Detection of specific bacterial agents by quantitative PCR assays in the bronchoalveolar lavage fluid of dogs with eosinophilic bronchopneumopathy vs. dogs with chronic bronchitis and healthy dogs

Original

Article

Detection

of

speci

fic

bacterial

agents

by

quantitative

PCR

assays

in

the

bronchoalveolar

lavage

fluid

of

dogs

with

eosinophilic

bronchopneumopathy

vs.

dogs

with

chronic

bronchitis

and

healthy

dogs

A.M.

Canonne

*

,

I.

Peters

,

E.

Roels

,

L.

Desquilbet

,

C.

Clercx

a

InternalMedicineUnit,NationalVeterinarySchoolofAlfort,94700Maisons-Alfort,France

b

TorranceDiamondDiagnosticServices,TheInnovationCentreUniversityofExeter,EX44RNExeter,UK

c

DepartmentofClinicalSciences,FacultyofVeterinaryMedicine,UniversityofLiège,4000Liège,Belgium

d

UnitofBiostatistics,NationalVeterinarySchoolofAlfort,94700Maisons-Alfort,France

ARTICLE INFO Articlehistory: Accepted16December2017 Keywords: Bordetellabronchiseptica Bronchoalveolarlavage Dogs Eosinophilicbronchopneumopathy Mycoplasmaspp. ABSTRACT

In humans, Mycoplasma pneumoniaeand Bordetella pertussis infectionsare suggested totriggeror

exacerbate asthma. Whether Mycoplasma or Bordetella are associated with chronic inflammatory

bronchialdiseasesindogshasnotbeeninvestigated.Theaimofthisstudywastoassessdetectionratesof

Mycoplasma canis(M.canis), M.cynosand Bordetellabronchiseptica(Bb), indogs witheosinophilic

bronchopneumopathy (EBP) and chronic bronchitis (CB), compared with healthy dogs. Specific

quantitative PCR(qPCR) analysisforM.canis,M.cynosand Bbwereretrospectivelyperformedon

bronchoalveolarlavagefluid(BALF)collectedfrom24dogswithEBP,21dogswithCBand15healthy

dogs.PossibleassociationsbetweenqPCRresultsandage,BALFcytologyorclinicalseverityscores(CSS)

indogswithEBPwereinvestigated.

TherewasnodifferenceinM.canis,M.cynosandBbdetectionratesindogswithEBP(n=6,n=2 and

n=6,respectively)anddogswithCB(n=2,n=2andn=2,respectively)comparedwithcontroldogs

(n=4,n=2andn=2,respectively).IndogswithEBP,theproportionthatwereqPCR-positiveforBbwas

higherindogswithhigherCSS(P=0.014)andBALFfromBb-positivedogshadhigherpercentageof

neutrophils(P<0.001).AmongdogsthatwereqPCR-positiveforBb,moderatetohighloadswereonly

detectedindogswithEBP.M.canisandM.cynosdetectionwasnotassociatedwithEBPorCB;higherBb

loadswereonlypresentindogswithEBPandhighCSS.Apossiblecauseandeffectrelationshipbetween

BbinfectionorloadandEBPremainsunclearandrequiresfurtherinvestigation.

©2017ElsevierLtd.Allrightsreserved.

Introduction

Idiopathiceosinophilicbronchopneumopathy(EBP)isachronic diseasecharacterizedbyeosinophilicinfiltrationofthelungand bronchialmucosainyoungadultdogs(Corcoranetal.,1991;Clercx etal.,2000;Rajamäkietal.,2002).Dogswithnon-specificchronic bronchitis(CB)alsopresentwithchroniccough,butareusually

older. The etiology of both inflammatory conditions remains

unclear. In EBP, an underlying immunologic hypersensitivity is

suspected,but theincitingantigensremainmostlyunidentified

(Clercxetal.,2002;Peetersetal.,2005;ClercxandPeeters,2007). Forbothofthesebronchialdiseases,currenttreatmentsconsistof long-termoraland/orinhaledglucocorticoids(Clercxetal.,2000; Bexfielsetal.,2006;Rozanski,2014;Canonneetal.,2016b).

In humans, infections with Mycoplasma pneumoniae (M.

pneumoniae) have been associated with asthma for decades

(Hansbro et al., 2004; Atkinson, 2013; Ye et al., 2014). M.

pneumoniae infection is associated with acute exacerbation of

adultasthmaticsandfuture developmentof asthmainchildren

(Hansbroetal.,2004)andspecifictreatmentimprovespulmonary function in asthmatics (Kraft et al., 2002).However, Bordetella pertussis(B.pertussis)hasbeenalsodiscussedaspotentialtrigger inhumaninflammatorybronchialdiseaseandasthma(Harjuetal., 2006;Wakashinetal.,2008;Nicolaietal.,2013;Yinetal.,2017).

