Proteinuria during dengue fever in children
Anne-Claire Andries
a, Veasna Duong
a, Julien Cappelle
b,c, Sivuth Ong
a, Alexandra Kerleguer
d, Sowath Ly
b, Arnaud Tarantola
b, Paul F. Horwood
a, Anavaj Sakuntabhai
e,f, Philippe Dussart
a, Philippe Buchy
a,g,*
aInstitutPasteurinCambodia,InternationalNetworkofPasteurInstitutes,VirologyUnit,PhnomPenh,Cambodia
bInstitutPasteurinCambodia,InternationalNetworkofPasteurInstitutes,EpidemiologyandPublicHealthUnit,PhnomPenh,Cambodia
cCentredecoopérationinternationaleenrechercheagronomiquepourledéveloppement(CIRAD),DépartementES,UnitéAGIRs,Montpellier,France
dInstitutPasteurinCambodia,InternationalNetworkofPasteurInstitutes,Medicallaboratory,PhnomPenh,Cambodia
eInstitutPasteur,FunctionalGeneticsofInfectiousDiseasesUnit,Paris,France
fCentreNationaldelaRechercheScientifique,UnitéderechercheAssociée3012,Paris,France
gGlaxoSmithKline,VaccinesR&D,ScientificAffairsandPublicHealth,Singapore
ARTICLE INFO Articlehistory:
Received27November2016
Receivedinrevisedform20December2016 Accepted21December2016
CorrespondingEditor:EskildPetersen, Aarhus,Denmark
Keywords:
Dengue proteinuria urine dipstick UPCR
urineelectrophoresis
ABSTRACT
Objectives:Thisstudyaimedtoinvestigateproteinuriaoccurringduringdenguediseaseinchildrenand assessifmeasurementofthisparametercanhelpphysiciansintheclinicalmanagementofpatients.
Methods:Proteinuriawasassessedbydipstickandquantifiedbyurineprotein:creatinineratio(UPCR)in samplesfrompatientshospitalizedwithaconfirmeddengueinfectionandinhealthycontrols.
Results:Thedipsticktestedpositivein42.9%ofthepatientspresentingathospitalwithdengueversus 20.0%inhealthycontrols.UPCRincreasedduringthecriticalphaseofthedisease;peakingoneweekafter feveronsetthendecreasingasthepatientsrecovered.Patientswithwarningssignsorseveredengue weremore likelytopresentwithproteinuriadetected byUPCRat thetimeofhospital admission comparedtopatientswithoutwarningsigns.Thesensitivityofthismarker,however,waslimitedasonly 16.1%ofthepatientswithwarningsignshadproteinuria.
Conclusions:UrinedipstickandUPCRdonotseemtobeveryvaluableforthetriageofthepatientsatthe timeoftheinitialconsultationbuttheobservationofadecreaseoftheUPCRduringthecourseofthe illnessappearstoindicateanevolutiontowardsrecovery.
©2017TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.
ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by- nc-nd/4.0/).
Background
Dengue is the most prevalent mosquito-borne viraldisease worldwide.1 A majority of the infections are asymptomatic or resultinamildfebrileillness,butthedenguevirus(DENV)isalso capableofproducingalife-threateningdisease.Themainformof severedengueischaracterizedbyplasmaleakagewithorwithout bleeding,which maylead tocirculatorycollapse, calleddengue shocksyndrome.Thecourseofdengueillnesscanbedividedinto threemainphases:thefebrilephase,thecriticalphase andthe recoveryphase.Severeclinicaldiseasemanifestationsoccurduring thecriticalphasewhichbeginsaroundday4–7aftertheonsetof feverandlastsusually48–72hours.Duringthecriticalphase,the
conditionof patientscan improveor worsen rapidly;requiring carefulmonitoringbycaregivers.Earlyclinicalmanagementbased onfluidreplacementtherapyreducesthemorbidityandmortality associatedwithseveredengue.2
The major obstacle for an effective clinical management of dengueistheinabilitytoaccuratelypredict,atanearlystageof infection,whichpatientsarelikelytodevelopasevereformofthe disease.Thereis aneedfor simple,effectiveandcheapteststo identifypatientsatriskandguidetriage.Willsetal.observedan increase of urinary protein clearance due to the increase in systemic vascular permeability that occurs in severe dengue.
Subsequently, ithasbeenproposed that asimple urine protein excretionscreeningtestcouldbeindicativeofthesevereformof dengue and therefore guide the triage and monitoring of the patientswithsuspecteddengueinfection.3
Theobjectiveofthisstudywastoinvestigatethepresenceof proteinuriaduringdenguediseaseinchildrenbysimpleurinalysis strip and byprotein:creatinineratio(UPCR) and assessif these parameters can help the physicians to improve the clinical
*Correspondingauthorat:GlaxoSmithKline,VaccinesR&D,ScientificAffairsand PublicHealth,Singapore.
E-mailaddresses:[email protected],[email protected] (P.Buchy).
http://dx.doi.org/10.1016/j.ijid.2016.12.022
1201-9712/©2017TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).
