Evaluation of resistance acquisition during tuberculosis treatment using whole genome sequencing

brazjinfectdis2016;20(3):290–293

w w w . e l s e v ie r . c o m / l o c a t e / b j i d

The

Brazilian

Journal

of

INFECTIOUS

DISEASES

Brief

communication

Evaluation

of

resistance

acquisition

during

tuberculosis

treatment

using

whole

genome

sequencing

Cinara

Silva

Feliciano

_,

_Jessica

_Rodrigues

_Plac¸a

_,

_Kamila

_Peronni

_,

Wilson

Araújo

Silva

Jr

_,

_Valdes

_Roberto

_Bollela

a_{DepartmentofInternalMedicine,FaculdadedeMedicinadeRibeirãoPreto,UniversidadedeSãoPaulo(USP),RibeirãoPreto,SP,Brazil}

b_{DepartmentofGenetics,FaculdadedeMedicinadeRibeirãoPreto,UniversidadedeSãoPaulo(USP),RibeirãoPreto,SP,Brazil}

c_{CenterforMedicalGenomics,HospitaldasClínicasdaFaculdadedeMedicinadeRibeirãoPreto,UniversidadedeSãoPaulo(USP),}

RibeirãoPreto,Brazil

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received28October2015

Accepted5January2016

Availableonline20March2016

Keywords: Tuberculosis

Drugresistanttuberculosis

Drugsusceptibilitytests

Wholegenomesequencing

a

b

s

t

r

a

c

t

Tuberculosis(TB)isstillconsidered amajorglobal publichealth problemintheworld

andthereisaconcernabouttheworldwideincreaseofdrug-resistance(DR).Thispaper

describesthe analysisofthree Mycobacteriumtuberculosisisolates froma singlepatient

collectedoveralongtreatmentperiodoftime.DRwasinitiallyinvestigatedthrough

pheno-typictesting,followedbylineprobeassays(LPAs)andwholegenomesequencing(WGS).It

presentsanintriguingsituationwhereamultidrug-resistant(MDR-)TBcasewasdiagnosed

andtreatedbasedonlyonlatephenotypicdrugsusceptibilitytestingofisolate1.During

thetreatment,anothertwoisolateswerecultivated:isolate2,ninemonthsafterstarting

MDR-TBtreatment;andisolate3,cultivatedfivemonthslater,duringregularuseofanti-TB

drugs.ThesetwoisolateswereevaluatedusingmolecularLPAandWGS,retrospectively.All

mutationsdetectedbyLPAwerealsodetectedintheWGS,includingconversionfrom

fluo-roquinolonessusceptibilitytoresistancefromisolate2toisolate3.WGSshowedadditional

mutations,includingsomewhichmayconferresistancetoother drugsnottested

(ter-izidone/cycloserine)andmutationswithnocorrespondentresistanceindrugsusceptibility

testing(streptomycinandsecond-lineinjectabledrugs).

©2016ElsevierEditoraLtda.ThisisanopenaccessarticleundertheCCBY-NC-ND

license.(http://creativecommons.org/licenses/by-nc-nd/4.0/)

Tuberculosis (TB) is still considered a major global public

health problem, and Brazil is the 16th country in

abso-lutenumberofcases.Despiteconsistentadvancesachieved

with control measures, there are still major challenges to

face the growing resistance to anti-tuberculosis drugs in

severalcountries,includingBrazil.1_{Molecularepidemiology}

∗ _{Correspondingauthor.}

E-mailaddress:vbollela@gmail.com(V.RobertoBollela).

studiesofM.tuberculosishavegainedemphasisamong

clini-calresearchers.NewknowledgeontheTBpathogenesisand

its causativeagentcould bethekeytothe developmentof

newcontrolstrategies.2,3_{Recently,theWorldHealth}

Organi-zationsettheGlobal“StopTBPlan”tofindpatientsharboring

resistant strains ofM. tuberculosis. This initiative has been

http://dx.doi.org/10.1016/j.bjid.2016.01.004

1413-8670/©2016ElsevierEditoraLtda.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense.(http://creativecommons.org/

brazj infect dis.2016;20(3):290–293

291

importanttotestfirst-lineTBdrugsandpromotingresearch

todevelopnewdrugs,vaccines,anddiagnosticstrategies.1,4

Amongthetechniquesused,wholegenomicsequencing

(WGS)isworthmentioning,especiallyfortheinvestigationof

bacilliresistance.Thisapproachenablesaccurateassessment

ofmutationsrelatedtobacilliresistancetoseveralTBdrugs.