*Correspondingauthor.

E-mailaddress:morgane.canonne-guibert@vet-alfort.fr(A.M. Canonne).

1

Presentaddress:InternalMedicineUnit,NationalVeterinarySchoolofAlfort,7 AvenueCharlesdeGaulles,94700Maisons-Alfort,France.

https://doi.org/10.1016/j.tvjl.2017.12.014

1090-0233/©2017ElsevierLtd.Allrightsreserved.

ContentslistsavailableatScienceDirect

The

Veterinary

Journal

Furthermore,inamousemodelofallergicasthma,priorB.pertussis infectionwasdocumentedtoexacerbateairwaysigns(Ennisetal., 2004;Kavanaghetal.,2010).

AlthoughMycoplasmacynos(M.cynos)hasbeenidentifiedasan

emerging,possiblycontagious,andlethalpathogenindogswith

canineinfectiousrespiratorydisease(CIRD,kennelcoughorcanine

infectious tracheobronchitis; Rycroft et al., 2007; Zeugswetter

etal.,2007;Manneringetal.,2009;Priestnalletal.,2014),butthe

exact role of M. canis and M. cynos as primary respiratory

pathogens still remains unclear (Chandler and Lappin, 2002;

Chalkeretal.,2004;Chanetal.,2013).Moreover,arecentstudy

demonstrated the impact of oral bacterial contamination on

Mycoplasma-specificPCR resultsof bronchoalveolarlavagefluid

(BALF;Chanetal.,2013),makingtheinvestigationofapotential

primary role of Mycoplasma spp. in canine lower respiratory

conditionsevenmorecomplex.

Bordetella bronchiseptica (Bb) is recognized as one of the

primary causative pathogen agents of CIRD. Bb can exist in

respiratorytractofdogsaseithercommensalorpathogen(Bemis etal., 1977;Bemis, 1992;Schulzetal.,2014).Currently,quantitative PCR(qPCR)assaysprovidehighlysensitiveandspecificdetection ofofBb,M.canisandM.cynos(Chanetal.,2013;Schulzetal.,2014; Lavan and Knesl, 2015; Canonne et al., 2016a). However, the potentialroleoftheseinfectiousagentsinchronicinflammatory

bronchial diseases in dogshave not yet been investigated and

whetherthesebacterialagentscouldactastriggersor

exacerbat-ingagentsindogswithEBPisunknown.

Theaimofthepresentstudywastoevaluatethepresenceand bacterialloadofM.canis,M.cynosandBbinsteroid-naïvedogs

newly-diagnosedwithEBPandcomparetheresultstothosefor

dogswithnon-specificCBandhealthydogs,usingspecificqPCR

analysis performed on BALF samples. Additionally, potential

associations between positive qPCR results for any of these

bacterialspeciesandtheclinicalseverityofEBPwereassessed.

Materialsandmethods

Dogs

Client-owned dogs presentedto the National VeterinarySchool of Alfort Companion Animal Hospital between March 2009 and February 2016 and diagnosedwithEBPorCBwereprospectivelyrecruited.

DefinitivediagnosisofidiopathicEBPandCBwasmadeinaccordancewith criteriapreviouslydescribed(Clercxetal.,2000;Rozanski,2014).Briefly,thiswas based on compatible respiratory clinical complaints, radiological changes, bronchoscopic findings, analysis of the BALF including culture and standard cytological examination and, in somecases, histopathologicalexamination of endoscopicbronchialmucosalbiopsies.IndogswithCB,bacterialinvolvementin neutrophilic airway inflammation was excluded based on cytology (lack of degenerativeneutrophils,absenceofintra-cellularbacteria),cultureandpositive responsetosteroids.IndogswithEBP,exclusionofothercausesofeosinophilic airway inflammation, including cardiopulmonary parasites, was required to definitivelydiagnoseidiopathicEBPbyusingfecalandBALFanalysis(Baermann method,cytologyandqPCRonBALF).Alldogswerenewly-diagnosedand steroid-naïve.

IndogswithafinaldiagnosisofidiopathicEBP,aclinicalseverityscore(CSS; 1-5/5)wasassessedatthetimeofdiagnosisbasedonseverityandfrequencyofcough, evidenceofexerciseintoleranceorlethargy,aspreviouslydefined(Canonneetal., 2016b). Coughwas scoredfrom 1-3/3depending onfrequencyand severity; presenceofretchingorexerciseintolerancewasrecordedwith1additionalpoint. Pointsobtainedfromcoughscoring,retchingandexerciseintolerancewereadded toamaximumtotalof5points.