ContentslistsavailableatScienceDirect
International Journal of Infectious Diseases
j o u r n a lh o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / i j i d
managementofdenguecases.Tobettercharacterizethedengue associatedproteinuriaprofiles,aurinaryproteinelectrophoresis wasperformedinasubsetofthehospitalizedpatients.
Methods Clinicalsamples
Atotalof241spoturinesamplesobtainedfrom108patients with laboratory-confirmed dengue virus infection (DVI) were randomlyselectedamongsamplesobtainedin2013duringthe DENgueresearchFrameworkforResistingEpidemicsinEurope (DENFREE)study.Aconfirmeddenguecasewasdefinedbythe detection of viral RNA by RT-PCR and/or the detection of the NS1 protein and/or an IgM seroconversion and/or a four-fold antibody titer increase measured by hemagglutination inhibi- tion assay (HIA) in paired plasma of patientspresenting with symptomssuggestiveofDVI.Mosturinesamples(225/241)were collected during hospitalization 1–11 days after the onset of fever. An additional 16 samples were collected one or three monthsafterdischargefromhospital.Patientswerecategorized as experiencing dengue “with” or “without” warning signs (WS+ or WS-, respectively) or severe dengue based on the 2009casedefinitionsestablished bytheWorldHealthOrgani- zation (WHO).2 The immune status of patients (primary or secondaryinfection)wasdeterminedbyHIAaccordingtoWHO criteria.4
Afurther15samplescollectedfromhealthychildrenrecruited duringacommunity-basedstudywerealsoaddedtothepanel.
Thesecontrolswerehouseholdmembersofsomeofthepatients identifiedduringtheDENFREEhospital-basedstudythathadno biologicalevidenceofDVI(RT-PCRnegative,NS1negative,noanti- DENVIgMorincreasingIgGtiter).
Ethicalstatement
TheDENFREEprojectwasapprovedbytheCambodianNational Ethics Committee for Health Research (authorization no.
063NECHR).Thechildren’slegalrepresentativessignedawritten consentbeforetheenrolmentofthepatient.
Urineanalysis
Thepresenceofproteininurinecollectedatthetimeofpatient admission at hospital was checked using semi-quantitative Mission urine dipsticks (Acon, San Diego, USA). The test was performedaccordingtothemanufacturer’sinstructions. Results weregradedas:noprotein;150mg/L(trace);300mg/L;1000mg/L;
3000mg/L; and 20,000mg/L. Presence of protein at any concentrationwas considered asa positiveresult. UPCRs were measuredinsamplescollectedduringthecourseofhospitaliza- tionusingaCobasintegra400plusanalyzer(RocheDiagnostics, Germany).ProteinuriawasdefinedasaUPCR45mg/mmol.5
Urineprotein electrophoresis (UPEP) was performed with a MINICAP capillary electrophoresis platform (Sebia, France).
Following themanufacturer’sinstructions, only sampleswitha totalproteinconcentration>100mg/Lweretested.
Statisticalanalysis
Statistical tests were performed using STATA version 11.0 (StataCorp,USA).The statisticaldifferencesbetweencategorical groupsweredetectedusingtheFisher’sexacttest.TheKruskal- Wallis rank test and the Mann Whitney U test were used for continuousindependentvariablesandtheWilcoxontestwasused forcontinuousdependantvariables.Thecorrelationbetweentwo
continuousvariableswasassessedbySpearman’srankcorrelation test.
A Generalized Additive Mixed Model (GAMM) was used to evaluatefactorsindependentlyassociatedwiththeUPCRvalues andwiththeoccurrenceofproteinuria.Fiveexplanatoryvariables wereincludedinthemodel:age,gender,samplingdayafteronset offever(daof), immunestatus(primaryorsecondaryinfection) and 2009 WHO diseaseclassification (denguewithoutwarning signs,denguewithwarningsigns,severedengue).AGAMMwas used because a non-linear relationship between the response variables (UPCRorproteinuria occurrence)andboth “daof” and
“age” explanatory variables was expected.A mixedmodel was usedwithpatientIDasarandomeffectexplanatoryvariableto takeintoaccountthenon-independenceofsamplescollectedfrom the same patients and potential random individual variations betweenpatients.Theanalysiswasperformedusingthe“gamm4” package (version 0.2-3) under the R statistical environment (RFoundation,Vienna,Austria).
Significancewasassignedatp<0.050.
Results
Atotalof108patientswereincludedinthisstudy.Asummary of the patients’ characteristics, clinical and virologic data is presentedinTable1.
Proteindetectionbydipstickinurinespecimenscollectedatadmission
Atotalof39patients(42.9%)testedpositiveforproteinuriaby dipstickatthetimeofadmissiontothehospital(Table2).Most positivesamples(71.8%, 28/39)onlycontainedtracesofprotein (150mg/L) while nine and two patients had an approximate proteinconcentrationof300mg/Land1000mg/L,respectively.A positivedipstickresultwasobservedfor38.1%(8/21)and32%(16/
50)ofthepatientsadmittedathospitalforadengueWS-andWS+, respectively(Fisher’sexact P=0.784).Proteinsweredetected in theurinesampleof75%(15/20)ofthepatientspresentingwitha severedengue.Threeurinesamplesofthecontrolgroup(20%,3/
15)containedtracesofprotein.Theprevalenceofproteinuriain denguepatientsWS-andWS+wasnotsignificantlydifferentfrom the one obtained for the control group (P=0.295 and 0.522, respectively). When considering the immune status of the patients, 53.2%(25/47)of thosewitha secondaryinfection and 26.9% (7/26) of the patients with a primary infection had proteinuria(P=0.048).
