Performance of rapid tests for carbapenemase detection among Brazilian Enterobacteriaceae isolates

h ttp : / / w w w . b j m i c r o b i o l . c o m . b r /

Clinical

Microbiology

Performance

of

rapid

tests

for

carbapenemase

detection

among

Brazilian

Enterobacteriaceae

isolates

Lisiane

Rech

Pancotto

_,

_Carolina

_Silva

_Nodari

_,

_Franciéli

_Pedrotti

_Rozales

_,

Tatiane

Soldi

_,

_Carolina

_Gomes

_Siqueira

_,

_Ana

_Lúcia

_Freitas

_,

_Afonso

_Luís

_Barth

a_{HospitaldeClínicasdePortoAlegre,LaboratóriodePesquisaemResistênciaBacteriana(LABRESIS),PortoAlegre,RS,Brazil} b_{UniversidadeFederaldoRioGrandedoSul,ProgramadePós-Graduac¸ãoemCiênciasFarmacêuticas,PortoAlegre,RS,Brazil} c_{UniversidadeFederaldoRioGrandedoSul,FaculdadedeFarmácia,PortoAlegre,RS,Brazil}

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received9April2017 Accepted7July2018

Availableonline16August2018 AssociateEditor:AnaLuciaDarini

Keywords: Carbapenemases Rapidtests CarbaNP Blue-Carba CIM

a

b

s

t

r

a

c

t

Theglobalemergenceofcarbapenemasesledtotheneedofdevelopingnewmethodsfor theirrapiddetection.Theaimofthisstudywastoevaluatetheperformanceoftherapid tests for carbapenemase-producing and non-producing Enterobacteriaceae. Carbapenem non-susceptibleEnterobacteriaceaefromasurveillancestudysubmittedtoamultiplexreal timePCRforcarbapenemasedetectionwereincludedinthisstudy.Theisolateswere sub-jectedtotherapidphenotypictestsCarbaNP,Blue-CarbaandCarbapenemInactivation Method(CIM).Atotalof83carbapenemase-producing(43)andnon-producing(40)isolates wereincludedinthestudy.Thesensitivity/specificitywere62.7%/97.5%,95.3%/100%,and 74.4%/97.5%forCarbaNP,Blue-CarbaandCIM,respectively.BothCarbaNPandBlue-Carba presentedtheirfinalresultsafter75minofincubation;thefinalresultsforCIMwereobtained onlyafter8h.FailuretodetectOXA-370carbapenemasewasthemainproblemforCarbaNP andCIMassays.AstheBlue-Carbapresentedthehighestsensitivity,itcanbeconsideredthe bestscreeningtest.Conversely,CIMmightbetheeasiesttoperform,asitdoesnotrequire specialreagents.Theearlydetectionofcarbapenemasesaidstoestablishinfectioncontrol measuresandpreventcarbapenemasestospreadreducingtheriskofhealthcareassociated infectionsandtherapeuticfailure.

©2018SociedadeBrasileiradeMicrobiologia.PublishedbyElsevierEditoraLtda.Thisis anopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/

licenses/by-nc-nd/4.0/).

Introduction

Nowadays carbapenem resistance among Enterobacteriaceae

isamajorconcern.Thesebacteriabecomeresistant tothis

∗ _{Correspondingauthor.}

E-mail:albarth@hcpa.edu.br(A.L.Barth).

antibioticclassmainlyduetothreemechanisms,either iso-lated or combined: efflux pumps; permeability loss; and carbapenemaseproduction.1_{Thelatterrepresentsagrowing}

concern,sincetheseenzymeshydrolyzenotonly carbapen-ems but also other ␤-lactams and their dissemination is usuallyplasmid-mediated.2