Bronchoscopy,BALFprocedureandprocessingwereperformedasdescribed earlier(Clercxetal.,2000).Briefly,dogswereanesthetizedusingvariousanesthetic protocolsaftera5minpre-oxygenationperiod.Fiveto20mLaliquots(depending onbodyweight)ofsterilesaline(NaCl0.9%)wereinstilledtwice intoasame bronchusand athirdtime into adifferent lung lobeviaaflexible pediatric endoscope(Fujinon,Pediatricvideo-bronchoscopeEB-530S),followedby immedi-ateaspirationbygentleautomatedsuction.TherecoveredBALFwasimmediately processed.AliquotsofnaïveBALFwereusedforsemi-quantitativebacterialculture (CollardLaboratories,accordingtomethodspreviouslydescribed;Peetersetal., 2000), total cell count determination using a hemocytometer, and cytospin

preparation(centrifugationat1400rpm,197g,for4minat20_{C,ThermoShandon}

Cytospin4).Differentialcellcountcalculationswereestablishedbycountingatotal of300cellsviewedonhighpowerfieldsonthecytospinpreparation;400–600/mL wasconsideredanormalcellcount.BALFcytologywasconsiderednormalifthere were<12%neutrophilsand<7%eosinophils.TheremainingrecoveredBALFwas centrifugedat1300gfor15minat4_{C.Theresultingsupernatantandcellpellet}

werestoredseparatelyat 80_{C.Additionally,angiostrongylosiswasexcludedinall}

dogswithEBPbasedonanegativequantitativePCR(qPCR)testresultsinBALF (Canonneetal.,2016b).Additionally,alldogsweretreatedatthetimeofdiagnosis with anthelmintics(fenbendazole50mg/kgq24hfor5days,PanacurIntervet International via MSD Animal Health, or milbemycin oxime 1mg/kg once, MilbemaxNovartis),whileresultswerepending.

Controlgroup

Stored BALFsamples collectedfrom 15healthy dogs wereavailable from previousstudiesforwhichethicalapprovalhadbeenpreviouslyobtainedfromthe UniversityLocalEthicalCommittee(UniversityLocalEthicalCommitteeApproval number1435;Approvaldate30thApril2013).Thestoredsampleswerecollected fromhealthydogsbelongingtoveterinarystafforstudents(n=13)orfromshelters (n=2), and none had either history orclinical signs of respiratory disease. Bronchoscopy,BALFandlaboratoryprocessingandstoragehadbeenperformedas fordogswithEBPandCBandbothtotalcellcountsandcytologicalexaminationof cytocentrifugedpreparationsofBALFwerewithinnormallimitsineachdog. QuantitativePCRanalysis

QuantitativePCRanalysiswasperformedonstoredBALFforM.canis,M.cynos andBbindogswithEBP,CBandhealthydogs.Thefrozenpelletedcellswerethawed and,afterre-suspensioninasmallvolumeofsterilesaline(0.9%NaCl),samples weresenttothelaboratoryforqPCRanalysis.TheqPCRresultswereexpressedasCt values.Ctvalueswerearbitrarilyfurthercategorizedintofivegroups:veryhighload (Ct<20),highload(20.1–24),moderateload(24.1–28),lowload(Ct28.1–32)and verylowload(>32.1).

Statisticalanalysis

Statisticalanalyseswereperformedwithacommerciallyavailablesoftware (XLstatsoftware).Datawereexpressedforcontinuousandcategoricalvariablesas medianwithrangeorasproportions,respectively.

Chi-squareorFisher’sexact(whennecessary)testswereusedtocomparethe proportionsofqPCRdetectionofM.canis,M.cynosandBbbetweendogswithEBP andhealthydogsandbetweendogswithEBPanddogswithCB.DogswithEBPwere grouped according to CSS into two groups:dogs with CSS3 (mildclinical severity)anddogswithCSS>3(moderate-to-severeclinicalsigns).Theproportion ofdogswithEBPandpositiveqPCRresultsforM.canis,M.cynosandBbwere comparedbetweendogswithCSS3anddogswithCSS>3byusingexactFisher’s exacttest(forn5).Foreachbacterialagent,themedianageofdogswithEBPwith positiveqPCRwascomparedtothemedianageofdogswithEBPwithnegativeqPCR usingWilcoxon–MannWhitneytests.Foreachbacterialagent,mediantotalcell countandpercentageofneutrophilsintheBALFofdogswithEBPthatwere qPCR-positivewerecomparedtoBALFparametersfromdogswithEBPthatwere qPCR-negativebyusingWilcoxon–MannWhitney’stest.

ValuesofP0.05wereconsideredstatisticallysignificant.

Results

Aconveniencesampleof24dogswithEBP(medianage,4.2y

[0.9–13.2])and21dogswithnon-specificCB(medianage,8.0[0.9– 14.0])wereselected.

Table 1 reports the proportions of dogs withpositive qPCR resultforM.canis,M.cynosandBbineachofthethreegroups.For each bacterialagent,nosignificantdifferencein qPCRdetection

rateswerefoundbetweendogsdiagnosedwithEBPandhealthy

dogs and between dogswith EBPand dogs withCB (Table 1).

