A novel view on the pathogenesis of complications after intravesical BCG for bladder cancer

Perspective

A novel view on the pathogenesis of complications after intravesical BCG for bladder cancer

Manu P. Bilsen

, Krista E. van Meijgaarden

, Hanna K. de Jong

, Simone A. Joosten

**, Corine Prins

, Lucia J.M. Kroft

, Jacqueline T. Jonker

, Stijn Crobach

, Rob C. Pelger

, Tom H.M. Ottenhoff

, Sandra M. Arend

*

aDepartmentofInfectiousDiseases,LeidenUniversityMedicalCentre,Leiden,TheNetherlands

bDepartmentofInternalMedicine,OnzeLieveVrouweGasthuis,Amsterdam,TheNetherlands

cDepartmentofRadiology,LeidenUniversityMedicalCentre,Leiden,TheNetherlands

dDepartmentofNephrology,LeidenUniversityMedicalCentre,Leiden,TheNetherlands

eDepartmentofPathology,LeidenUniversityMedicalCentre,Leiden,TheNetherlands

fDepartmentofUrology,LeidenUniversityMedicalCentre,Leiden,TheNetherlands

ARTICLE INFO

Articlehistory:

Received4December2017 Receivedinrevisedform9May2018 Accepted10May2018

CorrespondingEditor:EskildPetersen,Aar- hus,Denmark

Keywords:

BCG Bladdercancer Complications

Mycobacterialgrowthinhibitionassay Pathogenesis

ABSTRACT

IntravesicalbacillusCalmette-Guérin(BCG)iswidelyusedforhigh-risk,non-muscle-invasivebladder cancer.This reportdescribesfourcases thatillustrate thespectrumofBCG-induced complications, varying fromgranulomatous prostatitis tosepsis.There is considerable debate regardingwhether inflammationorinfectionisthepredominantmechanisminthepathogenesisofBCGdisease.Intwo patientswith asystemicillness,thesymptomsfirstresolved afteradding prednisone,indicatinga principal role for inflammation in systemic disease. In vitro testing of T-cell responses and a mycobacterialgrowthinhibitionassaywereperformedforthesepatientswithsystemicdisease.The patientwithmildsymptomsshowedmoreeffectiveinvitrogrowthreductionofBCG,whilethepatient withsepsisandorganinvolvementhadhighT-cellresponsesbutineffectivekilling.Whilethesefindings arepreliminary,itisbelievedthatimmunologicalassays,asdescribedinthisreport,mayprovideabetter insightintothepathogenesisofBCGdiseaseinindividualpatients,justifyingfurtherresearch.

©2018TheAuthor(s).PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.

ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by- nc-nd/4.0/).

Introduction

IntravesicalbacillusCalmette–Guérin(BCG)instillationfollow- ingtransurethralresection(TUR)isconsideredthegoldstandardin thetreatment of high-risk, non-muscle-invasive bladder cancer.

ComparedwiththeuseofTURalone,additionalintravesicalBCG significantlyreducesdiseaserecurrenceandtheriskofprogression

to invasive disease (Shelley et al., 2010).However, BCG-related complicationsarecommonandmaynecessitatetreatmentdiscon- tinuation. Sideeffects limitedto thegenitourinary tract include chemical cystitis (35%), bacterial cystitis (23.3%), macroscopic haematuria (22.6%), and symptomatic granulomatousprostatitis (0.9%).Systemiccomplicationsarepersistenthigh-gradefeverasan isolatedmanifestation(2.9%),pneumonitiswithorwithouthepatitis (0.7%),andsepsis(0.4%)(Brausietal.,2014;Lammetal.,1992).The International BladderCancer Group recommends deferring BCG instillation in the case of traumatic catheterization, concurrent urinarytractinfection,andTURinthepast2weeks,sincetheserisk factorsmayfacilitatehaematogenousdissemination(Witjesetal., 2008).Inarandomizedcontrolledtrialcomparingone-thirddoseto full-doseBCGfor1or3years,theincidenceofsideeffectsdidnot differbetweenthegroups(Brausietal.,2014).

Itisunclearwhetherinflammationorinfectionunderliesthese BCG-relatedcomplications,andtheirrelativeimportancemayvary betweenpatients.Sinceadiagnostictooltomakethisdistinctionis

* Correspondingauthorat:DepartmentofInfectiousDiseases,C5P-40,Leiden UniversityMedicalCentre,Albinusdreef2,2333ZALeiden,TheNetherlands.