ProteinuriadeterminedbytheUPCRduringthecourseof hospitalization
The UPCR was tested at admission in 52 patients (median daof=4,IQR=[3–4])andproteinuriawasdetectedin23.1%(n=12) of the cases. None of the WS- dengue cases had proteinuria compared with 16.1% (5/31) in WS+ cases and 50% (7/14) in patientsexperiencingseveredengue(Table3).Theprevalenceof proteinuriaatadmissionwasnotsignificantlydifferentbetween primary and secondary dengue infections (13.6% vs 28.6%, P=0.281)(Table3).
TheUPCRwasmeasuredin178urinesamplescollectedduring the course of hospitalization from nine WS-patients, 43 WS+
patientsand17patientswithaseveredengue.Ninesampleswere excludedfromtheanalysisduetoverylowcreatinineconcentra- tion(<1mmol/L)thatresultedinanoverestimationoftheUPCR.
Proteinuriawasobservedin33.3%and23.3%oftheWS-andWS+
patients, respectively, and 47.1% of the patients with severe dengue,(Table3).Differencesbetweenthethreegroupsofpatients werenotstatisticallysignificant.Atotalof26.1%oftheprimary
denguecasesand33.3%ofthesecondarycasesexperiencedatleast one episode of proteinuria during hospitalization (P=0.768).
Proteinuriawasmoreoftendetectedbetweenday5andday7after onsetoffeverwhenUPCRvalueswerethehighest(Figure1).None ofthecontrolshadaUPCR45mg/mmol.
SixteenpatientsincludedinthestudyweretestedforUPCRone (n=4)orthree(n=12)monthsafterdischargefromhospital.None hadaUPCR45mg/mmol.
UPCRsaccordingtothedayofsamplingaftertheonsetoffever areshowninFigure1.Nosignificantdifferencewasobservedin UPCRs between the different groups of patients. UPCRs were significantlyhigherineach categoryofdenguepatientsateach time-pointcomparedtocontrols.
ModelingfactorsindependentlyassociatedwithUPCRvaluesin denguecases
The“daof”variableandtheageofthepatienthadasignificant effect on UPCRs (P<0.001) (Table 4). The GAMM-estimated relationshipsbetweenUPCRandboth“age”and“daof”variables areshowninFigure2.ThemodelpredictedhigherUPCRinurine samplescollectedfromolderchildren(13years)andbetween DAOF6–8.TheimmunestatusofthepatienthadnoeffectonUPCR values. A binomial GAMM was also used to assess factors independently associatedwith theoccurrence ofproteinuria in denguepatients,usingthesamevariables.Nosignificantassocia- tionwasfound(Supplementaryfile2).
Table2
Resultsofurinaryproteindipsticktestperformedatthetimeofadmissiontohospital.
Positivedipstick,total Positivedipstick,150mg/L(trace) Positivedipstick,300mg/L Positivedipstick,1000mg/L
Control(n=15) 3(20.0%) 3(20.0%) 0(0.0%) 0(0.0%)
Totaldengue(n=91) 39(42.9%) 28(30.8%) 9(9.9%) 2(2.2%)
2009WHOdengueclassification
Denguewithoutwarningsigns(n=21) 8(38.1%) 6(28.6%) 2(9.5%) 0(0.0%)
Denguewithwarningsigns(n=50) 16(32.0%) 12(24.0%) 4(8.0%) 0(0.0%)
Severedengue(n=20) 15(75.0%) 10(50.0%) 3(15.0%) 2(10.0%)
Immunestatus
Primaryinfection(n=26) 7(26.9%) 7(26.9%) 0(0.0%) 0(0.0%)
Secondaryinfection(n=47) 25(53.2%) 16(34.0%) 7(14.9%) 2(4.3%)
Undetermineda(n=18) 7(38.9%) 5(27.8%) 2(11.1%) 0(0.0%)
aImmunestatusisundeterminedwhentheintervalbetweenadmissionandhospitaldischargeis<7daysandtheHItiteratthetimeofhospitaldischargeis2560.
Table1
Summaryofdemographic,clinicalandvirologicinformationofthepatients.