AmongdogswithpositiveqPCRfor M.cynosonBALF, Ctvalues

correspondingtovery highloadwereonly foundin dogswith

inflammatorybronchialdisease.AmongdogswithpositiveqPCR

forBbonBALF,moderateorhighloadwereonlyfoundindogswith EBP.

IndogswithEBP,CSSvariedbetweentwoandfiveatthetimeof diagnosis;totalBALFcellcountswere700–6000/mL;%eosinophils was20–85;and%neutrophilswas1–30(Table2).Theproportion ofdogswithBbintheBALFwassignificantlyhigherindogswith CSS>3(5/9,56%)thanindogswithCSS3(1/15,7%;P=0.015).

For the mycoplasmal species, there was no difference in CSS

betweendogswithqPCR-positiveqPCR-negativeresults(datanot

shown).IndogswithEBP,themedianageofthosethatwere qPCR-negativeforBbwasnotdifferentfromthatforqPCR-positivedogs (3.7years[0.9–13.0]and5.8years[0.9–13.2],respectively;P=0.44)

andmedianageforqPCR-negativeandqPCR-positiveforM.canis

andM.cynoswasnotdifferent(M.canis,4.3years[0.9–13.0]and 2.9 years [1.6–13.2], respectively; P=0.73; M. cynos, 4.1 years [0.9–13.0]and7.6years[0.9–13.2],respectively;P=0.71).

In dogswith EBP, for each bacterialagent taken separately,

medianBALFtotalcellcountsfromqPCR-negativedogswasnot

differentfrom that from qPCR-positive dogs (datanot shown).

Median%neutrophilsinBALFfromdogsqPCR-positiveforBband

M.cynosweresigni_ficantlyhigherthanmedianvaluesfromdogs

that were qPCR-negative (Bb, 15% [3–20] and 3.5% [1–30],

respectively;P<0.001;M.cynos,12.5%(10–15) and3.5%[1–30],

respectively;P=0.02).Median %neutrophilsin BALFfromdogs

withpositive and negative qPCR results for M. canis were not

different(datanotshown). Discussion

Theobjectivesofthepresentstudyweretodetectthepresence

andloadofM.canis,M.cynosandBbindogswithEBPcompared

withdogsdiagnosedwithnon-specificCBandhealthycontrols,

usingspecificqPCRonBALFsamples.Theresultsofthepresent

studysuggest thatanyassociationbetweeneither mycoplasmal

speciesandcanineEBPisunlikely.However,indogswithEBP,the probabilityofapositiveqPCRresultforBbincreasedstatistically

withclinicalseverity.Moreover,amongdogswithpositiveqPCR

resultsforM.cynosorBb,Ctvaluesindicatingmoderate,highor

very high load were found only in dogs with inflammatory

bronchialdisease.Thesmallgroupsizeofdogsthatwere

qPCR-positive precluded statistical comparisons of bacterial load

betweengroups.

HumanasthmaandcanineEBPhavedifferentcharacteristics;

forexample,bronchialhyper-responsivenessisnotahallmarkof

canineEBP.However,humanasthmaandcanineEBPareboth

TH2-driven inflammatory conditions characterized by eosinophilic

infiltration.Theroleofbacterialgeneraknowntobeimplicatedin

theinductionor exacerbationof humanasthmahave notbeen

investigated in dogs with EBP. In humans, infection with

M. pneumoniae can favour the development of asthma in

predisposedpatientsand specifictreatmentwithclarithromycin

improves pulmonary function (Martin et al.,2001; Kraft et al., 2002;Hansbroetal.,2004;ElSayedZakietal.,2009;Atkinson, 2013;Yeetal.,2014).Theproportionofasthmaticswhohaveboth

IgM and IgG forM. pneumoniaeincreases withclinicalseverity

(Iramainetal.,2016;Yinetal.,2017).Ina mousemodel,single

exposuretoaspecifictoxin ofM.pneumoniaewasshowntobe

sufficienttocauseasthma-likediseasewithhistologicalevidence ofpulmonaryeosinophilicinflammation(Medinaetal.,2012).Our

study failed to support thehypothesis of a similar association

betweenmycoplasmalinfectionorcarriageandEBPindogs.This

couldpossiblybeexplained bytherelativelysmallsample size

limitingthepowerof statisticaltestsfor comparison.Inhuman

asthma, the role of Mycoplasma spp. has sometimes been

investigated using PCR, but is mainly assessed using serum

Mycoplasma-specificantibodies(ElSayedZakietal.,2009;Iramain et al., 2016); such antibodieswere notevaluated in ourstudy.

