Identifying patients with axial spondyloarthritis On a mission to achieve timely recognition. Marloes van Onna

Transcription

1 Identifying patients with axial spondyloarthritis On a mission to achieve timely recognition Marloes van Onna

2 ISBN: PrintingwasfinanciallysupportedbyAbbvie,PfizerB.V,UCBPharmaB.V,JanssenCilagB.V. Coverdesign:Gettyimages,RaphaelSchneider Layout:TinyWouters Production:Datawyse,UniversitairePersMaastricht

3 Identifyingpatientswithaxialspondyloarthritis: onamissiontoachievetimelyrecognition PROEFSCHRIFT TerverkrijgingvandegraadvandoctoraandeUniversiteitMaastricht, opgezagvanderectormagnificus,prof.dr.l.l.gsoete volgenshetbesluitvanhetcollegevandecanen, inhetopenbaarteverdedigen opvrijdag20november2015om12.00uur door MarloesGertrudeBerdienevanOnna UUNIVERSITAIRE PERS MAASTRICHT P M

4 Promotores Prof.dr.A.E.R.C.HBoonen Prof.dr.R.B.M.Landewé,AcademischMedischCentrum,Amsterdam& ZuyderlandZiekenhuis,HeerlenSittard Copromotor Dr.A.M.vanTubergen Ledenvandebeoordelingscommissie Prof.dr.J.E.Wildberger(voorzitter) Prof.dr.R.A.deBie Prof.dr.F.vandenBosch,UniversitairZiekenhuis,Gent Prof.dr.J.A.Knottnerus Dr.A.E.A.MWeel,MaasstadZiekenhuis,Rotterdam

5

6

7 Contents Chapter1 Generalintroduction 9 PartI Epidemiologyofspondyloarthritis 25 Chapter2 Theglobalprevalenceofspondyloarthritis:asystematic 27 reviewandmetaregressionanalysis PartII TheuseofMRIindetectionofearlyaxialspondyloarthritis 71 Chapter3 HLAB27andgenderindependentlydeterminethelikelihoodof 73 apositivemriofthesacroiliacjointsinpatientswithearly inflammatorybackpain:atwoyearmrifollowupstudy Chapter4 BonemarrowedemaonMRIofthesacroiliacjointsis 85 associatedwithdevelopmentoffattylesionsonmri overaoneyearintervalinpatientswithearlyinflammatory lowbackpain:a2yearfollowupstudy Chapter5 GadoliniumcontrastenhancedMRIsequencedoesnothavean 101 incrementalvalueintheassessmentofsacroiliitisinpatients withearlyinflammatorybackpainbyusingmriincombination withpelvicradiographs:a2yearfollowupstudy Chapter6 Naturalcourseofbonemarrowedemaonmagneticresonance 113 imagingofthesacroiliacjointsinpatientswithearly inflammatorybackpain:a2yearfollowupstudy PartIII Timelyidentificationofspondyloarthritisinprimarycare 127 Chapter7 Generalpractitioners perceptionsoftheirabilitytoidentifyand 129 referpatientswithsuspectedaxialspondyloarthritis: aqualitativestudy Chapter8 Educationimprovesreferralofpatientssuspectedofhaving 141 spondyloarthritisbygeneralpractitioners.astudywith unannouncedstandardizedpatientsindailypractice Chapter9 Summaryandgeneraldiscussion 159 SamenvattinginhetNederlands 173 Valoriationaddendum 187 Dankwoord 193 Curriculumvitae 199

8

9 CHAPTER1 Generalintroduction

10

11 Generalintroduction 11 1.INTRODUCTION Spondyloarthritis comprises a group of clinically and genetically closely related inflammatory rheumatic disorders, including ankylosing spondylitis (AS), psoriatic arthritis,reactivearthritisandarthritisand/orspondylitisassociatedwithinflammatory boweldisease(ibd).[1] Dependingonthelocationofthepredominantclinicalfeatures,SpAisnowadaysalso dividedintoaxialandperipheraldisease.[2]axialspa(axspa)canbefurthersubdivided intoasandnonradiographic(nr)axspa,dependingonpresenceofabnormalitieson pelvicradiographs.[2]innraxspainflammationmayalreadybedetectedonanmriof the sacroiliac joints (SIJs). Patients with nraxspa may eventually progress to a radiographicstage,withanaverageprogressionrateof10%after2years,andupto 60%after10years.[35] Thefieldofspondyloarthritis(SpA)hasexperiencedmajorprogressinthelastdecade, especially with regard to development of new therapeutic options and the use of magneticresonanceimaging(mri)toestablishanearlydiagnosisofspa.[6]therecent developments in the field of SpA increased the need to timely recognize patients having the disease. However, several priorities on the research agenda to improve earlier diagnosis remain, such as further refining the role of MRI in the diagnosis, followupandprognosisofpatientswithspa.also,improvingrecognitionandreferral ofpatientswithearlyspainprimarycareisanimportantchallenge.theaimofthis thesis is to provide more insight into the role of MRI in SpA and recognition of this disease in primary care. This chapter first provides an overview of the current knowledgeaboutspaandthenelaboratesonthespecificobjectivesofthisthesis. 1.1Clinicalfeaturesofspondyloarthritis InpatientswithaxSpA,themostimportantclinicalfeatureisinflammatorybackpain (IBP)causedbysacroiliitisandspondylitis.[7]IBPistypicallycharacterizedbyawakening night pain, improvement of pain with exercise, no improvement with rest and a favourableresponse to nonsteroidalantiinflammatory drugs (NSAIDs).[8] Peripheral SpA is characterized by peripheral arthritis (predominantly of the lower limbs), enthesitis and dactylitis, as presenting symptom.[1] Extraarticular manifestations related toaxial andperipheralspa include psoriasis, anterioruveitis andibd.[1]the mainfeaturesofbothaxialandperipheralspaarelistedinbox1.[9]

12 12 Chapter1 1.2Pathophysiology Although the clinical symptoms of the SpA are typically heterogeneous, there is evidenceofsharedgeneticmarkersandlinkage.[10]thestrongestknowncontributing geneticfactorforspaisthemajorhistocompatibilitycomplex(mhc)classimolecule humanleukocyteantigen(hla)b27.hlab27ismoststronglyassociatedwithaxspa. TheriskofdevelopingaxSpAinHLAB27positiveindividualsisashighas5 7%.[11,12] Approximately 8090% of Northern European patients with axspa are HLAB27 positive.[11]othersubtypesofspahavelowerdegreesofassociationwithhlab27. For instance, the frequency of HLAB27 in patients with psoriatic arthritis and peripheralarthritisisaround20%,butthisincreasesto60%inpatientswithassociated sacroiliitis.[13] More recently, additional genetic links that contribute to the pathogenesis of SpA have been identified, e.g. polymorphisms of the interleukin23 receptor (IL23R) gene and endoplasmic reticulum aminopeptidase 1 (ERAP1) gene. GeneticvariantsofIL23RarealsolinkedtoconcomitantpsoriasisandIBD.[11,12] In patients with SpA, the genetic susceptibility is, at least in part, responsible for an autoinflammatory status. This autoinflammatory reaction may eventually initiate tissue remodelling which may eventually lead to ankylosis (immobilisation and consolidationofajoint).[12] 1.3Epidemiology TheprevalenceofSpAinWesternEuropeancountriesisestimatedtobeapproximately 0.51%,similarlytothatofrheumatoidarthritis.[14]TheestimatedprevalenceofAS, the most frequent subtype of SpA, ranges from %.[14] The incidence and prevalence rates of SpA are strongly dependent and directly correlated to the prevalence of HLAB27 in a given population.[14] Other important contributors to variationinprevalenceofspaincludedifferencesinselectionofthetargetpopulation, variationinthecriteriausedforcasedefinitionandstudydesign.[9] Unlikemostotherformsofarthritis,thefirstsymptomsofSpAusuallyappearbefore thefourthdecadeoflife.[15]malesareaboutthreetimesmoreoftenaffectedwithas than females.[1] However, in the whole group of axspa, the male : female ratio approaches1:1.[1,16]

13 Generalintroduction 13 Box1.SpAfeatures Inflammatorybackpain GoodresponsetoNSAIDs HLAB27positive ApositivefamilyhistoryforSpA IncreasedCRPconcentration Enthesitis Peripheralarthritis Dactylitis Priorurogenitalorgastrointestinalinfection Inflammatoryboweldisease Psoriasis Acuteanterioruveitis Sacroiliitisdetectedbyimaging(conventionalradiographyorMRI) SpA=spondyloarthritis; NSAID=nonsteroidal antiinflammatory drug; CRP=Creactive protein; B27=humanleukocyteantigenB27;MRI=magneticresonanceimaging. HLA 2.CLASSIFICATIONCRITERIAFORSPONDYLOARTHRITIS TheideathatthedisordersbelongingtoSpAweredifferentfromotherrheumatological disorders e.g. rheumatoid arthritis was first described by the American Rheumatism Association in the Nomenclature and Classification of Rheumatic diseases in 1963.[17,18]Theunifyingconceptof seronegativespondyloarthritides,nowknown as spondyloarthritisorspa,wasfirstintroducedbymollandwrightin1974.[19] Over the last decades, several sets of classification criteria for SpA have been developedinanattempttodefinesubgroupsofpatientstofacilitateresearchstudies. Figure 1 shows a historical timeline in which several classification criteria for SpA in general,foras,andforaxialandperipheralspaarepresented.[7,2023] ThemodifiedNewYork(mNY)criteriafortheclassificationofASrelyonclinicaland radiologicalfindings(figure1.1).[20]patientsfulfilthemnycriteriawhensacroiliitison aconventionalradiographispresentandatleastoneclinicalcriterionispresent.[20]a major disadvantage of the mny criteria is that advanced sacroiliitis on conventional radiography is required for fulfilment. Development of radiographic sacroiliitis may take up to several years. Some patients, especially females, may never develop radiographicsacroiliitis.[15]thismeansthatthemnycriterialacksensitivityespecially attheearlydiseasestage,whichmayresultinadiagnosticdelayupto9years.[15]

14 14 Chapter1 Figure1.1 Classificationcriteriaforaxialspondyloarthritis (SpA),historicaltimeline. AS,ankylosing spondylitis;ibp,inflammatorybackpain;nsaids,nonsteroidalantiinflammatorydrugs;hlab27,humanleukocyteantigenb27;ibd, inflammatoryboweldisease;mrisij,magneticresonanceimagingofthesacroiliacjoints. *Sacroiliitis onconventionalradiography:bilateralsacroiliitis grade24orunilateralsacroiliitis grade3or4.

15 Generalintroduction 15 TheAmorandEuropeanSpondyloarthropathyStudyGroup(ESSG)criteriawereboth developedinthe1990s,inordertoencompasstheentirespectrumofspa,including theaxialandperipheralmanifestations,andearlyaswellasmilddiseasestages.[21,22] Notwithstanding, the Amor and ESSG criteria lacked sensitivity and specificity, especiallyinpatientswithearlydisease.[24,25]inaddition,thesecriteriadonotallow foracleardifferentiationbetweenaxialandperipheraldisease,whilethisdistinctionis importantforetiologicstudiesandtestingtreatmentstrategies.[9,26] ToovercomethelimitationsofthemNY,ESSGandAmorcriteria,itwasnecessaryto develop new criteria sets. The most recently developed criteria sets are the AssessmentofSpondyloArthritisinternationalSociety(ASAS)classificationcriteriafor axialandperipheralspa.[7,23]theentrycriterionfortheasasaxspacriteriaisback pain3monthsandageatonset<45years(figure1.1).therearetwoarmsinthe ASASaxSpA criteria, an imaging and clinical arm. The imaging arm includes both sacroiliitis detected either by MRI or on a conventional radiograph; for classification one other additional SpA feature also needs to be present. HLAB27 is the main criterion of the clinical arm; for classification two other additional SpA features also need to be present (Figure 1.1).[4] The entry criteria for the ASAS peripheral SpA criteriaarearthritis,enthesitis,and/ordactylitis.forclassificationpurposes,presence ofadditionalfeaturesisneeded(figure1).theseadditionalfeaturesmayincludeeither oneormorefromthefollowinglist:psoriasis,ibd,precedinginfection,hlab27,uveitis and sacroiliitis on imaging or two or more of the following list: arthritis (past or present),enthesitis(pastorpresent),dactylitis(pastorpresent),historyofpreviousibp andapositivefamilyhistoryofspa.[23] 3.CRITERIAFORINFLAMMATORYBACKPAIN IBPisakeyfeaturefortheESSGandASASaxSpAclassificationcriteria.Todefinethis featureseveralcriteriasetshavebeendevelopedoverthelastyears.[8,27,28]single parameterslacksufficientdiscriminativeabilitytodefineibp,becauseallparameters mayalsobepresentinpatientswithothercausesofbackpain.[27]themostrecent criteria set for IBP is the ASAS classification criteria set for IBP, established in 2009 (Figure 1.2).[27] The ASAS IBP criteria were tested in avalidation cohort and havea sensitivityof77.6%andspecificityof72.4%,whenatleastfouroutoffiveparameters were fulfilled.[27] Other criteria sets for IBP are the Calin criteria and Berlin criteria (Figure1.2).[8,28]

16 16 Chapter1 Figure1.2. Classificationcriteriaforinflammatorybackpain(IBP). morn.,morning;asas,assessmentofspondyloarthritisinternationalsociety. 4.IMAGINGINAXIALSPONDYLOARTHRITIS 4.1Conventionalradiography Conventional radiography is the most widely used imaging technique for assessing structuralchangesofthesijs.[29,30]radiographsofthesijscandemonstratesclerosis, erosionsandankylosis.[31]thepresenceofsacroiliitisisgradedfrom0(normal)to4 (ankylosis)(table1.1).[20] The main disadvantage of conventional radiography in the clinical assessment of patients suspected for having axspa is its low sensitivity, especially in early disease.[7,29]inaddition,onlystructuraldamageofthesijs,whichistheconsequence of inflammation, can be detected. Other major challenges when using conventional radiographsofthesijsfordetectionofsacroiliitisarethelowintraandinterobserver agreement,projectionartefactsandpoorvisibility.[2933]

17 Generalintroduction 17 Table1.1 ComparisonofMRIandconventionalradiographyintheclassificationofsacroiliitis. Conventionalradiograpgy Magneticresonanceimaging(MRI) Standardprotocoland Anteroposterior(AP)pelvicradiograph Semicoronalsectionorientationalongaxisofthesacralbone,usingatleastaT1 orientation weightedturbosesequence,fst2weightedturbosesequenceorstirsequence* *Additionalinformationtoassesssacroliitis:FST1weightedsequenceafter administrationofgd Characteristicactivelesions Notapplicable Bonemarrowedema(BME)(T2/STIR/T1postgadolinium),capsulitis,synovitis, enthesitis Subchondralsclerosis,erosions,fatdeposition,bonybridges/ankylosis Widening/narrowingjointspace, erosions,sclerosis,ankylosis Characteristicstructural changes ASAS/OMERACTdefinitionofapositiveMRI: ModifiedNewYork(mNY)criteria: Atleast1activelesion(BME),highlysuggestiveofaxialSpA,ispresentinatleast2 successiveslicesorincaseof2lesions(bme)aredetectedin1slice** Grade0:normal Grade1:suspiciouschanges Grade2:minimalabnormalitieswith smallareasoferosionsorsclerosis withoutalterationinjointspacewidth Grade3:moderateoradvanced erosions,sclerosis,widening,narrowing, and/orpartialankylosisofthejoint Grade4:totalankylosis Gradingofchangesand definitionofsacroiliitis **Thepresenceofsynovitis,capsulitis,orenthesitisonlywithoutconcomitant subchondralbmeiscompatiblewithsacroliitisbutnotsufficientformakingadiagnosis ofactivesacroliitis **StructurallesionsonlywithoutconcomitantsubchondralBMEisnotsufficientfor makingadignosisofactivesacroliitis Bilateralsacroiliitisgrade24or unilateralsacroiliitisgrade3or4 MRI,magneticresonanceimaging;BME,bonemarrowedema;mNYcriteria,modifiedNewYorkcriteria;ASAS,AssessmentofSpondyloArthritisinternationalSociety; OMERACT, Outcome Measures in Rheumatoid Arthritis Clinical Trials; SE, spin echo; STIR, short tau inversion recovery; FS, fat saturated; Gd, gadolinium; SpA, spondyloarthritis.[20,34]

18 18 Chapter1 4.2Magneticresonanceimaging Nowadays,MRIoftheSIJsisconsideredanimportantimagingtechniqueformakinga diagnosis of axial SpA.[34] By using MRI, both active (inflammatory) lesions and structuralchangescanbevisualized(table1.1).[34]typicalactivelesionsinthesijsare subchondral bone marrow edema (BME) / osteitis, and to a lesser extent synovitis, enthesitisandcapsulitis.[34]activelesionsarebestvisualisedont2fatsaturated(fs) weightedspinecho (SE)sequence,T1 FS gadoliniumcontrastenhancedsequence or the short tau inversion recovery (STIR) sequence. Structural changes that can be detected by MRI are erosions, fat deposition (fatty lesions), sclerosis and ankylosis. ThesestructurallesionsarebestdetectedontheT1weightedSEsequence.[34] A positive MRI of the SIJs in the context of identifying sacroiliitis, is defined as the presenceofsubchondralbmethatishighlysuggestiveofaxialspa,andappearsastwo ormoredistinguishablelesionsononeslice,orasonelesiononatleasttwosuccessive slices (ASAS/ Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) definition).[34]thepresenceofsynovitis,enthesitisorcapsulitiswithoutbmedoesnot suffice. Only active lesions (BME) are currently considered for fulfilment of the ASAS/OMERACTdefinition.[34]Structuralchangesmayalsooccurinconjunctionwith BME,orwithoutBMEasasequelofpreviousinflammation.[34] Several scoring methods for assessment and quantification of active lesions on MRI have been proposed throughout the years.[3638] However, there is no preferred scoringsystem.[34]allofthesescoringsystemsareprimarilyconfinedtotheresearch setting. 5.EARLYREFERRALANDRECOGNITIONOFSPONDYLO ARTHRITISINPRIMARYCARE SpA isamongtheinflammatoryrheumaticdisorderswith the longest delaybetween theonsetofsymptomsandthemakingofadiagnosis.specificallyinthecaseofas,the delayupisupto9years.[15]anearlydiagnosisofaxspaisimportantnotonlytoavoid unnecessarydiagnosticproceduresandinappropriatetreatments,butashortdisease duration is also associatedwith a favourableresponseto NSAIDsandantiTNFalpha therapy.[39] Furthermore, several studies have shown that effective treatment of active axspa is associated with improved work capacity and quality of life [4043]. Providinganearlydiagnosiscanprovidereassurancetopatientsandmayoptimisethe

19 Generalintroduction 19 benefitofeducationandlifestylemodifications.recognizingpatientswithaxspaearly inthediseasecourseisthereforeimportant. Thegeneralpractitioner(GP)isakeyroleplayerintheidentificationandreferralof patientssuspectedofhavingspa.[44]however,knowledgeandawarenessaboutspa andtheassociatedextraarticularmanifestationsislowamonggps,resultinginpoor recognition and delayed referral of patients.[45] A recent study has shown that 1 in 4patients with chronic back pain of3months that had started before the age of 45yearsandwhowererecruitedfromprimarycare,canbeclassifiedashavingaxSpA aftercarefulevaluation.[46]thesepatientshadnotbeenrecognizedassuchbythegp. Several referral tools for patients with chronic back pain have been proposed a few yearsago,inordertoimproverecognitionandhencereferralofpatientssuspectedfor axspainprimarycare.[4751]thesetoolsincludedparameterssuchasibp,sacroiliitis on imaging and / or HLAB27. When patients fulfilled the referral criteria and were actuallyreferredtotherheumatologist,upto45%patientswerefoundtohaveaxspa afteradiagnosticworkup.[4852] 6.THECONTEXTOFRESEARCHIN THE FIELD OF SPONDYLOARTHRITIS ATTHESTARTOFTHIS THESIS ThisthesisdescribesourresearchwithregardtoassessingtheglobalprevalenceofSpA andtheearlyidentificationofpatientswithspa,bothinprimaryandsecondarycare. Until about the turn of the millennium, SpA took the back seat in rheumatology, becauseoflimitedtherapeuticoptions.earlydiagnosiswassimplynotanareaofmajor concern.thisdramaticallychangedwiththeintroductionofthehighlyeffectivetnf alpha blockers that significantly improved quality of life of patients dealing with SpA.[53]Theexpansionoftherapeuticarmamentariummakesitnecessarytoprovide reliabledataontheprevalenceofspa,inordertogaininsightinthesocietalimpact andcostsassociatedwithspa. In addition, new and effective therapeutic options make it even more important to diagnose SpA at an early stage. In the light of these developments, the Early SpondyloArthritisClinic(ESpAC)wasestablishedin2000.[54]Sixtyeightpatientswith IBPoflessthantwoyearsdurationwereincludedinthisprospectivecohortstudyand followed for two yearswith repeated clinicaland radiologicalexaminations. In2009, when starting this PhD program, MRI had already evolved as an important imaging modalityfordiagnosisandclassificationofearlyaxspabutseveralresearchquestions

20 20 Chapter1 remained,forexampleabouttheevolutionofactivelesionsandstructuralchangeson MRIovertime.Suchquestionscouldonlybeansweredusingcohortswithasubstantial followup period, like the ESpAC study, that was the basis for several studies in this thesis. Next,recentresearchprojectsfocussingondiagnosingandtreatingpatientswithearly SpA,presumeatimelyreferralofpatientsbytheirGPs.Severalreferraltoolshavebeen developedinordertoshortenthediagnosticdelaythatissocharacteristicforthese patients,butthesetoolscanonlybesuccessfulifknowledgeandawarenessaboutspa increases among GPs. Educational programmes with special focus on SpA might improvetherecognitionofspafeatures,andconsequentlyleadtoanearlierreferralof thosesuspectedofhavingspa. 7.AIMSOFTHISTHESIS Themainresearchquestionsunderlyingthisthesisareasfollows: Whatistheglobalprevalenceofspondyloarthritis(SpA)anditsphenotypesandwhich studycharacteristicsmightexplainheterogeneityinprevalenceestimates? HowcantheuseandinterpretationofMRIoftheSIJinpatientswithearlyaxialSpAbe improved?whichdeterminantspredictthepresenceofactiveandstructurallesionson MRIovertime? WhatisthecurrentlevelofknowledgeofGPsaboutaxSpA? HowcantherecognitionandreferralofSpAbyGPsinclinicalpracticebeimproved? 8.OUTLINEOFTHISTHESIS Thisthesisisdividedintothreerelatedparts.describestheglobalprevalenceof SpA(Chapter2).ofthisthesisfocusesontheuseofMRIoftheSIJsfordetection of sacroiliitis in 68 patients with recentonset IBP, included in the ESpAC. Chapter 3 describes how activelesions onmri of thesijs in patients included in ESpACevolve over time, and discusses which determinants are responsible for persisting active lesionsonmriofthesijs.chapter4describestheassociationbetweenbmeonmriof the SIJs and development of structural changes on both MRI and conventional

21 Generalintroduction 21 radiographsinpatientsincludedintheespac.toassessactivelesionsonmriofthe SIJs,severalMRIsequencescanbeused.InChapter5,thepotentialincrementalvalue ofthet1postgddtpamrisequenceofsijscomparedtothecombinationofthestir MRI sequence and conventional radiographs is discussed. Chapter 6 describes the naturalcourseofbmeonmriofthesijsovera2yearfollowupperiod. TheoptimizationofidentificationandreferralofpatientssuspectedofhavingSpAin primary care is described in of this thesis. Chapter 7 describes the level of knowledgeandexperiencesofgpswithregardtoaxspa.chapter8describesastudyin which the current practice performance with regard to referral and recognition of patientswithearlyspawasinvestigatedbyusingstandardizedpatients.inaddition,the influenceofeducationonthisperformancewasalsostudied.theresearchpresented in this thesis is summarized and discussed in Chapter 9 and 10. Chapter 10 also discusses how the researchpresented in this thesis canbe translated to clinical and societalbenefit.