Carbapenemases can beclassified as classA, Bor D ␤-lactamases,3 _{based on their molecular structure.In Brazil,}

differentcarbapenemaseclasseshaveemergedovertheyears

https://doi.org/10.1016/j.bjm.2018.07.002

1517-8382/©2018SociedadeBrasileiradeMicrobiologia.PublishedbyElsevierEditoraLtda.ThisisanopenaccessarticleundertheCC BY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

buttheKlebsiellapneumoniaecarbapenemase(KPC)isalready consideredendemic.2_{Wehavealsoobservedthespreadofthe}

NewDelhiMetallo-␤-lactamase(NDM)inrecentyears4,5_and

theemergenceofOXA-370insomeregions.5_Moreover,

scat-teredreportsofGES-5andIMP-1havebeendescribedoverthe country.6,7

Theglobalemergenceofcarbapenemasesbroughttolight theneedofdevelopingnewmethodsforitsrapiddetection. Theenzymaticactivityofcarbapenemasesisexploredasan alternative fortheirdetection insomerecent studies.This activitymightbedetectedbybiochemicalassayssuchasCarba NPandBlue-Carbamethods,8,9_{inwhichthehydrolysisofthe}

␤-lactamringisvisualizedbythecolourchangeofapH indi-cator,orusing massspectroscopy inwhichthe absence of thecarbapenemmolecular ionisobservedinthe presence ofacarbapenemase-producingbacteria.10,11_VanderZwaluw

etal.alsoproposedamethodology(theCarbapenem Inactiva-tionMethod–CIM)inwhichacarbapenem-susceptibleE.coli

strainisabletogrownearameropenemdiskafterits incuba-tionwithacarbapenemase-producingstrain.12_Recently,the

Clinical&LaboratoryStandardsInstitute(CLSI)proposedthe useofCarbaNPandtheCIMassaysasscreeningmethodsfor carbapenemasedetection.13

Theaimofthis study was toevaluatethe performance of the rapid tests Carba NP, Blue-Carba and CIM against carbapenemase-producingand-non-producing Enterobacteri-aceae.

Material

and

methods

Bacterialisolates

Carbapenemnon-susceptibleEnterobacteriaceae, from a pre-vious surveillance study, were identified by biochemical characterization and, when necessary, confirmed using VITEK2system(BioMérieux,France).Isolatesresistantto,at least, imipenem and/or meropenem were selectedfor this study.

Genotypicdetectionofcarbapenemases

Carbapenemasegenesweredetectedbyamultiplexreal-time PCRwithspecificprimersfortheblaKPC,blaGES,blaIMP,blaNDM,

blaVIM and blaOXA-48-like.14 PCRproducts ofisolatespositive

forblaGES and blaOXA-48-like were purifiedusing the ExoStar

kit(GEHealthcare)andsequencedusingaBigDye Termina-torkit(version3.1)andanABI3500GeneticAnalyzer(Applied Biosystems).

Phenotypicdetectionofcarbapenemases

CarbaNP

Thecarbapenemase detection byCarba NP wasperformed andinterpretedaspreviouslydescribedbyNordmannetal.8

Briefly,overnightculturedcoloniesgrowninMuellerHinton agarweresubmittedtoextractionwithB-PERII(Bacterial Pro-teinExtractionReagent, ThermoScientificPierce,Rockford, USA)andincubatedfor2hinasolutioncontainingphenol-red andimipenem.Thebacterialextractwasalsoincubatedina

phenol-redsolutionwithoutantibiotic.Theresultwas consid-eredpositivewhenthesolutioncontainingimipenembecame orangeoryellow.Theresultwasnotvalidatedifthesolution withoutantibioticpresentedacolourchangemoreexpressive thanthesolutioncontainingimipenem.AknownKPC-2 pro-ducerwasusedaspositivecontrol,andE.coliATCC25922was testedasnegativecontrol.