Additionally, the potential therapeutic benefit of a specific

antibiotic treatment, such as doxycycline, macrolide or

fluoro-quinolone,hasnotbeen evaluatedindogswithEBPor CBand

positive qPCR results for M.cynos. Lastly,involvementof other

mycoplasmalspeciesincaninechroniclowerairwayinflammation

could not be excluded;several otherMycoplasma species have

beeninvestigatedindogs(Chalkeretal.,2004),butuntilrecently,

accuratediagnosticmethodssuchasqPCRwerenotavailablefor

lesscommoncaninemycoplasmalspecies.

AlthoughBbwasnotmorefrequentlydetectedindogswithEBP thanindogswithCBorhealthydogs,thefrequencyofBb-positive

dogs with EBPwas positively associated with clinical severity.

AmongdogswithpositiveqPCRforBbonBALF,moderateorhigh

loadswerefoundindogswithEBPonly. WhetherBbisableto

triggereosinophiliclowerairwayinflammationandwhetherthe

inflammatory context in airways favours Bb growth in dogs is

unknown. In humans, Bb infections are unusual and almost

exclusively described in immunocompromized patients with

contactwithanimals.B.pertussisismorecommonandhasbeen

discussed as a triggering and exacerbating factor in chronic

inflammatory bronchial diseases; B. pertussis toxin favours

eosinophilic airway inflammation (Ennis et al., 2004; Harju

etal.,2006;Wakashinetal.,2008;Kavanaghetal.,2010;Nicolai etal.,2013;Yinetal.,2017).Despitealargeoverlapbetweenvalues,

Table1

FrequencyofpositiveqPCRperformedonbronchoalveolarlavagefluidforMycoplasmacanis,MycoplasmacynosandBordetellabronchisepticaindogswitheosinophilic bronchopneumopathy(EBP),chronicbronchitis(CB)andhealthydogs.

DogswithEBP DogswithCB Healthydogs P

Mycoplasmacanis 6/24(25%)a,b

2/21(10%)b

4/15(13%)a

1.00a

0.25b

Veryhighload(Ct<20) – –

Highload(20.1–24) – –

Moderateload(24.1–28) 1 –

Lowload(Ct28.1–32) 3 – 1

Verylowload(>32.1) 2 2 3

Mycoplasmacynos 2/24(8%)a,b

2/21(10%)b

2/15(13%)a

0.63a

1.00b

Veryhighload(Ct<20) 2 1 –

Highload(20.1–24) – – –

Moderateload(24.1–28) – – 1

Lowload(Ct28.1–32) – – –

Verylowload(>32.1) – 1 1

Bordetellabronchiseptica 6/24(25%)a,b

2/21(10%)b

2/15(13%)a

0.45a

0.25b

Veryhighload(Ct<20) – – –

Highload(20.1–24) 1 – –

Moderateload(24.1–28) 2 – –

Lowload(Ct28.1–32) – – –

Verylowload(>32.1) 3 2 2

a,b

themedian%neutrophilsintheBALFwashigherindogswithEBP withpositiveqPCRforBbthanindogswithEBPwithnegativeqPCR

results, while median total cell counts were not different.

Consequently,largerstudiesinvestigatingtherole ofBbin dogs

with EBP and potential clinical benefits of treatment for

Bb-positivedogsareneeded.

The potential influence of any previous Bb vaccination on

positivesamplescouldnotbecompletelyruledoutinourstudy.

However,amongdogswithknownvaccinationstatus,thenumber

ofvaccinateddogswasnotdifferentamongthethreegroups(data notshown).Inpuppiesvaccinatedwithasingledoseofmodified liveintranasalvaccine,BbwasdetectedbyqualitativePCRduring

1 month after vaccination (Ruch-Gallie et al., 2016). To our

knowledge,theimpactofpreviousvaccinationonqPCRperformed

onBALFhasnotbeeninvestigated.Nevertheless,noneofthedogs inourstudywerevaccinatedwithin2monthspriortodiagnosis.

Because of this, we assumed that the impact of previous

vaccinationonqPCRforBbinBALFwasminimal.

AsalldogswithEBPandCBwerereferredcases,theimpactof

previous antimicrobial treatment, such as doxycycline or

fluo-roquinolones,onourqPCRresultscannotbecompletelyruledout;

this mighthavecaused underestimatesof thepresenceofeach

agentindiseaseddogscomparedtohealthydogs.Thispotential

riskwasconsideredquitelowsincetherewasusuallyawash-out periodofatleast2daysbeforeBALprocedures.

Thelastminorlimitationofthisstudywasthelackofvalidation ofCSSthatwasusedtoscoreclinicalsignsindogswithEBP.Tolimit variability,CSSwasattributedtoeachdogwithEBPbyonlyone

author(AMC),whoremainedmaskedtoqPCRresults.Intra-and

inter-observervariabilitiesinCSSevaluationshouldbeassessed beforethisscoringsystemcanbeusedonalargerscaleindogs

withEBP.