** Correspondingauthorat:DepartmentofInfectiousDiseases,K05014A,Leiden UniversityMedicalCentre,Albinusdreef2,2333ZALeiden,TheNetherlands.

E-mailaddresses:s.a.joosten@lumc.nl(S.A. Joosten),s.m.arend@lumc.nl (S.M. Arend).

1 Present address: Department of Internal Medicine, Haaglanden Medisch Centrum,TheHague,TheNetherlands.

2 Presentaddress:DepartmentofInternalMedicine,AcademicMedicalCentre, Amsterdam,TheNetherlands.

https://doi.org/10.1016/j.ijid.2018.05.006

1201-9712/©2018TheAuthor(s).PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

ContentslistsavailableatScienceDirect

International Journal of Infectious Diseases

j o u r n a l h o m ep a g e : w w w . e l s e v i e r . c o m / l o c a te / i j i d

currentlynot available, treatment withtuberculostatic drugs is frequentlyinitiatedandanti-inflammatorydrugsareaddedinthe caseofsystemicdisease,withtheriskofsideeffects.

ThisreportdescribesfourcasesillustratingthespectrumofBCG disease.Intwopatientswithasystemicillness,immunologicalassays includinganovelmycobacterialgrowthinhibitionassay(MGIA)were performed,withtheaimofstudyingtheindividualpathogenesis.

Casereports

Theclinicalcharacteristicsofthepatientsaredescribedindetailin Table1.Inshort,patient1wasa66-year-oldmancomplainingof

anejaculation,dysuria,andfrequencyat1monthafterthesixthBCG instillation.Apelviccomputedtomography(CT)scanshowedprostate abnormalities(Figure1A),withchronicgranulomatousinflammation inabiopsy(Figure2).Thepatientwastreatedwithisoniazidand rifampicinfor6months,resultingintheresolutionofsymptoms.

Patient 2, a 70-year-old man, presented with weight loss, malaise, and abdominal pain at 2 months after the 18th BCG instillation. A CT scan revealed a leaking aneurysm of the abdominal aorta, which was repaired via an endovascular approach.Bloodcultureswerenegative(nomycobacterialculture was performed). After10months, a follow-up CTscan showed progression of the aneurysm with leakage (Figure 1B) and

Table1

Overviewofclinicalcharacteristics,diagnosticandtherapeuticapproaches,treatment,andoutcomesforfourpatientswithBCGdisease.

Patient number Age (years)/

sex

Symptoms Underlyingcondition(s) TypeofBCGcomplication Number ofTURs

TimebetweenlastTUR andfirstBCGdose (days)

Numberof instillations

TimebetweenlastBCGdoseand onsetofsymptoms(days)

1 66/M

Anejaculation Dysuria Frequency

COPD

Partialnephrectomy (renalcellcarcinoma)

Granulomatousprostatitis andseminalvesiculitis

1 47 6 30

2 70/M

Weightloss Malaise Abdominal pain

Hypertension EVAR

Mycoticaneurysm 2 21 18 60

3 67/M

Fever (39.2^C) Weightloss Malaise Feverafter secondBCG

Norelevanthistory BCGvaccination

Systemicreactionwithout organinvolvement

2 23 1 4

4 77/M

Fever (39.5^C) Cough Dyspnoea

COPD Gout Mitralvalve insufficiency

Pneumonitisandhepatitis 3 43 Sincelast

TUR:14 Total:59 Traumatic instillation

1

Patient number Age (years)/

sex

Laboratory results

Radiology Microbiology Histology Antibiotics Steroids Outcome

1 66/M

ESR25 PSA13.3 AST34 ALT49 ALP66 GGT43

PelvicCTscan:

Inhomogeneous enlargedprostatewith hyperdensityofright seminalvesicle

Urineculture negativefor commonbacteria andBCG

Chronic granulomatous inflammationin prostate

INH300mg/

day RIF600mg/

day (6months)

None ResolutionofBCGdisease Cystoprostatectomydueto tumourprogression

2 70/M

ESR33 AST52 ALT31 ALP234 GGT154

CTscan:

Progressionofaneurysm withtype1endoleak Para-aortic lymphadenopathy Splenomegaly

PCRonresected (para)aortictissue positiveforM.bovis

Chronic granulomatous inflammationinaortic wallandlymphnodes

None None Deathduetoaorticrupture1 weekafteraorticreconstruction, lymphnoderesection,and splenectomy