Thisinformationwouldallowthedevelopmentofnew

diag-nosticmethodsandtherapeuticstrategiesappliedfordisease

control.5

Outhredetal.advocatetheuseofWGSforall

multidrug-resistant M. tuberculosis isolates as an alternative plan to

improvepatientcare,monitorfortransmissionevents,and

contributetobetter understandingofresistance-associated

mutations.6

Thisstudydescribesachallengingcaseoftreatment

fail-ureofapatientunderMDR-TBtherapyandreportsphenotypic

and molecular drug resistance test results in correlation

to mutations identified with a whole genome sequencing

analysis.

ThreeisolatesofM.tuberculosisfromasinglepatient,

dur-ing different time points of his treatment, were evaluated

basedonphenotypingandgenotypingtesting:isolate1was

collectedbeforethepatientstartedthefollow-upinthe

ref-erencecenter,whenclinicalandmicrobiologicalfailurewas

diagnosed,despiteregularTBtreatmentwithrifampicin(R),

isoniazid(H),pyrazinamide(Z)andethambutol(E);isolate2

wascollectedintheninthmonthofMDR-TBtreatment(other

clinicalandmicrobiologicalfailure);andisolate3wasobtained

onthe13thmonthofMDR-TBtreatment.

A nonradiometric phenotypic susceptibility testing was

performed in liquid medium (MGIT 960; Becton Dickinson

DiagnosticSystems, Sparks, MD) forisolate 1.Besides this

susceptibilityphenotypictesting,twolineprobeassays(LPA),

GenotypeMTBDRplusandMTBDRsl(HainLifescience,GmbH,

Germany), were performed in isolates 2 and 3. Genotype

MTBDRplus evaluates the mainmutations associated with

rifampicin(rpoBgenemutations)andisoniazid(katGeinhA

mutations)resistance.GenotypeMTBDRsldetectsmutations

relatedwithresistancetofluoroquinolones(gyrAgene

muta-tions),second-lineinjectabledrugs(SLID)(rrsgenemutations)

and ethambutol (embBgene mutations).7,8 _{Finally, thetwo}

storedsamples(isolates2and3)weresubmittedtoWGS

anal-ysisusing IlluminaMiSeqSequecingSystem(Illumina,San

Diego,CA,USA). LPAsand WGStestswereperformed with

storedisolates2and3attheendofthepatient’streatment.

Therefore,thisinformationwasnotavailabletotheclinician

duringthetreatment.Isolate1wasnottestedagainbecause

itwasnotviable.

Generated reads with phred scale score superior to 30

was mapped withBWA v0.7.5a program (Burrows-Wheeler

AlignmentTool) usingthereference genomeM.tuberculosis

H37Rv.Conversionfromsequencealignmentmapformatto

sorted,indexedBAMfileswasdoneusingSAMtools(version

0.1.19). PCR-duplicates were removed using the

MarkDu-plicates option of the Picard software tools (version 1.61).

The variants were found according to the pipeline

SAM-tools/BCFtools v 0.1.18 and annotated with SnpEff v 4.0.

DatabasesTBDrugResistanceMutationDatabase9_andM.tbDrug

ResistanceDirectedSequencingDatabase10_{wereusedtoidentify}

mutationsdescribedfortheTBbacilli.Alldetectedmutations

wereconfirmedbasedonTBprofileronlinetool,describedbyColl

etal.,11_{toremovesinglenucleotidepolymorphisms(SNPs)at}

drugresistancelociwhichwerehistoricallymisclassifiedas

drugresistancemarkers.

ThisprojectwasapprovedbytheEthicsandResearch

Com-mitteeoftheHospitaloftheRibeirãoPretoMedicalSchoolof

theUniversityofSãoPaulo(protocolnumber:944117–

Febru-ary1,2015).

ThepatientwasinitiallytreatedwithRHZEforsixmonths.

During the last month clinical and microbiological failure

(acid-fastsmearpositiveandrespiratorysymptoms)was

diag-nosed.ThisM.tuberculosisisolateshowedRandHresistance

inthefirstphenotypicsusceptibilitytesting.Thepatientwas

referred to the regional TBdrug resistance center and his

treatment was switchedto streptomycin (Sm),E, ofloxacin

(FQ), Z, and terizidone (Tz), following the Brazilian

guide-linesforthemanagementofMDR-TB.Duringthefirst nine

months of treatment, there was a transient improvement

followedbyrecurrenceofsymptoms.Sputumculturewas

col-lectedandthesameMDR-TBtreatmentwasmaintaineduntil

the 13thmonth.Thetherapy wasempiricallyoptimizedby

adding ethionamide, extending the period ofstreptomycin

andincreasingthelocalsupportfordirectlyobserved

treat-ment,whiletheresultsofphenotypictestingfromisolate3

werestillpending.Thetreatmentwassuccessfullycompleted

after24months,withclinicalandmicrobiologicalcure.After

that, instigatedbythis unusualandintriguingcaseand its

outcome,wedecided tocarryon molecularstudies on the

patient’sisolates2and3.