Table2

Signalment,qPCRresultsfrombronchoalveolarlavagefluid(BALF)forMycoplasmacanis,M.cynos,Bordetellabronchiseptica(Bb),clinicalseverityscores(CSS)andBALF cytologyindogswitheosinophilicbronchopneumopathy(EBP).

Case Breed Sex Age

(years)

M.canisload M.cynosload Bbload CSS BALFcytology (totalcellcount/mL, %eosinophils–%neutrophils)

BALFbacterialculture

1 Mixed-breed M 13 – – – 2 1400/mL, 20%–5% Negative 2 Mixed-breed F 6 – – – 2 4000/mL, 85%–3% Negative 3 Sharpei F 3 – – – 2 2500/mL, 65%–5% Negative 4 Shitzu M 4 – – – 3 3000/mL, 60%–1% Negative 5 Shitzu M 1.2 – – – 3 1600/mL, 55%–30% Negative 6 Mixed-breed M 12.4 – VH – 3 1900/mL, 30%–10% Negative 7 Beauceron F 1.6 L – – 3 1800/mL, 55%–3% Negative 8 Sharpei F 13.2 M – M 3 2000/mL, 55%–3% Negative 9 Mixed-breed F 3.4 – – – 3 1400/mL, 60%–1% Negative 10 Tervuren M 2.0 VL – – 3 700/mL, 55%–16% Negative 11 Whippet F 3.0 L – – 3 1100/mL, 20%–3% Negative 12 Goldenretriever F 3.3 – – – 3 1300/mL, 80%–4% Negative 13 Brittanyspaniel F 7.1 VL – – 3 800/mL, 25%–3% Negative 14 Siberianhusky M 0.9 – – – 3 1500/mL, 80%–3% Negative 15 Mixed-breed F 5.2 – – – 3 4000/mL, 85%–7% Negative 16 Doberman M 0.9 – – VL 4 2000/mL, 40%–10% Negative 17 Siberianhusky F 1.0 – – – 4 1600/mL, 75%–15% Negative 18 Bassetartesien M 4.3 – – – 4 2500/mL, 35%–3% Negative 19 Whippet F 8 – – – 4 1700/mL, 40%–15% Negative 20 Brittanyspaniel F 5.7 – – M 5 2500/mL, 70%–20% Negative 21 Borderterrier M 2.9 L VH H 5 6000/mL, 20%–15% Negative 22 Foxterrier F 10.0 – – VL 5 1800/mL, 58%–20% Negative

23 JackRussellterrier M 4.4 – – – 5 3100/mL,

62%–1%

Negative

24 Tervueren M 6 – – VL 5 4500/mL,

60%–15%

Negative

Conclusions

ThepresentinvestigationbyqPCRonBALFdoesnotsupportany roleforM.canisandM.cynosinthepathogenesisofEBPindogs. Nevertheless,indogswithEBP,Bbwasdetectedmorefrequentlyin thosewithmoresevereclinicalsigns.Moreover,amongdogsthat

wereqPCR-positiveforBb,moderateorhighloadswereobserved

inonlydogswithEBP.Despitebothobservations,acauseandeffect

relationship between the presence and load of Bb and severe

canineEBPremainsunclear.However,sincedogswithEBPcould

potentially act as Bb carriers and a source of infection for

susceptibledogs,screeningforthispathogenshouldbeperformed indogsnewly-diagnosedwithEBP,particularlyincaseswherethe

clinical presentation is severe, there is a high percentage of

neutrophilsintheBALF,oriftherehasbeenanacuteexacerbation ofclinicalsignsinpreviouslydiagnosedcases.TheseBb-positive

dogsshould beappropriatelyaccordingly and respiratory signs

shouldbemonitoredclosely. Conflictofintereststatement

Noneoftheauthorsofthispaperhasafinancialorpersonal

relationship with other people or organisations that could

inappropriatelyinfluenceorbiasthecontentofthepaper. Acknowledgements

Thisresearchdidnotreceiveanyspecificgrantfromfunding agenciesinthepublic,commercialornot-for-profit sectors.We gratefullythankDrMaudGirod,DrEmilieVanGrinsven,DrElodie

Darnis, Dr Elodie Roels, Dr Valérie VanDerVorst, Dr Charlotte

Auquier, Dr Emilie Krafft, Dr Frédéric Billen and veterinary

technicians for their help in the inclusion of cases and the

collectionofsamples. References

Atkinson,T.P.,2013.Isasthmaaninfectiousdisease?Newevidence.CurrentAllergy andAsthmaReports13(December),702–709.

Bemis,D.A.,Greisen,H.A.,Appel,M.J.,1977.Pathogenesisofcaninebordetellosis. JournalofInfectiousDiseases135,753–762.

Bemis,D.A.,1992.BordetellaandMycoplasmarespiratoryinfectionsindogsandcats. VeterinaryClinicsofNorthAmerica:SmallAnimalPractice22,1173–1186. Bexfiels,N.H.,Foale,R.D.,Davison,L.J.,Watson,P.J.,Skelly,B.J.,Herrtage,M.E.,2006.