3 67/M

ESR103 AST32 ALT54 ALP75 GGT39

CTscan(day20after onsetofsymptoms):

Norelevantfindings

Urineandblood culturesnegativefor commonbacteria andBCG

ND Ciprofloxacin

500mgtwicea day

(2weeks)

Prednisone 15mg/day withtapering schedule (2months)

ResolutionofBCGdisease TUR+intravesicalKLHdueto recurrence,2yearsrecurrence- free

4 77/M

CRP48 AST67 ALT58 ALP153 GGT132

CTscan:

Miliarylungpattern Hepatomegaly Perihilar lymphadenopathy

Urineandblood culturesnegativefor commonbacteria andBCG

ND INH300mg/

day RIF600mg/

day ETH1600mg/

day (1month,due to

hepatotoxicity)

Prednisone 20mg/day withtapering schedule (6weeks)

ResolutionofBCGdisease IntravesicalKLH→recurrence anddeathafter2years

ALP(inU/l),alkalinephosphatase;ALT(inU/l),alanineaminotransferase;AST(inU/l),aspartateaminotransferase;BCG,bacillusCalmette–Guérin;COPD,chronicobstructive pulmonarydisease;CRP(inmg/l),C-reactiveprotein;CT,computedtomography;ESR(inmm),erythrocytesedimentationrate;ETH,ethambutol;EVAR,endovascular aneurysmrepair;GGT(inU/l),gamma-glutamyltransferase;INH,isoniazid;KLH,keyholelimpethemocyanin;M,male;M.bovis,Mycobacteriumbovis;ND,notdone;PSA(in mg/l),prostate-specificantigen;RIF,rifampicin;TUR,transurethralresection.

splenomegaly. The patient underwent an aortic repair with resection of the infected tissue, para-aortic lymph nodes, and thespleen. Histologyshowedchronicgranulomatousinflamma- tion,andPCRontheresectedtissuewaspositiveforMycobacterium bovis.However,beforeantibiotictreatmentcouldbeinitiatedthe patientdiedduetoaorticrupture.

Patient3,a67-year-oldmanwhohadbeenvaccinatedwithBCG inthepast,developedapersistenthigh-gradefever,weightloss, and malaiseat4daysafterhisfirstBCGinstillation.Pulmonary involvementwasruledoutandliverenzymeswerewithinnormal limits. The fever persisted despite ciprofloxacin and only disappearedafterprednisonewasstarted.Thehighfeverrecurred after the second BCG instillation and BCG treatment was discontinued. After a renewed TUR and intravesical keyhole limpethemocyanin(KLH)instillationsforlocaltumourrecurrence, thepatienthasbeenrecurrence-freefor2years.

Patient4,a77-year-old-man,presentedwithahigh-gradefever, cough, and dyspnoea at 1day after traumatic BCG instillation followingmanyuneventfulpreviousinstillations.ACT scanrevealed amiliarylungpattern,whileelevatedliverenzymesandhepato- megaly indicatedhepaticinvolvement(Figure1C,D). Treatment with isoniazid,rifampicin,and ethambutolwas initiated,buthis symptomsfirstresolvedafterprednisonewasadded2weekslater.

The antibioticswere discontinuedafter1 month dueto serious hepatotoxicity and were not reintroduced given the favourable clinicalcourse.KLHinstillationswere givenforrecurrenceanda nephroureterectomywasperformedaftertumourprogression.The patientdied2yearslaterfrommetastasizeddisease.

Immunologicalassays

Immunologicalassayswereperformedforpatients3and4who bothhadearlyonsetsystemiccomplications.Thestudyprotocol (P07.048)allowingexploratoryresearchofimmuneresponsesin Figure1.(A)PelvicCTscan(coronalplane)ofpatient1,displayinganinhomogeneous,enlargedprostate(P)andasymmetryoftheseminalvesicleswithhyperdensityofthe rightseminalvesicle(R)comparedtotheleftseminalvesicle(L).(B)TransverseabdominalCTscanofpatient2showingleakageofcontrastfluid(arrow)outsidethe endovascularaorticstent.(C)AbdominalCTscan(coronalplane)ofpatient4showinghepatomegaly(craniocaudaldiameter22cm)andperihilarlymphadenopathy(white arrow).(D)High-resolutionthoracicCTscan(maximumintensityprojection)ofpatient4demonstratingabilateralmiliarypatternwithmultiplesub-millimetrepulmonary nodules.