22 22 Chapter1 REFERENCES 1. Mercieca C, Landewé R, Borg AA. Spondyloarthropathies: pathogenesis, clinical features. In JWJ Bijlsma,ed.EULARtextbookonrheumaticdiseases.London:BMJPublishingGroup,2012: Rudwaleit M, Khan MA, Sieper J. The challenge of diagnosis and classification in early ankylosing spondylitis:doweneednewcriteria?arthritisrheum2005;52: PoddubnyyD,RudwaleitM,HaibelH,etal.Ratesandpredictorsofradiographicsacroiliitisprogression over2yearsinpatientswithaxialspondyloarthritis.annrheumdis2011;70: SampaioBarros PD, Bortoluzzo AB, Conde RA, et al. Undifferentiated spondyloarthritis: a longterm followup.jrheumatol2010;37: SampaioBarrosPD,BertoloMB,KraemerMH,etal.Undifferentiatedspondyloarthropathies:a2year followupstudy.clinrheumatol2001;20: Sieper J, Rudwaleit M, Baraliakos X, et al. The Assessment of SpondyloArthritis international Society (ASAS)handbook:aguidetoassessspondyloarthritis.AnnRheumDis2009;68:ii RudwaleitM,vanderHeijdeD,LandewéR,etal.ThedevelopmentofAssessmentofSpondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: Calin A, Porta J, Fries JF, et al. Clinical history as a screening test for ankylosing spondylitis. JAMA 1977;237: Van Tubergen A. The changing clinical picture and epidemiology of spondyloarthritis. Nat Rev Rheumatol.2015;11: VanTubergenA,WeberU.Diagnosisandclassificationinspondyloarthritis:identifyingachameleon. NatRevRheumatol2012;8: RobinsonPC,BrownMA.Thegeneticsofankylosingspondylitisandaxialspondyloarthritis.RheumDis ClinNorthAm2012;38: DougadosM,BaetenD.Spondyloarthritis.Lancet2011;377: BraunJ,SieperJ.Ankylosingspondylitis,otherspondyloarthritides,andrelatedconditions.InWarrell DA,CoxTM,ed.OxfordTextbookofMedicine.Oxford:OxfordUniversityPress,2010: StolwijkC,BoonenA,vanTubergenA,etal.Epidemiologyofspondyloarthritis.RheumDisClinNorth Am2012;38: FeldtkellerE,KhanMA,vanderHeijdeD,etal.AgeatdiseaseonsetanddiagnosisdelayinHLAB27 negativevs.positivepatientswithankylosingspondylitis.rheumatolint2003;23: Feldtkeller E, Bruckel J, Khan MA. Scientific contributions of ankylosing spondylitis patient advocacy groups.curropinrheumatol2000;12: Blumberg BS, Bunim JJ, Calkins E, et al. ARA nomenclature and classification of arthritis and rheumatism(tentative).arthritisrheum1964;7: Zeidler H, Calin A, Amor B. A historical perspective of the spondyloarthritis. Curr Opin Rheumatol 2011;23: Moll JM, Haslock I, Macrae IF, et al. Associations between ankylosing spondylitis, psoriatic arthritis, Reiter's disease, the intestinal arthropathies, and Behcet's syndrome. Medicine (Baltimore). 1974;53: VanderLindenS,ValkenburgHA,CatsA.Evaluationofdiagnosticcriteriaforankylosingspondylitis.A proposalformodificationofthenewyorkcriteria.arthritisrheum1984;27: AmorB,DougadosM,MijiyawaM.Criteriaoftheclassificationofspondylarthropathies.RevRhumMal Osteoartic1990;57: Dougados M, van der Linden S, Juhlin R, et al. The European Spondylarthropathy Study Group preliminarycriteriafortheclassificationofspondylarthropathy.arthritisrheum1991;34: RudwaleitM,vanderHeijdeD,LandewéR,etal.TheAssessmentofSpondyloArthritisInternational Societyclassificationcriteriaforperipheralspondyloarthritisandforspondyloarthritisingeneral.Ann RheumDis2011;70: Tomero E, Mulero J, de Miguel E, et al. Performance of the Assessment of Spondyloarthritis InternationalSocietycriteriafortheclassificationofspondyloarthritisinearlyspondyloarthritisclinics participatingintheesperanzaprogramme.rheumatology(oxford)2014;53:35360.

23 Generalintroduction Van den Berg R, de Hooge M, van Gaalen F, et al. Percentage of patients with spondyloarthritis in patientsreferredbecauseofchronicbackpainandperformanceofclassificationcriteria:experience fromthespondyloarthritiscaughtearly(space)cohort.rheumatology(oxford)2013;52: SieperJ,vanderHeijdeD.Review:Nonradiographicaxialspondyloarthritis:newdefinitionofanold disease?arthritisrheum2013;65: SieperJ,vanderHeijdeD,LandewéR,etal.Newcriteriaforinflammatorybackpaininpatientswith chronic back pain: a real patient exercise by experts from the Assessment of SpondyloArthritis internationalsociety(asas).annrheumdis2009;68:7848. Rudwaleit M, Metter A, Listing J, et al. Inflammatory back pain in ankylosing spondylitis: a reassessment of the clinical history for application as classification and diagnostic criteria. Arthritis Rheum2006;54: MaksymowychWP.Controversiesinconventionalradiographyinspondyloarthritis.BestPractResClin Rheumatol2012;26: BraunJ,SieperJ,BollowM.Imagingofsacroiliitis.ClinRheumatol2000;19:517. Van Tubergen A, HeuftDorenbosch L, Schulpen G, et al. Radiographic assessment of sacroiliitis by radiologistsandrheumatologists:doestrainingimprovequality?annrheumdis2003;62: VandenBerg,R,Lenczner,G,FeydyA,etal.ReadingOfTheSacroiliacJointsOnPlainRadiographs: Agreement Between Clinical Practice and Trained Central Reading Of The DESIRCohort. Abstract ACR/ARHPmeeting,2013(number:1718). RudwaleitM,JurikAG,HermannKG,etal.Definingactivesacroiliitisonmagneticresonanceimaging (MRI)forclassificationofaxialspondyloarthritis:aconsensualapproachbytheASAS/OMERACTMRI group.annrheumdis2009;68: WeberU,ZublerV,PedersenSJ,etal.Developmentandvalidationofamagneticresonanceimaging reference criterion for defining a positive sacroiliac joint magnetic resonance imaging finding in spondyloarthritis.arthritiscareres(hoboken)2013;65: Maksymowych WP, Inman RD, Salonen D, et al. Spondyloarthritis research Consortium of Canada magnetic resonance imaging index for assessment of sacroiliac joint inflammation in ankylosing spondylitis.arthritisrheum2005;53:7039. Madsen KB, Jurik AG. Magnetic resonance imaging grading system for active and chronic spondylarthritischangesinthesacroiliacjoint.arthritiscareres(hoboken)2010;62:118. Hermann KG, Braun J, Fischer T, et al. Magnetic resonance tomography of sacroiliitis: anatomy, histologicalpathology,mrmorphology,andgrading.radiologe2004;44: RudwaleitM,ClaudepierreP,WordsworthP,etal.Effectiveness,safety,andpredictorsofgoodclinical response in 1250 patients treated with adalimumab for active ankylosing spondylitis. J Rheumatol 2009;36:8018. KeatAC,GaffneyK,GilbertAK,etal.Influenceofbiologictherapyonreturntoworkinpeoplewith workdisabilityduetoankylosingspondylitis.rheumatology(oxford)2008;47:4813. PrinceDS,McGuiganLE,McGirrEE.Workinglifeandphysicalactivityinankylosingspondylitispreand postantitumornecrosisfactoralphatherapy.intjrheumdis2014;17: VanderHeijdeD,HanC,DeVlamK,etal.Infliximabimprovesproductivityandreducesworkdaylossin patients with ankylosing spondylitis: results from a randomized, placebocontrolled trial. Arthritis Rheum2006;55: Boonen A, van der Heijde D, Landewé R, et al. Direct costs of ankylosing spondylitis and its determinants:ananalysisamongthreeeuropeancountries.annrheumdis2003;62: Rudwaleit M, Sieper J. Referral strategies for early diagnosis of axial spondyloarthritis. Nat Rev Rheumatol2012;8: JoisRN,MacgregorAJ,GaffneyK.Recognitionofinflammatorybackpain andankylosingspondylitisinprimarycare.rheumatology(oxford)2008;47: JoisRN,MacgregorAJ,GaffneyK.Recognitionofinflammatorybackpainandankylosingspondylitisin primarycare.rheumatology(oxford)2008;47: VanHoevenL,LuimeJ,HanH,etal.Identifyingaxialspondyloarthritisindutchprimarycarepatients, ages2045years,withchroniclowbackpain.arthritiscareres(hoboken)2014;66: Rudwaleit M, Sieper J. Referral strategies for early diagnosis of axial spondyloarthritis. Nat Rev Rheumatol2012;8:2628. BrandtHC,SpillerI,SongIH,etal.Performanceofreferralrecommendationsinpatientswithchronic backpainandsuspectedaxialspondyloarthritis.annrheumdis2007;66:

24 24 Chapter PoddubnyyD,VahldiekJ,SpillerI,etal.Evaluationof2screeningstrategiesforearlyidentificationof patientswithaxialspondyloarthritisinprimarycare.jrheumatol2011;38: Hermann J, Giessauf H, Schaffler G, et al. Early spondyloarthritis: usefulness of clinical screening. Rheumatology(Oxford)2009;48:8126. Sieper J, Srinivasan S, Zamani O, et al. Comparison of two referral strategies for diagnosis of axial spondyloarthritis:therecognisinganddiagnosingankylosingspondylitisreliably(radar)study.ann RheumDis2013;72: Van den Berg R, de Hooge M, van Gaalen F, et al. Percentage of patients with spondyloarthritis in patientsreferredbecauseofchronicbackpainandperformanceofclassificationcriteria:experience fromthespondyloarthritiscaughtearly(space)cohort.rheumatology(oxford)2013;52: BraunJ,BrandtJ,ListingJ,etal.Biologictherapies in thespondyloarthritis:new opportunities,new challenges.curropinrheumatol2003;15: HeuftDorenboschL,LandewéR,WeijersR,etal.Performanceofvariouscriteriasetsinpatientswith inflammatorybackpainofshortduration;themaastrichtearlyspondyloarthritisclinic.annrheumdis 2007;66:928.

25 PARTI Epidemiologyofspondyloarthritis

26

27 CHAPTER2 Theglobalprevalenceofspondyloarthritis: asystematicreviewandmetaregressionanalysis CarmenStolwijk,MarloesvanOnna,AnneliesBoonen,AstridvanTubergen

28 28 Chapter2 ABSTRACT Objective Tosummarizetheprevalenceofspondyloarthritis(SpA)anditssubtypesinthegeneral population, and to identify demographical and methodological characteristics that mightexplainheterogeneityinprevalenceestimates. Methods Asystematicliteraturesearchwasperformedtoidentifyrelevantarticles.Riskofbias wasassessedanddatawereextracted.pooledprevalenceswerecalculated.potential sources of heterogeneity were explored by subgroup analysis and metaregression analysis. Results Intotal,84articleswereincluded.ThepooledprevalenceofSpA,basedonarandom effects model, was 0.55% (95% CI ); for ankylosing spondylitis (AS) 0.18% (95% CI ); and for psoriatic arthritis (PsA) 0.15% (95% CI ), but substantial heterogeneity was revealed (I 2 >99%). The following characteristics were significantlyassociatedwithvariationinprevalenceofspa,asand/orpsa:proportion of females, mean age of the sample, geographic area and setting (demographical characteristics); year of data collection, case finding, and case ascertainment (methodological characteristics). For the other SpA subgroups too few studies were available to conduct a metaanalysis, but prevalence estimates of reactive arthritis (range0.0%0.2%),sparelatedtoinflammatoryboweldisease(range0.0%0.1%),and undifferentiatedspa(range0.0%0.7%)weregenerallylow. Conclusion SpAisacommondisease,butwithlargevariationinreportedprevalenceestimates, which can partly be explained by differences in demographical and methodological characteristics. Particularly, geographic area as well as case finding account for a substantialpartoftheheterogeneity.

29 Theglobalprevalenceofspondyloarthritis 29 INTRODUCTION Sinceitsestablishmentinthe1970s,thediseaseconceptspondyloarthritis(SpA)has seen major developments with respect to identification and classification of the disease, measurement and prediction of outcome, and treatment options.[1] While SpA can be considered a condition itself, several subtypes can be distinguished, including ankylosing spondylitis (AS), psoriatic arthritis (PsA), SpA related to inflammatory bowel disease (IBDSpA), reactive arthritis (ReA) and undifferentiated SpA (uspa).[2] More recently, the Assessment of SpondyloArthritis international Society (ASAS) proposed to distinguish axial from peripheral SpA, depending on the predominantarticularmanifestationsofthedisease,anddevelopednewclassification criteria.[3,4] The treatment possibilities in patients with SpA have improved substantiallyinthelastdecadewiththeintroductionofbiologicals.onthisline,itis relevanttogaininsightintopatternsoftheprevalenceofspaanditssubtypes,asthis may contribute to our understanding of both the needs of healthcare systemsin termsofavailabilityofhealthcareresourcesandbudgetsandtheetiopathogenesisof thedisease. Considerable variation in the reported prevalence of SpA has already been recognized.[5] In particular, a wide range of estimates across geographic regions is found, which has classically been related to the presence of HLAB27.[6] However, other yet unknown demographical and methodological characteristics of the studies may also play a role in this variation. characteristics refer to for examplethemeanageofthesample,themale:femaleratio,orthegeographicregion. characteristics include for example year of data collection, sampling frame,andcasefinding. Systematic approaches to gain insight into the epidemiology of SpA in the general populationarelimited.[7]nostudieshavebeenperformedassessingandquantifying theeffectofdemographicalandmethodologicalcharacteristicsontheprevalenceof SpA.Therefore,theaimsofthepresentstudywere1)toperformasystematicreview andmetaanalysisoftheliteratureontheprevalenceofspaanditssubtypes,and2)to identify demographical and methodological characteristics that might explain heterogeneityinprevalence.

30 30 Chapter2 METHODS This systematic review was performed in accordance with the quality of reporting metaanalysesofobservationalstudies(moose)guidelines.[8] Searchstrategy MEDLINE (via PubMed) and EMBASE (OVID) were searched between July 1stforprimarystudiesinvestigatingtheprevalenceofSpAoroneofitssubtypes.The search strategy consisted of a combination of text words and controlled vocabulary terms (e.g. MeSH terms) relating to SpA and its subtypes, and to prevalence or epidemiology. The detailed search strategy is outlined in Supplementary File 1. Two reviewers independently reviewed titles and abstracts on eligibility criteria for inclusion, after which fulltext was read. In addition, hand search of references was performed. If the fulltext of the articles could not be retrieved, authors were contactedvia .incaseofanydiscrepanciesbetweenthetworeviewers,athird reviewerwasconsultedforfinaldecision. Selectioncriteria OnlyoriginalresearchstudiesreportingprimarydataontheprevalenceofSpAorits subtypes in the general population were included. Studies were excluded if: 1) the study was not observational; 2) the study was published in a language other than English,French,German,Dutch,SpanishorItalian;3)thestudywasonlydescribingthe prevalenceofyoungonset(<16years)spa;4)thearticlewasnotpublishedinfulltext; or5)thestudyhadasamplesize<1000subjects. Dataextraction Data extraction was performed by two reviewers using a standardized form and included:studyidentification(firstauthor,yearofpublication),anddemographicalas wellasmethodologicalcharacteristics.thecharacteristicsweremean ageofthesample,proportionoffemales,geographicareaandsetting.geographicarea wassubdividedintothefollowingcategories:europe,northamerica,southamerica, MiddleEastandNorthAfrica,SubSaharanAfrica,CentralAsiaandRussia,SouthAsia, SouthEastAsia,EastAsia,Oceania,and NorthernArticindigenouscommunities.The lastcategorywasseparatedsinceitisknownthattheprevalenceofhlab27ishigher in these populations.[9,10] The setting was subdivided into urban, rural, or a

31 Theglobalprevalenceofspondyloarthritis 31 combination of both. The characteristics were starting year of data collection (or alternatively publication year if not reported), sampling frame, case findingandcaseascertainment.samplingframewassubdividedintocensus,household register, convenience sample, general practitioner database, hospital database, register, or a list of specific group of subjects (e.g. employees of a company). Case finding was based on the procedure to identify cases and included selfreported symptoms, selfreported diagnosis, selfreported diagnosis followed by external confirmation(2stepapproachdiagnosis),selfreportedsymptomsfollowedbyexternal confirmation(2stepapproachsymptoms),diagnosisbyanexpert,andhospitalmedical recordsorinternationalclassificationofdisease(icd)codes.caseascertainmentwas basedonanexternalcriterionusedforcasedefinitionandwassubdividedintoclinical diagnosis (i.e. diagnosis by a physician) and classification criteria used for each SpA subtype, such as the modified New York criteria for AS, the European Spondyloarthropathy Study Group (ESSG) criteria, the ClASsification for Psoriatic Arthritis(CASPAR)criteria,orASASSpAcriteria.[3,1113] Finally, data related to prevalence were extracted (raw data were extracted or the numerator and denominator were calculated). If a study presented age and/or sex specificestimates,thesedatawereextracted,insteadofthetotalcount. Riskofbiasassessment Tworeviewersassessedindependentlytheriskofbiasforeachincludedstudy,usinga slightlymodifiedversionofastandardizedtoolbyhoyetal..[14]supplementaryfile2 shows the risk of bias tool including instructions on how each item was scored. We excludedtheitem wasthelengthoftheshortestprevalenceperiodfortheparameter ofinterestappropriate fromthechecklist,sincethisquestionisnotapplicablefora chronicdisease,suchasspa. Datasynthesisandanalysis Because prevalence estimates were expected to be below 1%, the values were transformed with the double arcsine transformation for metaanalysis and meta regression.[15] The pooled prevalence was estimated by combining the transformed prevalence estimates using a randomeffects model. Studies from Northern Artic indigenous communities were excluded from the metaanalysis, because the risk for SpAinthisgroupisclearlydifferentfromtherestoftheworld,whichwouldbias(i.e. overestimate) the results. The pooled prevalence and 95% confidence intervals (CI)

32 32 Chapter2 were back transformed to prevalence estimates for ease of interpretation. The heterogeneityamongstudieswastestedbythecochran sqtestandthei 2 statistic,the latterdescribingthepercentageofvariationacrossstudies.[16] Potential sources of heterogeneity were investigated by an exploratory subgroup analysis, using random effects analogous to oneway analysis of variance, in which groups of estimates were arranged according to potentially relevant demographical (mean age of the sample, proportion of females, geographic area, and setting) and methodologicalcharacteristics(yearofdatacollection,studysize,samplingframe,case finding,andcaseascertainment).studiesfromnorthernarticindigenouscommunities wereexcludedinthesubgroupanalysis. Univariable and multivariable metaregression analyses were performed to explore associations between demographical and methodological characteristics and the prevalence.variableswithapvalueof<0.20intheunivariableanalysiswereentered intothemultivariablemodel.abackwardprocedurewasused,removingvariableswith a pvalue of >0.05 in the multivariable model in order of significance. The following variables were tested: mean age of the sample, proportion of females, geographic area, setting, year of data collection, case finding, case ascertainment, and the dichotomized risk of bias criteria not yet covered by the previous variables. Due to collinearity,samplingframewasnotincludedinthemodel.ifdataonageorsexwere missing,respectivelythemeanor50%wereimputed.forthemetaregressionanalysis, SPSSmacroswereused(Metareg.spsandMetaF.sps).[17] RESULTS Searchresults The database search yielded 9,240 studies (Supplementary File 3). After removing duplicatesandtitle/abstractscreening,179articlesremainedforfullpaperreview.two papers could not be retrieved, despite contacting the authors.[18,19] After fulltext reading, 80 articles were included. With hand search of references, 4 papers were added,leadingtoatotalof84articles. Characteristicsofincludedstudies The prevalence of SpA was reported in 30 studies (100 age and/or sexspecific estimates),ofasin53studies(179estimates),ofpsain35studies(89estimates),of ReAin17studies(67estimates),ofIBDSpAin4studies(15estimates),andofuSpAin

33 Theglobalprevalenceofspondyloarthritis studies (88 estimates). A detailed overview of all included studies is provided in SupplementaryFile4.Acombinedmetaanalysisandmetaregressionanalysiscouldbe performed for SpA, AS andpsa. Onlya limited number of studieswereavailable for ReA, IBDSpA and uspa, therefore these results are described and summarized narratively. Table 2.1 shows a summary of the main characteristics of studies included in the largestgroups;spa,asandpsa. Riskofbias AnoverviewoftheriskofbiasassessmentisprovidedinSupplementaryFile5.High riskofbiaswasmostcommonforitem1(representativenessofsampleforthenational generalpopulation)anditem2(representativenessofsamplingframe). Prevalenceofspondyloarthritis The pooled population prevalence of SpA was 0.55% (95%CI ), with high heterogeneity(i 2 =99.9%).Figure2.1showstheprevalenceestimatesofSpAaccording todifferentdemographicalandmethodologicalcharacteristics.meanageofthestudy population and geographic area contributed significantly to the observed heterogeneity. The prevalence of SpA was higher in studies from North America (1.35%,95%CI ,n=1study)andEurope(0.54%,95%CI )compared with South Asia (0.22%, 95%CI ) and SouthEast Asia (0.20%, 95%CI ). No studies from SubSaharan Africa, Central Asia and Oceania were available (Figure 2.2a). With respect to the sampling frame, true population studies reported higher prevalence estimates compared with hospitalbased studies. The prevalenceofspawasalsohigherinmorerecentstudies(yearofdatacollectionfrom 2000onwards)andinsmallerstudies(<5000subjects).Further,prevalenceestimates werehigherifbasedontheessgcriteriacomparedwithclinicaldiagnosis.table2.2 showstheresultsofthemetaregressionanalysis.thefinalmodelexplained51.8%of theheterogeneity.inthismodel,casefinding(p<0.01),alowerproportionoffemales (p=0.01),geographicarea(p<0.01),andmorerecentyearofdatacollection(p=0.02) werepositivelyassociatedwiththeprevalenceofspa.

34 34 Chapter2 Table2.1.Characteristicsoftheincludedstudiesforspondyloarthritis,ankylosingspondylitis,andpsoriaticarthritis. Spondyloarthritis Ankylosingspondylitis Psoriaticarthritis Mean(range) Numberof subjects Mean(range) Nstudies (estimates)* Numberof subjects Mean(range) Nstudies (estimates)* Numberof subjects Nstudies (estimates)* Total 30(100) 104,466,975 53(179) 119,995,702 35(89) 14,105,141 17(67) 50.6 ( ) 35(157) 48.2 (19.585) Ageinyears 21(87) 49.4 (20.585) 20(33) 51.1 (0100) 39(154) 49.2 (0100) %female 25(84) 49.5 (0100) 1998 ( ) 1990 ( ) 2000 ( ) Startingyeardata collection(range) Geographicareasampled Europe 10(24) 3,281,474 21(43) 10,312,889 19(46) 10,532,446 NorthAmerica 1(2) 5,103 2(15) 109,414,800 4(4) 3,416,497 SouthAmerica 2(3) 8,022 4(6) 28,841 4(5) 100,958 SubSaharanAfrica 3(3) 102,467 MiddleEast 5(8) 23,316 5(7) 34,821 1(1) 7,670 EastAsia 2(15) 101,110,921 6(44) 43,379 2(15) 21,477 SouthEastAsia 2(4) 4,574 2(3) 5,600 1(1) 2,594 SouthAsia 4(4) 16,155 2(2) 13,305 1(1) 8,145 Indigenous 4(40) 17,410 8(56) 39,600 3(18) 15,354 *Numberofstudies(numberofestimates).Onestudycanprovideforexampletwoestimates:oneforfemaleandoneformalesubjects.:nodataavailable.

35 Theglobalprevalenceofspondyloarthritis 35 Figure2.1 Prevalence of spondyloarthritis grouped by demographical and methodological characteristics. Estimates of Northern Artic communities were not included in the subgroupanalysis. ESSG=European Spondyloarthropathy Study Group; ASAS= Assessment of Spondyloarthritis InternationalSociety. AxialandperipheralSpA Twostudiesreportedtheprevalenceofaxialand/orperipheralSpAaccordingtothe ASAS classification criteria.[20,21] In a large populationbased cohort consisting of 20,625 employees of the French national electricity and gas company, a crude prevalenceofspaof0.48%wasfound(0.36%foraxialspaand0.12%forperipheral SpA)(20).InastudyfromtheUS,inwhichasampleofmedicalrecordsofpatientswith chronicbackpainwerereviewedagainsttheasascriteriaandextrapolatedtootherus rheumatologypractices,theprevalenceofaxialspawasestimatedat0.70%.[21]

36 36 Chapter2 a b c Figure2.2 Maps showing the global prevalence of spondyloarthritis (a), ankylosing spondylitis (b), and psoriaticarthritis(c)

37 Theglobalprevalenceofspondyloarthritis 37 Table2.2 Univariableandmultivariablemetaregressionanalysisontheprevalenceofspondyloarthritis. Univariableanalysis Multivariableanalysis Characteristic B(95%CI) Pvalue R2 B(95%CI) Pvalue Meanage 0.4*103(0.9*103;0.2*102) NE %female 0.4*103(0.8*103;0.00) *103(0.6*103;0.1*103) 0.01 NorthAmerica 0.08(0.02;0.12) (0.01;0.18) 0.02 SouthAmerica 0.4*102(0.09;0.08) (0.15;0.00) 0.06 MiddleEast 0.04(0.09;0.02) (0.15;0.04) <0.01 EastAsia 0.03(0.02;0.08) (0.06;0.04) 0.66 SouthEastAsia 0.06(0.14;0.02) (0.16;0.01) 0.02 SouthAsia 0.05(0.13;0.02) (0.16;0.03) <0.01 Region Reference=Europe NorthernArticIndigenous 0.10(0.06;0.14) < (0.11;0.24) <0.01 Setting Rural 0.10(0.05;0.16) < NS Reference=urban Combination 0.01(0.05;0.03) 0.68 NS Startdatacollection 0.3*102(0.7*103;0.5*102) *102(0.7*103;0.7*102) 0.02 Riskofbiasitem1,representativenesstargetpopulation 0.03(0.05;0.10) NE Riskofbiasitem2,representativenesssamplingframe 0.2*102(0.04;0.05) NE Riskofbiasitem3,randomselection 0.01(0.08;0.07) NE Riskofbiasitem4,nonresponsebias 0.07(0.11;0.02) < NS Riskofbiasitem5,wasaproxyused? 0.8*103(0.05;0.05) NE Riskofbiasitem6,validityofcasedefinition* 0.06(0.12;0.00) NA Riskofbiasitem7,validandreliablestudyinstrument 0.02(0.03;0.07) NE Riskofbiasitem8,datacollectionmode 0.16(0.10;0.22) < NS Riskofbiasitem9,numerator/denominatorappropriate 0.06(0.14;0.02) NS Twostepdiagnosis 0.13(0.08;0.18) < (0.02;0.11) <0.01 Twostepsymptoms 0.01(0.03;0.05) (0.04;0.13) <0.01 Expertdiagnosis 0.11(0.02;0.20) (0.01;0.15) 0.03 Casefinding Reference=medicalrecords Selfreport 0.09(0.02;0.19) (0.01;0.18) 0.02 Caseascertainment ESSG 0.08(0.03;0.14) < NS Reference=clinicaldiagnosis ASAS 0.02(0.14;0.18) 0.80 NS *Riskofbiasitem6wasnotincludedinthemultivariableanalysisduetocollinearitywithcaseascertainment.ESSG=EuropeanSpondyloarthropathyStudyGroup criteria; ASAS=Assessment of SpondyloArthritis International Society criteria; ICD=international classification of disease; NE=not entered in multivariable model (becausep>0.20inunivariableanalysis);ns=notsignificant;na=notapplicable (becauseofcollinearity).