Managementof13casesofcaninerespiratorydiseaseusinginhaled corticosteroids.JournalofSmallAnimalPractice47,377–382.

Canonne,A.M.,Billen,F.,Tual,C.,Ramery,E.,Roels,E.,Peters,I.,Clercx,C.,2016a. QuantitativePCRandcytologyofbronchoalveolarlavagefluidindogswith Bordetellabronchisepticainfection.JournalofVeterinaryInternalMedicine30, 1204–1209.

Canonne,A.M.,Billen,F.,Peeters,D.,Bolen,G.,Roels,E.,Clercx,C.,2016b.Longterm followupindogswithidiopathiceosinophilicbronchopneumopathytreated withinhaledsteroidtherapy.JournalofSmallAnimalPractice57,537–542. Chalker,V.J.,Owen,W.M.,Paterson,C.,Barker,E.,Brooks,H.,Rycroft,A.N.,Brownlie,

J.,2004.Mycoplasmasassociatedwithcanineinfectiousrespiratorydisease. Microbiology150,3491–3497.

Chan,C.M.,Ridgway,M.D.,Mitchell,M.A.,Maddox,C.W.,2013.Associationbetween Mycoplasma-specificpolymerasechainreactionassayresultsandoralbacterial contaminationofbronchoalveolarlavagefluidsamplesfromdogswith respiratorytractdisease:121cases(2005–2012).JournaloftheAmerican VeterinaryMedicalAssociation243,1573–1579.

Chandler,J.C.,Lappin,M.R.,2002.Mycoplasmalrespiratoryinfectionsinsmall animals:17cases(1988–1999).JournaloftheAmericanAnimalHospital Association38,111–119.

Clercx,C.,Peeters,D.,Snaps,F.,Hansen,P.,McEntee,K.,Detilleux,J.,Henroteaux,M., Day,M.J.,2000.Eosinophilicbronchopneumopathyindogs.Journalof VeterinaryInternalMedicine14,282–291.

Clercx,C.,Peeters,D.,German,A.J.,Khelil,Y.,McEntee,K.,Vanderplasschen,A., Schynts,F.,Hansen,P.,Detilleux,J.,Day,M.J.,2002.Animmunologic investigationofcanineeosinophilicbronchopneumopathy.Journalof VeterinaryInternalMedicine16,229–237.

Clercx,C.,Peeters,D.,2007.Canineeosinophilicbronchopneumopathy.Veterinary ClinicsofNorthAmerica:SmallAnimalPractice37,917–935.

Corcoran,B.M.,Thoday,K.L.,Henfrey,J.I.,Simpson,J.W.,Burnie,A.G.,Mooney,C.T., 1991.Pulmonaryinfiltrationwitheosinophilsin14dogs.JournalofSmall AnimalPractice32,494–502.

Ennis,D.P.,Cassidy,J.P.,Mahon,B.P.,2004.PriorBordetellapertussisinfection modulatesallergenprimingandtheseverityofairwaypathologyinamurine modelofallergicasthma.Clinical&ExperimentalAllergy34,1488–1497. ElSayedZaki,M.,Raafat,D.,ElMetaal,A.A.,2009.Relevanceofserologyfor

MycoplasmapneumoniaediagnosiscomparedwithPCRandcultureinacute exacerbationofbronchialasthma.AmericanJournalofClinicalPathology131, 74–80.

Hansbro,P.M.,Horvat,J.C.,Gibson,P.G.,2004.Roleofatypicalbacterialinfectionof thelunginpredisposition/protectionofasthma.Pharmacology&Therapeutics 101,193–210.

Harju,T.H.,Leinonen,M.,Nokso-Koivisto,J.,Korhonen,T.,Räty,R.,He,Q.,Hovi,T., Mertsola,J.,Bloigu,A.,Rytilä,P.,etal.,2006.Pathogenicbacteriaandvirusesin inducedsputumorpharyngealsecretionsofadultswithstableasthma.Thorax 61,579–584.

Iramain,R.,Jara,A.,Jimenez,J.,Bogado,N.,Cardozo,L.,2016.Chlamydiapneumoniae, andMycoplasmapneumoniae:aretheyrelatedtosevereasthmainchildhood? JournalofAsthma53,618–621.

Kavanagh,H.,Noone,C.,Cahill,E.,English,K.,Locht,C.,Mahon,B.P.,2010.Attenuated BordetellapertussisvaccinestrainBPZE1modulatesallergen-inducedimmunity andpreventsallergicpulmonarypathologyinamurinemodel.Clinical& ExperimentalAllergy40,933–941.