Isolate1showedresistancetoRandHinthephenotypic

test.Phenotypictestingofisolate2showedresistancetothese

twodrugsandalsotoZ.TheLPAofthisisolateshowed

resis-tancetoR(lostwildtype8andgainedrpoBS450Lmutation),

H (lostwildtypeandgained KatGS315T1mutation),andE

(lostwildtype1andgainedembBM306Vmutation).Isolate3

showedresistancetoR,I,E(samemutationsdescribedin

iso-late2),andacquiredanewresistancepatterntoFQ(gyrAD94G

mutation)intheLPAtest.Phenotypictestingofisolate3,which

becameavailableclosetotheendofthepatient’streatment,

showedresistancetotheabovementioneddrugsandalsoto

Z.ThecriticalconcentrationsofBactec-MGIT960TM_reported

bythemanufacturer’sdrugsusceptibilitytesting(DST)

proto-colwereasfollows:H:0.10␮g/mL;R:1.0␮g/mL;EM:5.0␮g/mL;

Sm:1.0␮g/mL;Z:100␮g/mL;FQ(ofloxacin):2␮g/mL;Amikacin

1.0␮g/mL;Capreomycin2.5␮g/mL.Thesusceptibilityprofile

ofisolatesinthe phenotypictestand LPA aredescribed in

Table1.

TheWGSanalysisofthesetwostrainsgeneratedthetotal

of43,036,497reads(28,171,267forstrain1and14,865,230for

strain2),withanaveragecoverageof469×fortheisolate2

and244× fortheisolate3.Bioinformaticsanalysisreported

11mutationsalreadydescribedasassociatedwithresistance

in isolate 2 and 12 mutationsin isolate 3 (Table 2).

Addi-tionalmutationingyrAgene(D94G)wasidentifiedinisolate

3, whichisoneofthe mostfrequentmutationsassociated

withresistancetoFQ,andistestedbytheLPAGenotype

MTB-DRsl.AlthoughallmutationsshowedbyLPAwerevalidated

byWGS,additionalmutationsweredetected,includingthose

conferringresistancetootherdrugsdespitebacilli

292

Table1–ResultsofphenotypicandLPAtests:susceptibilityprofileoftesteddrugs.

Isolate1detectedresistance Isolate2detectedresistance Isolate3detectedresistance Phenotypicdrugsusceptibilitytesting(DST) Rifampicin;Isoniazida _{Rifampicin;Isoniazid; Rifampicin;Isoniazid}

Pyrazinamideb _{Ofloxacin;Pyrazinamide}

Ethambutolc

GenotypeMTBDRplus(R;H) NA Rifampicin Rifampicin Isoniazid Isoniazid GenotypeMTBDRsl(E;FQ;SLID) NA Ethambutol Ethambutol

FQ

WGS NA Available(seeTable2) Available(seeTable2) NA,notavailable;R,rifampicin;H,isoniazid;E,ethambutol;FQ,fluoroquinolones;SLID,secondlineinjectabledrugs.

a _{Isolate1collectedbeforeMDR-TBtreatment–susceptibletoEandSm.SLIDnottested.} b _{Isolate2collectedintheninthmonthofMDR-TBtreatment–susceptibletoE,SLIDandFQ.} c _{Isolate3collectedonthe13thmonthofMDR-TBtreatment–susceptibletoSmandSLID.}

Table2–Wholegenomicsequencingmutationsidentifiedinisolates2and3.

Gene Associated drug Genomic position Mutation Strain2 mutation Strain3 mutation Reference allele Mutated allele Varianttype

gyrA Fluoroquinolones 7362 E21Q Yes Yes G C Missense

gyrA Fluoroquinolones 7582 D94Ga _{No Yes A G Missense}

rpoB Rifampicin 759939 P45T Yes Yes C A Missense rpoB Rifampicin 761155 S450L Yes Yes C T Missense rpsL Streptomycin 781395 – Yes Yes T C Promotergene tlyA Aminoglycosides 1917972 L11L Yes Yes A G Synonymous katG Isoniazid 2154915 E399E Yes Yes A G Synonymous katG Isoniazid 2155168 S315T Yes Yes G C Missense pncA Pyrazinamide 2289039 W68L Yes Yes G T Missense alr Cycloserine 3840719 L234L Yes Yes A G Synonymous alr Cycloserine 3841403 E6D Yes Yes G T Missense embB Ethambutol 4247429 M306V Yes Yes A G Missense

a _{Mutationdetectedonlyintheisolate3,bywholegenomicsequencing.}

ExceptformutationD94G,theSNPsreadsforisolate2and3were100%identical,withnosignsofheteroresistance.