Figure2.Prostatebiopsyofpatient1showingprostaticglandssurroundedbya fibromuscularstromaandgranulomatousinflammationwithmultiplegiantcells (arrow).

Figure3. (A)ProliferationofT-lymphocytesinresponsetoMycobacteriumtuberculosispurifiedproteinderivative(PPD)orcontrol.Lymphocyteproliferationismeasuredas thecountperminute(cpm;³H-thymidineincorporation)forthreedifferentantigen(PPD)concentrations(mg/ml).Barsaremedianswiththestandarddeviationofthree measurements.Forpatient3,t=1isday25afterBCGinstillation(beforeinitiatingprednisone)andt=2isday109(followingthecessationofprednisone).Forpatient4,t=1is day57afterBCGinstillationaftertaperingofprednisone(dose2.5mgeveryotherday),t=2isday138(noprednisone),andt=3isday216(noprednisone).(B)LevelsofIFN-g productionbyperipheralbloodmononuclearcellsinresponsetoPPDorcontrolmedium;fortimepoints,seethecaptionforFigure3A.(C)Mycobacterialgrowthinhibition assay(MGIA):thecapacitytocontrolBCGgrowthisrepresentedaslog(CFU),ascalculatedfromthetimetopositivityoftheculture.Patients3and4arecomparedtofive

mycobacterialdiseasewas approvedbytheInstitutionalReview BoardofLeidenUniversityMedicalCentre,andwritteninformed consentwasobtainedfromthepatients.

Thelymphocytestimulationassaywasperformedasdescribed previously (Arend et al., 2000). In short, peripheral blood mononuclearcells (PBMC)were incubatedwith Mycobacterium tuberculosis purified protein derivative (PPD) (Statens Serum Institut, Copenhagen, Denmark) at 37^C in a humidified CO2

incubator. After 6 days, the supernatants were harvested for analysisbyinterferon-gamma (IFN-

g

FortheMGIA,PBMC(110⁶)wereincubatedwith100colony- formingunits(CFU)ofBCGonarotatorat37^Cinahumidified CO2 incubator. After 4 days, cultures were transferred to MycobacteriaGrowthIndicator Tubes(MGIT;BectonandDick- inson,FranklinLakes,NJ,USA)andplacedinaBACTECincubator untilgrowthwasdetected(Tanneretal.,2016).Themechanismof the control of BCG outgrowth has been published recently (Joostenetal.,2018).

Testresultsforpatient3,whohadahigh-gradefeverwithout organinvolvement,showedamoderatemycobacterial antigen- specificT-cellresponse,butenhancedinvitrogrowthinhibition ofBCGcomparedtohealthycontrols.Incontrast,thetestresults forpatient4,whopresentedwithsepsisandorganinvolvement, were characterized by a higher mycobacterial antigen T-cell responseassociatedwithalackofinvitrogrowthinhibitionof BCG(Figure3).

Discussion

Thebroadrangeofpossiblecomplicationsafterintravesical BCG instillation,illustrated by these four patient cases, has been described previously (Perez-Jacoiste Asin et al., 2014).

Theyvaryfromfocaldiseaselimitedtothegenitourinarytract, e.g.,granulomatous prostatitis (patient 1), cystitis, orepidi- dymo-orchitis,tofocaldiseaseoutsidethegenitourinarytract, e.g., arthritis, mycotic aneurysm (patient 2), or uveitis, to systemicreactionwithoutorganinvolvement(patient3),and finallytofull-blownsepsiswithpneumonitisand/orhepatitis (patient4).

Inviewofthediscussionregardingtherelativecontributionof inflammation andinfectiontothe pathogenesisof BCG-related complications,thefindingthatacid-faststaining,mycobacterial culture,andPCRoftenremainnegative(asfoundinthreeofthe fourpatientspresentedhere),favourstheinflammationhypoth- esis(Perez-Jacoiste Asinet al.,2014).Interestingly,bothofthe patientswithsystemicBCGdiseasedidnotrespondtoantibiotic therapyandsymptomsonlyresolvedafteraddingprednisone.A good outcome in patients with systemic disease treated exclusively with corticosteroids has been reported previously (LeMenseandStrange,1994).However,increasedmortalitywas seenaftertreatmentwithprednisonealoneinamousemodelof intraperitonealBCGinfection(Koukoletal.,1995),indicatingthat antibiotic therapy should not be withheld if infection is the predominantpathogenic mechanism. A randomizedcontrolled trial comparing intravesical BCG with and without isoniazid prophylaxisaroundthetimeofinstillationsshowednodifference in the incidence of focal or systemic complications, further supportinginflammationasanimportantpathogenicmechanism (Vegtetal.,1997).