Blue-Carba

CarbapenemasedetectionbyBlue-Carbawasperformedand interpreted as previously described by Pires et al.9 _Briefly,

overnightculturedcoloniesgrowninMuellerHintonagarwere incubatedfor2hinasolutioncontainingbromothymolblue andimipenem.Thebacterialcolonieswerealsoincubatedina bromothymolbluesolutionwithoutantibiotic.Theresultwas consideredpositivewhenthesolutioncontainingimipenem becamegreenoryellow,anditscolorwasdifferentfromthe oneobservedinthenegativecontrol.Theresultwasnot vali-datedifthesolutionwithoutantibioticpresentedthesameor strongercolourchangeasthesolutioncontainingimipenem. AknownKPC-2producerwasutilizedaspositivecontrol,and atesttubecontainingonlybacteriainoculumandBlue-Carba solutionwasusedasnegativecontrolforeachisolatetested.

Carbapenemaseinhibitionmethod(CIM)

TheCIMwasperformedaspreviouslydescribedbyvander Zwaluwetal.12_{Briefly,overnightculturedcolonieswere}

incu-batedfor2hwithameropenemdiskin400␮Lofwater,with zinc sulfate supplementation for metallo-beta-lactamase-producingisolates.Aftertheincubation,thediskwasplaced onaMueller-Hintonagarplatepreviouslyinoculatedwitha susceptibleE.coli(ATCC25922)andincubatedat35◦Cupto thegrowthofthestrain(around6h).Theresultwas consid-eredpositive whentheE. colipresentedaninhibitionzone ≤16mmaroundthemeropenemdisk,accordingtotheCLSI interpretation criteria.13 _{Aknown KPC-2 producerwas}

uti-lizedaspositivecontrol,andE.coliATCC25922wastestedas negativecontrol.

Results

Atotal of83carbapenem-non-susceptible Enterobacteriaceae

isolates were selected for this study: 12 KPC, 13 NDM, 12 OXA-370, four GES-5, and two IMP producers, as well as 40 carbapenemase non-producers. All isolates were sub-mitted to the rapid tests Carba NP, Blue-Carba and CIM. The results for carbapenemase-producing isolates, as well as the two carbapenemase-non-producing isolateswith at least one false positive result, are presented in Table 1. The38carbapenemase-non-producingisolatesthatwerenot includedinthetablepresentednegativeresultsforalltests performed.TheCarbaNPtestpresentedasensitivityof62.7% andaspecificityof97.5%.Thevastmajorityofisolateswith positiveresultspresentedaclearcolorchangewithin15min ofincubation; onlyoneisolaterequiredthemaximumtime of75mintobecomepositive(Table2).Thesensitivityof Blue-Carbawas95.3%anditsspecificitywas100%.Asobservedin CarbaNP,mostisolateswithpositiveresultsintheBlue-Carba presentedthefinalresultwithin15minofincubation;only

Table1–ResultsofrapidtestsCarbaNP,Blue-CarbaandCIMforcarbapenemase-producingandnon-producing Enterobacteriaceae.

IsolateID Carbapenemase CarbaNP Timetopositivity(min) Blue-Carba Timetopositivity(min) CIM

1101F KPC − NA + 30 + 1345F KPC + 15 + 15 + 1373F KPC + 15 + 15 + 1388F KPC + 15 + 15 + 1389F KPC + 45 + 15 + 1390F KPC + 15 + 15 + 3401F KPC + 15 + 15 + 3409F KPC + 15 + 15 + 3436F KPC + 15 + 15 + 3443F KPC + 15 + 15 + 3446F KPC + 15 + 15 + 3818F KPC + 15 + 15 + 821F NDM + 30 + 15 + 871F NDM + 15 + 45 + 1233F NDM + 15 + 30 + 2007F NDM + 15 + 15 + 2130F NDM + 15 + 15 + 2610F NDM − NA + 15 + 2612F NDM + 45 + 15 + 2748F NDM − NA + 15 + 3035F NDM − NA + 15 + 3304F NDM − NA + 15 + 3320F NDM + 15 + 15 + 3763F NDM + 15 + 15 + 3768F NDM + 15 + 15 + 1888F IMP + 15 + 15 + 3349F IMP + 15 + 15 + 1047F GES-5 + 60 − NA + 1597F GES-5 + 15 + 15 − 2818F GES-5 + 15 + 15 + 3691F GES-5 + 75 + 75 + 1534F OXA-370 − NA + 15 − 1636F OXA-370 − NA + 15 + 2169F OXA-370 − NA + 60 − 2246F OXA-370 − NA − NA − 2494F OXA-370 − NA + 60 − 2592F OXA-370 − NA + 15 + 2729F OXA-370 − NA + 15 − 2807F OXA-370 − NA + 30 − 3023F OXA-370 − NA + 30 − 3149F OXA-370 + 30 + 15 − 3284F OXA-370 − NA + 30 − 3704F OXA-370 − NA + 30 − 3413F Negative + 30 NV NA − 3452F Negative − NA − NA +