(gyrAE21Qinisolate2,rpsLpromotergeneingenomicposition

781395,tlyAL11Lforisolates2and3).

Therelationbetweengenomemutationsandphenotypic

resistanceisparticularlyimportantfortherapeuticdecision,

because different mutationscan cause different resistance

profilesornotevencauseanyphenotypicresistance.12 _Coll

etal.compiledalibraryofmutationspredictiveofdrug

resis-tance, and removed phylogenetic SNPs at drug resistance

loci,whichwerehistoricallymisclassifiedasdrugresistance

markers.11

WGShasgreatapplicationpotentialtodetectbacilli

resis-tanceinclinicalpractice.However,furtherworkisneededto

determineadditionalresistancepolymorphismsasitshould

benotedthathighpositivepredictivevaluesarecrucialfor

drugresistancetestswheretheconsequenceofafalse

posi-tiveresultmayleadtounnecessarytreatmentandprolonged

patientisolation.11

TheWGSofM.tuberculosisisagreatadvanceinthe

knowl-edgeofbacilliresistance,aswellasintheclinicalmanagement

ofTB.Thistechniquepresentsgreaterdiscriminatorypower,

enablinganalysisofadditionalmutations,notpossibletobe

assessedbyothermethods.WGShasthepotentialforclinical

use,asmentionedbyotherauthors6,11–14 _forfastand

accu-rateassessmentsincasesofillnesscausedbystrainsresistant

tomultipledrugs,withanimpactontherapeuticdecisions.

However,thisisahigh-costtechnologyanditisstillcritical

to understand and standardizecorrelations between

geno-typicandphenotypicresistancetoreallyoptimizeitsclinical

use.

Conflicts

of

interest

Theauthorsdeclarenoconflictsofinterest.

r

e

f

e

r

e

n

c

e

s

1.WorldHealthOrganization(WHO).Globaltuberculosisreport 2014.Geneva:WorldHealthOrganization;2014.

2.GomesHM,EliasAR,OelemannMAC,etal.Spoligotypesof

Mycobacteriumtuberculosiscomplexisolatesfrompatients

residentsof11statesofBrazil.InfectGenetEvol.

2012;12(4):649–56,http://dx.doi.org/10.1016/j.meegid.

2011.08.027.Epub2011Sep1.

3.GarcíaDeViedmaD,MokrousovI,RastogiN.Innovationsin

themolecularepidemiologyoftuberculosis.EnfermInfecc

MicrobiolClin.2011;29Suppl1:8–13,http://dx.doi.org/10.1016/

brazj infect dis.2016;20(3):290–293

293

4. WorldHealthOrganization(WHO).TheglobalplantostopTB 2011–2015.Geneva:WorldHealthOrganization;2010.

5. IlinaEN,ShitikovEA,IkryannikovaLN,etal.Comparative genomicanalysisofMycobacteriumtuberculosisdrug resistantstrainsfromRussia.PLoSOne.2013;8: e56577.

6. OuthredAC,JefsP,SulimanB,etal.Addedvalueofwhole genomesequencingformanagementofhighlydrug-resistant TB.JAntimicrobChemother.2015;70(4):1198–202.

7. WorldHealthOrganization(WHO).Molecularlineprobe assaysforrapidscreeningofpatientsatriskofmulti-drug resistanttuberculosis(MDR-TB).Geneva:WorldHealth Organization;2008.

8. BrossierF,VezirisN,AubryA,etal.Detectionbygenotype MTBDRsltestofcomplexmechanismsofresistanceto second-linedrugsandethambutolinmultidrug-resistant Mycobacteriumtuberculosiscomplexisolates.JClinMicrobiol. 2010;48:1683–9.

9.SandgrenA,StrongM,MuthukrishnanP,etal.Tuberculosis drugresistancemutationdatabase.PLoSMed.2009;6:e2.

10.BroadInstitute.Availableat:http://www.broadinstitute.org/

annotation/genome/mtbdrugresistance.1/MultiHome.htm. 11.CollF,McNerneyR,PrestonMD,etal.Rapiddeterminationof

anti-tuberculosisdrugresistancefromwhole-genome sequences.GenomeMed.2015;7:51.

12.WitneyAA,GouldKA,ArnoldA,etal.Clinicalapplicationof whole-genomesequencingtoinformtreatmentfor multidrug-resistanttuberculosiscases.JClinMicrobiol. 2015;53:1473–83.

13.LiuF,HuY,WangQ,etal.Comparativegenomicanalysisof Mycobacteriumtuberculosisclinicalisolates.BMCGenomics. 2014;15:469.

14.ClarkTG,MallardK,CellF,etal.Elucidatingemergenceand transmissionofmultidrug-resistanttuberculosisintreatment experiencedpatientsbywholegenomesequencing.PLoS One.2013;8:e83012.