstleprosyintheBrazilian Amazon:a clusterrandomisedtrial.PLoSNeglTropDis2008;2(2)e167,doi:http://dx.doi. org/10.1371/journal.pntd.0000167[publishedOnlineFirst:2008/02/14].

DouradoI,RiosMH,PereiraSM,CunhaSS,IchiharaMY,GoesJC,etal.Ratesof adversereactionstoﬁrstandseconddosesofBCGvaccination:resultsofalarge communitytrialinBrazilianschoolchildren.IntJTubercLungDis2003;7(4) 399–402[publishedOnlineFirst:2003/05/06].

DuppreNC,CamachoLA,daCunhaSS,StruchinerCJ,SalesAM,NeryJA,etal. EffectivenessofBCGvaccinationamongleprosycontacts:acohortstudy.Trans R Soc Trop Med Hyg 2008;102(7)631–8, doi:http://dx.doi.org/10.1016/j. trstmh.2008.04.015[publishedOnlineFirst:2008/06/03].

DuthieMS,PenaMT,EbenezerGJ,GillisTP,SharmaR,CunninghamK,etal.LepVax, a deﬁned subunit vaccine thatprovides effective pre-exposureand post-exposureprophylaxisofM.lepraeinfection.NPJVaccines2018;3:12,doi:http:// dx.doi.org/10.1038/s41541-018-0050-z[publishedOnlineFirst:2018/04/06].

FeenstraSG,PahanD, MoetFJ,Oskam L,RichardusJH.Patient-relatedfactors predictingtheeffectivenessofrifampicinchemoprophylaxisincontacts:6year followupoftheCOLEPcohortinBangladesh.LeprRev2012;83(3)292–304 [publishedOnlineFirst:2013/01/30].

Fine PE.Variationinprotectionby BCG: implicationsofand forheterologous immunity.Lancet1995;346(8986)1339–45[publishedOnlineFirst:1995/11/18].

FinePE,PonnighausJM.LeprosyinMalawi.2.Background,designandprospectsof theKarongaPreventionTrial,aleprosyvaccinetrialinnorthernMalawi.TransR SocTropMedHyg1988;82(6)810–7[publishedOnlineFirst:1988/01/01].

Geluk A. Correlates of immune exacerbations in leprosy. Semin Immunol 2018;39:111–8, doi:http://dx.doi.org/10.1016/j.smim.2018.06.003 [published OnlineFirst:2018/06/29].

Grange JM. Complications of bacille Calmette-Guerin (BCG) vaccination and immunotherapyandtheirmanagement.CommunDisPublicHealth1998;1(2) 84–8[publishedOnlineFirst:1998/06/30].

HWW.BCG-inducedactivations.IntJLepr1960;28(2):179–81.

Karonga PreventionTrial Group. Randomised controlled trial of single BCG, repeatedBCG,orcombinedBCGandkilledMycobacteriumlepraevaccinefor preventionofleprosyandtuberculosisinMalawi.KarongaPreventionTrial Group.Lancet1996;348(9019)17–24[publishedOnlineFirst:1996/07/06].

KleinnijenhuisJ,QuintinJ,PreijersF,JoostenLA,IfrimDC,SaeedS,etal.Bacille Calmette-Guerin induces NOD2-dependent nonspeciﬁc protection from reinfectionviaepigeneticreprogrammingofmonocytes.ProcNatlAcadSci USA2012;109(43)17537–42,doi:http://dx.doi.org/10.1073/pnas.1202870109 [publishedOnlineFirst:2012/09/19].

Krysztopa-Grzybowska K, Paradowska-Stankiewicz I, Lutynska A. The rate of adverseeventsfollowingBCGvaccinationinPoland.PrzeglEpidemiol2012;66 (3)465–9[publishedOnlineFirst:2012/12/13].

KumarS.Indiaresurrectsforgottenleprosyvaccine.Science2017;356(6342)999, doi:http://dx.doi.org/10.1126/science.356.6342.999 [published Online First: 2017/06/10].

LwinK,SundaresanT,GyiMM,BechelliLM,TamondongC,GarbajosaPG,etal.BCG vaccinationofchildrenagainstleprosy:fourteen-yearﬁndingsofthetrialin Burma.BullWorldHealthOrgan1985;63(6)1069–78[publishedOnlineFirst: 1985/01/01].

(8)

MerleCS,CunhaSS,RodriguesLC.BCGvaccinationandleprosyprotection:reviewof currentevidenceandstatusofBCGinleprosycontrol.ExpertRevVaccines 2010;9(2)209–22,doi:http://dx.doi.org/10.1586/erv.09.161[published Online First:2010/01/30].

MoetFJ,PahanD,SchuringRP,OskamL,RichardusJH.Physicaldistance,genetic relationship,age,andleprosyclassiﬁcationareindependentriskfactorsfor leprosyincontactsofpatientswithleprosy.JInfectDis2006;193(3)346–53, doi:http://dx.doi.org/10.1086/499278[publishedOnlineFirst:2006/01/03].

MoetFJ,SchuringRP,PahanD,OskamL,RichardusJH.Theprevalenceofpreviously undiagnosedleprosyinthegeneralpopulationofnorthwestbangladesh.PLoS Negl Trop Dis 2008a;2(2)e198, doi:http://dx.doi.org/10.1371/journal. pntd.0000198[publishedOnlineFirst:2008/02/28].

