HAL Id: hal-01191111
https://hal.archives-ouvertes.fr/hal-01191111
Submitted on 1 Sep 2015
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
High prevalence of Hepatitis E virus in French domestic pigs
Nicolas Rose, Aurélie Lunazzi, Virginie Dorenlor, Thiziri Merbah, Florent Eono, Marc Eloit, François Madec, Nicole Pavio
To cite this version:
Nicolas Rose, Aurélie Lunazzi, Virginie Dorenlor, Thiziri Merbah, Florent Eono, et al.. High prevalence
of Hepatitis E virus in French domestic pigs. Comparative Immunology, Microbiology and Infectious
Diseases, Elsevier, 2011, pp.419-427. �10.1016/j.cimid.2011.07.003�. �hal-01191111�
ContentslistsavailableatScienceDirect
Comparative Immunology, Microbiology and Infectious Diseases
jou rn a l h o m e pa g e:ww w . e l s e v i e r . c o m / l o c a t e / c i m i d
High prevalence of Hepatitis E virus in French domestic pigs
Nicolas Rose
a,∗, Aurélie Lunazzi
b, Virginie Dorenlor
a, Thiziri Merbah
b, Florent Eono
a, Marc Eloit
c, Franc¸ ois Madec
a, Nicole Pavio
baAnses,LaboratoiredePloufragan/Plouzané,BP53,22440Ploufragan,France
bAnses,Laboratoiredesantéanimale,ENVA,INRA,23AvenueduGeneraldeGaulle94706Maisons-Alfort,France
cAnses,ENVA,INRA,EcoleNationaleVétérinaired’Alfort,7AvenueduGeneraldeGaulle94704Maisons-Alfort,France
a rt i c l e i n f o
Articlehistory:
Received27May2011
Receivedinrevisedform28July2011 Accepted29July2011
Keywords:
HepatitisE Pig
Animalreservoir Zoonosis Publichealth
a b s t r a c t
TheimportanceofthedomesticpigreservoirforHepatitisEvirus(HEV)wasassessed byestimatingtheseroprevalenceandprevalenceofHEVcontaminatedliversinFrench slaughter-agedpigs.6565seraand3715liverswererandomlysampledfrom186pigfarms throughoutthecountry.Takingthesamplingdesignintoaccount,thefarm-levelsero- prevalencewas65%(95%CI57–74)and31%(95%CI24–38)oftheslaughter-agedpigs hadantibodiesagainstHEV.TheindividualprevalenceofHEVRNApositiveliverswas4%
(95%CI2–6)and24%(95%CI17–31)ofthefarmshadatleast1positiveliver.Mostiso- lateswereofgenotype3f(76.7%)withsmalleramountsof3c(18.6%)and3e(4.6%).The highprevalenceofHEVinpigsandthesimilaritiesbetweenHEVsubtypesfrompigsand humanscorroboratesthepossiblezoonoticoriginofsomeHEVautochthonousinfections.
© 2011 Elsevier Ltd. All rights reserved.
1. Introduction
Hepatitis E virusis a non-enveloped,positive sense, singlestrandedRNAvirusofapproximately7.2kilobases and sole member of the Hepevirus genus in the Hep- eviridae family [1]. In humans, it is responsible for an acuteenterically-transmittedhepatitissimilartoHepati- tisA.Somecasescanbeverysevereandleadtofulminant hepatitis (1–2%ofthecases).HepatitisE ismostly self- limitingandgenerallydoesnotprogresstochronicity[2,3]
althoughseveralchroniccaseshavebeenreported[4,5]in patientsunderimmunosuppressive treatment,whomay develop cirrhosis [5]. The four main genotypes of HEV [3]showadistinctgeographicaldistribution.Genotypes1 and2areexclusivelyrecoveredfromhumansassporadic casesorlargeoutbreaksinAsianandAfricancountriesand alsoinMexicoforgenotype2,whereasgenotypes3and 4aresharedbetweenhumansandanimals.Theselatter
∗ Correspondingauthor.Tel.:+33296016441;fax:+33296016295.
E-mailaddress:nicolas.rose@anses.fr(N.Rose).
genotypesarecommonlyassociatedwithlocally-acquired HepatitisEcasesinNorthAmerica,Europe,JapanandChina forgenotype3,andJapan,ChinaandTaiwanforgenotype 4.
InFrance,thenumberoflocally-acquired HepatitisE casesreportedtothenationalreferencecentreincreased between2002(9cases)and2009(184cases)whilethe numberofimportedcasesremainedstable[6,7].
InmostcasestheoriginofautochthonousHepatitisE is unknown,but foodborneinfection wasclearly estab- lishedintwocaseswhereHEVtransmissionfollowedthe consumptionofrawdeermeat[8]orundercookedwild boarmeat [9].In France,somesporadic cases,and also foodborne outbreaks,were reportedbetween 2008and 2009,andtheconsumptionofrawpigliversausageswas stronglysuspectedasthesourceofinfection[10].However thisevidenceoffoodbornesourcesisrelativelysparseand notalwayslinkeddirectlywiththelocally-acquiredcases reportedannuallyinFrance.Forsomecases,theNational ReferenceCentrequestionnairescitedafrequentconsump- tionofrawsaltedporkmeatandtheconsumptionofwater fromaprivatesource[6].InGermanyacase-controlstudy 0147-9571/$–seefrontmatter© 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cimid.2011.07.003
420 N.Roseetal./ComparativeImmunology,MicrobiologyandInfectiousDiseases34 (2011) 419–427
identifiedtheconsumptionofwildboarmeatandoffalas riskfactorsforHepatitisEinfectioninhumans[11].Inthe UnitedStates,thereportedHEVprevalenceinalargesero- prevalencestudywas21.0%andhavingapetinthehome andconsuming liverorotherorganmeatsweresignifi- cantlyassociatedwithincreasedoddsofHEVseropositivity [12].