38 38 Chapter2 Ankylosingspondylitis ThepooledpopulationprevalenceofASwas0.18%(95%CI ).Heterogeneity was high (I 2 =99.0%). Figure 2.3 shows the pooled prevalence of AS stratified by subgroups. The prevalence of AS was higher in males compared with females. The prevalencewashigherinruralpopulationsanddifferentacrossgeographicareas,with thehighestprevalenceratesinstudiesfromeuropeandnorthamerica(figure2.2b). Withregardtothemethodologicalcharacteristics,theprevalencewashigherinstudies with<5000subjects,differentamongsamplingframes,andhigherwhenpatientswere classifiedaccordingtothe(modified)newyorkorasascriteriacomparedwithclinical diagnosis. ThemetaregressionanalysisoftheprevalenceofASisprovidedinSupplementaryFile 6.Thefinalmodelexplained47.0%ofheterogeneity.TheprevalenceofASwashigher in samples with a lower percentage of females (p<0.01), different across geographic areas (higher prevalence rates in studies from North America, Europe and Northern Articindigenouscommunitiescomparedwithallothergeographicregions),andhigher insamplesfromruralareascomparedwithurbanareas(p<0.01).theprevalencewas lowerinsamplesinwhichcaseswerefoundbymedicalrecordscomparedwithtwo step symptoms approaches (p<0.01), but higher in studies with high risk of bias for validity/reliabilityofthestudyinstrument(p<0.01). Psoriaticarthritis ThepooledpopulationprevalenceofPsAwas0.15%(95%CI ).Heterogeneity washigh(i 2 =99.2%).TheresultsofthesubgroupanalysisareshowninFigure2.4.The prevalence of PsA was significantly different across age groups and related to geographic area (Figure 2.2c). The highest prevalence was found in Europe (0.19%, 95%CI ) and the lowest in the Middle East (0.01%, 95%CI ). The prevalenceofpsawasfurthersignificantlyrelatedtosamplingframeandcasefinding. The highest prevalence of PsA was found when diagnosis was based on selfreport (0.26%,95%CI ).

39 Theglobalprevalenceofspondyloarthritis 39 Figure2.3 Prevalence of ankylosing spondylitis grouped by demographical and methodological characteristics.

40 40 Chapter2 Figure2.4 Prevalenceofpsoriaticarthritisgroupedbydemographicalandmethodologicalcharacteristics.

41 Theglobalprevalenceofspondyloarthritis 41 SupplementaryFile7showsthemetaregressionanalysisoftheprevalenceofPsA.The final model explained 44.7% of the total heterogeneity. A higher mean age of the sample was positively related to the prevalence of PsA (p<0.01) and prevalence was significantly different across geographic areas. With respect to the methodological characteristics, prevalence was significantly higher when case finding was based on selfreportcomparedwithmedicalrecords. Reactivearthritis TheprevalenceofReAwasreportedin17studies:6fromEurope,1fromSouthEast Asia,1fromSubSaharanAfrica,and9fromNorthernArcticindigenouspopulations.In Europe, the prevalence of ReA ranged from 0.03% in Greece to 0.21% in Lithuania.[22,23]InastudyfromIndia,inwhich8,145individualswereinterviewed,no cases of ReA were identified (prevalence 0%).[24] In a hospitalbased study from Zimbabweaprevalenceof0.001%wasfound.[25]InNorthernArticindigenouspeople, theprevalencerangedfrom0.25%to1.0%.[26,27] Spondyloarthritisassociatedwithinflammatoryboweldisease TheprevalenceofIBDSpAwasreportedinonly4studies,allfromEurope.Prevalence rangedfrom0.0%ingreeceto0.09%initaly.[22,28]noformalcriteriaexisttoclassify IBDSpA. In these 4 studies, classification was based on the ASAScriteria, the ESSG criteriaincombinationwithibd,oricdcodes.[20,22,28,29] Undifferentiatedspondyloarthritis The prevalence of uspa was reported in 15 studies. Different criteria were used to classify patients with uspa. In most European studies, the prevalence ranged from 0.03%to0.10%.[20,22,23,2830]InanurbanpopulationfromTurkey,aprevalenceof 0.56%wasfound.[31]InastudyinblooddonorsfromGermanyaprevalenceof0.67% wasreported.[32]intwoasianstudies,thereportedprevalencewas0.15%inastudy from India [24] and 0.55% in a study from China.[33] In Northern Artic indigenous people,prevalencesrangedfrom0.20%to1.3%.[26,27,3436]

42 42 Chapter2 DISCUSSION Inthissystematicreview,pooledpopulationprevalenceestimateswerecalculatedfor SpA (0.55%, 95%CI ), AS (0.18%, 95%CI ) and PsA (0.15%, 95%CI ). Heterogeneity across studies was high, therefore estimates should be interpretedwithcaution.forrea,ibdspaanduspatoofewstudieswereavailableto conduct a metaanalysis and, therefore, results were only summarized. Prevalence estimates of ReA (range 0.0%0.2%), IBDSpA (range 0.0%0.1%), and uspa (range 0.0%0.7%)weregenerallylow. ThisstudyisthefirstthatpooledprevalencesofSpAanditssubtypesinthegeneral population, and additionally investigated demographical and methodological characteristics influencing the estimates. Geographic area was in the multivariable metaregression analysis one of the most important characteristics explaining heterogeneity in prevalence estimates of SpA. This variation might particularly be explained by genetic characteristics, such as HLAB27. Independent of other characteristics,thehighestprevalenceestimatesofspawerefoundinnorthernartic indigenouscommunities,inwhichupto50%ofpeoplehavebeenreportedtobehla B27positive.[9]Further,higherprevalenceestimateswerefoundinstudiesfromNorth America and Europe compared with Asia, Africa and the Middle East, corresponding with reported HLAB27 prevalences in these areas.[37] A relatively high pooled prevalenceofspawasfoundinnorthamerica,however,theestimatescamefroma single study with a high risk of bias.[38] In this study, the prevalence of SpA was estimatedaccordingtotheessgandtheamorcriteria,resultinginestimatesof1.4% and 0.9% respectively.[38] This finding also illustrates that the choice of different classification criteria in epidemiological studies have a large impact on the reported prevalence.forreasonsofcomparability,weusedestimatesfromthisstudybasedon theessgcriteriawhenreportingthepooledprevalenceinthepresentreview,because thesewerealsoappliedinmostotherstudies.[38]alsotheprevalenceofspaineast Asiawasunexpectedlyhigh(0.79%),andseemstobeexplainedbyahighprevalenceof uspainchina(0.55%).[33] Amongthemethodologicalcharacteristicsexplored,prevalenceestimatesofSpAwere positivelyandindependentlyrelatedtotheyearofdatacollectionwithmorerecent studies reporting higher prevalences. This may either be a true increase in the prevalence of SpA, or, more likely, an increased awareness and recognition of SpA. Further,prevalenceestimateswerehigherinstudieswherepopulationswerescreened forspacomparedwithstudiesinwhichcaseswereidentifiedfrommedicalrecords.

43 Theglobalprevalenceofspondyloarthritis 43 ThisfindingsuggeststhatthereisunderrecognitionofSpA.Inthesubgroupanalyses this is supported by the fact that sampling form census lists yielded the highest prevalence. With respect to case ascertainment, no significant difference was found betweenclassificationbasedontheessgcriteriacomparedwithclinicaldiagnoses.in thepresentreview,toofewstudiesusingthenewasascriteriawereavailabletodraw a conclusion. Future studies are needed to gain more insight into the impact of the ASAScriteriaontheprevalenceofSpAinpopulationstudies,althoughapplicationof these criteria might be less feasible in large epidemiological studies, because of inclusionofhlab27andmri.[39] SimilarlytoSpA,alsotheprevalenceofASwassignificantlydifferentamonggeographic areasinmultivariablemetaregressionandgenerallyhigherinregionsofthenorthern hemisphere. Further, the prevalence of AS was higher in the male population. Traditionally,ASisconsideredasadiseasepredominantlyoccurringinmales,although it has been shown that this may in part be an artefact induced by deficits in the diagnosisofasinfemales.[4042]nonradiographicaxialspa,ontheotherhand,isas common in female as in male subjects, indicating that females develop structural changes later or less frequently than males.[39,43] In contrast, no difference in the prevalence of PsA in gender distribution was found with the multivariable meta regression. The prevalence of PsA, however, was significantly related to age, and peakedintheagecategorybetween50and60years. Some limitations of the present study should be addressed. First, we applied a languagerestriction;thereforelanguagebiascannotbeexcluded.second,becausethe majorityofthestudiescamefromeurope,geographicalbiascouldhaveoccurred.this undoubtedlyinfluencedthepooledpopulationprevalence,whichshould,therefore,be interpretedwithcaution.further,evenwithinthedefinedgeographicareas,variation in prevalence might exist, for example between North and South Europe. However, further subdivision would have hampered the metaregression analysis. Third, direct comparisonsbetweentheresultsofthemetaanalysesandmetaregressionanalyses ofspa,as,andpsaarehampered,becauseoftenpopulationsfromdifferentstudies wereused.last,metaregressionanalysisitselfhassomelimitations.[44]resultsfrom metaregressions are observational, and therefore, can suffer from bias by confounding. Patients characteristics are based on groupaverages, and the relationshiponstudylevelmaynotbethesameastherelationshiponpatientlevel. Consistent with these concerns, high risk of bias was found for the items on representativeness of the sample and the sampling frame. Only a few studies were

44 44 Chapter2 trulyrepresentativeforthegeneralpopulation,whichmayhamperthegeneralization oftheresults. Inconclusion,thissystematicreviewwithmetaanalysissummarizedtheprevalenceof SpA and its subtypes. A large part of the heterogeneity could be explained by geographic characteristics. However, also other demographical and methodological characteristics, such as the proportion of females, year of data collection and case finding accounted for the observed variation. The results also suggested that there mightstillbeanunderdiagnosisofspa.itistobeexpectedthatbetterrecognitionof SpA will likely further increase the prevalence. High quality studies are needed to estimatetheprevalenceofaxialandperipheralspainthegeneralpopulation,andto estimatetheprevalenceofspaindevelopingcountries.

45 Theglobalprevalenceofspondyloarthritis 45 REFERENCES 1. Moll JM, Haslock I, Macrae IF, et al. Associations between ankylosing spondylitis, psoriatic arthritis, Reiter's disease, the intestinal arthropathies, and Behcet's syndrome. Medicine (Baltimore). 1974/09/01ed1974.p DougadosM,BaetenD.Spondyloarthritis.Lancet2011;377: RudwaleitM,vanderHeijdeD,LandeweR,etal.TheAssessmentofSpondyloArthritisInternational Societyclassificationcriteriaforperipheralspondyloarthritisandforspondyloarthritisingeneral.Ann RheumDis2011;70: RudwaleitM,vanderHeijdeD,LandeweR,etal.ThedevelopmentofAssessmentofSpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: StolwijkC,BoonenA,vanTubergenA,etal.EpidemiologyofSpondyloarthritis.RheumDisClinNAm 2012;38: MathieuA,PaladiniF,VaccaA,etal.TheinterplaybetweenthegeographicdistributionofHLAB27 allelesandtheirroleininfectiousandautoimmunediseases:aunifyinghypothesis.autoimmunrev 2009;8: DeanLE,JonesGT,MacDonaldAG,etal.Globalprevalenceofankylosingspondylitis.Rheumatology 2014;53: Stroup DF, Berlin JA, Morton SC, et al. Metaanalysis of observational studies in epidemiology: a proposalforreporting.metaanalysisofobservationalstudiesinepidemiology(moose)group.jama 2000;283: GoftonJP,ChalmersA,PriceGE,etal.HLA27andankylosingspondylitisinB.C.Indians.JRheumatol 1984;11: Benevolenskaia LI, Erdes S, Krylov M, et al. The epidemiology of spondyloarthropathies among the native inhabitants of Chukotka. 2. The prevalence of HLAB27 in the population and among spondyloarthropathypatients.terarkh1994;66: VanderLindenS,ValkenburgHA,CatsA.Evaluationofdiagnosticcriteriaforankylosingspondylitis.A proposalformodificationofthenewyorkcriteria.arthritisrheum1984;27: Dougados M, van der Linden S, Juhlin R, et al. The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis and rheumatism 1991;34: RudwaleitM,vanderHeijdeD,LandeweR,etal.ThedevelopmentofAssessmentofSpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: HoyD,BrooksP,WoolfA,etal.Assessingriskofbiasinprevalencestudies:modificationofanexisting toolandevidenceofinterrateragreement.jclinepidemiol2012;65: Barendregt JJ, Doi SA, Lee YY, et al. Metaanalysis of prevalence. J Epidemiol Community Health 2013;67: HigginsJP,ThompsonSG.Quantifyingheterogeneityinametaanalysis.StatMed2002;21: Wilson D. METAREG for SPSS/Win 6.1 or higher MACRO. 2001; Available from: ChopraA,PatilJ,BillempellyV,etal.Prevalenceofrheumaticdiseasesinaruralpopulationinwestern India:aWHOILARCOPCORDStudy.JAssocPhysiciansIndia2001;49: MürdenKD,ValkenburgHA,HopperJ.TheepidemiologyofrheumaticdiseasesinAustralia.APLARJ Rheumatol1992: CostantinoF,TalpinA,SaidNahalR,etal.PrevalenceofspondyloarthritisinreferencetoHLAB27in thefrenchpopulation:resultsofthegazelcohort.annrheumdis2013.doi: /annrheumdis [Epubaheadofprint] 21. Strand V, Rao SA, Shillington AC, et al. Prevalence of axial spondyloarthritis in United States rheumatology practices: Assessment of SpondyloArthritis International Society criteria versus rheumatologyexpertclinicaldiagnosis.arthritiscareres(hoboken)2013;65:

46 46 Chapter Trontzas P, Andrianakos A, Miyakis S, et al. Seronegative spondyloarthropathies in Greece: a populationbased study of prevalence, clinical pattern, and management. The ESORDIG study. Clin Rheumatol2005;24:5839. Adomaviciute D, Pileckyte M, Baranauskaite A, et al. Prevalence survey of rheumatoid arthritis and spondyloarthropathyinlithuania.scandjrheumatol2008;37:1139. Joshi VL, Chopra A. Is there an urbanrural divide? Population surveys of rheumatic musculoskeletal disordersinthepuneregionofindiausingthecopcordbhigwanmodel.jrheumatol2009;36: Lutalo SK. Chronic inflammatory rheumatic diseases in black Zimbabweans. Ann Rheum Dis 1985;44:1215. BoyerGS,LanierAP,TemplinDW,etal.SpondyloarthropathyandrheumatoidarthritisinAlaskanYupik Eskimos.JRheumatol1990;17: Boyer GS, Templin DW, CornoniHuntley JC, et al. Prevalence of spondyloarthropathies in Alaskan Eskimos.JRheumatol1994;21: DeAngelisR,SalaffiF,GrassiW.PrevalenceofspondyloarthropathiesinanItalianpopulationsample:a regionalcommunitybasedstudy.scandjrheumatol2007;36:1421. Haglund E, Bremander AB, Petersson IF, et al. Prevalence of spondyloarthritis and its subtypes in southernsweden.annrheumdis2011;70:9438. SarauxA,GuilleminF,GuggenbuhlP,etal.PrevalenceofspondyloarthropathiesinFrance:2001.Ann RheumDis2005;64: OnenF,AkarS,BirlikM,etal.Prevalenceofankylosingspondylitisandrelatedspondyloarthritidesin anurbanareaofizmir,turkey.jrheumatol2008;35:3059. Braun J, Bollow M, Remlinger G, et al. Prevalence of spondylarthropathies in HLAB27 positive and negativeblooddonors.arthritisrheum1998;41:5867. LiaoZT,PanYF,HuangJL,etal.Anepidemiologicalsurveyoflowbackpainandaxialspondyloarthritis inachinesehanpopulation.scandjrheumatol2009;38:4559. BenevolenskayaLI,BoyerGS,ErdeszS,etal.Spondylarthropathicdiseasesinindigenouscircumpolar populationsofrussiaandalaska.revrhumengled1996;63: BoyerGS,LanierAP,TemplinDW.Prevalenceratesofspondyloarthropathies,rheumatoidarthritis,and otherrheumaticdisordersinanalaskaninupiateskimopopulation.jrheumatol1988;15: BoyerGS,TemplinDW,LanierAP.RheumaticdiseasesinAlaskanIndiansofthesoutheastcoast:high prevalenceofrheumatoidarthritisandsystemiclupuserythematosus.jrheumatol1991;18: BaklandG,NossentHC.Epidemiologyofspondyloarthritis:areview.CurrRheumatolRep2013;15:351. ReveilleJD,WitterJP,WeismanMH.PrevalenceofaxialspondylarthritisintheUnitedStates:estimates fromacrosssectionalsurvey.arthritiscareres(hoboken)2012;64: Van Tubergen A. The changing clinical picture and epidemiology of spondyloarthritis. Nat Rev Rheumatol2014. Gran JT, Husby G, Hordvik M. Prevalence of ankylosing spondylitis in males and females in a young middleagedpopulationoftromso,northernnorway.annrheumdis1985;44: Masi AT, Wilkins WR. Does male:female sex ratio in ankylosing spondylitis change with age? J Rheumatol1996;23:9478. Feldtkeller E, Bruckel J, Khan MA. Scientific contributions of ankylosing spondylitis patient advocacy groups.curropinrheumatol2000;12: SieperJ,vanderHeijdeD.Review:Nonradiographicaxialspondyloarthritis:newdefinitionofanold disease?arthritisandrheumatism2013;65: ThompsonSG,HigginsJP.Howshouldmetaregressionanalysesbeundertakenandinterpreted?Stat Med2002;21:

47 Theglobalprevalenceofspondyloarthritis 47 SUPPLEMENTARYFILE1 Searchstrategy 1. Spondylarthropathies/ 2. Spondylarthr* 3. Spondyloarthr* 4. Spondyloarthritis.tw 5. (1OR2OR3OR4) 6. Bechtere*.tw. 7. (ankylos*andspondyl*).tw. 8. (MarieANDstruempell*).tw. 9. (6OR7OR8) 10. Arthritis,Psoriatic/ 11. (psoria*and(arthriti*orarthropath*)).tw. 12. (10OR11) 13. Arthritis,Infectious/ 14. reactivearthritis.tw. 15. (reiter*and(diseaseorsyndrome)).tw. 16. ((sexual*orchlamydiaoryersiniaorpostyersiniaorpostdysentericor salmonellaorshigellaorb27orpostinfectiousorpostinfectious)and arthrit*).tw. 17. (13OR14OR15OR16) 18. InflammatoryBowelDiseases/ 19. exparthritis/ 20. (18AND19) 21. COPCORD 22. (5OR9OR12OR17OR20OR21) 23. Prevalence.tw. 24. Prevalence/ 25. Incidence.tw. 26. Incidence/ 27. Epidemiology/ 28. Epidemiolog*.tw. 29. (23OR24OR25OR26OR27OR28) AND29

48 48 Chapter2 SUPPLEMENTARYFILE2 Riskofbiastool ThisriskofbiastoolisbasedontheriskofbiastoolbyHoyetal.[1],adaptedtoassess theriskofbiasinprevalencestudiesofspondyloarthritis(spa). Riskofbiasitem Criteriaforanswers Additionalnotesandexamples 1.Wasthetarget populationaclose representationofthe nationalpopulationin relationtorelevant variables,e.g.age,sex, occupation? Yes(LOWRISK):Thestudy s targetpopulationwasaclose representationofthenational population. No(HIGHRISK):Thestudy s targetpopulationwasclearly NOTrepresentativeofthe nationalpopulation Thetargetpopulationreferstothegroupof peopleorentitiestowhichtheresultsofthe studywillbegeneralised.examples: Thestudywasanationalhealthsurveyofpeople 15yearsandoverandthesamplewasdrawn fromalistthatincludedallindividuals inthepopulationaged15yearsandover.the answeris:yes(lowrisk). Thestudywasconductedinoneprovinceonly, anditisnotclearifthiswasrepresentativeofthe nationalpopulation.theansweris: No(HIGHRISK). Thestudywasundertakeninonevillageonlyand itisclearthatthiswasnotrepresentativeofthe nationalpopulation.theansweris: No(HIGHRISK). 2.Wasthesampling frameatrueorclose representationofthe targetpopulation? Yes(LOWRISK):The samplingframewasatrueor closerepresentationofthe targetpopulation. No(HIGHRISK):Thesampling framewasnotatrueorclose representationofthetarget population. Thesamplingframeisalistofthesamplingunits inthetargetpopulationandthestudysampleis drawnfromthislist.examples: Thesamplingframewasalistofalmostevery individualwithinthetargetpopulation.the answeris:yes(lowrisk). Theclustersamplingmethodwasusedandthe sampleofclusters/villageswasdrawnfromalist ofallvillagesinthetargetpopulation.the answeris:yes(lowrisk). Thesamplingframewasalistofjustone particularethnicgroupwithintheoveralltarget population,whichcomprisedmanygroups. Theansweris:No(HIGHRISK). 3.Wassomeformof randomselectionused toselectthesample? Yes(LOWRISK):Acensus wasundertaken,or,someform ofrandomselectionwasusedto selectthesample(e.g.simple randomsampling,stratified randomsampling,cluster sampling,systematicsampling). No(HIGHRISK):Acensuswas NOTundertaken,ANDsome formofrandomselectionwas Acensuscollectsinformationfromeveryunitin thesamplingframe.inasurvey,onlypartofthe samplingframeissampled.inthese instances,randomselectionofthesamplehelps minimisestudybias. Examples: Thesamplewasselectedusingsimplerandom sampling.theansweris:yes(lowrisk). Thetargetpopulationwasthevillageandevery personinthevillagewassampled.theansweris:

49 Theglobalprevalenceofspondyloarthritis 49 NOTusedtoselectthesample. Yes(LOWRISK). Thenearestvillagestothecapitalcitywere selectedinordertosaveonthecostoffuel.the answeris:no(highrisk). 4.Wasthelikelihoodof nonresponsebias minimal? Acensuscollectsinformation fromeveryunitinthesampling frame.inasurvey,onlypartof thesamplingframeissampled. Inthese instances,randomselectionof thesamplehelpsminimising studybias.examples: Thesamplewasselectedusing simplerandomsampling.the answeris:yes(lowrisk). Thetargetpopulationwasthe villageandeverypersoninthe villagewassampled.theanswer is:yes(lowrisk). Thenearestvillagestothe capitalcitywereselectedin ordertosaveonthecostoffuel. Theansweris:No(HIGHRISK). Acensuscollectsinformationfromeveryunitin thesamplingframe.inasurvey,onlypartofthe samplingframeissampled.intheseinstances, randomselectionofthesamplehelpsminimising studybias. Examples: Thesamplewasselectedusingsimplerandom sampling.theansweris:yes(lowrisk). Thetargetpopulationwasthevillageandevery personinthevillagewassampled.theansweris: Yes(LOWRISK). Thenearestvillagestothecapitalcitywere selectedinordertosaveonthecostoffuel.the answeris:no(highrisk). 5.Weredatacollected directlyfromthe subjects? Yes(LOWRISK):Alldatawere collecteddirectlyfromthe subjects. No(HIGHRISK):Insome instances,datawerecollected fromaproxy. Examples: Alleligiblesubjectsinthehouseholdwere interviewedseparately.theansweris:yes(low RISK). Arepresentativeofthehouseholdwas interviewedandquestionedaboutthepresence oflowbackpainineachhouseholdmember. Theansweris:No(HIGHRISK). Datawerecollectedfrommedicalrecordsora dataset:no(highrisk) 6.Wasanacceptable casedefinitionused Yes(LOWRISK):An acceptablecasedefinitionwas used. No(HIGHRISK):An acceptablecasedefinitionwas NOTused. Examples: SpA:casedefinitionwasbasedontheASAS criteria,essgcriteria,expertopinion,medical records,oricdcodes:yes(lowrisk). AS:casedefinitionwasbasedonthe(modified) NewYorkcriteria,theRomecriteria,medical records,oricdcodes:yes(lowrisk). PsA:casedefinitionwasbasedon:theCASPAR criteria,essgcriteria+psoriasis,asascriteria+ psoriasis,medicalrecords,oricdcodes:yes (LOWRISK) Therewasnodescriptionofclassificationcriteria used:no(highrisk) 7.Wasthestudy instrumentreliableand valid? Yes(LOWRISK):Thestudy instrumentisshowntobe reliableandvalid(ifthiswas TheauthorsusedtheCOPCORDquestionnaire, whichhadpreviouslybeenvalidated.theyalso testedtheinterraterreliabilityofthe

50 50 Chapter2 necessary),e.g.testretest, piloting,validationinaprevious study,etc. No(HIGHRISK):thestudy instrumentisnotshowntobe reliableandvalid(ifthiswas necessary) questionnaire.theansweris:yes(lowrisk). Theauthorsdevelopedtheirownquestionnaire anddidnottestthis forvalidityorreliability.theansweris:no(high RISK). 8.Wasthesamemode ofdatacollectionused forallsubjects?, Yes(LOWRISK):Thesame modeofdatacollectionwas usedforallsubjects. No(HIGHRISK):Thesame modeofdatacollectionwas NOTusedforallsubjects. Themodeofdatacollectionisthemethodused forcollectinginformationfromthesubjects.the mostcommonmodesarefacetofaceinterviews, telephoneinterviewsandselfadministered questionnaires.examples: Alleligiblesubjectshadafacetofaceinterview. Theansweris:Yes(LOWRISK). Somesubjectswereinterviewedoverthe telephoneandsomefilledinpostal questionnaires.theansweris:no(highrisk). 9.Werethenumerator anddenominatorfor theparameterof interestappropriate? Yes(LOWRISK):Thepaper presentedappropriate numerator(s)and denominator(s)forthe parameterofinterest(e.g.the prevalenceofspa). No(HIGHRISK):Thepaper didpresentnumerator(s)and denominator(s)forthe parameterofinterestbutoneor moreof Theremaybeerrorsinthecalculationand/or reportingofthenumeratorand/ordenominator. Examples: Therewerenoerrorsinthereportingofthe numerator(s)anddenominator(s)forthe prevalenceofspa.theansweris: Yes(LOWRISK). Inreportingtheoverallprevalenceoflowback pain(inbothmenandwomen),theauthors accidentallyusedthepopulationofwomenasthe denominatorratherthanthecombined population.theansweris:no(highrisk). Reference 1. HoyD,BrooksP,WoolfA,etal.Assessingriskofbiasinprevalencestudies:modificationofanexisting toolandevidenceofinterrateragreement.jclinepidemiol2012;65:9349.