NA,notapplicable;NV,notvalidated.

oneisolaterequiredthemaximumtimeof75mintobecome

positive(Table2).TheCarbapenemase InactivationMethod presentedasensitivityof74.4%andspecificityof97.5%.The CIMrequiredastandardtimeofatleast480minbetweenthe beginningandthefinalresultsofthetest(Table2).

Discussion

Latelyavarietyofnovelmethodsforcarbapenemase detec-tion have been proposed.9,12,15–17 _{All methods explore the}

carbapenem-hydrolysingactivityofthe␤-lactamasesandare

Table2–Summaryofthecharacteristicsoftherapid testsforcarbapenemasedetectionCarbaNP,Blue-Carba andCIM.

Sensitivity(%) Specificity(%) Maximumtime untilfinalresult (min)

CarbaNP 62.7 97.5 75

Blue-Carba 95.3 100.0 75

supposedtopresenthighsensitivityandspecificitywithina shortperiodoftime,withresultsinterpretableinasingleday ofwork.Inthisstudy,weevaluatedtherapidtestsCarbaNP, Blue-CarbaandCIM.

CarbaNPpresentedhighspecificity,butitssensitivitywas thelowestamongtheothermethodologies.Thereduced sen-sitivityofCarbaNPwasmostlyduetoOXA-370isolates.This findingwasalsoreportedinotherstudiesthatevaluatedthe performanceofCarbaNPwithOXA-48-producingisolates.18,19

TheCarba NP method proved to bea very fast methodol-ogyasmostpositiveresultswereobservedin15min.Infact, althoughthebacterialextractandthecarbapenemwere incu-batedfor120min,nosignificantcolourchangewasobserved after75min,whichwasconsideredthefinaltimeresultsfor thistest.

TheBlue-Carbamethodologypresentedthehighest sensi-tivityandspecificityamongthethreemethodologies.Onlyone OXA-370positiveisolateandoneGESpositiveisolatewerenot detectedbyBlue-Carba.Pasteranetal.observedverysimilar resultsofsensitivityandspecificityforthisassaywhenthey testedisolatesfromverydiverselocationsinLatinAmerica.21

AsfoundintheCarbaNPtest,mostisolatespresented posi-tiveresultsin15minandthefinalresultforBlue-Carbawas observedafter75minofincubation.

RegardingCIM,wehavemodifiedtheoriginalprotocolof vanderZwaluwetal.12_{andaddedzincsulphateinthe}

sus-pensionofbacteriawith themeropenem disk toachieve a highersensitivityofthemethodregardingthedetection of metallo-beta-lactamase.AsobservedwithCarbaNP,alower sensitivitywasobtainedamongtheOXA-370producers.The majorsetbackofthismethodology,comparedtotheothertwo evaluatedinthisstudy,isthetimeneededforthefinalresult, consideringthatatleast6harenecessaryforapropergrowth ofthecarbapenem-susceptibleE.coli,whichismandatoryfor thetestinterpretation.Ontheotherhand,theCIMseemsto bethesimplesttest,consideringthatnospecialreagentsare necessaryforitsexecution.