Denguewithoutwarningsigns Denguewithwarningsigns Severedengue
Numberofpatients(numberofsamples) 24(45) 62(143) 22(53)
Medianageinyears[iqr] 7[6–12] 8.5[7–12] 9[7–11]
Sex(Male) 15(62.5%) 37(59.7%) 15(68.2%)
Mediandayoffeverathospitaladmission[iqr] 2[2–3] 4[3–4] 4[4–5]
Medianlengthofstayathospital[iqr] 5[5–5.5] 5[4–6] 5[4–5]
Denguediagnostic
RT-PCRandNS1positive 19(79.2%) 46(74.2%) 9(40.9%)
RT-PCRpositiveandNS1negative 5(20.8%) 11(17.7%) 8(36.4%)
RT-PCRnegativeandNS1positive 0(0%) 3(4.8%) 1(4.5%)
Othera 0(0%) 2(3.2%) 4(18.2%)
Immunestatus
Primaryinfection 10(41.7%) 20(32.2%) 0(0%)
Secondaryinfection 11(45.8%) 32(51.6%) 21(95.5%)
Undeterminedb 3(12.5%) 10(16.1%) 1(4.5%)
Clinicalmanifestationsatadmission
Drybleedingc 4(16.7%) 15(24.2%) 6(27.3%)
Wetbleedingd 1(4.2%) 10(16.1%) 7(31.8%)
Abdominalpain 0(0%) 52(83.9%) 22(100%)
Persistentvomiting 0(0%) 39(62.9%) 22(100%)
Lethargy/restlessness 1(4.2%) 19(30.6%) 15(68.2%)
Hepatomegaly 0(0%) 29(46.8%) 15(68.2%)
Ascites 0(0%) 15(24.2%) 8(36.4%)
Pleuraleffusion 0(0%) 12(19.4%) 6(27.3%)
Facialedema 0(0%) 2(3.2%) 3(13.6%)
Coldextremities 0(0%) 1(1.6%) 19(86.4%)
Cyanosis 0(0%) 0(0%) 11(50%)
Hypotension,weakorundetectablepulse 0(0%) 0(0%) 12(50.5%)
Narrowpulsepressure 0(0%) 0(0%) 16(72.7%)
Delayedcapillaryrefilltime 0(0%) 0(0%) 14(63.6%)
iqr:interquartilerange.
aIgMseroconversioninpairedplasmaand/orfourfoldincreaseoftheantibodytitermeasuredbyhemagglutinationinhibitionassay(HIA)inpairedplasma.
b Immunestatusisundeterminedwhentheintervalbetweenadmissionandhospitaldischargeis<7daysandtheHItiteratthetimeofhospitaldischargeis2560.
cPetechiae,purpuraand/orecchymosis.
d Nose,gumsbleeding,hematemasis,melenaand/orconjunctivalhemorrhage.
AssociationbetweenserumtotalproteinandUPCR
SerumtotalproteinlevelsandUPCRswereassessedinparallel in51denguepatients.UPCRstendedtobehigherinpatientswith hypoproteinemia (median=31.6mg/mmol, IQR=[22.1–48.9]) comparedtopatientswithanormalserumproteinconcentration (median=25.9mg/mmol,IQR=[18.3–32.1])butthedifferencewas notsignificant(P=0.096).Atotalof38.1%(8/21)ofthepatients withhypoproteinemiahadaUPCR45mg/mmol,whereas10%(3/
30) of the patients with a normal total protein concentration presentedsuchasignificantproteinuria(P=0.035).
Urineproteinelectrophoresis
UPEPwas performedon27admission samplesinwhichthe totalproteinconcentrationwas100mg/L.Resultsarepresented inTable5.
Table3
UPCRsinsamplescollectedatthetimeofadmissionandduringthecourseofhospitalization.
Total dengue
2009WHOdiseaseclassification Immunestatus Denguewithout
warningsigns
Denguewith warningsigns
Severe dengue
Primary infection
Secondary infection
Undetermineda
Hospitaladmission
Numberofpatientstested 52 7 31 14 22 21 9
NumberofpatientswithanUPCR45mg/mmol 12 (23.1%)
0(0.0%) 5(16.1%) 7(50.0%) 3(13.6%) 6(28.6%) 3(33.3%) Allsamples(admissionandduringhospitalization)
Numberofpatientsincluded 69 9 43 17 23 33 13
NumberofpatientswithanUPCR45mg/mmolin atleastonesample
21 (30.4%)
3(33.3%) 10(23.3%) 8(47.1%) 6(26.1%) 11(33.3%) 4(30.8%)
MedianUPCR(mg/mmol)[irq] 30.0
[22.1- 42.2]
30.8 [20.5-43.0]
28.5 [21.6-38.5]
34.3 [25.7- 48.8]
29.6[22.1- 43.3]
29.3[21.6- 48.8]
33.1[23.6- 40.9]
MediandayofillnesswhenUPCR45mg/mmol [irq]
6[5–7] 6[5–6.5] 6[4–7] 6[5–7] 5.5[4–6.5] 6[4–8] 6[5–7]
UPCR:Urineprotein:creatinineratio.
iqr:interquartilerange.
aImmunestatusisundeterminedwhentheintervalbetweenadmissionandhospitaldischargeis<7daysandtheHItiteratthetimeofhospitaldischargeis2560.
Figure1.UPCRaccordingtotimeofsamplinganddiseaseseverity.