51 Theglobalprevalenceofspondyloarthritis 51 SUPPLEMENTARYFILE3 Selection of studies for the systematic review on the prevalence of spondyloarthritis

52 52 Chapter2 SUPPLEMENTARYFILE4 Characteristicsofincludedstudies

53 Theglobalprevalenceofspondyloarthritis 53 Characteristicsofincludedstudies Prevalence(%) Diagnosticmethod Total USpA Country Targetpopulation Samplingframe/sampling method Author, publication year population SpA AS PsA ReA IBD SpA 4, Twostep:interview medicalexamination Lithuania Vilnius&Kaunus,18+ Randomsampleofphone numbers 488, Diagnosisfrommedical record Identificationfromgeneral hospitalandprivate rheumatologistmedicalrecords Greece NorthwestGreece, 16+ Adomaviciute, 2008[1] Alamanos, 2004[2] 488, Diagnosisfrommedical record Identificationfromgeneral hospitalandprivate rheumatologistmedicalrecords Greece NorthwestGreece, 16+ Alamanos, 2003[3] 7, Kuwait Kuwait,15+ Randomsampleofhouseholds Twostep:interview medicalexamination Alawadhi, 2004[4] 6, Randomsampleofhouseholds Interview+medical examination Iraq 10/16countiesof Iraq,16+ Alrawi, 1978[5] 3, Mexico Yucatánregion,18+ Randomsampleofhouseholds Twostep:interview medicalexamination AlvarezNegemgyei, 2011[6] 1, Twostep:questionnaire medicalexamination Greece CentralGreece,18+ Randomsampleofpoll catalogues Anagnostopoulos, 2010[7] ICDcodes 3,200, Identificationfromhealthcare deliverysystem USA NorthernCalifornia, 18+ Asgari, 2013[8] Datafrommedicalrecords 2, Identificationofpatientsreferred torheumatologistfrommedical recordsandinterviews Canada NuuChahNulth NativeIndians,agens Atkins, 1988[9] Datafrommedicalrecords 217, Identificationfromhospital diagnosticdatabase Norway NorthernNorway, 16+ Bakland, 2005[10] Notspecified 10, Croatia Osijekregion,agens Randomsampleofinhabitants, notfurtherspecified BarisicDrusko, 1994[11] 5, Twostep:medicalrecords examination Identificationfromhospital medicalrecords Identificationfromhospital medicalrecords Greenland Egedesmindeand Jakobshavn,15+ Bardin, 1985[12] 12, Twostep:medicalrecords examination Greenland Inuitsfromfour Greenlandcities Bardin, 1987[13]

54 54 Chapter2 Prevalence(%) Diagnosticmethod Total USpA Country Targetpopulation Samplingframe/sampling method Author, publication year population SpA AS PsA ReA IBD SpA (Siberia); 3,786(Alaska) Interview+examination (Siberia);twostep:medical recordsinterview+ examination(alaska) Communitysurvey(Siberia), identificationfrompatient registriesandpatientcare database(alaska) SiberianEskimo, chuckiresidents, inupiaseskimo,yupik Eskimo,20+ Russia& Alaska Benevolenskaya, 1996[14] Reviewofmedicalrecords 2, Boyer,1988[15] Alaska InupiatEskimos,20+ Identificationfrompatientcare informationsystem Reviewofmedicalrecords 7, Boyer,1990[16] Alaska YupikEskimos,20+ Identificationfromcomputerized patientrecordsystem 3, < Twostep:identificationfrom medicalrecordsinterview +examination Identificationfromrheumatic diseaseregistriesandpatientcare informationdatabase Boyer,1994[17] Alaska InupiatandYupik Eskimos,20+ Medicalrecords 5, Identificationfrompatientcare informationsystem Boyer,1991[18] Alaska Tlingit,Haida, TsimshianIndians, Twostep:Questionnaire examination Braun,1998[19] Germany Berlin HLAB27positiveblooddonors from5bloodbanks,matched withhlab27negativeblood donors Interview+examination 1, FamilymembersofB27positive twinsfromdatabase(215)and 900randomlyselectedmales,not furtherspecified Brown,1997[20] Gambia Fulaethnicgroup, , EveryhouseholdinHavsa Threestep:interview examinationhealthcare centreuniversityhospital Cakir,2012[21] Turkey Havsaregion (westernturkey), 20+ 2, Twostep:questionnaire+ medicalevaluation Cardiel,2002[22] Mexico MexicoCity,18+ Stratifiedsampleofsubjectsfrom census ,000 (1950); 52,000(1973) Medicalrecordslinkagesystem Identificationofpatientswith diagnosis USA Rochester, Minnesota,15+ Carter,1979[23]

55 Theglobalprevalenceofspondyloarthritis 55 Prevalence(%) Author, publication year Country Targetpopulation Samplingframe/sampling method Diagnosticmethod Total Chaaya,2012[24] USpA Chaiamnuay, 1998 [25] Lebanon Lebanon,15+ Nationalrepresentativestratified sampleofhouseholds population SpA AS PsA ReA IBD SpA 3, Twostep:interview examination Thailand RuralThailand,15+ Randomselectionof3villages,all inhabitants,notfurtherspecified Chou,1994[26] Taiwan HenSan(rural),Sien Dien(suburban),Cu Tien(urban),20+ 2, Twostep:interview examination Constantino, 2013 [27] France Cohortamong employeesofthe Frenchnational electricityandgas company 8, Twostep:questionnaire medicalevaluation Dai,2003[28] China 4communitiesin Shanghai,15+ Randomselectionofsubjects from3districtunits 6, SubjectsofcohortwithDNA sampleavailable Twostep:questionnaire phoneinterview+ evaluation Allpersonsintheselected communities Dans,1997[29] Philippines Urbancommunityin Manilla,15+ 6, Twostep:questionnaireand physicalexamination evaluationbyrheumatologist Randomselectionof8barangays every5 th houseineachofthe8 barangays Davatchi,2008[30] Iran Tehran,15+ Randomselectionof50addresses (clusterhead)from22districts adjacenthousestoclusterhead 3, Twostep:interview+ examinationevaluationby rheumatologist Davatchi,2009[31] Iran 5villagesin TuyserkanCounty (NorthwestIran),15+ De Angelis, 2007 [32] ,291 Twostep:interview examination 1, Twostep:interview examination Italy Marcheregion (centralitaly),18+ Randomselectionofhouseholds ineachvillage 2, Twostep:questionnaire rheumatologicevaluation Eaton,2010[33] Denmark Denmark (nationwide) Randomsampleofindividuals, selectedfromlistof16general practices ICDcodes 5,506, IdentificationfromDanishCivil RegistrationsystemandNational hospitalregister

56 56 Chapter2 Prevalence(%) Diagnosticmethod Total USpA Country Targetpopulation Samplingframe/sampling method Author, publication year population SpA AS PsA ReA IBD SpA 1, Clustersamplingofhouseholds Twostep:questionnaire examination Farooqi,1998[34] Pakistan Sagri(rural), MohallahHukumDad (innercitypoor), Islamabad(affluent urban) 220, Diagnosisfrommedical records Identificationbyscreeningof hospitalrecordsandprivate rheumatologyservices Geirsson,2010[35] Iceland Iceland(nationwide), , Selfreportedphysicianmade diagnosis Gelfand,2005[36] USA ContinentalUSA,18+ Randomsampleofphone numbers Questionnaire+examination 6, Sevensamplesofgeneral population,notspecified Gömör,1977[37] Hungary Hungary,not specified,15+ 14, Twostep:questionnaire examination Twostep:questionnairetoall inhabitantsrandomsampleof subjectswhoreportedbackpain Gran,1985[38] Norway Tromso,northern Norway, ,973 Twostep:COPCORD questionnaire examination HomesinthecommunityofLas Cocuizas,notfurtherspecified Venezuela Urbancommunityin MonagasState,18+ Granados,2014 [39] Skanehealthcareregister ICDcode 849, Haglund,2011[40] Sweden Southernmost county, Diagnosisbyrheumatologist 154, Referredpatientsto rheumatologists Hanova,2010[41] Czech CityofCeske Budejovice&district ofcheb,16+ 5, Twostep:questionnaire examination Allinhabitantsofthethree clusters Haq,2005[42] Bangladesh Bhargaon(rural), Mohammadpur (UrbanSlumand UrbanAffluent),15+ 50, Twostep:questionnaire medicalrecordsreview ParticipantsoftheNord TrondelagHealthStudy3 Hoff,2013[43] Norway NordTrondelag county(central Norway),20+

57 Theglobalprevalenceofspondyloarthritis 57 Prevalence(%) Diagnosticmethod Total USpA Country Targetpopulation Samplingframe/sampling method Author, publication year population SpA AS PsA ReA IBD SpA Medicalrecordsreview 101,100, * Medicalrecordsfromallhospitals andclinicsinjapan Hukuda,2001[44] Japan Japan(nationwide), , Allhouseholdsinarea Twostep:interview examination Joshi,2009[45] India NarayanPethareain Punemetropolis,15+ Finland Finland,30+ representativepopulationsample interview+examination 1,000, Kaipiainen Seppänen, 1997 [46] 4, Twostep:interview evaluation Atrandomfromprimaryhealth centerhouseholdsregistration records Karkucak,2010[47] Turkey EasternBlackSea Region,5cities,20+ Medicalrecords 347, MedicalrecordsofGeneral MilitaryHospitalofAthens Kassimos,2014[48] Greece Militaryserviceof younggreekmales, 1830 Medicalrecords 82, Identificationfromhospital, healthcareclinicsandfamily practitioners Koko,2014[49] Albania RegionofGjirokaster, , Randomselectionofsubjects Twostep:questionnaire clinicalexamination Li,2011[50] China Randomselectionof 10/42villageswithin thelongquanand Yongdingareas,16+ 10, Twostep:interview evaluation Liao,2009[51] China DalangTown,16+ Allofficiallyregisteredresidents ofdalang ICDcodes 1,055, Sweden Skaneregion,0+ IdentificationfromSkane HealthcareRegister Löfvendahl, 2014 [52] Interviewandexamination 134, Love,2007[53] Iceland Reykjavik,18+ IdentificationofpossiblePsA patientsfromdatabaseof patientswithverifiedpsoriasis, andfromhospitalelectronic registry

58 58 Chapter2 Prevalence(%) Diagnosticmethod Total USpA Country Targetpopulation Samplingframe/sampling method Author, publication year population SpA AS PsA ReA IBD SpA Medicalrecords 100, Identificationfrommedical records Lutalo,1985[54] Zimbabwe Gwerycityand surroundingarea ICDcodes 321, Identificationfromdiagnostic codesatthe4rheumatology centers Madland,2005[55] Norway CountyofHordaland (WesternNorway), 20+ 1, Twostep:interview examination Mahajan,2003[56] India Jammu,15+ Randomselectionofhouseholds withincertainruralandurban localitiesofjammu 2, Twostep:questionnaire examination AllresidentsofTrungLiet Commune Vietnam TrungLietCommune (urban),15+ Minh Hoa, 2003 [57] 5, Randomselectionof100clusters Twostep:COPCORD questionnaire examination Moghimi,2013[58] Iran Sanandaj(Kurdish), 15+ ICDcodes 224, Identificationfromhospital patientregistries Nossent,2009[59] Norway Twonorthernmost countiesofnorway Interview+examination 2, Identificationofpossiblepatients frommedicalrecords Oen,1986[60] Canada Inuitpeopleinthe KeewatinDistrictof thenorthwest Territories,1565 4,785, ReadcodeforPsAin database UK,1890 IdentificationfromTheHealth ImprovementNetwork(THIN)GP database Ogdie,2013[61] United Kingdom 2, Twostep:interview examination Randomselectionof26/845 clusters Turkey Urbanpopulationin Izmir,20+ Onen,2008[62] 4, Twostep:questionnaire examination Randomselectionof2/6 subcentersintheunit,entire population Paul,2013[63] India RuralunitofCalicut DistrictinNorth Kerala,15+

59 Theglobalprevalenceofspondyloarthritis 59 Prevalence(%) Diagnosticmethod Total USpA Country Targetpopulation Samplingframe/sampling method Author, publication year population SpA AS PsA ReA IBD SpA 74,618? 6 Twostep:questionnaire interview+examination, Danishnationalpatient registry TwonationwideDanishtwin cohortsandonenationwide Norwegiantwincohort TwinsfromDenmark andnorway,20+ Pedersen,2008[64] Denmarkand Norway 4, Allhouseholds Twostep:questionnaire examination Mexico Urbanpopulation (Cuajimalpa,Mexico City),18+ PeláezBallestas, 2013[65] 19, Twostep:interview examination Mexico 5regions,18+ Stratifiedrandomsampleof households PelaezBallestas, 2011[66] Interview&examination 5, Reveille,2012[67] USA USA,2069 Complexmultistageprobability designtoobtainnationally representativesample 3, Cuba Lawton(Havana),15+ Randomsample Twostep:interview+ examinationconfirmation byrheumatologist ReyesLlerena, 2009[68] 4, Twostep:questionnaire examination Stratifiedrandomsampleof entirestate Mexico StateofNuevoLeon, 18+ RodriguezAmado, 2011[69] 2, Twostep:COPCORD interviewexamination Randomselectionofonehousein eachcluster(20),100right neighboringconsecutivehouses ineachcluster Iran CityofZahadan, southeastiran,15+ Sandoughi, 2013 [70] 2, Twostep:interview+ confirmationby rheumatologist(interviewor examination) Randomsampleofphone numbers Saraux,1999[71] France RegionofBrittany (westfrance),18+ 9, Twostep:interview confirmationby rheumatologist(interviewor examination) Randomsampleofphone numbers Saraux,2005[72] France 20countiesacross France

60 60 Chapter2 Author, publication year Country Targetpopulation Samplingframe/sampling method Prevalence(%) Sheeb,2000[73] USA Olmstedcounty, Minnesota Solomon,1975[74] SouthAfrica Northwestern Transvaal(rural), Johannesburg (urban),15+ Diagnosticmethod Total USpA population SpA AS PsA ReA IBD SpA Soriano,2011[75] Argentina BuenosAires, membersofhospital, 18+ Identificationfrommedical recordsfromrochester EpidemiologyProject Medicalrecords 65, Randomselectionofblocksof houses,completecensusof peoplelivingthere Identificationfrommedical recordsinuniversityhospital Examination 1, Steven,1992[76] Scotland Highlandregion,all Patientsregisteredwith29 generalpractitioners Medicalrecord 138, DiagnosismadebyGP 35, Strand,2013[77] USA Nationwide Randomselectionof rheumatologypractices,random selectionofmedicalrecords Medicalrecords 109,377, Trontzas,2005[78] Greece Nationwide,19+ Allresidentsinsevenareasin mainlandgreeceandrandom selectionofsubjectsinonerural andonesuburbanarea 8, Tsujimoto, 1978 [79] Twostep:interview examination Interview+examination 7, Van der Linden, 1984[80] Japan SuburbsofOsakaand Kamitonadaarea Inhabitantsofdistricts,not furtherspecified 2, Veerapen, 2007 [81] Netherlands Zoetermeer,44+ 2districtsofthetown,notfurther specified Twostep:xrayclinical examination 2, Malaysia Semiruralsuburbon theoutskirtsof Bantingtown,15+ Wigley,1994[82] China Beijing(north)and Shantou(south) areas,20+ Housetohouseregistryof individuals Twostep:questionnaire examination 9, Subjectswereselectedfromthe villageregisters Twostep:questionnaire examination

61 Theglobalprevalenceofspondyloarthritis 61 Prevalence(%) Diagnosticmethod Total USpA Country Targetpopulation Samplingframe/sampling method Author, publication year population SpA AS PsA ReA IBD SpA 124, Identificationfromdatabase andconfirmationinmedical records DatafromRochester epidemiologyproject Wilson,2009[83] USA Olmstedcounty, Minnesota,18+ 2, Twostep:COPCORD questionnaire examination Zeng,2004[84] China ChenghaiCity,16+ Allresidentsinareasofthe neighborhoodcommittees1,2,3 and19oftheeastadministration SpA=spondyloarthritis,AS=ankylosingspondylitis,PsA=psoriaticarthritis,ReA=reactivearthritis,USpA=undifferentiatedspondyloarthritis,ICD=internationalclassificationofdisease, HLAB27=humanleucocyteantigenB27,GP=generalpractitioner

62 62 Chapter2 REFERENCES Adomaviciute D, Pileckyte M, Baranauskaite A, et al. Prevalence survey of rheumatoid arthritis and spondyloarthropathyinlithuania.scandjrheumatol2008;37:1139. AlamanosY,PapadopoulosNG,VoulgariPV,etal.EpidemiologyofankylosingspondylitisinNorthwest Greece, Rheumatology(Oxford)2004;43:6158. Alamanos Y, Papadopoulos NG, Voulgari PV, et al. Epidemiology of psoriatic arthritis in northwest Greece, JRheumatol2003;30: AlAwadhiAM,OlusiSO,MoussaM,etal.Musculoskeletalpain,disabilityandhealthseekingbehavior inadultkuwaitisusingavalidatedarabicversionofthewhoilarcopcordcorequestionnaire.clin ExpRheumatol2004;22: AlRawi ZS, AlShakarchi HA, Hasan F, et al. Ankylosing spondylitis and its association with the histocompatibility antigen HLA B27: an epidemiological and clinical study. Rheumatol Rehabil 1978;17:725. AlvarezNemegyei J, PelaezBallestas I, Sanin LH, et al. Prevalence of musculoskeletal pain and rheumatic diseases in the southeastern region of Mexico. A COPCORDbased community survey. J RheumatolSuppl2011;86:215. AnagnostopoulosI,ZinzarasE,AlexiouI,etal.TheprevalenceofrheumaticdiseasesincentralGreece: apopulationsurvey.bmcmusculoskeletdisord2010;11:98. Asgari MM, Wu JJ, Gelfand JM, et al. Validity of diagnostic codes and prevalence of psoriasis and psoriatic arthritis in a managed care population, Pharmacoepidemiol Drug Saf 2013;22: AtkinsC,ReuffelL,RoddyJ,etal.RheumaticdiseaseintheNuuChahNulthnativeIndiansofthePacific Northwest.JRheumatol1988;15: Bakland G, Nossent HC, Gran JT. Incidence and prevalence of ankylosing spondylitis in Northern Norway.ArthritisRheum2005;53:8505. BarisicDrusko V, Dobric I, Pasic A, et al. Frequency of psoriatic arthritis in general population and among the psoriatics in Department of Dermatology. Acta Derm Venereol Suppl (Stockh) 1994;186:1078. BardinT,EnelC,LathropM,etal.Reiter'ssyndromeinGreenland.Aclinicalandepidermiologicalstudy oftwocommunities.scandjrheumatol1985;14: Bardin T, Enel C, Lathrop MG. [FiessingerLeroyReiter syndrome in Greenland. Clinical and epidemiologicalaspects].revrhummalosteoartic1987;54:3743. BenevolenskayaLI,BoyerGS,ErdeszS,etal.Spondylarthropathicdiseasesinindigenouscircumpolar populationsofrussiaandalaska.revrhumengled1996;63: BoyerGS,LanierAP,TemplinDW.Prevalenceratesofspondyloarthropathies,rheumatoidarthritis,and otherrheumaticdisordersinanalaskaninupiateskimopopulation.jrheumatol1988;15: BoyerGS,LanierAP,TemplinDW,etal.SpondyloarthropathyandrheumatoidarthritisinAlaskanYupik Eskimos.JRheumatol1990;17: Boyer GS, Templin DW, CornoniHuntley JC, et al. Prevalence of spondyloarthropathies in Alaskan Eskimos.JRheumatol1994;21: BoyerGS,TemplinDW,LanierAP.RheumaticdiseasesinAlaskanIndiansofthesoutheastcoast:high prevalenceofrheumatoidarthritisandsystemiclupuserythematosus.jrheumatol1991;18: Braun J, Bollow M, Remlinger G, et al. Prevalence of spondylarthropathies in HLAB27 positive and negativeblooddonors.arthritisrheum1998;41:5867. BrownMA,JepsonA,YoungA,etal.AnkylosingspondylitisinWestAfricansevidenceforanonHLA B27protectiveeffect.AnnRheumDis1997;56:6870. CakirN,PamukON,DervisE,etal.Theprevalences ofsomerheumatic diseasesinwesternturkey: Havsastudy.RheumatolInt2012;32: Cardiel MH, RojasSerrano J. Community based study to estimate prevalence, burden of illness and help seeking behavior in rheumatic diseases in Mexico City. A COPCORD study. Clin Exp Rheumatol 2002;20: Carter ET, McKenna CH, Brian DD, et al. Epidemiology of Ankylosing spondylitis in Rochester, Minnesota, ArthritisRheum1979;22:36570.

63 Theglobalprevalenceofspondyloarthritis ChaayaM,SlimZN,HabibRR,etal.HighburdenofrheumaticdiseasesinLebanon:aCOPCORDstudy. IntJRheumDis2012;15: ChaiamnuayP,DarmawanJ,MuirdenKD,etal.EpidemiologyofrheumaticdiseaseinruralThailand:a WHOILARCOPCORDstudy.CommunityOrientedProgrammefortheControlofRheumaticDisease.J Rheumatol1998;25: ChouCT,PeiL,ChangDM,etal.PrevalenceofrheumaticdiseasesinTaiwan:apopulationstudyof urban,suburban,ruraldifferences.jrheumatol1994;21:3026. CostantinoF,TalpinA,SaidNahalR,etal.PrevalenceofspondyloarthritisinreferencetoHLAB27in thefrenchpopulation:resultsofthegazelcohort.annrheumdis2013. DaiSM,HanXH,ZhaoDB,etal.Prevalenceofrheumaticsymptoms,rheumatoidarthritis,ankylosing spondylitis,andgoutinshanghai,china:acopcordstudy.jrheumatol2003;30: DansLF,TankehTorresS,AmanteCM,etal.TheprevalenceofrheumaticdiseasesinaFilipinourban population: a WHOILAR COPCORD Study. World Health Organization. International League of Associations for Rheumatology. Community Oriented Programme for the Control of the Rheumatic Diseases.JRheumatol1997;24: DavatchiF,JamshidiAR,BanihashemiAT,etal.WHOILARCOPCORDStudy(Stage1,UrbanStudy)in Iran.JRheumatol2008;35:1384. DavatchiF,TehraniBanihashemiA,GholamiJ,etal.Theprevalenceofmusculoskeletalcomplaintsina rural area in Iran: a WHOILAR COPCORD study (stage 1, rural study) in Iran. Clin Rheumatol 2009;28: DeAngelisR,SalaffiF,GrassiW.PrevalenceofspondyloarthropathiesinanItalianpopulationsample:a regionalcommunitybasedstudy.scandjrheumatol2007;36:1421. EatonWW,PedersenMG,AtladottirHO,etal.Theprevalenceof30ICD10autoimmunediseasesin Denmark.ImmunologicResearch2010;47: Farooqi A, Gibson T. Prevalence of the major rheumatic disorders in the adult population of north Pakistan.BrJRheumatol1998;37:4915. GeirssonAJ,EyjolfsdottirH,BjornsdottirG,etal.Prevalenceandclinicalcharacteristicsofankylosing spondylitisinicelandanationwidestudy.clinexprheumatol2010;28: GelfandJM,GladmanDD,MeasePJ,etal.Epidemiologyofpsoriaticarthritisinthepopulationofthe UnitedStates.JAmAcadDermatol2005;53:573. Gomor B, Gyodi E, Bakos L. Distribution of HLA B27 and ankylosing spondylitis in the Hungarian population.jrheumatolsuppl1977;3:335. Gran JT, Husby G, Hordvik M. Prevalence of ankylosing spondylitis in males and females in a young middleagedpopulationoftromso,northernnorway.annrheumdis1985;44: GranadosY,CedenoL,RosilloC,etal.Prevalenceofmusculoskeletaldisordersandrheumaticdiseases inanurbancommunityinmonagasstate,venezuela:acopcordstudy.clinrheumatol2014. Haglund E, Bremander AB, Petersson IF, et al. Prevalence of spondyloarthritis and its subtypes in southernsweden.annrheumdis2011;70:9438. Hanova P, Pavelka K, Holcatova I, et al. Incidence and prevalence of psoriatic arthritis, ankylosing spondylitis,andreactivearthritisinthefirstdescriptivepopulationbasedstudyintheczechrepublic. ScandJRheumatol2010;39:3107. HaqSA,DarmawanJ,IslamMN,etal.Prevalenceofrheumaticdiseasesandassociatedoutcomesin ruralandurbancommunitiesinbangladesh:acopcordstudy.jrheumatol2005;32: HoffM,GulatiAM,RomundstadPR,etal.Prevalenceandincidenceratesofpsoriaticarthritisincentral Norway:datafromtheNordTrondelagHealthStudy(HUNT).AnnRheumDis2013. HukudaS,MinamiM,SaitoT,etal.SpondyloarthropathiesinJapan:nationwidequestionnairesurvey performedbythejapanankylosingspondylitissociety.jrheumatol2001;28:5549. Joshi VL, Chopra A. Is there an urbanrural divide? Population surveys of rheumatic musculoskeletal disordersinthepuneregionofindiausingthecopcordbhigwanmodel.jrheumatol2009;36: KaipiainenSeppanen O, Aho K, Heliovaara M. Incidence and prevalence of ankylosing spondylitis in Finland.JRheumatol1997;24:4969. KarkucakM,CakirbayH,CapkinE,etal.TheprevalenceofankylosingspondylitisinTheEasternBlack Sea region of Turkey. [Turkish]Turkiye'de dogu Karadeniz Bolgesinde ankilozan spondilit (as) prevalansi{dotless}.europeanjournalofgeneralmedicine2011;8:405.