MoetFJ,PahanD,OskamL,RichardusJH.Effectivenessofsingledoserifampicinin preventingleprosyinclosecontactsofpatientswithnewlydiagnosedleprosy: clusterrandomisedcontrolledtrial.BMJ2008b;336(7647)761–4,doi:http://dx. doi.org/10.1136/bmj.39500.885752.BE[publishedOnlineFirst:2008/03/12].

Ottenhoff TH. New pathwaysof protective and pathological host defense to mycobacteria. Trends Microbiol 2012;20(9)419–28, doi:http://dx.doi.org/ 10.1016/j.tim.2012.06.002[publishedOnlineFirst:2012/07/13].

RichardusJH,OskamL.Protectingpeopleagainstleprosy:chemoprophylaxisand immunoprophylaxis(vaccination).ClinDermatol2015;33(1):19–25.

RichardusRA,AlamK,PahanD,FeenstraSG,GelukA,RichardusJH.Thecombined effectofchemoprophylaxiswithsingledoserifampicinand immunoprophy-laxiswithBCGtopreventleprosyincontactsofnewlydiagnosedleprosycases: a cluster randomized controlled trial (MALTALEP study). BMC Infect Dis 2013;13:456,doi:http://dx.doi.org/10.1186/1471-2334-13-456[published On-lineFirst:2013/10/04].

Richardus RA, Butlin CR, Alam K, Kundu K, GelukA, Richardus JH. Clinical manifestationsofleprosyafterBCGvaccination:anobservationalstudyin Bangladesh. Vaccine 2015;33(13)1562–7, doi:http://dx.doi.org/10.1016/j.vac-cine.2015.02.017[publishedOnlineFirst:2015/02/24].

RichardusR,vanHooijA,vandenEedenSJF,WilsonL,AlamK,RichardusJH,etal. BCGandadverseeventsinthecontextofleprosy.FrontImmunol2018;9:629, doi:http://dx.doi.org/10.3389/ﬁmmu.2018.00629 [published Online First: 2018/04/20].

SchuringRP,RichardusJH,PahanD,OskamL.Protectiveeffectofthecombination BCGvaccinationandrifampicinprophylaxisinleprosyprevention.Vaccine 2009;27(50)7125–8,doi:http://dx.doi.org/10.1016/j.vaccine.2009.09.054 [pub-lishedOnlineFirst:2009/09/30].

SetiaMS,SteinmausC,HoCS,RutherfordGW.TheroleofBCGinpreventionof leprosy:ameta-analysis.LancetInfectDis2006;6(3)162–70,doi:http://dx.doi. org/10.1016/s1473-3099(06)70412-1[publishedOnlineFirst:2006/02/28].

SharmaP,MukherjeeR,TalwarGP,SarathchandraKG,WaliaR,ParidaSK,etal. Immunoprophylactic effectsof theanti-leprosyMw vaccinein household contactsofleprosypatients:clinicalﬁeldtrialswithafollowupof8-10years. LeprRev2005;76(2)127–43[publishedOnlineFirst:2005/07/26].

SteinmannP,CavalieroA,AertsA,AnandS,ArifM,AySS,etal.TheLeprosy Post-ExposureProphylaxis(LPEP)programme:updateandinterimanalysis.LeprRev 2018;89:102–16.

TurnbullFM,McIntyrePB,AchatHM,WangH,StapledonR,GoldM,etal.National studyofadversereactionsaftervaccinationwithbacilleCalmette-Guerin.Clin InfectDis2002;34(4)447–53,doi:http://dx.doi.org/10.1086/338462[published OnlineFirst:2002/01/18].

UNAIDS https://aidsdatahub.org/sites/default/ﬁles/country_review/Bangladesh-2019-UNAIDS-data.pdf(Accessed30September2019).

vanHooijA,TjonKonFatEM,RichardusR,vandenEedenSJ,WilsonL,deDoodCJ, etal.Quantitativelateralﬂowstripassaysasuser-friendlytoolstodetect biomarkerproﬁlesforleprosy.SciRep2016;6:34260,doi:http://dx.doi.org/ 10.1038/srep34260[publishedOnlineFirst:2016/09/30].

vanHooijA,TjonKonFatEM,BatistadaSilvaM,CarvalhoBouthR,CunhaMessias AC,GobboAR,etal.EvaluationofImmunodiagnosticTestsforLeprosyinBrazil, China and Ethiopia. Sci Rep 2018;8(1)17920, doi:http://dx.doi.org/10.1038/ s41598-018-36323-1[publishedOnlineFirst:2018/12/19].

vanHooijA,vandenEedenS,RichardusR,TjonKonFatE,WilsonL,FrankenK,etal. Applicationofnewhostbiomarkerproﬁlesinquantitativepoint-of-caretests facilitatesleprosydiagnosisintheﬁeld.EBioMedicine2019;,doi:http://dx.doi. org/10.1016/j.ebiom.2019.08.009[publishedOnlineFirst:2019/08/20].

ZhangW,ZhangY,ZhengH,PanY,LiuH,DuP,etal.Genomesequencingand analysisofBCGvaccinestrains.PLoSOne2013;8(8)e71243,doi:http://dx.doi. org/10.1371/journal.pone.0071243[publishedOnlineFirst:2013/08/27].