Thedifficultyfordetectionofisolatesharboring OXA-48-like is probably related to the least expressive hydrolytic capacityofclassDcarbapenemases.20_{Also,itisstillnotclear}

ifOXA-370(avariantofOXA-48)isatruecarbapenemase, con-sideringthatlowcarbapenemMICvaluesareusuallyobserved inclinical samplespresentingthis enzyme.In this matter, thedetectionofimipenemhydrolysis,inwhichthetestsare based,isevenharder.IfOXA-370producingisolateswerenot includedinthisstudy,thesensitivityforCarbaNPtestwould beof83.8%andof96.7%forCIM.However,theOXA-370seems tobethe variantofOXA-48thatisprevalent inBrazil and, therefore, it isimportantthat the rapidtests beevaluated againsttheseisolates.

Noteworthy,minimalcolorchanges,yieldedbylower car-bapenemhydrolysislevels,wereeasiertoobserveusingthe Blue-Carbatest,consideringthatblue-greenchangesare eas-iertoobserve,comparedtored-orangeones.Insomecases, especiallyregardingOXA-370-andGES-5-producingisolates, theinterpretationoftestsresultswerechallenging,anda sec-ondopinionwasrequested.Thus,thecorrectinterpretation reliesoncapacitatedtechnicians.

Theglobalspreadofcarbapenemaseshighlightsthe impor-tanceoftheirrapiddetection.Consideringthatcarbapenems

arethelastresortforseriousEnterobacteriaceaeinfections, con-trollingthedisseminationofthemechanisms ofresistance totheseantibioticsismandatory.22_{AlthoughPCRis}

consid-eredthegoldstandardforcarbapenemasedetection,16_most

laboratoriesarenotabletoperformthismethodologydueto costsandthespecialequipmentrequired.Themethodologies evaluatedinthisstudyprovidereliableresultsinasingle work-day.Thecarbapenemasedetectionmightalsobedonedirectly fromclinicalsamples,asdescribedbyDortetetal.18

Insummary,theBlue-Carbapresentedthehighest sensi-tivity and,therefore, can beconsidered the besttest tobe usedasascreeningphenotypicmethodology,consideringthat themisdetectionofcarbapenemaseproducershasagreater impactforcliniciansthanafalsepositiveresult,and differ-encesobservedbetweensensitivityvaluesaremuchgreater thanthedifferencesobservedinspecificity.Ontheotherhand, CIMmightbetheeasiesttesttoperform,asitdoesnotrequire any specialreagent and it is easier tobe interpreted. The earlydetectionofcarbapenemaseshelpstoestablish infec-tion controlmeasuresandmaypreventcarbapenemasesto spreadreducing theriskofhealthcareassociatedinfections andtherapeuticfailure.

Conflicts

of

interest

Theauthorsdeclarenoconflictsofinterest.

r

e

f

e

r

e

n

c

e

s

1.PatersonDL.Resistanceingram-negativebacteria:

Enterobacteriaceae.AmJMed.2006;119:S20–S28[discussion S62–S70].

2.NordmannP,NaasT,PoirelL.Globalspreadof

Carbapenemase-producingEnterobacteriaceae.EmergInfect

Dis.2011;17:1791–1798.

3.AmblerRP.Thestructureofbeta-lactamases.PhilosTransR

SocLondBBiolSci.1980;289:321–331.

4.RozalesFP,RibeiroVB,MagagninCM,PaganoM,LutzL,Falci

DR,etal.EmergenceofNDM-1-producingEnterobacteriaceae

inPortoAlegre,Brazil.IntJInfectDis.2014;25:79–81.

5.MagagninCM,RozalesFP,AntochevisL,NunesLS,Martins

AS,BarthAL,etal.Disseminationofbla(OXA-370)gene

amongseveralEnterobacteriaceaespeciesinBrazil.EurJClin

MicrobiolInfectDis.2017[Epubaheadofprint].

6.PicaoRC,SantosAF,NicolettiAG,FurtadoGH,GalesAC.

DetectionofGES-5-producingKlebsiellapneumoniaeinBrazil.J

AntimicrobChemother.2010;65:796–797.