64 64 Chapter Kassimos DG, Vassilakos J, Magiorkinis G, et al. Prevalence and clinical manifestations of ankylosing spondylitisinyounggreekmales.clinrheumatol2014. KokoV,NdrepepaA,SkenderajS,etal.Anepidemiologicalstudyonankylosingspondylitisinsouthern Albania.MaterSociomed2014;26:269. LiR,SunJ,RenLM,etal.EpidemiologyofeightcommonrheumaticdiseasesinChina:alargescale crosssectionalsurveyinbeijing.rheumatology(oxford)2012;51:7219. LiaoZT,PanYF,HuangJL,etal.Anepidemiologicalsurveyoflowbackpainandaxialspondyloarthritis inachinesehanpopulation.scandjrheumatol2009;38:4559. LofvendahlS,TheanderE,SvenssonA,etal.Validityofdiagnosticcodesandprevalenceofphysician diagnosedpsoriasisandpsoriaticarthritisinsouthernswedenapopulationbasedregisterstudy.plos One2014;9:e Love TJ, Gudbjornsson B, Gudjonsson JE, et al. Psoriatic arthritis in Reykjavik, Iceland: prevalence, demographics,anddiseasecourse.jrheumatol2007;34: Lutalo SK. Chronic inflammatory rheumatic diseases in black Zimbabweans. Ann Rheum Dis 1985;44:1215. MadlandTM,ApalsetEM,JohannessenAE,etal.Prevalence,diseasemanifestations,andtreatmentof psoriaticarthritisinwesternnorway.jrheumatol2005;32: MahajanA,JasrotiaDS.PrevalenceofmajorrheumaticdisordersinJammu.JKScience2003;5:636. MinhHoaTT,DarmawanJ,ChenSL,etal.PrevalenceoftherheumaticdiseasesinurbanVietnam:a WHOILARCOPCORDstudy.JRheumatol2003;30: MoghimiN,DavatchiF,RahimiE,etal.WHOILARCOPCORDstudy(stage1,urbanstudy)inSanandaj, Iran.ClinRheumatol2013. Nossent JC, Gran JT. Epidemiological and clinical characteristics of psoriatic arthritis in northern Norway.ScandJRheumatol2009;38:2515. Oen K, Postl B, Chalmers IM, et al. Rheumatic diseases in an Inuit population. Arthritis Rheum 1986;29:6574. Ogdie A, Langan S, Love T, et al. Prevalence and treatment patterns of psoriatic arthritis in the UK. Rheumatology(Oxford)2013;52: OnenF,AkarS,BirlikM,etal.Prevalenceofankylosingspondylitisandrelatedspondyloarthritidesin anurbanareaofizmir,turkey.jrheumatol2008;35:3059. PaulBJ,RahimAA,BinaT,etal.Prevalenceandfactorsrelatedtorheumaticmusculoskeletaldisorders inruralsouthindia:whoilarcopcordbjdindiacalicutstudy.intjrheumdis2013;16:3927. Pedersen OB, Svendsen AJ, Ejstrup L, et al. The occurrence of psoriatic arthritis in Denmark. Ann RheumDis2008;67: PelaezBallestas I, NavarroZarza JE, Julian B, et al. A communitybased study on the prevalence of spondyloarthritisandinflammatorybackpaininmexicans.jclinrheumatol2013;19:5761. PelaezBallestas I, Sanin LH, MorenoMontoya J, et al. Epidemiology of the rheumatic diseases in Mexico.Astudyof5regionsbasedontheCOPCORDmethodology.JRheumatolSuppl2011;86:38. ReveilleJD,WitterJP,WeismanMH.PrevalenceofaxialspondylarthritisintheUnitedStates:estimates fromacrosssectionalsurvey.arthritiscareres(hoboken)2012;64: ReyesLlerena GA, GuibertToledanoM, PenedoCoello A, et al. Communitybased study to estimate prevalenceandburdenofillnessofrheumaticdiseasesincuba:acopcordstudy.jclinrheumatol 2009;15:515. RodriguezAmado J, PelaezBallestas I, Sanin LH, et al. Epidemiology of rheumatic diseases. A communitybased study in urban and rural populations in the state of Nuevo Leon, Mexico. J RheumatolSuppl2011;86:914. Sandoughi M, Zakeri Z, Tehrani Banihashemi A, et al. Prevalence of musculoskeletal disorders in southeasterniran:awhoilarcopcordstudy(stage1,urbanstudy).intjrheumdis2013;16: Saraux A, Guedes C, Allain J, et al. Prevalence of rheumatoid arthritis and spondyloarthropathy in Brittany,France.SocietedeRhumatologiedel'Ouest.JRheumatol1999;26: SarauxA,GuilleminF,GuggenbuhlP,etal.PrevalenceofspondyloarthropathiesinFrance:2001.Ann RheumDis2005;64: ShbeebM,UramotoKM,GibsonLE,etal.TheepidemiologyofpsoriaticarthritisinOlmstedCounty, Minnesota,USA, JRheumatol2000;27:

65 Theglobalprevalenceofspondyloarthritis SolomonL,BeightonP,ValkenburgHA,etal.RheumaticdisordersintheSouthAfricanNegro.PartI. Rheumatoidarthritisandankylosingspondylitis.SAfrMedJ1975;49: Soriano ER, Rosa J, Velozo E, et al. Incidence and prevalence of psoriatic arthritis in Buenos Aires, Argentina: a 6year health management organizationbased study. Rheumatology (Oxford) 2011;50: StevenMM.PrevalenceofchronicarthritisinfourgeographicalareasoftheScottishHighlands.Ann RheumDis1992;51: Strand V, Rao SA, Shillington AC, et al. Prevalence of axial spondyloarthritis in United States rheumatology practices: Assessment of SpondyloArthritis International Society criteria versus rheumatologyexpertclinicaldiagnosis.arthritiscareres(hoboken)2013;65: Trontzas P, Andrianakos A, Miyakis S, et al. Seronegative spondyloarthropathies in Greece: a populationbased study of prevalence, clinical pattern, and management. The ESORDIG study. Clin Rheumatol2005;24:5839. TsujimotoM.Epidemiologicalresearchontheprevalenceofankylosingspondylitis.MedJOsakaUniv 1978;28: VanDerLindenSM,ValkenburgHA,DeJonghBM,etal.Theriskofdevelopingankylosingspondylitisin HLAB27 positive individuals. A comparison of relatives of spondylitis patients with the general population.arthritisandrheumatism1984;27:2419. Veerapen K, Wigley RD, Valkenburg H. Musculoskeletal pain in Malaysia: a COPCORD survey. J Rheumatol2007;34: Wigley RD, Zhang NZ, Zeng QY, et al. Rheumatic diseases in China: ILARChina study comparing the prevalence of rheumatic symptoms in northern and southern rural populations. J Rheumatol 1994;21: Wilson FC, Icen M, Crowson CS, et al. Time trends in epidemiology and characteristics of psoriatic arthritisover3decades:apopulationbasedstudy.jrheumatol2009;36:3617. ZengQY,ChenR,XiaoZY,etal.LowprevalenceofkneeandbackpaininsoutheastChina;theShantou COPCORDstudy.JRheumatol2004;31:

66 66 Chapter2 SUPPLEMENTARYFILE5 Resultsofriskofbiasassessment Author,year ROB1* ROB2* ROB3* ROB4* ROB5* ROB6* ROB7* ROB8* ROB9* Adomaviciute,2008 High High Low High Low Low Low Low Low Alamanos,2004 High High Low Low High Low Low Low Low Alamanos,2003 High High Low Low High Low Low Low Low Alawadhi,2004 High Low Low Low Low High Low Low Low Alrawi,1978 Low Low Low Low Low High High Low Low AlvarezNemegyei,2011 High High Low Low Low High Low Low Low Anagnostopoulos,2010 High Low Low High Low Low Low Low Low Asgari,2013 High High Low Low High Low Low Low Low Atkins,1988 High High Low Low Low Low Low High Low Bakland,2005 High High Low Low High Low Low Low Low BarisicDrusko,1994 High High Low High Low High High Low Low Bardin,1985 High High Low Low Low Low Low Low Low Bardin,1987 High High Low Low Low Low Low Low Low Benevolenskaya(Russia),1996 High Low Low Low Low Low Low Low Low Benevolenskaya(Alaska),1996 High High Low Low High Low Low Low Low Boyer,1988 High High Low Low High Low Low Low Low Boyer,1990 High High Low Low Low Low Low Low Low Boyer,1994 High High Low Low Low Low Low High Low Boyer,1991 High High Low Low High Low Low Low Low Braun,1998 High High High Low Low Low Low Low High Brown,1997 High High High High Low High Low High Low Cakir,2012 High Low Low Low Low Low Low Low Low Carter,1979 High High Low Low High Low Low Low High Cardiel,2002 High High Low High Low High Low Low Low Chaaya,2012 Low Low Low Low Low High Low Low Low Chaiamnuay,1998 High High Low Low Low Low Low Low Low Chou,1994 High Low Low Low Low Low Low Low Low Constantino,2013 High High High High Low Low Low Low Low Dai,2003 High High Low Low Low Low Low Low Low Dans,1997 High High Low Low Low High Low Low Low Davatchi,2008 High Low Low High Low Low Low Low Low Davatchi,2009 High High Low Low Low Low Low Low Low DeAngelis,2007 High High Low High Low Low High Low Low Eaton,2010 Low High Low Low High Low Low Low High Farooqi,1998 High High Low Low Low Low Low Low Low Geirsson,2010 Low High Low Low High Low Low High Low Gelfand,2005 High High Low High Low High High Low Low Gomor,1977 High High High High Low Low Low Low Low Gran,1985 High Low Low Low Low Low High Low Low Granados,2014 High High High Low Low Low Low Low Low Haglund,2011 High High Low Low High Low Low Low Low Hanova,2010 High High High High High Low Low Low Low Haq,2005 High Low Low Low Low High Low Low Low Hoff,2013 Low Low Low High Low Low Low Low Low

67 Theglobalprevalenceofspondyloarthritis 67 Author,year ROB1* ROB2* ROB3* ROB4* ROB5* ROB6* ROB7* ROB8* ROB9* Hukuda,2001 High High Low High High Low Low Low High Joshi,2009 High High High High Low High Low Low Low KaipiainenSeppanen,1997 High High Low Low Low Low Low Low Low Karkucak,2010 High High Low Low Low Low High Low Low Kassimos,2014 High High Low Low High Low Low Low Low Koko,2014 High High Low Low High Low Low Low Low Li,2011 High High High High Low Low Low Low Low Liao,2009 High Low Low Low Low Low High Low Low Löfvendahl,2014 High High Low Low High Low Low Low Low Love,2007 High High Low High High Low Low Low Low Lutalo,1985 High High Low Low High High Low Low High Madland,2005 High High Low Low High Low Low Low Low Mahajan,2003 High High Low High Low Low Low Low Low MinhHoa,2003 High High Low High Low High Low Low Low Moghimi,2013 High High Low High Low High Low Low Low Nossent,2009 High High Low Low High Low Low Low Low Oen,1986 High High Low Low High Low Low Low Low Ogdie,2012 Low High Low Low High Low Low Low Low Onen,2008 High High Low Low Low Low Low Low Low Paul,2013 High High Low High Low High Low Low High Pedersen,2008 High Low Low Low Low Low Low Low Low PelaezBallestas,2013 High High Low High Low Low Low Low Low PelaezBallestas,2011 High High Low High Low Low Low Low High Reveille,2012 Low High High Low Low High High Low Low ReyesLlerena,2009 High High Low High Low High Low Low Low RodriguezAmado,2011 High Low Low Low Low High Low Low Low Sandoughi,2013 High Low Low Low Low Low Low Low High Saraux,1999 High High Low Low Low Low High Low Low Saraux,2005 High High Low High Low Low Low Low Low Sheeb,2000 High High Low Low High Low Low Low High Solomon,1975 High High Low High Low High High Low Low Soriano,2011 High High Low Low High Low Low Low Low Steven,1992 High High Low High Low High High Low High Strand,2013 High High High High High Low Low Low High Trontzas,2005 Low Low Low Low Low Low Low Low Low Tsujimoto,1978 High Low Low Low Low High High Low Low VanderLinden,1984 High High Low High Low Low Low Low Low Veerapen,2007 High Low Low High Low High Low Low Low Wigley,1994 High High Low Low Low High Low Low Low Wilson,2009 High High Low Low High Low Low Low Low Zeng,2004 High High Low Low Low Low Low Low Low *ROB=Riskofbias. ROB1=Wasthetargetpopulationrepresentativeforthegeneralpopulation?ROB2=Wasthesampling frameacloserepresentationofthetargetpopulation?rob3=wassomeformofrandomselectionusedto select the sample? ROB4 =Was thelikelihood of nonresponse bias minimal (participation >75% or non responderanalysis)rob5=weredatacollecteddirectlyfromthesubjects?rob6=wasanacceptablecase definitionused(i.e.validatedclassificationcriteriaoradiagnosisbyahealthprofessional)?rob7=wasthe studyinstrumentreliableandvalid?rob8=wasthesamemodeofdatacollectionusedforallsubjects? ROB9=Werethenumeratoranddenominatorfortheparameterofinterestappropriate?

68 68 Chapter2 SUPPLEMENTARYFILE6 Univariableandmultivariablemetaregressionanalysisontheprevalence ofankylosingspondylitis Variable Univariableanalysis Multivariableanalysis B(95%CI) Pvalue R 2 B(95%CI) Pvalue Meanage 0.4*10 3 (0.2*10 3 ;0.1*10 2 ) NE %female 0.5*10 3 (0.7*10 3 ;0.3*10 3 ) < *10 3 (0.6*10 3 ;0.3*10 3 ) <0.01 NorthAmerica 0.01(0.04;0.02) (0.01;0.05) 0.18 SouthAmerica 0.03(0.07;0.02) (0.07;0.00) 0.04 SubSaharanAfrica 0.07(0.13;0.01) (0.12;0.03) <0.01 MiddleEast 0.03(0.07;0.01) (0.10;0.04) <0.01 EastAsia 0.02(0.04;0.01) (0.08;0.03) <0.01 SouthEastAsia 0.05(0.11;0.02) (0.13;0.03) <0.01 SouthAsia 0.05(0.12;0.02) (0.11;0.01) 0.02 Region Reference=Europe Indigenous 0.02(0.01;0.04) (0.00;0.05) 0.02 Setting Rural 0.04(0.02;0.07) < (0.07;0.05) 0.01 Reference=urban Combination 0.01(0.03;0.00) (0.03;0.01) 0.18 Startdatacollection 0.00(0.00;0.00) 0.81 NE Riskofbiasitem1,targetpopulation 0.02(0.02;0.07) 0.31 NE Riskofbiasitem2,samplingframe 0.02(0.03;0.00) 0.10 NS Riskofbiasitem3,randomselection 0.01(0.04;0.03) 0.74 NE Riskofbiasitem4,nonresponsebias 0.01(0.04;0.01) 0.30 NE Riskofbiasitem5,wasaproxyused? 0.01(0.03;0.01) 0.27 NE Riskofbiasitem6,casedefinition* 0.04(0.06;0.01) <0.01 NA Riskofbiasitem7,studyinstrument 0.03(0.01;0.05) < (0.02;0.06) <0.01 Riskofbiasitem8,datacollectionmode 0.02(0.07;0.03) 0.49 NE Riskofbiasitem9,numerator/denominator 0.01(0.03;0.02) 0.51 NE Twostepdiagnosis 0.02(0.01;0.06) (0.00;0.05) 0.09 Twostepsymptoms 0.01(0.01;0.03) (0.02;0.07) <0.01 Casefinding** Reference=medicalrecords 0.02(0.02;0.07) (0.01;0.07) 0.10 Diagnosishealth professional Caseascertainment (Mod)NewYork,Rome 0.04(0.01;0.02) < NS Reference=clinicaldiagnosis ASAS 0.05(0.04;0.15) 0.28 NS Totalnumberofstudies=53,numberofageand/orsexspecificestimates=179.*Riskofbiasitem6wasnotincludedinthemultivariableanalysisduetocollinearitywith caseascertainment.(mod)newyork=modifiednewyorkcriteria;asas=assessmentofspondyloarthritisinternationalsocietycriteria;icd=internationalclassification ofdisease;ne=notenteredinmultivariablemodel(becausep>0.20);ns=notsignificant;na=notapplicable(becauseofcollinearity).

69 Theglobalprevalenceofspondyloarthritis 69 SUPPLEMENTARYFILE7 Univariableandmultivariablemetaregressionanalysisontheprevalence ofpsoriaticarthritis Variable Univariableanalysis Multivariableanalysis B(95%CI) P R2 B(95%CI) Pvalue Meanage 0.9*10 3 (0.5*10 3 ;0.1*10 2 ) < *10 3 (0.3*10 3 ;0.1*10 2 ) <0.01 %female 0.00(0.2*10 3 ;0.2*10 3 ) NE Region NorthAmerica 0.01(0.04;0.01) (0.04;0.1*10 2 ) 0.03 Reference=Europe SouthAmerica 0.03(0.06;0.01) (0.05;0.6*10 2 ) 0.01 MiddleEast 0.06(0.12;0.01) (0.10;0.2*10 3 ) 0.05 EastAsia 0.03(0.06;0.01) (0.05;0.1*10 2 ) 0.01 SouthEastAsia 0.04(0.10;0.02) (0.09;0.02) 0.28 SouthAsia 0.04(0.09;0.02) (0.08;0.01) 0.17 Indigenous 0.04(0.07;0.02) < (0.06;0.8*10 2 ) 0.01 Setting Rural 0.01(0.05;0.03) NS Reference=urban Combination 0.02(0.00;0.03) 0.07 NS Startdatacollection 0.6*10 3 (0.3*10 3 ;0.1*10 2 ) 0.18 NS Riskofbiasitem1,targetpopulation 0.02(0.04;0.04*10 1 ) NS Riskofbiasitem2,samplingframe 0.01(0.01;0.03) NS Riskofbiasitem3,randomselection 0.4*10 2 (0.03;0.02) NE Riskofbiasitem4,nonresponsebias 0.01(0.3*10 2 ;0.03) NS Riskofbiasitem5,wasaproxyused? 0.01(0.2*10 3 ;0.03) NS Riskofbiasitem6,casedefinition* 0.02(0.04;0.2*10 2 ) NE Riskofbiasitem7,studyinstrument 0.01(0.04;0.01) NE Riskofbiasitem8,datacollectionmode 0.03(0.08;0.02) NE Riskofbiasitem9,numerator/denominator 0.01(0.03;0.05) NE Casefinding Selfreport 0.05(0.01;0.09) (0.02;0.09) <0.01 Reference=medicalrecords Twostepdiagnosis 0.01(0.03;0.01) *10 2 (0.02;0.01) 0.75 Twostepsymptoms 0.02(0.06;0.10) *10 2 (0.01;0.02) 0.52 Diagnosisexpert 0.02(0.07;0.04) (0.01:0.14) 0.13 Caseascertainment ESSG+psoriasis 0.01(0.01;0.01) < NS Reference=clinicaldiagnosis CASPAR 0.6*10 2 (0.01;0.2*10 2 ) <0.01 NS ASAS+psoriasis 0.01(0.04;0.01) 0.13 NS Numberofstudiesn=35,numberofageand/orsexspecificestimatesn=89.*Riskofbiasitem6wasnotincludedinthemultivariableanalysisduetocollinearitywith diagnostic approach and case definition. CASPAR=ClASsification Psoriatic Arthritis criteria; ASAS=Assessment of Spondyloarthritis International Society criteria; ICD= internationalclassificationofdisease;ne=notenteredinmultivariablemodel(becausep>0.20);ns=notsignificant;na=notapplicable(becauseofcollinearity).

70

71 PARTII TheuseofMRIindetectionof earlyaxialspondyloarthritis

72

73 CHAPTER3 HLAB27andgenderindependentlydeterminethe likelihoodofapositivemriofthesacroiliacjointsin patientswithearlyinflammatorybackpain:atwo yearmrifollowupstudy MarloesvanOnna,AnneGretheJurik,DésiréevanderHeijde,AstridvanTubergen, LiesbethHeuftDorenbosch,RobertLandewé

74 74 Chapter3 ABSTRACT Objectives TodescribehowinflammationonMRIofthesacroiliacjointsinpatientswithrecent onset inflammatory back pain (IBP) evolves over time, and to study determinants of activityonmriofthesacroiliacjoint. Methods A 2year followup study with annual MRI of the sacroiliac joints was conducted in patientswithibpoflessthan2years duration.imageswerescoredforbonemarrow edema on short inversion recovery and enhancement after administration of gadoliniumont1. Results Ofthe68patients(38%male;meanage34.9±10.3years)enrolled,44hadanegative baselinemri.ofthese44patients,39patientshadatleast1followupmriofwhom 6patients (15%) developed activity on MRI during followup. Twentyfour patients (35%)hadanabnormalMRIatbaseline.In23ofthese24patientsfollowupMRIwas available.themribecamenegativein7ofthese23patients(30%)duringfollowup. HLAB27 positivity and male gender determined independently the likelihood of a positive MRI at any time point. In an HLAB27 positive patient the likelihood of a positive MRI during followup is 88% if the baseline MRI is positive and 27% if the baseline MRI is negative. In an HLAB27 negative patient with a negative MRI at baselinethelikelihoodofapositivemriduringfollowupislessthan5%. Conclusions ApositiveMRIatbaselinepredictsapositiveMRIduringfollowupinHLAB27positive patients.anegativemriatbaselineinhlab27negativepatientsstronglypredictsa negativemriduringfollowup.

75 ActivelesionsonMRIofthesacroiliacjoints 75 INTRODUCTION The concept of spondyloarthritis (SpA) covers a family of seronegative rheumatic disorders that share distinct clinical and genetic characteristics. General symptoms include inflammatory back pain (IBP), peripheral arthritis, enthesitis, dactylitis and anterioruveitis.ankylosingspondylitis(as)istheprototypeofspawithpredominant axial involvement.[1,2] Other subtypes of SpA are characterised by predominant peripheralinvolvement,suchaspsoriaticarthritis.toclassifypatientswithspa,several criteriasetsareavailable.thesearetheeuropeanspondyloarthropathystudygroup criteria(essg)andamorcriteriaforspaandthemodifiednewyorkcriteriaspecifically foras.[35]recently,anewcriteriasetforclassifyingaxialspahasbeendeveloped, mainly because available classification criteria were considered too insensitive, especially in early and mild disease.[6] Although not intended for this purpose, classification criteria are frequently used in daily practice. As a consequence, the diagnosis of AS is often delayed for 10 years, since according to modified New York criteria sacroiliitis should be demonstrated on a conventional radiograph, which is a relativelylatefeatureofthedisease.[7] Magneticresonanceimaging(MRI)canplayanimportantroleinthenonradiographic stageofaxialspa.ithasbeenshownthatmricandetecttheearlyinflammatorystages of sacroiliitis, months to years before structural damage can be detected on a conventionalradiograph.[8] Although activity (inflammation) detected on MRI is considered an important manifestationinearlyspa,thereisstilluncertaintyabouthowactivityonmrievolves over time. MRI followup studies in patients with a short duration of symptoms are limited.oostveenetal.suggestedthatchronicchangesonradiographsarepreceded by inflammatory lesions on MRI.[8] More recently, Bennett et al. showed that both severe sacroiliitis on MRI at baseline as well as HLAB27 positivity predict the developmentofabnormalitiesonradiographs8yearslater.[9] Itis,however,unclear whetheractivityinthesacroiliac(si)jointsasseenonmriisstableovertime,becomes quiescent after an initialperiod ofactivity, or fluctuatesover time. Moreknowledge about this is important with regard to the reliability of the new ASAS classification criteriainwhichactivityonmriisconsideredoneoftheentrycriteria.furthermore,it isextremelyimportantforclinicalpracticetoknowifonesinglemriissufficientorthat itisnecessarytorepeatthemriincertainpatientsduringthediagnosticprocess.