Kraft,M.,Cassell,G.H.,Pak,J.,Martin,R.J.,2002.Mycoplasmapneumoniaeand Chlamydiapneumoniaeinasthma:effect ofclarithromycin.Chest121,1782–1788. Lavan,R.,Knesl,O.,2015.Prevalenceofcanineinfectiousrespiratorypathogensin asymptomaticdogspresentedatUSanimalshelters.JournalofSmallAnimal Practice56,572–576.

Mannering,S.A.,McAuliffe,L.,Lawes,J.R.,Erles,K.,Brownlie,J.,2009.Straintypingof Mycoplasmacynosisolatesfromdogswithrespiratorydisease.Veterinary Microbiology135,292–296.

Martin,R.J.,Kraft,M.,Chu,H.W.,Berns,E.A.,Cassell,G.H.,2001.Alinkbetween chronicasthmaandchronicinfection.JournalofAllergyandClinical Immunology107,595–601.

Medina,J.L.,Coalson,J.J.,Brooks,E.G.,Winter,V.T.,Chaparro,A.,Principe,M.F., Kannan,T.R.,Baseman,J.B.,Dube,P.H.,2012.MycoplasmapneumoniaeCARDS toxininducespulmonaryeosinophilicandlymphocyticinflammation. AmericanJournalofRespiratoryCellandMolecularBiology46,815–822. Nicolai,A.,Nenna,R.,Stefanelli,P.,Carannante,A.,Schiavariello,C.,Pierangeli,A.,

Scagnolari,C.,Moretti,C.,Papoff,P.,Bonci,E.,etal.,2013.Bordetellapertussisin infantshospitalizedforacuterespiratorysymptomsremainsaconcern.BMC InfectiousDiseases13,526–532.

Peeters,D.,McKiernan,B.C.,Weisiger,R.M.,Schaeffer,D.J.,Clercx,C.,2000. Quantitativebacterialculturesandcytologicalexaminationofbronchoalveolar lavagespecimensindogs.JournalofVeterinaryInternalMedicine14,534–541. Peeters,D.,Clercx,C.,Day,M.J.,2005.Distributionofleucocytesubsetsinbronchial

mucosafromdogswitheosinophilicbronchopneumopathy.Journalof ComparativePathology133,128–135.

Priestnall,S.L.,Mitchell,J.A.,Walker,C.A.,Erles,K.,Brownlie,J.,2014.Newand emergingpathogensincanineinfectiousrespiratorydisease.Veterinary Pathology51,492–504.

Rajamäki,M.M.,Järvinen,A.K.,Sorsa,T.,Maisi,P.,2002.Clinicalfindings, bronchoalveolarlavagefluidcytologyandmatrixmetalloproteinase-2and-9in caninepulmonaryeosinophilia.VeterinaryJournal163,168–181.

Rozanski,E.,2014.Caninechronicbronchitis.VeterinaryClinicsofNorthAmerica: SmallAnimalPractice44,107–116.

Ruch-Gallie,R.,Moroff,S.,Lappin,M.R.,2016.Adenovirus2,Bordetella

bronchiseptica,andparainfluenzamoleculardiagnosticassayresultsinpuppies aftervaccinationwithmodifiedlivevaccines.JournalofVeterinaryInternal Medicine30,164–166.

Rycroft,A.N.,Tsounakou,E.,Chalker,V.,2007.SerologicalevidenceofMycoplasma cynosinfectionincanineinfectiousrespiratorydisease.Veterinary Microbiology120,358–362.

Schulz,B.,Kurz,S.,Weber,K.,Balzer,H.,Hartmann,K.,2014.Detectionofrespiratory virusesandBordetellabronchisepticaindogswithacuterespiratorytract infections.VeterinaryJournal201,365–369.

Wakashin,H.,Maezawa,Y.,Kagami,S.,Suto,A.,Watanabe,N.,Saito,Y.,Hatano,M., Tokuhisa,T.,Iwakura,Y.,Puccetti,P.,etal.,2008.IL23andTh17cellesenhance TH2-cellmediatedeosinophilicairwayinflammationinmice.AmericanJournal ofRespiratoryandCriticalCareMedicine178,1023–1032.

Ye,Q.,Xu,X.J.,Shao,W.X.,Pan,Y.X.,Chen,X.J.,2014.Mycoplasmapneumonia infectioninchildrenisariskfactorfordevelopingallergicdiseases.The ScientificWorldJournal2014,1–11ID9865277.

Yin,S.S.,Ma,F.L.,Gao,X.,2017.AssociationofMycoplasmapneumoniaeinfection withincreasedriskofasthmainchildren.ExperimentalandTherapeutic Medicine13,1813–1819.

Zeugswetter,F.,Weissenböck,H.,Shibly,S.,Hassan,J.,Spergser,J.,2007.Lethal bronchopneumoniacausedbyMycoplasmacynosinalitterofgoldenretriever puppies.VeterinaryRecord161,626–627.