7.LincopanN,McCullochJA,ReinertC,CassettariVC,Gales

AC,MamizukaEM.Firstisolationof

metallo-beta-lactamase-producingmultiresistantKlebsiella

pneumoniaefromapatientinBrazil.JClinMicrobiol.

2005;43:516–519.

8.NordmannP,PoirelL,DortetL.Rapiddetectionof

carbapenemase-producingEnterobacteriaceae.EmergInfect

Dis.2012;18:1503–1507.

9.PiresJ,NovaisA,PeixeL.Blue-carba,aneasybiochemicaltest

fordetectionofdiversecarbapenemaseproducersdirectly

frombacterialcultures.JClinMicrobiol.2013;51:4281–4283.

10.CarvalhaesCG,CayoR,ViscondeMF,BaroneT,FrigattoEA,

OkamotoD,etal.Detectionofcarbapenemaseactivity

quickanswerfortherightdecision.JAntimicrobChemother.

2014;69:2132–2136.

11.LasserreC,DeSaintMartinL,CuzonG,BogaertsP,LamarE,

GlupczynskiY,etal.Efficientdetectionofcarbapenemase

activityinEnterobacteriaceaebymatrix-assistedlaser

desorptionionization-timeofflightmassspectrometryin

lessthan30minutes.JClinMicrobiol.2015;53:2163–2171.

12.vanderZwaluwK,deHaanA,PluisterGN,BootsmaHJ,de

NeelingAJ,SchoulsLM.Thecarbapeneminactivation

method(CIM),asimpleandlow-costalternativeforthe

CarbaNPtesttoassessphenotypiccarbapenemaseactivity

ingram-negativerods.PLoSONE.2015;10:e0123690.

13.Clinical&LaboratoryStandardsInstitute–CLSI.Performance

StandardsofAntimicrobialSusceptibilityTesting;27th

InformationalSupplement;M100-S27.Wayne,PA:Clinicaland

LaboratoryStandardsInstitute;2017.

14.MonteiroJ,WidenRH,PignatariAC,KubasekC,SilbertS.

Rapiddetectionofcarbapenemasegenesbymultiplex

real-timePCR.JAntimicrobChemother.2012;67:906–909.

15.DortetL,PoirelL,NordmannP.Rapididentificationof

carbapenemasetypesinEnterobacteriaceaeandPseudomonas

spp.byusingabiochemicaltest.AntimicrobAgentsChemother.

2012;56:6437–6440.

16.MartinoMD,KogaPC,PasternakJ,DoiAM,CiolaCS,daSilva

CB,etal.Evaluationofanewrapidtestforcarbapenemase

detectionincarbapenemresistantEnterobacteriaceae.J

MicrobiolMethods.2015;115:20–21.

17.SaitoR,KoyanoS,DorinM,HigurashiY,MisawaY,NaganoN,

etal.Evaluationofasimplephenotypicmethodforthe

detectionofcarbapenemase-producingEnterobacteriaceae.J

MicrobiolMethods.2015;108:45–48.

18.DortetL,BrechardL,PoirelL,NordmannP.Rapiddetectionof

carbapenemase-producingEnterobacteriaceaefromblood

cultures.ClinMicrobiolInfect.2014;20:340–344.

19.HuangTD,BerhinC,BogaertsP,GlupczynskiY.Comparative

evaluationoftwochromogenictestsforrapiddetectionof

carbapenemaseinEnterobacteriaceaeandinPseudomonas

aeruginosaisolates.JClinMicrobiol.2014;52:3060–3063.

20.EvansBA,AmyesSG.OXAbeta-lactamases.ClinMicrobiolRev.

2014;27:241–263.

21.PasteranF,VelizO,CerianaP,LuceroC,RapoportM,

AlbornozE,etal.Evaluationoftheblue-carbatestforrapid

detectionofcarbapenemasesingram-negativebacilli.JClin

Microbiol.2015;53:1996–1998.

22.WalshTR.Emergingcarbapenemases:aglobalperspective.