76 76 Chapter3 TheaimofthepresentpaperistodescribetheevolutionofactivechangesonMRIof the SI joints in a group of patients presenting with inflammatory back pain (IBP) of shortdurationthatwerefollowedfor2yearswithrepeatedmris. PATIENTSANDMETHODS Designandstudypopulation Patients with IBP of less than 2 years duration were included in the Early SpondyloArthritisCohort(ESpAC)study.Thedetailsofthisstudyhavebeendescribed previously.[10] In brief, IBP was defined according to the Calin criteria, and was consideredpresentif4outof5ofthefollowingcriteriawerepresent:onsetofback painbeforetheageof40yearswithapersistenceofatleast3months,insidiousonset, association with morning stiffness and improvement with exercise.[11] Patients that fulfilledonly3outof5ofthecalincriteriabutreportednightpain,werealsoincluded inespac.presenceofothercommonspafeaturesasdefinedintheessgcriteria,such as a positive family history, anterior uveitis or a history of psoriasis or inflammatory boweldisease(ibd),waspreferredbutnotmandatory. Thestudywasdesignedasaprospectivecohortstudythatconsistedofthreeidentical clinicalandradiologicalexaminationroundsperformedatbaselineandafter1and2 years.dataobtainedatbaselinehavebeenpublishedpreviously.[10,1213] The study was approved by the institutional review board and all patients have providedwritteninformedconsent. MRI MRIoftheSIjointswasobtainedwitha1.5TeslaPhilipsGyroScanACSNT(Philips,The Netherlands).Patientswereexaminedinasupinepositionwiththeirkneesbent.The following sequences were used in an oblique coronal orientation: T1weighted spin echo(se),shortinversionrecovery(stir),t2weightedfastsewithfatsaturationand T1weighted SE with fat suppression (FS) after administration of the intravenous contrast agent gadolinium diethylenetriaminepentate (GdDTPA, 0.1 mmol/kg body weight).theslicethicknesswas4mmwithintervalsof0.4mm.thematrixwas512for thet1fspostgddtpasequenceand256fortheremainingsequences.themrisof eachpatientwerescoredasonesetwithrandomtimesequence,byoneexperienced radiologist(agj),whowasblindedtoclinical,laboratoryandradiographicresults.

77 ActivelesionsonMRIofthesacroiliacjoints 77 The MR images were scored using a combination of the Spondyloarthritis research Consortium of Canada (SPARCC) method and a modified version of the Aarhus MRI scoring method.[14,15] The SI joints were divided into four quadrants: upper iliac, loweriliac,uppersacralandlowersacral.incontrasttotheoriginalsparccsystem, therewasnomaximumtothenumberofevaluatedslices,inordertomaximisethe detectionofabnormalmrfindings.thetotalnumberofevaluatedslicesinindividual patients remained stable. Inflammation in the cartilagenous part of the joint was scoredpersliceinadichotomousmanner(presentvs.absent).bothsubchondralbone marrowedemaandenhancementafteradministrationofgddtpawerescored.bone marrow edema was defined as areas of increased signal intensity on STIR images comparedwithnormalbonemarrow.itspresencewasestimatedineachofthefour quadrants.lesionsextendingatleast1cmfromthejointspaceordemonstratinghigh signalintensitycomparabletothatofspinalfluid,weregivenanadditionalscoreona perjointandperslicebasis. Subchondralenhancementwasdefinedasareasofincreasedsignalintensitycompared withnormalbonemarrowonpostgddtpaimages.scoringwascomparabletoscoring edema,includingadditionalscoresfordepthandhighsignalintensitycomparableto enhancedvessels,respectively. AnMRIwasconsideredpositiveforactivesacroiliitisaccordingtotheASAS/Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) MRI working group definition:thepresenceofatleastoneactivelesioninatleasttwosuccessiveslicesor the presenceof more thanonelesion inonlyone slicewhich is highly suggestive of sacroiliitis.[16] Statisticalanalysis The MRI scores were analyzed and described in three different manners. First, we describedthemristatusovertime(positiveornegative).second,wehaveinvestigated thelikelihoodofhavingapositivemriatanytimeduringfollowup,aswellasfactors determining MRIpositivity, making use of generalized estimating equation (GEE) analysis for binomial outcome variables, with MRI status as dependent variable, and HLAB27statusandgenderasindependentexplanatoryvariables.Thecontributionof Creactiveprotein(CRP)inexplainingapositiveMRIwastestedbyaddingCRPinthe modelasacovariate.geeisanappropriatetechniquetostudytimetrendsindatasets with missing values while adjusting for withinpatient correlation. Thirdly, we have investigated the likelihood of finding a positive MRI if the baseline MRI is either negativeorpositive.

78 78 Chapter3 SPSSsoftwareversion15.0wasusedforalldescriptivestatisticsandtheGEEanalysis. AllPvaluesweretwotailedandstatisticalsignificancewassetatthe0.05level. RESULTS Patientcharacteristics Intotal,68patientswithIBPwereincludedinthisstudy.Baselinecharacteristicsare shown in Table 3.1. Details of these patients have been described previously.[10] Ninetyseven percent of all patients fulfilled at least four out of five Calin criteria. Almost half of the patients was HLAB27 positive. Anterior uveitis and IBD were reportedin10(15%)patientseach,and16(24%)patientshadahistoryofpsoriasis. Twentyfive patients had a positive family history for AS. Table 3.1 also provides informationaboutthe44patientsthatcompletedthe2yearfollowup.thesepatients didnotdifferfromthecompletegroupatbaseline. Table3.1 Characteristic Baselinecharacteristicsof68patientsincludedintheESpACstudy. Allpatients (n=68) Patientswithcomplete followupexaminations (n=44) Malegender 26(38) 15(34) Meanage(years)(SD) 34.9(10.3) 36.0(11.7) Mediansymptomduration(months)[IQR] 18[1224] 18[1224] HLAB27positive 31(46) 17(39) Historyofinflammatoryboweldisease 10(15) 7(16) Historyofanterioruveitis 10(15) 8(18) Historyofpsoriasis 16(24) 12(27) Historyofperipheralarthritis 19(28) 12(27) PositivefamilyhistoryofAS 25(37) 17(39) CRP(inmg/l)(mean(median)) 9(7) 9(7) ESR(inmm)(mean(median)) 13(6) 13(6) Patientswithelevatedacutephasereactants(in%) The values are expressed as number (percentage) of patients or as stated otherwise. ESpAC=Early Spondyloarthritis Clinic; AS=ankylosing spondylitis; IQR=interquartile range; SD=standard deviation; CRP= Creactiveprotein;ESR=erythrocytesedimentationrate. MRIfindings MRIoftheSIjointswasperformedinallpatientsatbaseline.Sixtytwopatientshadat least one followup MRI after 1 or 2 years followup and 44 patients completed all

79 ActivelesionsonMRIofthesacroiliacjoints 79 3examinations. Table 3.2 shows the MRI status (positive or negative) at baseline accordingtotheasas/omeractmriworkinggroupdefinitionaswellashowthese findingsdevelopovertime. Intotal,44patients(65%)hadanegativeMRIatbaseline,ofwhich15patients(34%) were HLAB27 positive. Of these44 patients, 39 (89%)completedeither one or two followupmris.insixofthese39patients(15%),themribecamepositiveat1or2 yearsfollowup.inonehlab27positivepatientofthesesixpatientsthemriremained positive at 2 years followup, in one HLAB27 negative patient the MRI became negativeagainat2yearsfollowup,intwohlab27negativepatientsthemribecame positiveat2yearsfollowup,andin2hlab27positivepatientsnomriswereavailable at2yearsfollowup. Table3.2 MRI status over time according to the ASAS/OMERACT MRI working group definition in 68 patients with early inflammatory low back pain who were included in the Early SpondyloArthritisCohort(ESpAC). Numberofpatients MRIavailability HLAB27+ Baseline 1year 2years NA NA NA NA NA NA NA 2 5 NA NA 3 +=positive MRI;=negative MRI, ASAS=Assessment of SpondyloArthritis international Society; ESpAC=Early Spondyloarthritis Cohort; HLAB27=human leukocyte antigen B27; NA=not available; OMERACT=Outcome MeasuresinRheumatoidArthritisClinicalTrials. Twentyfourpatients(35%)hadapositiveMRIatbaseline,ofwhich16(66%)patients werehlab27positive.intotal23ofthese24patientscompletedeitheroneortwo followupmris.themribecamenegativeinsevenofthese23patients(30%)during followupatoneoratbothassessments.fiveofthesesevenpatientswerehlab27 negative and in three of them the MRI was only weakly positive at baseline. In the

80 80 Chapter3 remainingtwopatientsastronglypositivemriwasdetectedatbaseline,butthemri became negative during followup. In two HLAB27 positive patients with a negative MRIat1yearfollowup,theMRIwaspositiveagainat2yearsfollowup. FactorsdeterminingapositiveMRI In the GEE analysis, both male gender (odds ratio (OR) 3.0 (95% confidence interval (CI); ), p=0.035) and HLAB27 positivity (OR 5.1 (95% CI ), p<0.001) independentlydeterminedthelikelihoodofapositivemriatanytimepoint.crpdid notcontributetoexplainingvariationinthemodel.theeffectsofthesedeterminants are visualised in an absolute manner in Figure 3.1, showing that the likelihood of a positivemriinhlab27negativefemaleswithibpisonlyabout10%ascomparedto HLAB27positivemaleswithIBPinwhomthislikelihoodiscloseto70%. Figure3.1 Likelihood of a positive MRI at any time point in patients with shortstanding inflammatory backpainassessedatbaseline,1yearand2yearsoffollowupinfunctionofhlab27status andgender.hlab27,humanleukocyteantigenb27. LikelihoodofapositiveMRIduringfollowup ThelikelihoodoffindingapositiveMRIifthebaselineMRIiseithernegativeorpositive was also investigated. Both HLAB27 status (OR 8.1 (95% CI ), p<0.001) and MRIstatusatbaseline(OR22.0(95%CI ),p<0.001)appearedtobestrongly andindependentlycontributorytoapositivemriofthesijointsovertime.figure3.2

81 ActivelesionsonMRIofthesacroiliacjoints 81 showsthelikelihoodofapositivemriinrelationtohlab27status.thelikelihoodofa positivemriisnegligible(<5%)incaseofanhlab27negativepatientwithanegative MRIatbaseline.Thelikelihoodiscloseto90%incaseofanHLAB27positivepatient withapositivemriatbaseline.thelikelihoodisintermediateinpatientswitheither HLAB27positivitybutanegativeMRIatbaseline(27%)orapositiveMRIbutnegative HLAB27statusatbaseline(49%). Figure3.2 Likelihood of a positive MRI at 1 or 2years of followup in patients with shortstanding inflammatorybackpainassessedatbaseline,1yearand2yearsoffollowup,infunctionofthe result of the baseline MRI (negative or positive) and HLAB27 status. HLAB27, human leukocyteantigenb27. DISCUSSION In this cohort of patients with early IBP, almost twothirds of the patients had a negativemriatbaseline.interestingly,inonlyaminorityofthesepatients(15%),the MRIstatuschangedduringfollowup.AnegativeMRIatbaselineincombinationwitha negativehlab27statushadalikelihoodof95%offindinganegativemriinthenext2 years. This finding may be of clinical relevance, since it suggests that in HLAB27 negativepatientswithanegativemrisacroiliitiscanbeexcludedwithahighlevelof certainty. AsimilarconclusioncanbedrawnfromthegroupofpatientswithapositiveMRI.One can conclude that if sacroiliitis is detected by MRI, there is a high likelihood that sacroiliitisremainspresentinhlab27positivepatients(88%).thisfindingtherefore

82 82 Chapter3 addstothecredibilityofmriasapivotalmeasureintheaxialspacriteria.itisofnote thattherewasnouseofbiologicalsinthiscohortofpatientswithearlyibpduringthis 2yearperiod. In general, both male gender and HLAB27 positivity independently determined the likelihoodofapositivemriatanytimepoint.thisisnotsurprising,sincespaandin particular AS is strongly associated with HLAB27 and has an overall male predominance.inthisstudyitwasfoundthathlab27positivemalepatientswithibp havethehighestchanceofapositivemriatanytime. ItisnowwellrecognisedthatMRIissuperiortoconventionalradiographyindiagnosing earlyspa.however,insightintheevolutioninmrifindingsinpatientswithearlyspa andsacroiliitisislimited,sincefollowupstudiesarescarce.toourknowledge,thisis the first study in patients with early IBP with three successive MRI examinations spanning 2 years of followup. Puhakka etal. performeda 1year followupstudy to describe changes in MRI findings of the SI joints in 34 patients with SpA of recent onset.[17] In this study, 30 out of 34 patients already had signs of inflammation on theirmriatbaseline,whichisahigherproportionthaninourcohort.resultsfromthis studyshowedthatthegeneralmriactivityscoredidnotchangeat1year;however, bonemarrowedemadecreasedsignificantlyduringfollowup. TheMRIscoringmethodusedinthepresentstudyisacombinationofthepublished SPARCC and the Aarhus grading methods.[14, 15] Both systems have proven to be reliable and reproducible. With the SPARCC scoring method, a maximum of six consecutiveslicesareevaluated.instead,wechosetoscoretheentiresijointonall qualitativelyoptimalslices.thiswasdonetomaximisethedetectionofactivelesions. In practice, it implied that in the majority of patients between 5 and 8 slices were evaluated,whiletakingcarethatthesame(numberof)sliceswerescoredperpatient over time. The scoring of active lesions was done dichotomously as by the SPARCC method.duetotheunlimitednumberofslicesevaluated,thepresentscoringmethod demandsmrscansofthesameanatomicalareaofeachpatientexaminedatfollowup tosecurethatscoringstartsandendsatthesameanatomiclevel.thismaycausesome limitation of the method if the technique for obtaining oblique coronal slices is not standardized. With a scoring based on both oblique coronal and axial slices this problemdoesnotexistasthesijointsthencanbeevaluatedthreedimensionally.[18] Thepresentstudyhasseverallimitationsthatneedtobeacknowledged.Althoughwe hadmrisofallpatientsatbaseline,asubstantialnumberofmriswasmissingatoneor at both followup visits. Missing data may have influenced the results, since it is possiblethatpatientswithmissingmrisrepresentaproportionofpatientsinwhom

83 ActivelesionsonMRIofthesacroiliacjoints 83 the symptoms that made them eligible for inclusion in the ESpAC study have disappeared.althoughsuchabiascannotbeexcluded,geeadjuststosomeextentfor such spurious effects, by adjusting for withinpatient correlation and not excluding patientswithmissingtimepoints.anotherlimitationofthisstudyisthatthemrisets were scored by one reader. However, the scoring system used in the present study uses a combination of scoring systems that already have proven to be reliable with respecttointerreaderagreement.[19]ageneralissueregardingassessmentofmris ofthesijoints,isthatinapreviousstudyinupto30%ofthesijointsofindividuals with chronic low back pain or healthy controls abnormalities were detected [20] Suchlesionsmayhavecontributedtosomeextenttotheresultswehavefound,butit is likely that these abnormalities will disappear in noninflammatory conditions over timeincontrasttomalepatientswithibpthatarehlab27positive.atlast,theespac cohort in which the analyses were performed is a rather small cohort with characteristics that do not necessarily resemble the average patient presenting with complaints suggestive of SpA. Extrapolation of findings should therefore be done cautiously. Inconclusion,MRIisaveryusefulimagingtechniqueinestablishinginflammationin thesijointsandmaycontributetomakingadiagnosisofaxialspainpatientswithibp. InthiscohortofpatientswithIBPofrecentonset,itseemsthatduringa2yearfollow upperiod,theproportionofpatientswithandwithoutsignsofmriactivityremains stabletoagreatextent.thisisanimportantfinding,sinceitsuggeststhatincaseof bothanegativemriandanegativehlab27status,aclassifyingdiagnosisofaxialspa canbeexcludedwithgreatcertainty.inthiscohortwefoundthatthecombinationofa positive MRI scan and HLAB27 positivity is associated with a high likelihood of a persistentpositivemriduringfollowup.inhlab27positivepatientswithanegative MRIatbaseline,inwhomaclassifyingdiagnosisofaxialSpAcannotbemadebythe routeofhlab27presenceplustwoadditionalspafeatures,afollowupmricanbe considered since a reasonable proportion of these patients will develop MRI activity overtime,especiallyinmalepatients.

84 84 Chapter3 REFERENCES Song IH, Sieper J, Rudwaleit M. Diagnosing early ankylosing spondylitis. Curr Rheumatol Rep 2007; 9: RudwaleitM.Newapproachestodiagnosisandclassificationofaxialandperipheralspondyloarthritis. CurrOpinRheumatol2010;22: Dougados M, van der Linden S, Juhlin R, et al. The European spondyloarthropathy study group preliminarycriteriafortheclassificationofspondyloarthropathy.arthritisrheum1991;34: AmorB,DougadosM,MijiyawaM.Criteriafortheclassificationofspondyloarthropathies(French).Rev RhumMalOsteoartic1990;57:859. VanderLindenS,ValkenburgHA,CatsA.Evaluationofdiagnosticcriteriaforankylosingspondylitis.A proposalformodificationofthenewyorkcriteria.arthritisrheum1984;27:3618. RudwaleitM,VanderHeijdeD,LandewéR,etal.ThedevelopmentofAssessmentofSpondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: FeldtkellerE,KhanMA,vanderHeijdeD,etal.AgeatonsetanddiagnosisdelayinHLAB27negativevs positivepatientswithankylosingspondylitis.rheumatolint2003;23:616. OostveenJ,PrevoR,DenBoerJ,etal.Earlydetectionofsacroiliitisonmagneticresonanceimagingand subsequent development of sacroiliitis on plain radiograpghy. A prospective, longitudinal study. J Rheumatol1999;26: BennettAN,McGonagleD,O'ConnorP,etal.Severityofbaselinemagneticresonanceimagingevident sacroiliitis and HLAB27 status in early inflammatory back pain predict radiographically evident ankylosingspondylitisateightyears.arthritisrheum2008;58: HeuftDorenboschLH,LandewéR,WeijersR,etal.Performanceofvariouscriteriasetsinpatientswith inflammatorybackpainofshortduration;themaastrichtearlyspondyloarthritisclinic.annrheumdis 2007;66:928. Calin A, Porta J, Fries JF, et al. Clinical history as a screening test for ankylosing spondylitis. JAMA 1977;237: HeuftDorenbosch LH, Landewé R, Weijers R, et al. Combining information obtained from magnetic resonance imaging and conventional radiographs to detect sacroiliitis in patients with recent onset inflammatorybackpain.annrheumdis2006;65:8048. HeuftDorenboschL,WeijersR,LandewéR,etal.Magneticresonanceimagingchangesofsacroiliac jointsinpatientswithrecentonsetinflammatorybackpain.arthritisresther2006;8:r11. Maksymowych WP, Inman RD, Salonen D, et al. Spondyloarthritis research consortium of Canada magnetic resonance imaging index for assessment of sacroiliac joint inflammation in ankylosing spondylitis.arthritisrheum2005;53:7039. PuhakkaKB,JurikAG,EgundN,etal.Imagingofsacroiliitisinearlyseronegativespondyloarthropathy. Assessment of abnormalities by MR in comparison with radiography and CT. Acta Radiol 2003;44: RudwaleitM,JurikAG,HermannKGA,etal.Definingactivesacroiliitisonmagneticresonanceimaging (MRI)forclassificationofaxialspondyloarthritis:aconsensualapproachbytheASAS/OMERACTMRI group.annrheumdis2009;68: Puhakka KB, Jurik AG, SchiottzChristensen B, et al. MRI abnormalities of sacroiliac joints in early spondyloarthropathy:a1yearfollowupstudy.scanjrheumatol2004;33: Madsen KB, Jurik AG. Magnetic resonance imaging grading system for active and chronic spondylarthritischangesinthesacroiliacjoint.arthritiscareres2010;62:118. LandewéRB,HermannKG,vanderHeijdeDM.Scoringsacroiliacjointsbymagneticresonanceimaging. Amultiplereaderreliabilityexperiment.JRheumatol.2005;32: Weber U, Lambert RGW, Ostergaard M, et al. The diagnostic utility of MRI in spondyloarthritis: An international multicentre evaluation of one hundred and eightyseven subjects. Arthritis Rheum 2010;62:

85 CHAPTER4 BonemarrowedemaonMRIofthesacroiliacjointsis associatedwithdevelopmentoffattylesionsonmri overa1yearintervalinpatientswithearly inflammatorylowbackpain:a2yearfollowupstudy MarloesvanOnna,AstridvanTubergen,DésiréevanderHeijde,AnneGretheJurik, RobertLandewé

86 86 Chapter4 ABSTRACT Objective Toassesswhetherbonemarrowedema(BME)detectedonMRIofthesacroiliacjoints (MRISIJ)isassociatedwithdevelopmentofstructuralchangesonbothMRIandpelvic radiographsinpatientswithearlyinflammatorybackpain(ibp). Methods PatientswithIBP2yearswerefollowedfor2yearswithannualMRISIJ.MRIswere scoredforbmeandstructuralchanges(erosionsandfattylesions).pelvicradiographs were graded according to the modified New York (mny) criteria. With generalized estimated equation analysis, a time trend in the structural change scores was investigated. Results Sixtyeightpatients(38%male;mean(SD)age34.9(10.3)years)wereincluded.During the 2year followup, pelvic radiograph grading remained constant. On MRI, the numberoferosionsperpatientincreasedsignificantly(meanscore2.5atbaselineand 3.5at2yearsfollowup;p=0.05).Atrendwasfoundforanincreaseinthenumberof fatty lesions per patient (mean score 5.4 at baseline and 8.5 at 2years followup; p=0.06).overall,bmewasassociatedwiththedevelopmentoffattylesions(rightsij: oddsratio(or)3.13,95%confidenceinterval(ci) ;leftsij:or22.13,95%ci ),preferentiallyinquadrantsshowingresolutionofBME.Incontrast,BME (ortheresolutionthereof)wasnotassociatedwiththedevelopmentoferosions. Conclusion BMEatbaseline,especiallywhenitdisappearsovertime,resultsinthedevelopmentof fattylesions,butanassociationwitherosionscouldnotbedemonstrated.

87 StructuralchangesonMRIofthesacroiliacjoints 87 INTRODUCTION MRIofthesacroiliacjoints(MRISIJ)andpelvicradiographsplayanimportantrolein the diagnosis and classification of axial spondyloarthritis (axspa). Both imaging techniquesareincludedinthe imagingarm oftheassessmentinspondyloarthritis internationalsociety(asas)criteriaforaxspa.[1]mrisijhasespeciallyproventobe usefulintheearlystageofaxspa,becausemrimaydetectsacroiliitisyearsbeforeitis seen on a pelvic radiograph. Moreover, MRI can detect both active lesions and structural changes, in contrast to pelvic radiographs which only detect structural changes.[2] Patients without radiographic sacroiliitis, but with active lesions on MRI suggestiveforsacroiliitis,arelabellednonradiographicaxspa(nraxspa).[1] Active (inflammatory) lesions that can bedetected on MRI are bone marrowedema (BME),capsulitis,synovitisandenthesitis.StructuralchangesonMRIareerosions,fat deposition(fattylesions),sclerosisandankylosis.[1,3]todate,onlybmeisconsidered mandatory for fulfilment of the ASAS/Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) working group definition of a positive MRI.[2] However, Weberetal.suggestedthatalsothedetectionoferosionsonMRISIJmaybehelpfulin makingadiagnosisinearlyaxspa.[3]thereis,however,limitedknowledgeaboutthe development of structural changes as detected on MRISIJ in patients with early IBP who may possibly have or develop axspa. Furthermore, data on the association between BME and development of structural changes are scarce. Only one study suggestedthatfattylesionsonmrisijmaybethefirstsignofstructuraldamageafter previousactiveinflammation.[4] Theaimsofthisstudyweretoassessthepresence1)ofstructuralchangesonboth MRISIJandpelvicradiographsand2)whetherBMEdetectedonMRISIJisassociated withthedevelopmentofstructuralchangesonbothmrisijandpelvicradiographsin patientswithrecentonsetibpovera2yearfollowupperiod. PATIENTSANDMETHODS Studypopulation Patients with IBP of less than 2 years duration were enrolled in the Early SpondyloArthritisClinic(ESpAC)study.Inthisprospectiveinceptioncohortstudy,three identicalclinicalandradiologicalexaminationswereperformedatbaselineandafter1 and2years.amoredetaileddescriptionofthestudypopulationandinclusioncanbe

88 88 Chapter4 foundinpreviousreports.[57]thecalincriteriawereusedtodefinethepresenceof IBP.Patientshadtofulfilatleast4ofthecriteria:onsetbeforetheageof40years, duration of back pain more than 3 months, insidious onset, morning stiffness, and improvementwithexercise.[8]patientswhofulfilledonly3outof5ofcalincriteria, but reported night pain were also included. Presence of other SpA features was preferredbutnotobligatory.patientswerenottreatedwithbiologicaltherapyduring theentirestudyperiod,theuseofnonsteroidalantiinflammatorydrugs(nsaids)was allowed. The study was approved by the ethics committee from the Maastricht UniversityMedicalCenter.Allpatientsgavewritteninformedconsent. MRIprotocol MRISIJ was performed using a 1.5 Tesla Philips Gyro scan ACSNT (Philips, The Netherlands). Patients were placed in supine position in a spine surface coil. The following sequences were used in an oblique coronal plane: T1weighted spin echo (SE), short tau inversionrecovery (STIR) and T2weighted fast SEwith fat saturation. TheMRIsetofeachindividualwasscoredindependentlyandinarandomtimeorder by1experiencedradiologist(agj),withoutknowledgeofclinicalorlaboratoryfindings. TheSIJswerescoredusingacombinationoftheSpondyloarthritisResearchConsortium ofcanada(sparcc)methodandamodifiedversionoftheaarhusmri scoring system.[9,10] In contrast to the original SPARCC system, there was no maximumtothenumberofevaluatedslices,butthenumberofsliceswithinapatient was kept the same for all time points. First, each SIJ was divided into 4 quadrants (upper/lower sacral and upper/lower iliac quadrant) and both BME and structural changes were scored separately per slice in a dichotomous manner (absent vs. present). Second, a total count of both BME and structural changes per SIJ was performed. The method for scoring BME has been published.[11] The structural changes scores included scoring of both erosions and fatty deposition of the bone marrow (fatty lesions). Erosions, defined as cortical defects of the SIJ lining, are detectedashypointensesignalonthet1sequence.subcorticalfattylesions,definedas replacement of normal bone marrow by fatty tissue, are detected as an increased signalonthet1sequence.accordingtotheasas/omeractworkinggroupdefinition for active lesions on MRISIJ, an MRISIJ is considered positive for active sacroiliitis whenatleast1activelesionispresentinatleast2successiveslicesorwhen2lesions aredetectedin1slice.[2]analogoustotheasas/omeractworkinggroupdefinition foractivelesionsonmrisij,wedefinedanmrisijaspositiveforstructuralchanges

89 StructuralchangesonMRIofthesacroiliacjoints 89 whenatleast1erosionorfattylesionispresentinatleast2successiveslicesorwhen 2erosionsorfattylesionsaredetectedin1slice. Pelvicradiographs AnteroposteriorpelvicradiographsoftheSIJwereobtainedandindependentlyscored in a random time order by 2 readers (AT and RL) who were blinded to clinical and laboratoryfindingsandwerenotinvolvedinthemrireading.incaseofdisagreement, thejudgmentofathirdreader(dh)wasdecisive.thepelvicradiographswerescored accordingtothemodifiednewyork(mny)criteriaforsacroiliitis,inwhichsacroiliitisof atleastgrade2bilaterallyorgrade34unilaterallymustbepresentforfulfillmentof themnycriteria.[12] Statisticalanalysis Descriptive statistics were used to describe the presence or absence of structural changesonmrisijandpelvicradiographsduringthe2yearfollowupperiod. ThepresenceofBMEatbaselineandduringfollowupinrelationtothedevelopment offattylesionsand/orerosionsatthesiteofbmewasinvestigatedperquadrantofsij (8 quadrants per patient) using descriptive analysis. Generalized estimated equation (GEE) analysis was performed to investigate a time trend in the MRI and pelvic radiographstructuralchangesscores.geeisamodelthatallowsstudyingtimetrends while taking the withinsubject correlation into account in a dataset with missing values. Furthermore, this technique was used to test whether BME on MRI is associatedwitherosionsandfattylesionsonmriandstructuralchangesonapelvic radiograph1yearlater.thiswasdoneusingageeautoregressivetimelagmodelthat correlatesthepresenceofbmeonmritoeachstructuralchangescoreoneithermri or pelvic radiograph 1 year later. On a per SIJ level, BME was used as independent variableandthecontinuous(sumof)mriorthepelvicradiographmnygradinginthe samelocationwereusedasdependentvariables.spsssoftwareversion18.0wasused forallstatisticalanalyses.allpvaluesweretwotailedandstatisticalsignificancewas setat0.05.