Serological studies in France evidenced significant differencesaccordingtogeographicallocation.Thepreva- lence of anti-HEV IgG in blood donors was 16.4% in southwestFrance[13] compared to3.2% in theregions Ile deFrance and Pays de la Loirewhich are in north- westFrance[14].Furthermore,anationalsurveyhasshown anincreasingNorth-to-SouthgradientofacuteHepatitis E[15]inagreementwithamorefrequentHEVexposure insouthern France. Thesedifferences according togeo- graphicalareahavenotbeenclearlyexplainedbutmight berelatedtoamorefrequentconsumptionofrawpigliver- basedproductsintheSouthofFrance[10].
HepatitisEvirusiswidespreadinthedomesticpigpop- ulation.Serologicalstudiescarriedoutinmanycountries showedahighprevalenceatthefarmlevel[16–20]with largevariationsattheindividuallevel[16,21–23]andalso betweenstudiesaccordingto geographicallocation and theserologicaltestsused.A few dataarealsoavailable forviremiaandviruspresenceinslaughter-agedpigs.The prevalenceoflivercontaminationinsamplescollectedin grocerystoreswas11.2%intheUS[24]and 6.5%inthe Netherlands[25].However,those figureswereobtained fromarelativelysmallsample.
IntheUnitedStatesandSweden,studiesofHEVpreva- lenceinpighandlersandveterinaryworkersrevealeda higherthan normalantibodyprevalencein thesepopu- lations[26,27].SwineveterinariansintheUSstudywere 1.5times(95%CI:1.03–2.20)moreatriskofbeingHEV seropositivethannormalblooddonorsinthematchedcon- trolpopulation.Thus, frequentcontactswithpigsmight increasetheprobabilityofHEVinfectionthroughexposure tocontaminatedmaterials.
Theaimofthepresentstudywasthereforetoquantita- tivelyassesstheimportanceofthedomesticpigpopulation inFranceasapotentialreservoirforHepatitisEvirus.A nationwideprevalencesurveywascarried outtodeter- mine(i)HEVseroprevalenceatthefarm-andpig-levels, (ii)theprevalenceofHEVinpigliversatslaughterand(iii) theHEVsubtypescirculatinginpigpopulation.
2. Materialsandmethods 2.1. Samplingdescription
Datafromtheliterature,inwhichtheobservedfarm- levelprevalencewashigh,wereusedtodeterminethetotal numberofherdstosample.Previousdataobtainedfroma limitednumberofslaughterhousesinFrancehadindicated afarm-levelprevalencecloseto70%[28]whichisconsis- tentwithdatafromothercountries.Thenumberofherds requiredtoestimate70%with10%relativeprecisionand 95%confidence,was165.Thisnumberwasincreasedto 186toallowforuncontrolledevents.Theobservedmini- mumwithin-herdprevalenceinthissamestudywasclose
to10%andthisvaluewasretainedastheminimumwithin- herdtargetprevalencetobedetected.Giventheexpected 50%sensitivityand98%specificityofthecommercialsero- logicaltests[28],thisledtosamplingof30pigsinflocks withlessthan50pigs,40pigsinflocksof50–100pigsand 50pigsinflockswithmorethan100pigs,withaconfidence levelof95%.Giventhetotalnumberofherdstobesampled, serologicalresultsfrommorethan5500pigswerethere- foreexpected.Takingintoaccountthemultistagecluster design(sampledpigsasprimarysamplingunitsarenested withinherdsandwithinslaughterhouses)thiswouldallow anexpectedindividualprevalenceof50%tobedetermined with5%relativeprecision.Thenumberofliverstobesam- pledperherdwasfixedat20whatevertheherdsize.This samplesizewouldleadtotheanalysisof3700liversand thusdeterminationofanexpected11%individualpreva- lencewitharelativeprecisionof20%,takingintoaccount themultistageclusterdesign.
The herds to be sampled were determined by ran- domselectionofalistofslaughterdatesandtimesfrom adatabasetable.Thisdatabasewasconstitutedbycom- pilingallpossibleslaughterdateandtimefromMay1st 2008toNovember30th2009forthe35selectedslaughter- houses(whichrepresentedmorethan95%ofthenational production according toa preliminarycensussurvey in Frenchslaughterhouses).Thenumberofherdstobesam- pledperslaughterhousewasdeterminedfromthenumber of pigs slaughtered/year and taking a minimum of 4 herds/slaughterhouse.TheSURVEYSELECTprocedure(SAS 9.1)wasusedtorandomlydefinethelistofdatesandtimes totaketheherd, accordingtothenumber requiredper slaughterhouseandstratifyingaccordingtoseason.
2.2. Biologicalsamples
The herds to sample were chosen according to the randomly selected date and time. Care wastaken that pigsfromaselectedherdwerenotmingledwithothers froma differentherd. Bloodsamplesweretaken atthe bleedingpost.Tominimizecross-contaminationofsam- ples at thebleeding post,empty blood collection tubes withoutanyadditivewerekeptbyanoperatorina test tube rack, holding them apart from the bleeding post.