Horizontalsolidlinesrepresentthemedianvaluemeasuredincontrolpatients.Horizontaldashlinesrepresentthequartiles.UPCR:Urineprotein:creatinineratio,DAOF:Day AfterOnsetofFever,WS-:denguewithoutWarningsigns,WS+:denguewithWarningsigns,SD:Severedengue
Table4
Effectsofage,gender,diseaseclassification,immunestatusanddayofdiseaseon UPCR.
Linearterms Estimate p-value
Male reference
Female 0.141 0.454
dengueWS- reference
dengueWS+ 0.146 0.604
severedengue 0.269 0.417
primarydengue reference
secondarydengue 0.029 0.899
Nonlinearterms edf p-value
s(daof) 4.921 <0.001
s(age) 3.414 <0.001
edf:effectivedegreesoffreedomofthesmoothfunctionterms(edf>1indicate nonlinearrelationships).
FvalueisanapproximateF-test.
daof:dayafteronsetoffever.
Discussion
Increasedurinaryexcretionofproteininpatientsinfectedby DENV is thoughttobe a hallmark of vascularendothelial cells defectandplasmaleakageassociatedwithcomplicatedformsof dengueandisconsideredapossibleprognosticmarker.3Protein- uria has been reported in up to 74% of patients with dengue hemorrhagic fever (DHF)6 and cases of self-limiting nephrotic- rangeproteinuria in patientswithDHFpresenting no manifes- tationsof renal damage have beendescribed elsewhere.7,8 The mechanism underlying the hypothesis of increased protein excretion in urine is that during DVI the glycocalyx of the endothelialcellsisdisruptedeitherbydirectactionofthevirusor bytheNS1 antigen,3,9,10causing plasmaleakage. Atthekidney level,alterationoftheglycocalyxlayerthatcoatstheglomerular endothelialcells enablesthepassageof macromolecules inthe primaryurine.Innormalconditions,thispassageisrestrictedin order to maintain the intravascular albumin concentration. As macromoleculescannotbereabsorbedbytubules, a glomerular proteinuria,mainlycharacterizedbythepresenceofalbuminin thefinalurine,occurs.Increasedclearanceofproteinsinurinedue todisruptionoftheglycocalyxhasbeendescribedduringdiabetes mellitus,11 childhood nephrotic syndrome12 and meningococcal septicemia.13
We first investigated thepresenceof proteinuria by usinga dipstick,oneoftheeasiestandcheapestmethodsforthedetection ofproteinsin urine.Weobservedthat urinarydipstickusedon samplescollectedatthetimeofinitialmedicalconsultationbefore admissiontothehospitalwasnotareliabletoolforthetriageof denguepatients.Theproportionsofpositivetestsincontrols(20%) andindengueWS-(38.1%)andWS+(32%)werenotstatistically different.Similarresultswereobtainedwhenthepatientswere classifiedaccordingtothe1997WHOguidelines(datanotshown).
Vasanwala et al. and Lumpaopong et al. also assessed the prognostic value of urine dipstick and demonstrated that this rapid test was not able to discriminatebetween dengue fever (DF) and DHF.14,15 In the present study, thecomparison of the dipstickresultswiththeUPCRresultsindicatedthatonly43.5%
of the patients who tested positive for proteinuria by dipstick hadaproteinuriaconsideredassignificantasdefinedbyaUPCR 45mg/mmol(datanotshown).Aproteinuriadetectedbydipstick and associatedwithaUPCR <45mg/mmolwas probablyminor and/ortheresultofaurineconcentrationassociatedwithamild dehydrationcausedbyfeverorvomitingaspreviouslyreported.16 TheUPCRisactuallyabettertooltoassesstheproteinuriathan simpledipstickasitintendstoadjustforfluctuationsinfluidintake andhydrationstatus.
The gold standard to determinethe proteinuria is provided bythemeasurementoftotalproteinoralbumininurineovera 24-hour period (timed urine). However, this method is time- consuming and cumbersome. Measurement of urine albumin:
creatinine or protein:creatinine ratios in random spot urine sampleshasbeendemonstrated tobeanacceptablealternative totimedurinesamples.17,18Sinceitwasalreadydemonstratedthat urinealbumin:creatinineratiosindenguepatientswerenotuseful toolstopredicttheevolutiontowardsseveredengueandtheneed forhospitalization,19weinsteadevaluatedthepotentialvalueof theUPCRwhichwaspreviouslyinvestigatedinadultsbutnotin children.14Usingamoreconservativedefinitionof45mg/mmolas recommendedbytheRoyalCollegeofPhysiciansofLondon5,20we demonstrated that patientswithWS+ and severedengue were more likely to have an elevated UPCR at the time of hospital admissioncomparedtoWS-patients.Nevertheless,thesensitivity ofthismarkerwaslimitedasonly16.1%ofthepatientsWS+had suchsignificantproteinuria.
Our multivariate analysisof UPCR during the course of the diseaseindicatedthatthisratiowassignificantlyhighersixtoeight daysaftertheonsetoffeverwhichisinagreementwithprevious reports.14,19,21Ratherthanbeingamarkerthatcouldbeusedfor patient triage at the time of first medical consultation, UPCR measurement seemstobemore usefulfor monitoring patients duringthecourseofthedisease.Indeed,thepeakUPCRusually occurredat theend ofthecriticalphase ofthediseaseandthe decreaseintheratioseemedtocorrespondwiththebeginningof therecoveryphase.