IntravesicalBCGresultsinmarkedleukocyturia,theinfluxof immunecellsintothebladderwall,andthereleaseofcytokines intheurine,whichismorepronouncedinthecaseofprevious BCGexposure,reflectingrecallimmuneactivation(Redelman- Sidi et al., 2014). On the other hand, ofloxacin given after intravesicalBCGsignificantlyreducedtheincidenceofcompli- cationscomparedtoplacebo(Colombeletal.,2006).Together withthemicrobiologicalidentificationofM.bovisinsomecases, predominantly with focal disease, this argues in favour of infection.

Inagreementwithpreviouslargecaseseriesstudies(Gonzalez etal.,2003;Perez-JacoisteAsinetal.,2014),theintervalbetween the last instillation and onset of symptoms was shorter for patients3and4withsystemiccomplicationsthanforpatients1 and 2with focaldisease. In the caseseries by Gonzalez et al.

(2003), a late presentation with mostly focal disease was microbiologically confirmed in two-thirds of the patients, comparedtoonlyone-thirdof thepatientswithearlysystemic disease.Thissuggeststhatacuteinflammatorymechanismsplay arelativelylargerroleinthedevelopmentofsystemicreactions, whereas focal disease is mainly caused by infection-induced chronicinflammation.

Immunetestsshowedthatmycobacterialantigen-specificT- cell responses were highest in patient 4, who presented with sepsisandorganinvolvement.Intriguingly,thebloodcellsofthis patienthadlimitedcapacityforinvitrogrowthinhibitionofBCG, while patient 3, who had fever in the absence of organ involvement,showedenhancedinvitro BCGgrowthinhibition.

Whilenoconclusionscanbedrawnfromonlytwo cases,these immunologicalassaysmightprovideaclueaboutthepathogen- esisofBCGcomplicationsinindividualpatients.Highmycobac- terial antigen-specificT-cellresponseswouldpointtosystemic inflammation as the predominant mechanism, in accordance withthebeneficialeffectofsteroids.Inpatientswithlowinvitro growth inhibition of BCG, infection might be more important, explainingwhythesepatientsbenefitmorefromtuberculostatic drugs,althoughthedurationof suchtreatmentisdebatable,as illustratedbypatient4.

A preliminarymodel forthemechanism of actionof BCGin bladdercancerhasbeenproposed(Redelman-Sidiet al.,2014), consistingofinternalizationofliveBCGbymalignanturothelial cellsfollowedbyMHC-IIup-regulationininfectedbladdercancer cells and the presentation of a BCG- and/or tumour-specific antigen to T-cells. This leads to T-cell-mediated cytotoxicity specific to malignant urothelial cells. Since the patient with enhancedinvitrogrowthinhibitionofBCGisnowrecurrence-free whilethepatientwithlimitedinvitrogrowthinhibitiondieddue to tumour progression, one could speculate that differences betweenindividualsinthehandlingofliveBCGbyimmunecells andperhapsalsomalignanturothelialcellsmightberelatedtothis observation.Thiswarrantsfurtherresearch.

In conclusion, the complications of intravesical BCG are manifold and the contribution of inflammation and infection variesbetweenpatients.Immunologicalassayshavepotentialto provide insight into the pathogenesis of BCG disease. Future studiesshouldalsoexploreimmunologicalmechanismsinvolved intumourcontrolfollowingBCGinstillations.

Financialsupport None.

healthynon-BCGvaccinatedorM.tuberculosis-exposed,PPD-negativecontrols(fortimepoints,seethecaptionforFigure3A).EffectiveinvitrogrowthinhibitionofBCG correspondswithlowlog(CFU).MGIAwerealwaysperformedinduplicate.Forcontrols,eachdotrepresentsthemeanofduplicatevalues.Forpatients3and4,duplicatesare shownasseparatedots.Thelinesrepresentthemedianvalue.