Emptytubesweregiventothesecondoperatorforeach pigseparatelytobefilledinjustafterexsanguination.The filledtubewasimmediatelycorkedandstoredinasepa- ratebagbeforeanewonewasfilledagain.Eachsampled pig was identified with an ear tag to ensure that liv- ersweresampledfrommatchedcarcasses.Smallsections (2cm×1cm×1cm)werecutfromtheareaimmediately abovethegall-bladder(leftmediallobe)ofselectedlivers changingglovesandbladesbetweeneachlivertoprevent fromcross-contaminationfromonelivertoanother.These sampleswereimmediatelyplacedat4◦C,frozenandthen storedat−80◦Cuntilanalysis.
2.3. Serologicalanalysis
Serum samplescollectedattheslaughterhousewere testedwiththeAnti-HEVtotalimmunoglobulinforHuman diagnosisEIAgenHEVAbKit®Adaltis(Ingen,France).All
themanufacturer’sinstructionswerefollowed,exceptthat the secondary antibody was replaced by a peroxidase- conjugatedrabbitpolyclonalanti-pigIgG(H+L)(Abcam, France)diluted1:8000inthedilutionbufferIDVet3which isusedforthedilutionofsecondaryantibodiesconjugated toperoxidasein ELISAtest (IDVet,Montpellier,France).
For dataanalysis, thecut-off value (Co)was calculated as the mean optical density (OD450nm) value of the negative control (NC)+0.350. Theexperiment wasvali- datedwithavalueofNC−blankcontrol(reagentsonly)
<0.05. Results were interpreted as the ratio of sample (S) OD450nm tothecut-off value (orS/Co) as follows:
S/Co<1.1=negative;S/Co>1.1=positive.Eachsamplewas subjectedtomonocupuleanalysis.Thenegativecontrolisa serumobtainedfromaSpecificPathogenFreepigfromthe Level3animalfacilityattheAnsesLaboratoryinPloufra- gan(France)[29].Thepositivecontrolisapoolofpositive swineseraidentifiedpositiveinapreviousstudyonevalu- atingtheperformanceofavirus-likeparticles-basedELISA forserologyofHEVinswineandtheElisatestusedinthe presentstudy[28].
2.4. MoleculardetectionofHEV
RNAwasextractedfromliversamplesusingtheRNeasy MiniKit(Qiagen,France)accordingtothemanufacturer’s instructions.Thirtymilligramsoflivertissueswerehomog- enized in lysisbuffer (RNeasyMiniKit, Qiagen,France) usingtheFastPrep24System(MP Biomedicals,France) in Lysismatrix D tubes (MP Biomedicals,France). After elutingtheRNAinwater,RT-nestedPCRwasperformed onfivemicrolitersaccordingtotheprotocoldescribedby Cooperandcollaborators[17].Theamplifiedfragmentof 348 nucleotides is located in the 5 region of theHEV ORF-2.Toavoidcontaminationswiththesensitivenested RT-PCR methodused, sampleswere analyzed following theGoodLaboratoryPracticeandtheunidirectionalwork- flow in 4 separaterooms. In addition, allPCR products weresequencedanddifferentHEVstrainswereidentified ineachherdconfirmingtheabsenceofcrosscontamina- tion[30].RepeatabilityofHEVRNAisolationwasassessed byindependentanalysis(extractionandmoleculardetec- tion)of140liversamplesfrom7herds.Thesameresults (positive and negative samples) were obtained in both experiments(datanotshown).Then,liversampleswere analyzed in singledetection and positive sampleswere confirmedusingthereal-timeRT-PCRprocedurepublished
byJothikumarandcollaborators[31](datanotshown).To controlthe extractionand detection procedures,a pos- itive liversample from anHEVexperimentally infected swine,withagenotype3HEV(Genbankaccessionnum- berEF494700),wasrunsimultaneouslywitheveryseries of 20 liversamples fromthe sameherd. Negativecon- trols (Reverse transcription, first PCR and nested PCR) were included in every series. The presence of RT-PCR inhibitor was investigated per series of 20 samples, by addingHEVRNAextractedfromapositiveliverfroman experimentally-infectedpigtooneofthenewlyextracted sample.Positivesamplesweredispatchedforsequencing toEurofinsMWGOperon(Ebersberg,Germany).
Foreachpositiveliversample,moleculardetectionof HEVwasperformedonthecorrespondingsera.RNAextrac- tionwasperformedontwohundredmicrolitersofserum usingtheQiampViralRNAextractionKit(Qiagen,France).
HEVamplificationwasperformedon5laccordingtothe protocoldescribedbyCooperandcollaborators[17].Neg- ativeandpositive(serumfromanexperimentally-infected pig or negative control) controls (reverse transcrip- tion, first PCR and nested PCR) were included in each experiment.
2.5. Statisticalanalysis
The design of the survey (multistage cluster sam- pling,unequalweightingofobservations)wastakeninto account in determining the seroprevalenceof fattening pigsandtheprevalenceofHEVcontaminatedlivers.The sampling rate was calculated for each slaughterhouse usingpreviouslycollecteddataonnumberofpigsslaugh- tered/year/slaughterhouse updated for years 2008 and 2009.TheTaylorexpansionmethodprovidedinProcSUR- VEYFREQ[32]wasusedtoestimatethesamplingerrorof estimatorsbasedoncomplexsampledesigns[33,34].For theseroprevalenceresults,theindividualsensitivity(Se) andspecificity(Sp)ofthetestestimatedpreviously[28]
wasusedtocorrecttheestimateattheindividuallevel.