Hypoproteinemiain denguecasesusuallyindicatesa plasma leakage.4Hereweobservedthatsignificantproteinuriawasmore Figure2.RelationshipbetweenUPCRanddayofsamplingafterfeveronset(A)or
patientage(B).
AGAMM-modelwasusedtoestimatetherelationship.Thefigureinthey-axisisthe effectivedegreesoffreedomofthesmoothfunctionterms.daof:dayafteronsetof fever
Table5
Urineproteinelectrophoresisprofilesindenguepatients.
Pattern Numberofobservation MeanUPCR
Total (n=27)
WS- (n=2)
WS+
(n=13)
SD (n=12)
Pattern1:Normal 3(11.1%) 0(0.0%) 2(15.3%) 1(8.3%) 27.3(min=17.;max=33.10)
Pattern2:Alpha-1/Alpha-2 14(51.9%) 2(100%) 7(53.8%) 5(41.7%) 38.8(min=18.5;max=73.5)
Pattern3:Albumin 6(22.2%) 0(0.0%) 2(15.3%) 4(33.3%) 85.2(min=27.7;max=130.2)
Pattern4:“Serum-like” 4(14.8%) 0(0.0%) 2(15.3%) 2(16.7%) 298.12(min=114.5;max=664.4)
UPCR:Urineprotein:creatinineratio.
WS-:Withoutwarningsigns;WS+:Withwarningsigns;SD:SevereDengue.
frequentinpatientswithhypoproteinemiabuta UPCR45mg/
mmol was observed in only 38.1% of the patients with hypo- proteinemia.Theseresultsmaysuggestthattheplasmaleakage doesnotnecessarilyalwayscausesignificantproteinuriaorthatit isquicklycompensated.
Threemaincategoriesofproteinuriacanbedistinguishedby UPEP: overflow proteinuria,tubular proteinuria and glomerular proteinuria.23 Overflow proteinuria is a pre-renal proteinuria characterized by an abnormal peak of one of the five protein fractions.UPEPintubularproteinuriademonstratesalphaand/or betaglobulinpeaks,withorwithoutaminorfractionofalbumin.
UPEPanalysisofglomerularproteinuriashowseitheradominant albuminfractioninselectiveglomerularproteinuriaorapattern similartoserumproteinsinnon-selectiveglomerularproteinuria.
Wedidnotobserveanycaseofoverflowproteinuriainthelimited numberofcasesincludedinourstudy.Patternscompatiblewitha tubularproteinuria,aselectiveglomerularproteinuriaandanon- selectiveglomerular proteinuria, wereobservedin 51.9%, 22.2%
and 14.8% of the samples tested by UPEP, respectively. The glomerularproteinuriawasassociatedwithahighvalueofUPCR andweshowedinthisstudythataUPCR75mg/mmolwas100%
specificofa glomerularproteinuria. Ifwesetavalueof75mg/
mmolfor theUPCR as a hallmark of a glomerular proteinuria, 12 patients beyond the 21 that experienced a significant proteinuriacouldhavebeenclassifiedinthatgroup.Amongthose 12patients,onewasclassifiedasdenguewithoutwarningsigns, sixasdenguewithwarningsignsandfiveasseveredengue.The hypothesisthattheproteinuriaindenguediseasewassolelythe resultofdisruptionoftheglycocalyxcanonlybepartiallytrueas we detected both tubular and glomerular proteinuria cases in almost the same proportions.Renal impairments associated to both glomerular and tubular syndromes have already been described.6Kidneybiopsiesfrompatientswithdengueofvarying degrees ofseverity revealedacutetubular necrosis resultingin tubularimpairments.24–28Ischemicprocesses,whicharecommon inpatientswithseverehypovolemicshock,havebeensuggestedas anexplanationtotubularnecrosisin DSSpatients.25Interstitial edemaandmononuclearinfiltrationcanalsocontributetotubular necrosisinpatientswithoutshocksyndrome.24,27
Alimitation of ourstudywas the smallnumber of patients withoutwarningsignsasclinicalspecimenswereonlycollected fromhospitalizedchildrenand notfromambulatorycases.This resultedinanunderrepresentationofthemildestformofdengue diseaseand mayexplainwhywedidnotobservedifferencesin proteinuria between patients with or without warning signs.
AnotherlimitationwastheabsenceofdailyUPCRmeasurementfor each patient: the median number of samples tested for each patientwasonlytwoforamedianlengthofhospitalizationoffive days.Itisthereforepossiblethattheoccurrenceofmoresignificant proteinuriaepisodeswerenotcapturedinsomepatients.Another limitationofthestudyisthatwewerenotabletoexplorepotential confoundingfactorsasnumberofotheretiologyofproteinuriado exist,includingprimaryviralinfectionsthatarenotuncommonin children (e.g., adenovirus, cytomegolavirus, Epstein-Barr virus, enterovirus,etc.).