Conflictofinterest

RP has no conflicts of interest, including specific financial interests,relevanttothesubmittedwork.Hisfinancialactivities outsidethesubmittedworkincludebeingontheadvisoryboardof JanssenBiologicsBVandAmgenEuropeBV,andprovidingguest lectures for AstraZeneca Netherlands. TO has no conflicts of interest, including specific financial interests, relevant to the submitted work. His financial activities outside the submitted workincludebeingamemberoftheseniormanagementteamof the TB Vaccine Initiative (TBVI; www.tbvi.eu), receiving many externalresearchgrants(governmental,EuropeanCommission), reimbursement for travel costs for attending meetings, and holdingtwoEUpatents.Allotherauthorsreportnoconflictsof interest.

References

ArendSM,GelukA,vanMeijgaardenKE,vanDisselJT,TheisenM,AndersenP,etal.

AntigenicequivalenceofhumanT-cellresponsestoMycobacteriumtuberculosis- specificRD1-encodedproteinantigensESAT-6andculturefiltrateprotein10 andtomixturesofsyntheticpeptides.InfectImmun2000;68:3314–21.

BrausiM,OddensJ,SylvesterR,BonoA,vandeBeekC,vanAndelG,etal.Sideeffects ofBacillusCalmette-Guerin(BCG)inthetreatmentofintermediate-andhigh- riskTa,T1papillarycarcinomaofthebladder:resultsoftheEORTCgenito- urinarycancersgrouprandomisedphase3studycomparingone-thirddose with full dose and 1year with 3 years of maintenance BCG. Eur Urol 2014;65:69–76.

ColombelM,SaintD,ChopinB,NicolasL,RischmannP.Theeffectofofloxacinon bacilluscalmette-guerininducedtoxicityinpatientswithsuperficialbladder

cancer:resultsofarandomized,prospective,double-blind,placebocontrolled, multicenterstudy.JUrol2006;176:935–9.

GonzalezOY, MusherDM,Brar I,FurgesonS,BoktourMR,SeptimusEJ,etal.

SpectrumofbacilleCalmette-Guerin(BCG)infectionafterintravesicalBCG immunotherapy.ClinInfectDis2003;36:140–8.

JoostenSA,vanMeijgaardenKE,ArendSM,PrinsC,OftungF,KorsvoldGE,etal.

Mycobacterialgrowthinhibitionisassociatedwithtrainedinnateimmunity.J ClinInvest2018;128:1837–51.

KoukolSC,DeHavenJI,RiggsDR,LammDL.DrugtherapyofbacillusCalmette- Guerinsepsis.UrolRes1995;22:373–6.

LammDL,vanderMeijdenPM,MoralesA,BrosmanSA,CatalonaWJ,HerrHW,etal.

IncidenceandtreatmentofcomplicationsofbacillusCalmette-Guerinintra- vesicaltherapyinsuperficialbladdercancer.JUrol1992;147:596–600.

LeMenseGP,StrangeC.GranulomatouspneumonitisfollowingintravesicalBCG.

Whattherapyisneeded?.Chest1994;106:1624–6.

Perez-JacoisteAsinMA,Fernandez-RuizM,Lopez-MedranoF,LumbrerasC,Tejido A, San Juan R, et al. Bacillus Calmette-Guerin (BCG) infection following intravesical BCG administrationasadjunctive therapy for bladdercancer:

incidence,riskfactors,andoutcomeinasingle-institutionseriesandreviewof theliterature.Medicine(Baltimore)2014;93:236–54.

Redelman-SidiG,GlickmanMS,BochnerBH.ThemechanismofactionofBCG therapyforbladdercancer–acurrentperspective.NatRevUrol2014;11:153–62.

ShelleyMD,MasonMD,KynastonH.Intravesicaltherapyforsuperficialbladder cancer:asystematicreviewofrandomisedtrialsandmeta-analyses.Cancer TreatRev2010;36:195–205.

TannerR,O’SheaMK,FletcherHA,McShaneH.Invitromycobacterialgrowth inhibition assays: a tool for the assessment of protective immunityand evaluationoftuberculosisvaccineefficacy.Vaccine2016;34:4656–65.

VegtPD,vanderMeijdenAP,SylvesterR,BrausiM,HoltlW,deBalincourtC.Does isoniazidreducesideeffectsofintravesicalbacillusCalmette-Guerintherapyin superficial bladder cancer? Interim results of European Organization for ResearchandTreatmentofCancerProtocol30911.JUrol1997;157:1246–9.

WitjesJA,PalouJ,SolowayM,LammD,BrausiM,SpermonJR,etal.Clinicalpractice recommendationsforthepreventionandmanagementofintravesicaltherapy- associatedadverseevents.EurUrolSuppl2008;7:667–74.