Theherd-levelseroprevalenceestimatewasalsocorrected bycalculatingvaluesforherd-sensitivity(HSe)andherd- specificity(HSp)asdefinedinDohooetal.[35].
Theinfluenceof season,geographical region,typeof herdandrelationshipsbetweenserologicalandvirologi- calresultswereassessedbylogisticregression,takinginto accountthesurveydesignandunequalweightingofobser- vationswiththeSURVEYLOGISTICprocedure[32].
Table1
SeroprevalenceandHEVprevalenceestimatesfromtheNationalprevalencestudy(186farms,6565bloodsamples,3715livers,France,2008–2009).
Samplesize Number
positive
Prevalence estimate(%)a
95%Confidence intervalb
Design effectc HEVserology
Individualpig-level 6565 1069 31 24–38 12.4
Farmlevel 186 137 65 57–74 1.2
HEVvirology(livers)
Individualpig-level 3715 128 4 2–6 9.1
Farmlevel 186 43 24 17–31 1.4
aBasedonsampledesign,appliestothetargetpopulation.
bBasedonvarianceestimateusingtheTaylorseriesexpansionmethod.
cRatiooftheactualvariance(estimatedbasedonthesampledesign)tothevarianceofasimplerandomsamplewiththesamenumberofobservations.
422 N.Roseetal./ComparativeImmunology,MicrobiologyandInfectiousDiseases34 (2011) 419–427
Fig.1.Distributionoftheobservedwithin-farmseroprevalence(186farms,2008–2009,France).Chartrepresentingthenumberoffarms(Yaxis)basedon thepercentageofpositiveseraperfarm(Xaxis).
3. Results
3.1. HEVseroprevalence
Takingthesamplingdesignandsensitivityandspeci- ficityoftheserologicaltest intoaccount,thefarm-level seroprevalencewas65%with[57–74]as95%confidence interval and 30% of the individual slaughter-aged pigs on average had antibodies against HEV (Table 1). The observedwithin-farmseroprevalencerangedfrom3%to 88% (medianat 10%, Fig.1). Five regions were defined accordingtothelocationofthesampledfarms(Fig.2).More than60%ofthenationalpigproductioncomesfromwest- ern France (North-West+Center-West). When different parameterssuchasthegeographicalregionoforigin,sea- sonandfarmtypewereconsidered,asignificantregional effectwasonlyfoundatthefarm level(p=0.03), farms locatedinwesternFrance,themainFrenchpigproducing area,beingmorelikelytobeseropositive(OR=2.4[1–5]) (Table2).Theeffectsofseasonandfarmtypewerenon- significant(p=0.28andp=0.51,respectively).
3.2. HEVprevalenceinlivers
Amongthe3715liverssampledattheslaughterhouse, 128were found positive for HEVRNA (Fig. A1). When thesamplingdesign,clusteringandunequalweightingof observationsweretakenintoaccount,theestimatedindi- vidualprevalencewas4% with[2–6]as 95%confidence interval(Table1).Asignificantregionaleffectwasfound, theNorth-Westareabeingmoreatriskthanotherregions (Table 2). This effect was more pronounced when the westernareas(North-West+Center-West)werecombined and compared with the other areas (OR=3.9 [1,5–11]).
Noseasonal effectwas found(p=0.96) and thetype of
farm(farrow-to-finishversusfinishingfarm)wasalsonon- significant(p=0.09)(Table2).
Prevalenceestimatesatthefarm-levelwereobtained byconsideringthatthefarmwaspositiveifatleast1liver testedpositive.Fortythreeofthe186sampledfarmshad at least1HEVpositive sample,leading toanestimated farm-levelprevalenceof24%with[17–31]as95%confi- denceintervalandtakingintoaccountthesamplingdesign, clusteringandunequalweighting(Table1).Theobserved within-farm prevalenceranged from 5 to 75% with an extremelyleft-skeweddistribution(Fig.3).Theprobabil- ityofafarmhavingatleast1HEVpositivepigwasgreater in thewesternarea(North-West+Center-West)than in therestofFrance(OR=3.7[2,7,9]),andinfinishingfarms comparedwithfarrow-to-finishfarms(OR=2.6[1–6]).In agreementwithindividualprevalencefindings,theeffect ofseasonwasnon-significant(p=0.86)(Table2).Astrong relationshipwasobservedbetweentheprobabilitythata liverwouldbeHEVpositive andwithin-farmseropreva- lence.Theoddsofaliverbeingcontaminatedwerealmost 7timeshigherwhenthewithin-farmseroprevalencewas greaterthan25%(Table2).Attheindividuallevel,6.6%of theliversweredetectedHEVpositiveinseropositiveani- malsversus2.6%inseronegativepigs.
SequencingofthePCRproducts,tocharacterizetheviral straincirculatinginthepigreservoir,revealedthatallviral strainsidentifiedintheliversamplesbelongedtogenotype 3and moreparticularlytosubtypes3e,3cand3f.Most wereofgenotype3f,76.7%(33/43)while3cwasidentified in18.6%(8/43)oftheliversamplesand3ein4.6%(2/43).
75.8%ofgenotype3f(n=33)camefromherdslocatedinthe North-West(mainlyBrittany)whereas3cgenotypeswere morefrequentlyfoundintheSouth-WestandNorth-East (62.5%,n=8).