90 90 Chapter4 RESULTS Patientcharacteristics Baselinecharacteristicsofthe68patientsincludedinESpACareshowninTable4.1. MRISIJ and pelvic radiographs were obtained in all patients at baseline. Sixtytwo (91%) patients had at least 1 followup MRI and 44 (65%) patients completed both followupmris.sixtyfive(96%)patientshadatleast1followuppelvicradiographand 48 (71%) patients completed both followup pelvic radiographs. In 10 (15%) of 68 patients adjudication of pelvic radiographs was considered necessary because of disagreementbetweenthefirst2readers. Table1.Baselinecharacteristicsof68patientsincludedintheESpAC. Characteristic Allpatients (n=68) PatientswithcompleteMRIfollow updata (n=44) Malesex 26(38) 15(34) Meanage(SD)[years] 34.9(10.3) 36.0(11.7) Mediansymptomduration(IQR)[months] 18(1224) 18(1224) HLAB27positive 31(46) 17(39) Historyofinflammatoryboweldisease 10(15) 7(16) Historyofanterioruveitis 10(15) 8(18) Historyofpsoriasis 16(24) 12(27) Historyofperipheralarthritis 19(28) 12(27) FamilyhistoryofSpA 37(54) 26(59) MeanCRP(SD)[mg/l] 9(11) 9(12) ElevatedCRP a 16(24) 10(22) MeanESR(SD)[mm] 13(15) 13(16) ElevatedESR a 24(36) 13(30) PresenceofBMEonMRISIJ 24(35) 14(32) FulfillmentESSGcriteria 58(85) 39(89) FulfillmentAmorcriteria 48(71) 31(70) FulfillmentASASaxSpAcriteria 40(59) 22(50) FulfillmentmNYcriteria 15(22) 9(20) The values are expressed as number (percentage) of patients unless stated otherwise. ESpAC=Early Spondyloarthritis Clinic; SpA=spondyloarthritis; IQR=interquartile range; SD=standard deviation; CRP=C reactive protein; ESR=erythrocyte sedimentation rate; BME=bone marrow edema; MRISIJ=magnetic resonance imaging of the sacroiliac joints; ESSG=European Spondyloarthropathy Study Group; ASAS= AssessmentinSpondyloArthritisinternationalSociety;axSpA=axialspondyloarthritis;mNYcriteria=modified New York criteria. a In 66 of 68 patients baseline CRP and ESR measurements were available. ESR normal range:7mmformales;12mmforfemales.crpcutoffvalue,normalrange:<10mg/l. At baseline, 64 (94%) out of 68 patients fulfilled the European Spondyloarthropathy Study Group (ESSG) and/or Amor and/or ASAS axspa classification criteria. Sixtysix (97%)patientsfulfilled4ofthe5Calincriteria.Theremaining2(3%)patientsfulfilled3

91 StructuralchangesonMRIofthesacroiliacjoints 91 ofthe5calincriteriaandreported nightpain aswell.forty(59%)patientsfulfilledthe ASASaxSpAcriteriaatbaseline,ofwhich22fulfilledtheimagingandclinicalarmofthe ASAS axspa criteria, 9 patients fulfilled only the imaging arm, and 9 other patients fulfilled only the clinical arm. Fifteen (22%) patients fulfilled the mny criteria at baseline. StructuralchangesonMRISIJ ScoresforBMEonMRISIJhavebeenpresentedinapreviousreport.[11]Table4.2and 4.3 show the baseline and followup MRI findings for erosions and fatty lesions, respectively and the relation with fulfillment of the mny criteria and ASAS axspa criteriaatbaseline,andhlab27status. Table4.2 ErosionsonMRISIJatbaselineandduringfollowup. Numberof MRISIJ mnycriteria+ a patients Baseline 1Year 2Years ASASaxSpA criteria+ b HLAB27+ c NA NA NA NA NA NA NA NA NA ThenumberofpatientswithpresentorabsentsignsoferosionsonMRISIJatbaselineandduringfollowup isshown. a NumberofpatientsthatfulfilthemNYcriteriaatbaseline. b Numberofpatientsthatfulfilthe ASAS axspa criteria at baseline. c Number of patients that are HLAB27 positive. +=structural changes present;=structural changes absent, NA=MRI not available, MRISIJ=magnetic resonance imaging of the sacroiliac joints; HLAB27=human leukocyte antigen B27, mny criteria=modified New York criteria, ASAS axspacriteria=assessmentinspondyloarthritisinternationalsocietyaxialspondyloarthritiscriteria. Atbaseline,erosionsand/orfattylesionsonMRIweredetectedin17(25%)patients,of which71%werehlab27positiveand9(53%)fulfilledthemnycriteria.erosionswere detectedin12(18%)outof68patientsofwhich8(66%)patientsalsofulfilledthemny criteria(table4.2).fattylesionsweredetectedin13(19%)outof68patientsofwhich 7(54%)patientsalsofulfilledthemNYcriteria(Table4.3).Coexistenceofbotherosions andfattylesionsonmrioccurredin8(9%)outof68patientsofwhich6(75%)patients alsofulfilledthemnycriteria.

92 92 Chapter4 Table4.3 FattylesionsonMRISIJatbaselineandduringfollowup. Number of MRISIJ mnycriteria+ a patients Baseline 1Year 2Years ASASaxSpA criteria+ b HLAB27+ c NA NA NA NA NA NA NA NA NA NA NA ThenumberofpatientswithpresentorabsentsignsoffattylesionsonMRISIJatbaselineandduringfollow upisshown. a NumberofpatientsthatfulfilthemNYcriteriaatbaseline. b Numberofpatientsthatfulfilthe ASAS axspa criteria at baseline. c Number of patients that are HLAB27 positive. +=structural changes present;=structural changes absent, NA=MRI not available, MRISIJ=magnetic resonance imaging of the sacroiliac joints; HLAB27=human leukocyte antigen B27, mny criteria=modified New York criteria, ASAS axspacriteria=assessmentinspondyloarthritisinternationalsocietyaxialspondyloarthritiscriteria. Duringfollowup,4(8%)outof51patientswithnosignsoferosionsatbaseline,andat least1followupmripresent,havedevelopederosions,ofwhich2patientswerehla B27positiveand3patientsalreadyfulfilledthemNYcriteriaatbaseline(Table4.2).In2 of these 4 patients, the erosions were accompanied by simultaneous appearance of fattylesionsonmri.five(10%)outof50patientswithoutfattylesionsatbaselineand atleast1followupmridevelopedfattylesionsduringfollowup,fourofthemwere HLAB27positiveand3patientsfulfilledthemNYcriteriaatbaseline(Table4.3). In1(8%)outof12patientswitherosionsonMRIatbaseline,theerosionswerenot detectedanymoreatthe2yeartimepoint(table4.2).in3(23%)outof13patients withfattylesionsonmriatbaseline,thefattylesionswerenotdetectedduringfollow up(table4.3). Structuralchangesonpelvicradiographs Atbaseline,15(22%)patientsfulfilledthemNYcriteriaforradiographicsacroiliitisof which80%werehlab27positive.eight(53%)ofthese15patientsalsohadbmeat baselineandin9(60%)patientserosionsand/orfattylesionsweredetectedonmri. During followup no changes in mny grading were found in these 15 patients. Four patientswhofulfilledthemnycriteriaatbaselinedevelopednewerosionsand/orfatty

93 StructuralchangesonMRIofthesacroiliacjoints 93 lesionsonmri(table4.2and4.3).nopatientswerenewlyclassifiedwithradiographic sacroiliitisaccordingtothemnycriteriaatfollowup. AssociationbetweenBMEandstructuralchangesonMRISIJ ThefirstanalysiswasaGEEanalysis(scoresof58patientswithatleast2successive time points available included) showing that the number of erosions per patient increasedsignificantlyduringfollowup(estimatedmarginal(em)meanscoreof2.5at baseline,3.3at1yearfollowupand3.5at2yearfollowup;p=0.05).therewasalsoa trendforanincreaseinthenumberoffattylesions(emmeanscoreof5.4atbaseline, 7.7at1yearfollowupand8.5at2yearfollowup;p=0.06). ThesecondanalysiswasaGEEanalysisperSIJ(2perpatient)showinganassociation betweenbmeandthedevelopmentoffattylesionsonmri(table4.4).anassociation betweenbmeandsubsequentdevelopmentoferosionsonmricouldnotbeproven (Table4.4). Table4.4 GEEanalysis(continuousMRISIJscore)showingassociationbetweenbonemarrowedemaon MRISIJandthedevelopmentofstructuralchangesonMRISIJduringfollowup. Perjointanalysis,structuralchangesMRISIJ Oddsratio(95%CI) pvalue MRIdevelopmentof Fattydegeneration(perunit) RightSIJ 3.13( ) 0.04 structuralchanges LeftSIJ 22.13( ) 0.03 (perunitchange) Erosions(perunit) RightSIJ 0.24( ) 0.43 LeftSIJ 1.24( ) 0.88 N=58.GEE=generalizedestimatedequation,MRISIJ= magneticresonanceimagingofthesacroiliacjoints. 95%CI=95%confidenceinterval ThethirdanalysiswasadetaileddescriptiveanalysisatthelevelofquadrantsofSIJ,in order to investigate the association between the presence of BME and the developmentoferosionsand/orfattylesionsatthesamesite.thisanalysiswasdone to investigate how an inflammatory lesion at a particular site associates with the developmentofastructurallesionatthesamesite.everypatienthad8quadrants(4 leftand4right)availableforcomparison. ThestartingpointofthisanalysiswasthepresenceofBME.Continuouspresenceof BMEovertimevs.resolutionofBMEwastakenintoconsideration.Theresultsofthis analysis which are presented in Table 4.5 show that an increase of fatty lesions preferentially occurs in quadrants in which BME has resolved over time (55%) in comparisonwithquadrantsinwhichbmehaspersistedovertime(26%).withregard tothedevelopmentoferosionssuchadisparitycouldnotbeconfirmed(26%vs.30%).

94 94 Chapter4 Table4.5 Presence of bone marrow edema on MRI at baseline and followup in relation to the subsequentdevelopmentoffattylesionsanderosionsatthesamesacroiliacjointquadrant. Baseline LastfollowupMRI* BME(present) BME(present) BME(present) ANDincreaseof FLatthesame quadrant BME(absent) BMEabsent ANDincreaseof FLatthesame quadrant Allquadrants (26) 42 23(55) BME(present) BME(present) BME(absent) BME(present) ANDincreaseof erosionsatthe samequadrant BMEabsent ANDincreaseof erosionsatthe samequadrant Allquadrants (30) 42 11(26) n=62 patients.* Followup at 1 or 2 years, depending on last MRI. SIJ=sacroiliac joint; BME=bone marrow edema,fl=fattylesion,mri=magneticresonanceimaging. DISCUSSION The present study has demonstrated that BME on MRISIJ is associated with the developmentoffattylesionsinacohortofpatientswithearlyibp.suchanassociation could not be demonstrated for erosions, despite a significant overall numerical increase of erosions over time. Furthermore, approximately 10% of the patients without erosions and fatty lesions at baseline developed new structural changes on MRIduringthe2yearfollowupperiod. An increase in the number of structural changes on MRI over time has been demonstrated in a previous study evaluating MRISIJ abnormalities in patients with earlyaxspa.[13]inthisstudybymadsenetal.80outof94(85%)patientswithaxspa had erosions and fatty lesions present on MRI at baseline. After a mean followup periodof51months,boththenumberoferosionsandfattylesionshadsignificantly increased.[13]wehavealsofoundasignificantincreaseinthenumberoferosions,as wellasastrongtrendforanincreaseinthenumberoffattylesions.inthestudyby Madsen et al., the proportion of patients with structural changes was higher at baselinecomparedwithourcohort(85%versus25%),andalsotheirfollowupduration waslonger,whichmightexplainthedifferenceintimetrendsfoundforfattylesions. Thepresenceoffattylesionsmightalsobeunderestimatedinourstudyduetothefact that thepresence of intense BME mayprevent the detection of fatty lesions due to counteracting MR signals on the T1 MRI sequence (BME low signal intensity; fatty lesionshighsignalintensity).[4]

95 StructuralchangesonMRIofthesacroiliacjoints 95 In thepresent study, wehave statisticallydemonstratedanassociation between the presenceofbmeandthesubsequentdevelopmentoffattylesionsonmri.evenmore importantistheindicationthatthesefattylesionsseemtohavepreferentiallyoccurred atthesites(quadrants)inwhichexistingbmehasresolvedovertime.thisisentirelyin line with existing theories about the role of fatty lesions as a repair reaction in responsetoinflammatorytriggers.astudybysongetal.hasdescribedanassociation betweendisappearanceofbmeonmriinboththesijandspineandthesubsequent appearanceoffattylesionsonmriinpatientswithearlyaxialspatreatedwitheither etanerceptorsulfasalazineovera1yearfollowupperiod.inpatientsinwhombme resolved,fattylesionsoccurredin10.5%ofthesijquadrants1yearlater,butinthose inwhombmepersisted,fattylesionsoccurredinonly2.4%ofthesijquadrants1year later.[4]inastudybymaksymowychetal.,76patientswithaxspawerefollowedfor1 yearwithrepeatedmrisofthespine.[14]thechanceofdevelopingnewfattylesions was significantly higher at vertebral corners with BME at baseline as compared to vertebral corners without BME at baseline (18% vs. 3%).[14] A correlation between BME and the subsequent development of fatty lesions on MRISIJ could not be demonstrated in the SPondyloArthritis Caught Early (SPACE)cohort over a 3month followup period.[15] This last finding suggests that a longer followup period is necessarybeforefattylesionscanbedetectedinresponsetosubsidingbme. Whilst we have found an increase in the number of erosions in this study, we have failedtodemonstrateasignificantassociationbetweenthepresenceofbmeandthe development of erosions. A possible explanation is that the erosions occur independently of inflammation. Larger cohorts are necessary to provide sufficient statisticalpowertoclarifytherelationbetweenbmeanddevelopmentofbothfatty lesionsanderosions. TherateofdevelopmentfromnonradiographictoradiographicaxSpAinpatientswith aclinicaldiagnosisofaxspawithoutsignsofradiographicsacroiliitisisestimatedtobe approximately10%per2years.[16]inpatientswithbmeonmrisij,thispercentage increases to around 20% per 2 years.[17] Our study has failed to demonstrate an associationbetweenbmeanddevelopmentofstructuralchangesdetectedonpelvic radiographs,norcouldwedemonstrateachangeinthelevelofsacroiliitisaccordingto the mny criteria during the 2year followup. The short followup period and the relativelysmallsamplesizemayhavehamperedtodetectthesechanges. ThereareseveralconcernswhenassessingstructuralchangesoftheSIJoneitheran MRIorpelvicradiographanditisdebatablewhichimagingmodalityisapproachingthe truth or whether they are complementary. Both imaging modalities are subject to

96 96 Chapter4 observererrorsandespeciallytheevaluationofpelvicradiographsmaybehampered byprojectionartifactsandpoorvisibility.[18,19]erosionsonmrimaynotbereliably detectediftheslicethicknessisnotsmallenoughormovementartifactsmaylimitan accurateimageinterpretation.thenaturalirregularshapeofthecorticalliningofthe SIJmayalsolimitdetectionoferosions.Inthepresentstudy,ofthe15patientswith sacroiliitis on pelvic radiographs according to independent readers, erosions and/or fattylesionsonmriweredetectedinonly9(60%)patientsatbaseline.themajorityof thepatients,whodevelopednewerosionsand/orfattylesionsonmriduringfollow up,alreadyfulfilledthemnycriteriaatbaseline. Some limitations of the present study need to be addressed. First, the MRIs were scored by one reader only,which may intheory influence the reliability of the data. However, the reader was very experienced and the MRI scores showed a high consistencyovertimedespitescoringeachmrisetindependentlyandblindingofthe reader for time order. Furthermore, and in contrast to scoring pelvic radiographs, a numberofstudieshaveshownaratherhighinterobserveragreementwhenscoring MRISIJs.[9,20] Second, in the present study, the T1 and STIR sequence were simultaneously scored for active lesions and structural changes. This scoring method couldhaveresultedinreaderbias,sincethepresenceofbmeonthestirsequence maypossiblyhavetriggeredthereadertoscreenmorecarefullyforstructuralchanges on the T1 sequence or vice versa. This may influence the sensitivity for scoring structural changes on MRISIJ. In 16 of the 17 patients with erosions and/or fatty lesions at baseline, concomitant BME was also found. However, one could also postulatethatconcomitantbmedecreasesthespecificityofscoringstructuralchanges onmri,sinceareaderismorelikelytoscoreanindeterminatelesionasastructural change.scoringthet1andstirsequenceindependentlycouldpossiblylowertherisk ofreaderbiaswhenevaluatingthepresenceofactivelesionsandstructuralchangeson MRISIJ. Third, in the MRI scoring system used in the present study, we applied an unlimitednumberofevaluatedslices.serialimageacquisitionofthesameanatomical regionoffersadvantagestomonitorchangesovertime.however,thisscoringmethod may be limited by the possibility of misalignment between 2 successive MRI examinations, which may cause measurement error. Last, ESpAC is a relatively small cohortwithselectedpatients.patientswerereferredby(related)medicalspecialties (i.e. dermatology, gastroenterology) and through family members of the local ankylosing spondylitis society. This referral strategy may explain the relative high proportionofpatientswithextraaxialmanifestationsand/orpositivefamilyhistoryfor SpA. As a consequence, a high proportion of patients fulfilled at least one of the

97 StructuralchangesonMRIofthesacroiliacjoints 97 classificationcriteriaforspa.furthermore,theproportionoffemalepatientsinespac isrelativelyhigh(62%)andtheproportionofpatientswithapositivehlab27statusis relativelylow(46%).however,thesepercentagesareinaccordancewithothercohorts thatincludedpatientswithearlyibp.[21,22]nevertheless,extrapolationofthestudy findingsshouldbedonewithcaution. In conclusion, in this cohort of patients presenting with IBP of recent onset and suspectedforaxspa,thenumberofpatientswitherosionsandfattylesionsdetected onmrisijremainedrelativelystableduringthe2yearsoffollowup,buttheoverall number of MRI erosions in patients who already had erosions at baseline has increased.signsofbmeonmri,andespeciallytheresolutionofit,weresignificantly correlatedwiththedevelopmentoffattylesionsonmri,butnotwiththedevelopment ofstructuralchangesvisibleonpelvicradiographs.

98 98 Chapter4 REFERENCES RudwaleitM,vanderHeijdeD,LandewéR,etal.ThedevelopmentofAssessmentofSpondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: RudwaleitM,JurikAG,HermannKG,etal.Definingactivesacroiliitisonmagneticresonanceimaging (MRI)forclassificationofaxialspondyloarthritis:aconsensualapproachbytheASAS/OMERACTMRI group.annrheumdis2009;68: WeberU,ZublerV,PedersenSJ,etal.DevelopmentandvalidationofanMRIreferencecriterionfor definingapositivesijmriinspondyloarthritis.arthritiscareres(hoboken)2013;65: SongIH,HermannKG,HaibelH,etal.Relationshipbetweenactiveinflammatorylesionsinthespine and sacroiliac joints and new development of chronic lesions on wholebody MRI in early axial spondyloarthritis:resultsoftheesthertrialatweek48.annrheumdis2011;70: HeuftDorenboschL,LandewéR,WeijersR,etal.Performanceofvariouscriteriasetsinpatientswith inflammatorybackpainofshortduration;themaastrichtearlyspondyloarthritisclinic.annrheumdis 2007;66:928. HeuftDorenboschL,LandewéR,WeijersR,WandersA,HoubenH,vanderLindenS,etal.Combining information obtained from magnetic resonance imaging and conventional radiographs to detect sacroiliitisinpatientswithrecentonsetinflammatorybackpain.annrheumdis2006;65:8048. HeuftDorenboschL,WeijersR,LandewéR,vanderLindenS,vanderHeijdeD.Magneticresonance imagingchangesofsacroiliacjointsinpatientswithrecentonsetinflammatorybackpain.arthritisres Ther2006;8:R11. Calin A, Porta J, Fries JF, Schurman DJ. Clinical history as a screening test for ankylosing spondylitis. JAMA1977;237: Maksymowych WP, Inman RD, Salonen D, Dhillon SS, Williams M, Stone M, et al. Spondyloarthritis research consortium of Canada magnetic resonance imaging index for assessment of sacroiliac joint inflammationinankylosingspondylitis.arthritisrheum2005;53:7039. PuhakkaKB,JurikAG,EgundN,SchiottzChristensenB,StengaardPedersenK,vanOvereemHansenG, etal.imagingofsacroiliitisinearlyseronegativespondyloarthropathy.assessmentofabnormalitiesby MRincomparisonwithradiographyandCT.ActaRadiol2003;44: VanOnnaM,JurikAG,vanderHeijdeD,vanTubergenA,HeuftDorenboschL,LandewéR.HLAB27 andgenderindependentlydeterminethelikelihoodofapositivemriofthesacroiliacjointsinpatients withearlyinflammatorybackpain:atwoyearmrifollowupstudy.annrheumdis2011;70: VanderLindenS,ValkenburgHA,CatsA.Evaluationofdiagnosticcriteriaforankylosingspondylitis.A proposalformodificationofthenewyorkcriteria.arthritisrheum1984;27:3618. Madsen KB, SchiøttzChristensen B, Jurik AG. Prognostic significance of magnetic resonance imaging changesofthesacroiliacjointsinspondyloarthritisafollowupstudy.jrheumatol2010;37: MaksymowychWP,MorencyN,ConnerSpadyB,LambertRG.Suppressionofinflammationandeffects on new bone formation in ankylosing spondylitis: evidence for a window of opportunity in disease modification.annrheumdis2013;72:238. DeHoogeM,VandenBergR,NavarroCompánMV,ReijnierseM,VanGaalenF,FagerliKM,etal.Do bonemarrowedema(bme)lesionsinthesacroiliacjoint(sij)changeintofattylesionsovera3month periodinpatientswithaxialspondyloathritis(axspa)[abstract]?annrheumdis2013;72:656. Poddubnyy D, Rudwaleit M, Haibel H, Listing J, MärkerHermann E, Zeidler H, et al. Rates and predictorsofradiographicsacroiliitisprogressionover2yearsinpatientswithaxialspondyloarthritis. AnnRheumDis2011;70: SieperJ,vanderHeijdeD.Review:Nonradiographicaxialspondyloarthritis:newdefinitionofanold disease?arthritisrheum2013;65: BraunJ,SieperJ,BollowM.Imagingofsacroiliitis.ClinRheumatol2000;19:517. Van Tubergen A, HeuftDorenbosch L, Schulpen G, Landewé R, Wijers R, van der Heijde D, et al. Radiographic assessment of sacroiliitis by radiologists and rheumatologists: does training improve quality?annrheumdis2003;62:51925.