Thekidneyisoneofthemajororgansaffectedduringdengue butthemechanism andtheprevalenceof renalinvolvementin childrenwithdenguefeverremainsunclear.Renalfunctionisnot routinely assessed in patients hospitalized for suspected or confirmeddengue and thus the impactof dengue infection on the kidneys could well be under-recognized. This is a third limitation of our study as the proteinuria was measured retrospectivelyand therenal functionwas not assessed by the clinicians. There is a clear need, for a future study, to assess simultaneouslythe various parameters associated with a renal dysfunction as the hematuria, the proteinuria, an electrolyte
disturbance, the glomerular filtration rate, and eventually to conduct renal biopsies. In conclusion, this study indicates that proteinexcretioninurine,estimatedbyUPCR,increasesduringthe courseofdenguediseasetoreachamaximumlevelapproximately one weekafter theonset offever. Wefoundthat neitherurine dipsticknorUPCRmeasurementwereusefulforpatienttriageat thetimeoffirstvisittohospital.Monitoringtheurinaryprotein excretionduringthecourseofthedisease,however,mightbea factorindicatingtheevolutiontowardsdiseaserecovery.
Competinginterests
Anne-Claire Andries is currently a Biosynex employee and PhilippeBuchyaGSKvaccinesemployee.
Acknowledgements
Wewouldliketothankthepatients,nursesanddoctorswho participatedinthisstudy.Theresearchleadingtotheseresultshas received fundingfromtheEuropean UnionSeventhFramework Programme(FP7/2007/2011)underGrantAgreementn282378.
AppendixA.Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttp://dx.doi.org/10.1016/j.ijid.2016.12.022.
References
1.BradyOliverJ,GethingPeterW,BhattSamir,MessinaJaneP,BrownsteinJohn S,HoenAnneG,etal. Refining theglobalspatial limitsofdenguevirus transmissionbyevidence-based consensus.PLoSNegl TropDis2012;6(8):
e1760.doi:http://dx.doi.org/10.1371/journal.pntd.0001760.
2.SpecialProgrammeforResearchandTraininginTropicalDiseases,World HealthOrganization.Dengue:guidelinesfordiagnosis,treatment,prevention, andcontrol.Newed.Geneva:TDR:WorldHealthOrganization;2009.
3.WillsBridgetA,OraguiEmmanuelleE,DungNguyenMinh,LoanHaThi,Chau NguyenVinh,FarrarJeremyJ,etal.Sizeandchargecharacteristicsofthe proteinleakindengueshocksyndrome.JInfectDis2004;190(4):810–8.doi:
http://dx.doi.org/10.1086/422754.
4.WorldHealthOrganization.Denguehaemorrhagicfever:diagnosis,treatment, prevention,andcontrol.2ndedGeneva:WorldHealthOrganization;1997.
5.RichardBurden,CharlieTomson.GuidelineDevelopmentCommittee,Joint SpecialtyCommitteeonRenalDiseaseoftheRoyalCollegeofPhysiciansof LondonandtheRenalAssociation.Identification,managementandreferralof adultswithchronickidneydisease:conciseguidelines.ClinMedLondEngl 2005;5(6):635–42.
6.LizarragaKarloJ,AliNayer.Dengue-associatedkidneydisease.JNephropathol 2014;3(2):57–62.doi:http://dx.doi.org/10.12860/jnp.2014.13.
7.VasanwalaFF,PuvanendranR,NgJM,SuhailSM.Twocasesofself-limiting nephropathies secondarytodengue haemorrhagicfever.SingaporeMed J 2009;50(7):e253–5.
8.SakaraHutspardol,OlarnPrommalikit,NuttapholUpiya,JintanaChataroop- wijit, Khemika Khemakanok, Kesara Assadamongkol. Heavy proteinuria followingdenguehemorrhagicfever.SoutheastAsianJTropMedPublicHealth 2011;42(3):579–82.
9.AvirutnanPanisadee,ZhangLijuan,PunyadeeNuntaya,ManuyakornAnanya, PuttikhuntChunya,KasinrerkWatchara,etal.SecretedNS1ofdenguevirus attachestothesurfaceofcellsviainteractionswithheparansulfateand chondroitinsulfate.E.PLoS Pathog2007;3(11):e.183. doi:http://dx.doi.org/
10.1371/journal.ppat.0030183.
10.SimmonsCameronP,FarrarJeremyJ,NguyenvanVinhChau,WillsBridget.
Dengue.NEnglJMed2012;366(15):1423–32.doi:http://dx.doi.org/10.1056/
NEJMra1110265.
11.NieuwdorpMax,MooijHansL,KroonJojanneke,AtaseverBektas,SpaanJos AE,InceCan,etal.Endothelialglycocalyxdamagecoincideswithmicro- albuminuriaintype1diabetes.Diabetes2006;55(4):1127–32.
12.CengizNurcan,BayazitAysunK,NoyanAytul,AnaratRuksan,AnaratAli.