To confirm that the amplification of viral RNA in the liver samples was correlated with the presence of
Fig.2.Definitionofregionsaccordingtothelocationofsampledfarmsandnumberoftestedfarmspercounty(186sampledfarms,2008–2009,France).
infectiousHEVparticles,onepositiveliversamplewasused inanexperimentalmodelofpiginfection.Pigsinoculated intravenouslywiththisliversamplestartedtoshedHEV intheirfeces2daysafterinfection(datanotshown)and seroconversionwasalsoobserved31dayspost-inoculation (datanotshown),thusconfirmingHEVinfection.
ThepresenceofHEVRNAintheserumsamplescor- respondingtopositiveliversampleswasthenexamined toevaluateifthepresenceofHEVRNAwasconcomitant
withviremiaandpossiblevirusdisseminationinthewhole organism.HEVRNAwasamplifiedin17ofthe128sera (21.1%)correspondingtopigswithanHEVpositiveliver.
4. Discussion
Anaccuratequantitativeassessmentoftheroleofpigs as potential HEV reservoirs is lacking although several figureshavebeenreportedfordifferentcountriesinthe
424 N.Roseetal./ComparativeImmunology,MicrobiologyandInfectiousDiseases34 (2011) 419–427
Table2
FactorsaffectingHEVseroprevalenceandprevalenceinlivers(186farms,6565bloodsamples,3715livers,France,2008–2009).
Outcomevariableand categories
HEVserology HEVvirology(livers)
Individualpig-level Farm-level Individualpig-level Farm-level
OR(95%CI)a pvalue OR(95%CI) pvalue OR(95%CI) pvalue OR(95%CI) pvalue
Region 0.22 0.03 0.001 0.06
North-West 0.8(0.4–1.4) 0.7(0.1–4.8) 4.4(1.2–15.9) 2.0(0.5–7.8)
Center-West 0.7(0.3–1.7) 0.4(0.05–2.9) 2.6(0.5–14.2) 1.0(0.2–5.9)
North-East 0.5(0.2–0.9) 0.2(0.02–1.2) 0.4(0.08–1.8) 0.4(0.07–1.9)
South-East 0.5(0.2–1.6) 0.3(0.02–4.2) 3.1(0.6–17.1) 1.2(0.2–7.6)
South-West 1 1 1 1
Areaofproduction 0.72 0.03 0.01 0.003
WesternFrance 1.1(0.6–2.1) 2.4(1.1–5.4) 3.9(1.4–11.0) 3.7(1.6–8.7)
Other 1 1 1
Season 0.34 0.28 0.96 0.86
Summer 0.8(0.5–1.4) 0.4(0.1–1.1) 1.0(0.3–3.2) 1.1(0.4–3.3)
Autumn 1.1(0.7–1.9) 0.7(0.2–1.7) 1.2(0.3–5.2) 0.8(0.3–2.1)
Winter 1.5(0.8–3.0) 0.8(0.2–2.9) 1.4(0.4–4.9) 1.2(0.4–3.8)
Spring 1 1 1
Farmtype 0.37 0.51 0.09 0.03
Finishing 1.3(0.7–2.2) 1.4(0.5–3.7) 2.3(0.9–5.9) 2.6(1.1–6.0)
Farrow-to-finish 1 1 1 1
HEVindividualserology NAb NA NA NA 0.01 NA NA
Positive 2.8(1.2–6.2)
Negative 1
HEVwithin-farm seroprevalence(%)
NA NA NA NA 0.006 0.07
>25 6.7(2.1–21.6) 3.7(1.1–12.1)
[0–25] 3.2(0.9–11.2) 1.8(0.6–5.3)
Negative 1 1
Boldfaceindicatesasignificantresult.
aOddsratioand(95%confidenceinterval).
b Notapplicable.
Fig.3. Distributionoftheobservedwithin-farmprevalenceofHEVcontaminatedlivers(43HEVpositivefarms,2008–2009,France).Chartrepresenting thenumberoffarms(Yaxis)basedonthepercentageofpositiveliverswithineachfarm(Xaxis).
literature. In most of those studies both the sampling strategyandestimatesprecisionwerepoorbecause“con- venient”samplesweretaken,ingeneral,andfromalimited numberoffarmsandanimals[24,25,36–38].Tothebest ofourknowledge,thepresentresultsarethefirstavail- able froma representativenationwide surveyin which HEVantibodiesandviruswereexploredsimultaneously inliversfromslaughter-agedpigs.Inviewofthecomplex samplingdesign, the obtainedprevalenceestimates are applicabletothetargetpopulation.Thisisofconsiderable interestwhenconductingaquantitativeriskassessmentto estimatethenumberofHEVcasesinhumansthatcanbe attributedtotheconsumptionofpork-derivedfoodprod- ucts.
Themainriskfor publichealth consistsofdelivering pigstotheslaughterhousewhicheitherharbourthevirus indifferentorganssuchasgallbladder,intestineormus- clesorareviremic aspreviouslyshowninexperimental HEVinfectionofpigs oratslaughterhouse[39,40].Posi- tiveHEVliverswerefoundinseropositivepigs(6.6%)but alsoinseronegativeanimals(2.6%),which suggeststhat positiveandnegativeserologicalresultscouldnotexclude the presence of thevirus in slaughter-aged pigs in the case of recent infection [39,41]. Oneinteresting finding from ourstudy is the positive association betweenthe probabilityof liversbeingHEVpositive andthewithin- herdHEVseroprevalence.Thisclearlysuggeststhatspecific on-farm conditions favor virus spread and increase the likelihoodofdeliveringinfectedslaughter-agedpigs.Pigs fromfinishingfarmsweremorelikelytobeHEVpositive than those from farrow-to-finish farms,which suggests thatcharacteristicsrelatedtothisrearingsystem(suchas thepossible minglingof pigs of differentorigins)could be a potential risk factor. This suggests that one way of controlling the overall level of HEV infection in the pigreservoirwouldbetodeveloparearingmanagement plan.
TheobservedfrequenciesofHEVpositiveliversinother countries(TheNetherlands[25]orUSA[24])werehigher thantheobservedprevalenceinourstudy.Inthestudyper- formedintheUSA,thesamedetectionmethodwasused, thus,thisdifferencecouldbeexplainedbythesampling strategyaswellasthedifferentfarmingsystemsinsuch countries,thefarrow-to-finishsystembeingdominantin Franceandthewean-to-finishorfinishingsystemswith collectivefarrowingunitsbeingthemainsystemsencoun- teredinTheNetherlandsandUSA.ThepresenceofHEV RNAinserumsamplesalsosuggeststhattheconsumption
ofpigorgansotherthanliver(e.g.muscles)mightconsti- tutearisk.Manypork-derivedfoodproductsareconsumed afterabriefdryingprocess(dryham)thatmaynotinacti- vateHEV.InvestigationofHEVpresenceinotherpigorgans shouldbeconsideredaswellastheimpactofprocesseson HEVinactivation.
The regional effect showed that the probability of obtainingHEVpositiveliverswashigherinwesternFrance, whichisthiscountry’smainproductionarea.Thisobser- vationisincontradictiontotheobserveddistributionof humancasesinFrance(mostlylocatedintheSouth-East andSouth-West[15]).However,pigsfromwesternFrance areexportedandprocessedalloverthecountry.Incontrast totheeatinghabitsinnorthernFrance,thereisastrongtra- ditionspecifictotheSouthFranceforfoodproductsbased onrawporkliver.Theseproducts(Figatelli,smokedliver sausage,driedliversausage)aremanufactured,soldand consumedlocally[10].
InFrance,thenon-importedHepatitisEaremainlydue tosubtypes3f(upto88%),3cand3e[42,43],thelattertwo alsobeingthedominantsubtypesinpigsinmanyEuro- peancountries[44].HEVdetectionin pigliversausages showedamajorityof3f,followedbythe3cand3esub- types[10],whichisinlinewithourfindingsinliverstaken fromslaughter-aged pigs (3f=76.7%). Evidence of food- borneinfectionsthroughconsumptionofrawliver-based sausages[10] or undercookedoffal [45]as wellashigh seroprevalenceinprofessionalsexposedtopigs[26,27],all suggestapotentialinvolvementofthepigpopulationinthe epidemiologyofhumanHEVinfections.Furthermore,both thehighprevalenceinthedomesticpigpopulationandthe similardistributionofgenotypesinautochthonoushuman casessuggestahighpotentialforzoonotictransmission.
Giventhewidespreadofthevirusinthepigpopulation andtheamountofporkproductsconsumedinFrance,one wouldexpectmanymorecasesthanhavebeenobserved.
ViralhepatitisEisverylikelyunderdiagnosedsincetheHEV seroprovalenceinFrenchblooddonorsishigh(3.2–16.4%
dependingonthearea).Furtherdataareneededasregards theinfectiousdosesandtheeffectoffoodproductprocess- ingandcookingonsurvivalofthevirusinpork-derived products.
Conflictofintereststatement
The authors declare that they have no competing interests.
Fig.A1.DetectionbyRT-nestedPCRofpartialORF-2ofHEVinliversamplesfromoneselectedherd.Liversamplesarenumbered1–20.Positivesamples (fragmentof348basepair)areindicatedwithawhitearrow.InbC+:controlfortestingtheabsenceofinhibitor.ExtC+:positiveliverextractedandanalyzed simultaneously.RTC-,PCRC-andNPCRC-arerespectivelynegativecontroloftheRT,first-andnestedPCR.M:molecularweightmarker,sizeareindicated ontheleftinbasepair.
426 N.Roseetal./ComparativeImmunology,MicrobiologyandInfectiousDiseases34 (2011) 419–427 Acknowledgements
This study was supported by a national grant from AgenceNationale de la Recherche (PNRA07-008 HEVE- ZOONEPI).ALwassupportedbythisgrantfor24months (PNRA07-008HEVEZOONEPI).
AppendixA.
SeeFig.A1.
References
[1]EmersonSU,AndersonD,ArankalleA,MengXJ,PurdyM,Schlauder GG,TsarevSA.Genushepevirus.In:FauquetCM,MayoMA,Maniloff J,DesselbergerU,BallLA,editors.Virustaxonomy:eighthreport oftheInternationalCommitteeonTaxonomyofViruses.London:
Elsevier/AcandemicPress;2004.p.853–7.
[2]JameelS.MolecularbiologyandpathogenesisofhepatitisEvirus.
ExpRevMolMed1999;1:1–16.
[3] PurcellRH,EmersonSU.HepatitisEvirus.In:KnipeD,HoweP,edi- tors.Fieldsvirology.NewYork:Ravenpress;2001.p.3051–61.
[4]GerolamiR,MoalV,ColsonP.ChronichepatitisEwithcirrhosisina kidney-transplantrecipient.NEnglJMed2008;358:859–60.