Glycosaminoglycanexcretioninchildrenwithnephroticsyndrome.Pediatr NephrolBerlGer2005;20(4):486–90.doi:http://dx.doi.org/10.1007/s00467- 004-1739-y.
13.Oragui EE, Nadel S, Kyd P, Levin M. Increased excretion of urinary glycosaminoglycansinmeningococcalsepticemiaandtheirrelationshipto proteinuria.CritCareMed2000;28(8):3002–8.
14.VasanwalaFarhadF,TheinTun-Linn,LeoYee-Sin,GanVictorC,HaoYing, LeeLindaK,etal.Predictivevalueofproteinuriainadultdengueseverity.
PLoS Negl Trop Dis 2014;8(2):e2712. doi:http://dx.doi.org/10.1371/journal.
pntd.0002712.
15.Adisorn Lumpaopong, Pinyada Kaewplang, Veerachai Watanaveeradej, Prapaipim Thirakhupt, Sangkae Chamnanvanakij, Konggrapun Srisuwan, etal.Electrolytedisturbancesandabnormalurineanalysisinchildrenwith dengueinfection.SoutheastAsianJTropMedPublicHealth2010;41(1):72–6.
16.MarksMI,McLainePN,DrummondKN.Proteinuriainchildrenwithfebrile illnesses.ArchDisChild1970;45(240):250–3.
17.MorgensternBruceZ,ButaniLavjay,WollanPeter,WilsonDavidM,Larson TimothyS.Validityofprotein-osmolalityversusprotein-creatinineratiosin theestimationofquantitativeproteinuriafromrandomsamplesofurinein children.AmJKidneyDisOffJNatlKidneyFound2003;41(4):760–6.
18.JafarTazeenH,ChaturvediNish,HatcherJuanita,LeveyAndrewS.Useof albumincreatinineratioandurinealbuminconcentrationasascreeningtest foralbuminuriainanIndo-Asianpopulation.NephrolDialTransplantOffPubl EurDialTransplAssoc–EurRenAssoc2007;22(8):2194–200.doi:http://dx.doi.
org/10.1093/ndt/gfm114.
19.HanhTienNguyenThi,LamPhungKhanh,DuyenHuynhThiLe,NgocTranVan, HaPhanThiThanh,KieuNguyenTanThanh,etal.Assessmentofmicro- albuminuriaforearlydiagnosisandriskpredictionindengueinfections.PloS One2013;8(1):e54538.doi:http://dx.doi.org/10.1371/journal.pone.0054538.
20.KeaneWF,EknoyanG.Proteinuria,albuminuria,risk,assessment,detection, elimination(PARADE):apositionpaperoftheNationalKidneyFoundation.Am JKidneyDisOffJNatlKidneyFound1999;33(5):1004–10.
21.VasanwalaFarhadF,PuvanendranRukshini,Fook-ChongStephanie,NgJoo- Ming, SuhailSufiM,Lee Kheng-Hock.Could peakproteinuriadetermine whetherpatient withdenguefever developdenguehemorrhagic/dengue
shocksyndrome?–aprospectivecohortstudy.BMCInfectDis2011;11:212.doi:
http://dx.doi.org/10.1186/1471-2334-11-212.
23.AbueloJG.Proteinuriadiagnosticprinciplesandprocedures.AnnInternMed 1983;98(2):186–91.
24.AyeKhin Saw,Charngkaew Komgrid,Win Ne,Wai KyawZin,MoeKyaw, PunyadeeNuntaya,etal.Pathologichighlightsofdenguehemorrhagicfeverin 13autopsycasesfromMyanmar.HumPathol2014;45(6):1221–33.doi:http://
dx.doi.org/10.1016/j.humpath.2014.01.022.
25.PóvoaTiagoF,AlvesAdaMB,OliveiraCarlosAB,NuovoGerardJ,ChagasVera LA,PaesMarcianoV.ThePathologyofSevereDengueinMultipleOrgansof HumanFatalCases:Histopathology,UltrastructureandVirusReplication.PLoS ONE2014;9(4):e83386.doi:http://dx.doi.org/10.1371/journal.pone.0083386.
26.KamolwishLaoprasopwattana,PornpimolPruekprasert, PornsakDissanee- wate,AlanGeater.VachvanichsanongPrayongOutcomeofdenguehemor- rhagicfever-causedacutekidneyinjuryinThaichildren.JPediatr2010;157 (2):303–9.doi:http://dx.doi.org/10.1016/j.jpeds.2010.02.008.
27.NikitaMehra,AmishPatel,GeorgiAbraham,ReddyYogesh,ReddyYuvaram NV.AcutekidneyinjuryindenguefeverusingAcuteKidneyInjuryNetwork criteria:incidenceandriskfactors.TropDoct2012;42(3):160–2.doi:http://dx.
doi.org/10.1258/td.2012.120023.
28.MohsinN,MohamedE,GaberM,ObaidaniI,BudruddinM,AlBusaidyS.
Acute tubular necrosis associated with non-hemorrhagic Dengue fever:
a case report. Ren Fail 2009;31(8):736–9. doi:http://dx.doi.org/10.3109/
08860220903003404.