[5] KamarN,SelvesJ,MansuyJM,OuezzaniL,PéronJM,GuitardJ,Coin- taultO,EspositoL,AbravanelF,DanjouxM,DurandD,VinelJP, IzopetJ,RostaingL.HepatitisEvirusandchronichepatitisinorgan- transplantrecipients.NEnglJMed2008;358(8):811–7.
[6] Nicand E, Bigaillon C, Tessé S. Hepatitis E in France: surveil- lance data for human cases, 2006–2008. Bull Epidémiol Hebd 2009;3(1–32):337–42.
[7]Nicand E, Delaune C, Tessé S. Données de surveillance des cas humains d’hépatite E en France 2006–2009. Feuill Biol 2011;298:19–24.
[8]Tei S, Kitajima N, Takahashi K, Mishiro S. Zoonotic transmis- sion of hepatitis E virus from deer to human beings. Lancet 2003;362(9381):371–3.
[9] Li TC, Chijiwa K, Sera N, Ishibashi T, Etoh Y, Shinohara Y, KurataY,IshidaM,SakamotoS,TakedaN,MiyamuraT.Hepati- tisEvirustransmissionfromwild boarmeat.EmergInfectDis 2005;11(12):1958–60.
[10] ColsonP,BorentainP,QuyriauxB,KabaM,MoalV,GallinaP,Heyries L,RaoultD,GerolamiR.PigliversausageasasourceofHepatitisE virustransmissiontohumans.JInfectDis2010;202(6):825–34.
[11] WichmannO,SchimanskiS,KochJ,KohlerM,RotheC,PlentzA,Jilg W,StarkK.Phylogeneticandcase-controlstudyonhepatitisEvirus infectioninGermany.JInfectDis2008;198(12):1732–41.
[12]KuniholmMH,PureellRH,McQuillanGM,EngleRE,WasleyA,Nel- sonKE.EpidemiologyofhepatitisEvirusintheUnitedStates:results fromthethirdnationalhealthandnutritionexaminationsurvey, 1988–1994.JInfectDis2009;200(1):48–56.
[13]MansuyJM,Legrand-AbravanelF,CalotJP,PeronJM,AlricL,Agudo S,RechH,DestruelF,IzopetJ.Highprevalenceofanti-hepatitisE virusantibodiesinblooddonorsfromSouthWestFrance.JMedVirol 2008;80(2):289–93.
[14]BoutrouilleA,Bakkali-KassimiL,CrucièreC,PavioN.Prevalenceof anti-hepatitisEvirusantibodiesinFrenchblooddonors.JClinMicro- biol2007;45(6):2009–10.
[15]RenouC,MoreauX,ParienteA,CadranelJ-F,MaringeE,MorinT, CausseX,PayenJ-L, IzopetJ,NicandE,BourlièreM,Penaranda G, Hardwigsen J, Gerolami R, Peron J-M, Pavio N. A national surveyofacute hepatitis Ein France.AlimentPharmacol Ther 2008;27(11):1086–93.
[16]BlacksellSD,MyintKSA,KhounsyS,PhruaravanhM,MammenJr MP,DayNPJ,NewtonPN.PrevalenceofhepatitisEvirusantibodies inpigs:implicationsforhumaninfectionsinvillage-basedsub- sistencepigfarmingintheLaoPDR.TransRSocTropMedHyg 2007;101(3):305–7.
[17]CooperK,HuangFF,BatistaL,RayoCD,BezanillaJC,TothTE,Meng XJ.Identificationofgenotype3hepatitisEvirus(HEV)inserumand fecalsamplesfrompigsinThailandandMexico,wheregenotype1 and2HEVstrainsareprevalentintherespectivehumanpopulations.
JClinMicrobiol2005;43(4):1684–8.
[18] Garkavenko O, Obriadina A, Meng J, Anderson DA, Benard HJ, SchroederBA,KhudyakovYE,FieldsHA,CroxsonMC.Detectionand
characterisationofswinehepatitisEvirusinNewZealand.JMed Virol2001;65(3):525–9.
[19]MengX-J,PurcellRH,HalburPG,LehmanJR,WebbDM,Tsareva TS,HaynesJS,ThackerBJ,EmersonSU.Anovelvirusinswineis closelyrelatedtothehumanhepatitisEvirus.ProcNatlAcadSci USA1997;94(18):9860–5.
[20] SeminatiC,MateuE,PeraltaB,deDeusN,MartinM.Distributionof hepatitisEvirusinfectionanditsprevalenceinpigsoncommercial farmsinSpain.VetJ2008;175(1):130–2.
[21]GuimaraesFR,SaddiTM,VitralCL,PintoMA,GasparAMC,SoutoFJD.
HepatitisEvirusantibodiesinswineherdsofMatoGrossoState, CentralBrazil.BrazJMicrobiol2005;36(3):223–6.
[22] MunneMS,VladimirskyS,OteguiL,CastroR,BrajtermanL,Soto S,GuarneraE,MolinaV,MonfellanoM,SchlauderGG,Gonzalez JE.Identification ofthefirst strainofswinehepatitisEvirusin SouthAmericaandprevalenceofanti-HEVantibodiesinswinein Argentina.JMedVirol2006;78(12):1579–83.
[23] TakahashiM,NishizawaT,MiyajimaH,GotandaY,LitaT,TsudaF, OkamotoH.SwinehepatitisEvirusstrainsinJapanformfourphylo- geneticclusterscomparablewiththoseofJapaneseisolatesofhuman hepatitisEvirus.JGenVirol2003;84(4):851–62.
[24] Feagins AR, Opriessnig T, Guenette DK, Halbur PG, Meng X-J.
DetectionandcharacterizationofinfectiousHepatitisEvirusfrom commercialpigliverssoldinlocalgroceryintheUSA.JGenVirol 2007;88(3):912–7.
[25]BouwknegtM,Lodder-VerschoorF,VanDerPoelWHM,RutjesSA.
HumanAMDRHepatitisEvirusRNAincommercialporcineliversin TheNetherlands.JFoodProt2007;70(12):2889–95.
[26]MengXJ,WisemanB,ElvingerF,GuenetteDK,TothTE,EngleRE, Emerson SU,PurcellRH.PrevalenceofantibodiestohepatitisE virusinveterinarians workingwithswineandinnormalblood donorsintheUnitedStatesandothercountries.JClinMicrobiol 2002;40(1):117–22.
[27]OlsenB,Axelsson-OlssonD,ThelinA,WeilandO.Unexpectedhigh prevalenceofIgG-antibodiestohepatitisEvirusinSwedishpigfarm- ersandcontrols.ScandJInfectDis2006;38(1):55–8.
[28]RoseN,BoutrouilleA,FabletC,MadecF,EloitM,PavioN.Theuse ofBayesianmethodsforevaluatingtheperformanceofavirus-like particles-basedELISAforserologyofhepatitisEvirusinfectionin swine.JVirolMethods2010;163(2):329–35.
[29]CarioletR,LeDiguerherG,EcobichonP,JulouP,JollyJP,MadecF.
Productionoflongterm,low-costspecificpathogenfreepigs.In:
ISPAIA,editor.SymposiumoftheInternationalSocietyforAnimal Hygiene;2004October11–13.2004.p.149.
[30]BouquetJ,TesséS,LunazziA,EloitM,RoseN,NicandE,PavioN.
ClosesimilaritybetweensequencesofHepatitisEVirusrecovered fromHumanandSwineinFrancebetween2008and2009.Emerg InfectDis2011,inpress.
[31]Jothikumar N,CromeansTL,RobertsonBH,MengXJ,HillVR.A broadly reactiveone-step real-timeRT-PCRassayfor rapidand sensitive detection of hepatitis E virus. J Virol Methods 2006 Jan;131(1):65–71.
[32]SASInstituteInc.SAS/STATUser’sGuide.Cary,NC,USA:SASInstitute;
2002.
[33] Fuller WA. Regression analysis for sample survey. Sankhy 1975;37:117–32[SeriesC(3)].
[34]WoodruffRS.Asimplemethodforapproximatingthevarianceofa complicatedestimate.JAmStatAssoc1971;66:411–4.
[35]DohooI,MartinW,StryhnH.VeterinaryEpidemiologicResearch.
Charlottetown,PrinceEdwardIsland,Canada:AVCInc.;2003.
[36]Di Bartolo I, Martelli F, Inglese N, Pourshaban M, Caprioli A, OstanelloF,RuggeriFM.Widespreaddiffusionofgenotype3hep- atitisEvirusamongfarmingswineinNorthernItaly.VetMicrobiol 2008;132(1–2):47–55.
[37]BreumSO,HjulsagerCK,deDeusN,SegalésJ,LarsenLE.HepatitisE virusishighlyprevalentintheDanishpigpopulation.VetMicrobiol 2010.
[38] DiBartoloI,PonterioE,CastelliniL,OstanelloF,RuggeriFM.Viral andantibodyHEVprevalenceinswineatslaughterhouseinItaly.
VetMicrobiol2011;149:3–4,330-8.
[39]Bouwknegt M, RutjesSA, Reusken CB, Stockhofe-ZurwiedenN, FrankenaK,deJongMC,deRodaHusmanAM,PoelWH.Thecourse ofhepatitisEvirusinfectioninpigsaftercontact-infectionandintra- venousinoculation.BMCVetRes2009;5(7).
[40]LeblancD,PoitrasE,GagnéMJ,WardP,HoudeA.HepatitisEvirus loadinswineorgansandtissuesatslaughterhousedeterminedby real-timeRT-PCR.IntJFoodMicrobiol2010;139(3):206–9.
[41]KanaiY,TsujikawaM,YunokiM,NishiyamaS,IkutaK,HagiwaraK.
Long-termsheddingofhepatitisEvirusinthefecesofpigsinfected
naturally,borntosowswithandwithoutmaternalantibodies.JMed Virol2011;82(1):69–76.
[42] Legrand-AbravanelF,KamarN,Sandres-SauneK,GarrousteC,Dubois M,MansuyJM,MuscariF,SallustoF,RostaingL,IzopetJ.Charac- teristicsofautochthonoushepatitisEvirusinfectioninsolid-organ transplantrecipientsinFrance.JInfectDis2010;202(6):835–44.
[43] Legrand-AbravanelF,MansuyJ-M,DuboisM,KamarN,PeronJ-M, RostaingL,IzopetJ,Hepatitis.Evirusgenotype3diversity,France.
EmergInfectDis2009;15(1):110–4.
[44]LuL,LiC,HagedornCH.PhylogeneticanalysisofglobalhepatitisE virussequences:geneticdiversity,subtypesandzoonosis.RevMed Virol2006;16(1):5–36.
[45]YazakiY,MizuoH,TakahashiM,NishizawaT,SasakiN,GotandaY, OkamotoH.SporadicacuteorfulminanthepatitisEinHokkaido, Japan, may be food-borne, as suggested by the presence of hepatitis E virusin pig liveras food.J Gen Virol 2003;84(9):
2351–7.