99 StructuralchangesonMRIofthesacroiliacjoints WeberU,MaksymowychWP,JurikAG,etal.Validationofwholebodyagainstconventionalmagnetic resonance imaging for scoring acute inflammatory lesions in the sacroiliac joints of patients with spondylarthritis.arthritisrheum2009;61:8939. van den Berg R, de Hooge M, Rudwaleit M, Sieper J, van Gaalen F, Reijnierse M, et al. ASAS modification of the Berlin algorithm for diagnosing axial spondyloarthritis: results from the SPondyloArthritis Caught Early (SPACE)cohort and from the Assessment of SpondyloArthritis internationalsociety(asas)cohort.annrheumdis2013;72: DougadosM,d'AgostinoMA,BenessianoJ,BerenbaumF,BrebanM,ClaudepierreP,etal.TheDESIR cohort: a 10year followup of early inflammatory back pain in France: study design and baseline characteristicsofthe708recruitedpatients.jointbonespine2011;78:

100

101 CHAPTER5 GadoliniumcontrastenhancedMRIsequencedoes nothaveanincrementalvalueintheassessmentof sacroiliitisinpatientswithearlyinflammatoryback painbyusingmriincombinationwithpelvic radiographs:a2yearfollowupstudy MarloesvanOnna,AstridvanTubergen,DésiréevanderHeijde,AnneGretheJurik, RobertLandewé

102 102 Chapter5 ABSTRACT Objective To evaluate the potential incremental value in detecting sacroiliitis of the T1 post gadolinium diethylenetriaminepentaacetic acid (GdDTPA) MRI sequence of the sacroiliac joints (SIJ) compared with the combination of short tau inversion recovery (STIR) MRI sequence and pelvic radiographs in patients with inflammatory back pain (IBP)suspectedforaxialspondyloarthritis. Methods A 2year followup study was conducted in patients with IBP of less than 2 years duration.annualmriofthesij(mrisij)wasperformedandscoredforbonemarrow edema (BME). Pelvic radiographs were scored according to the modified New York (mny)criteria.agreementonthepresenceofbmedetectedbythestirandpostgd DTPAsequenceandtheincrementalvalueofpostGdDTPAsequenceoverSTIRplus radiographswasanalysedbydescriptivemethodsandkappastatistics. Results Atbaseline,20(29%)outof68patients(38%male;mean(SD)age34.9(10.3)years) enrolledhadbmebothonthestirandpostgddtpasequences;4patients(6%)on thestirsequenceonly;noneonthepostgddtpasequenceonly(kappavalue:0.87). Fifteen(22%)patientsfulfilledthemNYcriteriaatbaseline. Sixtytwo (91%) patients had at least 1 followup MRISIJ. At 2year followup, 2patients had BME on the postgddtpa sequence without BME on the STIR sequence.these2patientsalreadyfulfilledthemnycriteriaatbaseline. Conclusion InthiscohortofpatientswithearlyIBP,thepostGdDTPAsequenceoftheMRISIJdid nothaveanincrementalvalueinthedetectionofsacroiliitiscomparedwiththestir sequencepluspelvicradiographs.

103 ValueofGadoliniumcontrastenhancedMRIsequence 103 INTRODUCTION Pelvic radiographs and magnetic resonance imaging (MRI) are important imaging techniques todetectsacroiliitis in patientswitha suspicion of axial spondyloarthritis (axspa). In the Assessment in SpondyloArthritis international Society (ASAS) axspa classificationcriteria,sacroiliitisoneithermriorpelvicradiographisusedastheentry criterion for fulfillment of the imaging arm.[1]. With MRI, both active lesions and structuralchangescanbedetected,incontrasttopelvicradiographsthatonlyvisualize structuralchanges.typical activelesions inthesacroiliacjoints(sij)detectedbymri are subchondral bone marrow edema (BME), as well as synovitis, enthesitis and capsulitis.[2]differentmritechniquescanbeusedtodetectactivelesions;theseare shorttauinversionrecovery(stir)thatsuppressesthesignalintensityoffat,andt1 withorwithoutfatsuppressionafteradministrationofthecontrastagentgadolinium diethylenetriaminepentaaceticacid(gddtpa).onlybmeonthestirorpostgddtpa sequence is considered for the definition of active sacroiliitis according to the ASAS/Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) consensus.[2] The ASAS/OMERACT MRI working group further states that the STIR sequencealoneisusuallysufficienttodetectsacroiliitis.thisstatementissupportedby several studies that found no additional value of the postgddtpa sequence compared with the STIR sequence in detecting sacroiliitis in patients with early or established axspa.[3,4] However, these studies were crosssectional[3] or had a 3month followup period only.[4] Furthermore, these studies did not include pelvic radiographs.besidesthatintheasasclassificationcriteriabothpelvicradiographsand MRISIJareconsideredforfulfilmentoftheimagingarm,alsoindailypractice,usually thefirststepformakingadiagnosisistoperformapelvicradiograph.whennegative ofequivocal,thenextstepistomakeanmrisij.althoughontheshortterm,thepost GdDTPAMRIsequencedoesnotseemtobeofadditionalvalue,itisunknownwhatits valueisoveralongerfollowupperiod,especiallyinpatientswithearlydisease,and takingthepelvicradiographintoaccount.fluctuatingorsubsidingbmeoneitherthe STIR or postgddtpa sequence may affect the sensitivity and specificity of MRI in detectingsacroiliitisandhamperthediagnosticprocess.[5]combiningtheinformation onpelvicradiographswiththeinformationonmrisijmayyieldanhigherprobability ofdetectingsacroiliitis.[6] TheaimofthisstudywastoevaluatethepotentialincrementalvalueofthepostGd DTPAsequencefordetectingsacroiliitiscomparedwiththecombinationofSTIRMRI sequence and pelvic radiographs in patients presenting with IBP of short duration

104 104 Chapter5 suspected for axspa. These patients were followed for 2 years with repeated radiologicalexaminations. MATERIALSANDMETHODS Studypopulation Patients with IBP of less than 2 years duration were enrolled in the Early SpondyloArthritis Clinic (ESpAC) study. In this prospective cohort study, systematic clinical and radiological examinations were performed at baseline and after 1 and 2years. A more detailed description of the study population has been reported previously.[7] For IBP to be present, patients had to fulfil 4 of the following 5 Calin criteria: onset of symptoms before the age of 40 years, duration of back pain more than3months,insidiousonset,morningstiffnessandimprovementwithexercise.[8] Patientswhofulfilledonly3outof5oftheCalincriteriabutreportednightpain,were also eligible. Presence of extraaxial manifestations of SpA was preferred but not obligatory. Patients were not treated with biological therapy during the entire study period. The use of nonsteroidal antiinflammatory drugs (NSAIDs) was allowed. The studyhasbeenapprovedbytheethicscommitteeofthemaastrichtuniversitymedical Center.Allpatientshavegivenwritteninformedconsent. MRIprotocol MRIoftheSIjointswasobtainedwitha1.5TeslaPhilipsGyroScanACSNT(Philips,The Netherlands). Patients were examined while lying in a supine position. By using an obliquecoronalsliceorientation,thefollowingsequenceswereobtained: T1weightedspinecho(SE),256x256matrix STIR,256x256matrix T1weighted SE with fat suppression after administration of the intravenous contrastagentgddtpa(0.1mmol/kgbodyweight),512x256matrix The slice thickness was 4 mm with 0.4 mm intervals. Each MRI set was scored with unknowntimesequencebyoneexperiencedradiologist(agj),withoutknowledgeof clinicalorlaboratoryfindings.

105 ValueofGadoliniumcontrastenhancedMRIsequence 105 MRIandpelvicradiographscoring The MR images were scored using a combination of the Spondyloarthritis Research Consortium of Canada (SPARCC) method and a modified version of the Aarhus MRI scoring method.[9,10] In contrast to the original SPARCC system, there was no maximum to the number of evaluated slices, in order to maximise the detection of abnormalmrifindings.thenumberofevaluatedsliceswithinaindividualpatientwas keptthesameforalltimepoints.eachsijwasdividedinto4quadrants:upperiliac, lower iliac, upper sacral and lower sacral. All images (STIR and postgddtpa) were scored for the presence of subchondral BME with the corresponding T1 sequence withoutcontrastsimultaneously.bmepresentinthecartilaginouspartofthejointwas scored per slice in a dichotomous manner (present vs. absent). BME was defined as areas of increased signal intensity on both the STIR and postgddtpa images comparedwithnormalbonemarrow,anditspresencewasestimatedineachofthe 4quadrants.Sincesynovitis,capsulitisandenthesitisarenotconsideredsufficientfor thedefinitionofapositivemriaccordingtotheasas/omeractdefinition,onlythe MRIscoresofBMEweretakenintoaccount.[2] AnMRIwasconsideredpositivewhenatleastoneBMElesionwaspresentinatleast twoconsecutiveslices,orwhentwoormorebmelesionsweredetectedinoneslice, followingtheasas/omeractworkinggroupdefinitionforbmelesionsonmrisij.[2] Anteroposterior pelvic radiographs of the SIJ were obtained from all patients at baseline and during one and 2year follow up. Two readers (AT and RL), who were blinded to the clinical and laboratory findings and were not involved in the MRI reading,independentlyscoredallradiographswithunknowntimesequenceaccording tothemodifiednewyork(mny)criteria.[11]incaseofdisagreement,judgmentofa thirdreader(dh)wasconclusive. Statisticalanalysis Descriptive statistics were used to analyse the presence of BME on MRISIJ and structural changes on pelvic radiographs at baseline and during the2year followup period.agreementonthepresenceofbmesuggestiveforsacroiliitisdetectedbystir andpostgddtpasequenceswasanalyzedonaperpatientbasisbykappastatistics.a kappavalueof00.20indicatedpooragreement, indicatedfairagreement, indicated moderate agreement, indicated substantial agreement and indicated(almost)perfectagreement.[12]descriptivestatisticswereused

106 106 Chapter5 tocomparethepresenceofsacroiliitisdetectedoneitherpelvicradiographsorstir MRIwithfindingsonthepostGdDTPAMRIsequences.SPSSsoftwareversion18.0was usedforallstatisticalanalyses. RESULTS Patientcharacteristics BaselineMRISIJandpelvicradiographswereavailableinall68patientsincludedinthe ESpAC.BaselinecharacteristicsareshowninTable5.1. Sixtytwo(91%)patientshadatleast1followupMRIand44(65%)patientscompleted bothfollowupmris.sixtyfive(96%)patientshadatleast1followuppelvicradiograph and48(71%)patientscompletedbothfollowuppelvicradiographs. At baseline, 64 (94%) out of 68 patients fulfilled the European Spondyloarthropathy Study Group (ESSG) and/or Amor and/or ASAS axspa classification criteria. Fifteen (22%)patientsfulfilledthemNYcriteriaatbaselineandin24(35%)patientsBMEwas detected on MRISIJ. Eight (53%) of 15 patients who fulfilled the mny criteria at baselinehadsignsofbmeonboththestirandpostgddtpasequenceatbaseline. AgreementbetweentheSTIRandpostGdDTPAMRIsequences Atbaseline,agoodagreementbetweentheSTIRandpostGdDTPAsequenceonaper patientbasiswasfound(kappa=0.87).twenty(29%)patientsshowedbmeonmrisij suggestiveforsacroiliitisonbothstirandpostgddtpasequences.in4(6%)patients, BME was detected on the STIR sequence only, but this was minimal in 3 of these 4patients.NoneofthepatientshadsignsofBMEonthepostGdDTPAsequenceonly. Twentythree (96%) of 24 patients with BME on MRI at baseline had at least one followupmri.amoderatetogoodagreementbetweenthestirandpostgddtpa sequence on a per patient basis was also found at 1year (kappa=0.83) and 2year (kappa=0.75)followup.

107 ValueofGadoliniumcontrastenhancedMRIsequence 107 Table5.1 Baselinecharacteristicsof68patientsincludedintheESpAC. Characteristic Allpatients (n=68) PatientswithcompleteMRIfollow updata (n=44) Malesex 26(38) 15(34) Meanage(SD)[years] 34.9(10.3) 36.0(11.7) Mediansymptomduration(IQR)[months] 18(1224) 18(1224) HLAB27positive 31(46) 17(39) Historyofinflammatoryboweldisease 10(15) 7(16) Historyofanterioruveitis 10(15) 8(18) Historyofpsoriasis 16(24) 12(27) Historyofperipheralarthritis 19(28) 12(27) FamilyhistoryofSpA 37(54) 26(59) MeanCRP(SD)[mg/l] 9(11) 9(12) ElevatedCRP a 16(24) 10(22) MeanESR(SD)[mm] 13(15) 13(16) ElevatedESR a 24(36) 13(30) PresenceofBMEonMRI 24(35) 14(32) FulfillmentESSGcriteria 58(85) 39(89) FulfillmentAmorcriteria 48(71) 31(70) FulfillmentASASaxSpAcriteria 40(59) 22(50) FulfillmentmNYcriteria 15(22) 9(20) The values are expressed as number (percentage) of patients unless stated otherwise. ESpAC=Early Spondyloarthritis Clinic; SpA=spondyloarthritis; IQR=interquartile range; SD=standard deviation; CRP= Creactive protein; ESR=erythrocyte sedimentation rate; BME = bone marrow edema; MRI=magnetic resonance imaging; ESSG=European Spondyloarthropathy Study Group; ASAS=Assessment in Spondylo ArthritisinternationalSociety;axSpA=axialspondyloarthritis;mNYcriteria=modifiedNewYorkcriteria. a In 66of68patientsbaselineCRPandESRmeasurementswereavailable.ESRnormalrange:7mmformales; 12mmforfemales.CRPcutoffvalue,normalrange:<10mg/l DetectionofBMEontheSTIRandpostGdDTPAsequenceatfollowup Table 5.2 shows that in 5 (26%) of 19 patients with BME on both STIR and post GdDTPAsequencesatbaseline,BMEcouldnolongerbedetectedonSTIRsequenceat 2yearsfollowup.In2ofthese5patients,however,BMEwasstillvisibleonthepost GdDTPAsequence.Intheremaining3patientsBMEhasdisappearedonboththeSTIR andpostgddtpasequenceat2yearsfollowup.these3patientswereallhlab27 negative.in2of4patientswithbmeonthestirsequenceonlyatbaseline,bmewas nolongerpresentat2yearsfollowup.bothpatientswerehlab27negative. Five (13%) of 39 patients without signs of BME on both the STIR and postgddtpa sequences at baseline, developed BME at followup. None of these 5 patients developedbmeonthepostgddtpasequenceonly.

108 108 Chapter5 Table5.2 Baseline DetectionofBMEontheSTIRandpostGdDTPAsequencesperpatientatbaselineandfollow up. Followup * BME on both STIR andpostgddtpa (n=19) BMEonSTIRonly (n=4) NoBME (n=39) BMEonbothSTIR andpostgddtpa BMEonSTIR only BMEonpostGd GTPAonly NoBME Bone marrow edema (BME) on the Short Inversion Recovery (STIR) and postgadolinium diethylenetriaminepentaaceticacid(postgddtpa)sequencein62patientswithatleast1followupmri.* FollowupMRIat1or2years,dependingonlastavailableMRI Incremental value of the postgddtpa sequence compared with the combinationofstirsequenceandpelvicradiograph Atbaseline,31(46%)patientshadBMEonMRISIJand/orfulfilledthemNYcriteriafor radiographicsacroiliitis.noneofthese31patientshadsignsofbmeonthepostgd DTPA sequence only at baseline. Thirty (97%) of 31 patients had both at least one followupmrisijandatleastonefollowuppelvicradiograph.atfollowup,nonew patients fulfilled the mny criteria for radiographic sacroiliitis. BME on the postgd DTPAsequenceonlywasfoundin2(7%)ofthese30patientsatfollowup.However, bothpatientsalreadyfulfilledthemnycriteria. DISCUSSION The present study showed that the postgddtpa MRI sequence does not have an incrementalvalueinthedetectionofsacroiliitisinacohortofpatientswithearlyibp whowerefollowedfor2yearscomparedtothecombinationofstirmrisequenceand pelvicradiographs. Both MRI sequences can be used to detect sacroiliitis with similar efficiency, as is reflectedinthehighkappavaluesfoundinourstudy.earlierstudiesalsocomparedthe concordance between STIR and postgddtpa sequences in detecting BME on MRISIJ.[3,4] De Hooge et al. found a 100% agreement between the STIR and post GdDTPAMRIsequenceindetectingBMEonMRISIJinaprospectivecohortstudyof 127patients with chronic back pain of less than 2 years duration with onset below

109 ValueofGadoliniumcontrastenhancedMRIsequence years.[4].In8(6%)ofthe127patients,synovitisand/orcapsulitisand/orenthesitis was detected on the postgddtpa sequence, but without corresponding BME. However,thesefindingsarenotsolelyconsideredintheASAS/OMERACTdefinitionof activesacroiliitis.alsointhisstudy,itwasconcludedthatthepostgddtpasequence doesnothaveanadditionalvalueintheassessmentofactivesacroiliitisoverthestir sequence.[4] Madsen et al. found in 40 patients with established axspa, who were assessedbyanobliquetransaxialmriofthesij,thatthestirsequencecanreplacethe postgddtpasequence.[3] Therewasagreementbetweenbothimagingsequencesin 60(75%)ofthe80SIJs.WiththeSTIRsequencemoreBME,mainlyintheperipheryof structural changes, was detected. However, the authors suggested that the post GdDTPAsequencemightbesuperiortotheSTIRsequencewithrespecttodetecting smallsubcorticallesions.[3]incontrasttothesetwostudies,ourlongitudinalstudyalso comparedthemrifindingstothefindingsofpelvicradiographsovera2yearfollowup period. In a small subset of patients, BME could be detected on the postgddtpa sequence only during followup. However, these patients already fulfilled the mny criteria. This suggests that postgddtpa sequence does not provide additional diagnostic information in the detection of sacroiliitis in this cohort of patients with early IBP when information from the STIR sequence and pelvic radiographs are combined. The present study shows that in case of discordance between the STIR and post GdDTPA sequences, BME was mainly detected on the STIR sequence. In a previous study in the same cohort we have demonstrated that the combination of a positive MRIscanforBMEandapositiveHLAB27statusisassociatedwithahighlikelihoodof persistent signs of BME on MRI during followup.[13] Five patients in our cohort showed subsiding BME on the STIR sequence without BME on the postgddtpa sequence during followup, and all of them were HLAB27 negative. Three of these 5patients did not fulfil the mny criteria. Whether the MRI in these 5 patients gave falsepositive results or whether these patients showed fluctuating disease activity remainsunclearbecauseinespacpatientswerenotgivenaclinicaldiagnosisofaxspa thatcouldserveasanexternalstandard. TheMRIscoringmethodusedinthepresentstudyisacombinationoftheSPARCCand Aarhus grading method.[9,10] Both scoring systems have proved to be reliable with respecttointerreaderagreement.[14]incontrasttothesparccmethod,weapplied anunlimitednumberofslicestobeevaluated,andthesame(numberof)sliceswere scored per MRI examination per patient over time. The advantage of this method is that all qualitatively optimal slices are scored, thereby maximising the chance of

110 110 Chapter5 detecting active lesions. A general concern when scoring MRIs of the SIJs is the possibility of misalignment between two successive MRI examinations, which may cause measurement error. We ensured that scoring started and ended at the same anatomicalleveltominimizethechanceofmisalignmentbetweentwosuccessivemri examinations. Somelimitationsofthepresentstudyneedtobeaddressed.First,theMRIsetswere scored by one reader. However, the reader was experienced and the MRI scores showedhighconsistencyovertimedespiteindependentscoringofeachmrisetand blinding of the reader for time order. Second, the STIR sequence was not scored independentlyofthepostgddtpasequence.thismighthavecontributedtothehigh perpatientkappavaluesatbaselineandfollowup.third,anumberofmrisandpelvic radiographs was missing at followup. Baseline MRIs and pelvic radiographs were neverthelesscompleteandthebaselineresultsalreadyledtotheconclusionthatthe postgddtpa sequence can be omitted, a conclusion that did not change when assessingthefollowupmriscombinedwiththeinformationfrompelvicradiographs. Fourth,discordancesbetweentheSTIRsequenceandthepostGdDTPAsequencecan beduetomricoilartefacts,whichmayhavecontributedtoanoverestimationofthe presence of BME on the STIR sequence.[15] Fifth, in ESpAC, the use of NSAIDs was allowed. The actual use of NSAIDs per patient was not recorded. It is possible that continuousorondemandtreatmentwithnsaidsmayhaveledtosubsidingbme.[16] However, a number of BME lesions may also have subsided due to the natural fluctuatingcourseofthedisease.[17]last,patientsincludedinespacwererecruited via local rheumatologists, (related) medical specialties (i.e. dermatology, gastroenterology) and through family members of the local ankylosing spondylitis society.thisselectiverecruitmentmayexplaintherelativehighproportionofpatients that fulfilled a least one of the classification criteria for axspa. Furthermore, the proportionoffemalepatientsinespacisrelativelyhigh(62%)whereastheproportion of patients with a positive HLAB27 status is relatively low (46%). However, these percentages are in accordance with other cohorts that included patients with early IBP.[18,19]Nevertheless,extrapolationofthestudyfindingsshouldbedonecautiously. In conclusion, combined use of pelvic radiographs and the STIR MRI sequence is sufficient for detecting sacroiliitis in this early IBP cohort suspected for axspa. The postgddtpa MRI sequence does not have an incremental value in detecting sacroiliitis,neitheratbaselinenorduring2yearsoffollowup.

111 ValueofGadoliniumcontrastenhancedMRIsequence 111 REFERENCES RudwaleitM,VandeHeijdeD,Landewé,etal.ThedevelopmentofAssessmentofSpondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: RudwaleitM,JurikAG,HermannKG,etal.Definingactivesacroiliitisonmagneticresonanceimaging (MRI)forclassificationofaxialspondyloarthritis:aconsensualapproachbytheASAS/OMERACTMRI group.annrheumdis2009;68: Madsen KB, Egund N, Jurik AG. Grading of inflammatory disease activity in the sacroiliac joints with magnetic resonance imaging: comparison between shorttau inversion recovery and gadolinium contrastenhancedsequences.jrheumatol2010;37: De Hooge M, Van den Berg R, NavarroCompán V, et al. Magnetic resonance imaging (MRI) of the sacroiliac joints in the early detection of spondyloarthritis (SpA): No added value of gadolinium comparedtoshorttauinversionrecovery(stir)sequence.rheumatology(oxford),inpress. oussoue,jurikag.difficultiesforthedetectionofpositivesignsofsacroiliitisinspondyloarthritidesby magneticresonanceimaging(mri)ineverydayclinicalpractice.resultsfromanauditcircle(auditand reaudit).clinexprheumatol2011;29:5945. SalvadoriniG,BandinelliF,DelleSedieA,etal.Ankylosingspondylitis:howdiagnosticandtherapeutic delayhavechangedoverthelastsixdecades.clinexprheumatol2012;30:5615. HeuftDorenboschL,LandewéR,WeijersRetal.Performanceofvariouscriteriasetsinpatientswith inflammatorybackpainofshortduration;themaastrichtearlyspondyloarthritisclinic.annrheumdis 2007;66:928. Calin A, Porta J, Fries JF, et al. Clinical history as a screening test for ankylosing spondylitis. JAMA 1977;237: Maksymowich WP, Inman RD, Salonen D, et al. Spondyloarthritis research consortium of Canada magnetic resonance imaging index for assessment of sacroiliac joint inflammation in ankylosing spondylitis.arthritisrheum2005;53:7039. PuhakkaKB,JurikAG,EgundN,etal.Imagingofsacroiliitisinearlyseronegativespondyloarthropathy. Assessment of abnormalities by MR in comparison with radiography and CT. Acta Radiol 2003;44: VanderLindenS,ValkenburgHA,CatsA.Evaluationofdiagnosticcriteriaforankylosingspondylitis.A proposalformodificationofthenewyorkcriteria.arthritisrheum1984;27:3618. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33: Van Onna M, Jurik AG, Van der Heijde D, et al. HLAB27 and gender independently determine the likelihoodofapositivemriofthesacroiliacjointsinpatientswithearlyinflammatorybackpain:atwo yearmrifollowupstudy.annrheumdis2011;70: LandewéRB,HermannKG,VanderHeijdeDMetal.Scoringsacroiliacjointsbymagneticresonance imaging.amultiplereaderreliabilityexperiment.jrheumatol2005;32: AlthoffCE,FeistE,BurovaE,etal.Magneticresonanceimagingofactivesacroiliitis:dowereallyneed gadolinium?eurjradiol2009;71:2326. JarettSJ,SiveraF,CawkwellLS,etal.MRIandclinicalfindingsinpatientswithankylosingspondylitis eligible for antitumour necrosis factor therapy after a short course of etoricoxib. Ann Rheum Dis 2009;68: Barkham N, Keen HI, Coates LC,et al. Clinical and imaging efficacy of infliximab in HLAB27Positive patients with magnetic resonance imagingdetermined early sacroiliitis. Arthritis Rheum 2009;60: Van den Berg R, De Hooge M, Rudwaleit M, et al. ASAS modification of the Berlin algorithm for diagnosingaxialspondyloarthritis:resultsfromthespondyloarthritiscaughtearly(space)cohortand from the Assessment of SpondyloArthritis international Society (ASAS)cohort. Ann Rheum Dis doi: /annrheumdis Dougados M, D Agostino MA, Benessiano J, et al. The DESIR cohort: a 10year followup of early inflammatory back pain in France: study design and baseline characteristics of the 708 recruited patients.jointbonespine2011;78:

112

113 CHAPTER6 Naturalcourseofbonemarrowedemaon magneticresonanceimagingofthesacroiliacjoints inpatientswithearlyinflammatorybackpain: a2yearfollowupstudy MarloesvanOnna,AstridvanTubergen,DésiréevanderHeijde,AnneGretheJurik, RobertLandewé

114 114 Chapter6 ABSTRACT Objective Todescribethedistributionandevolutionovertimeofbonemarrowedema(BME)on MRIofthesacroiliacjoints(MRISIJ)inpatientswithrecentonsetinflammatoryback pain(ibp)suspectedforaxialspondyloarthritis(axspa). Methods A 2year followup study with annual MRISIJ was conducted in patients with IBP of duration2years.eachsijwasdividedinto4quadrantsandmriscoreswereanalyzed onaperpatientandpersijquadrantbasis.thepresenceofbmeineachsijquadrant wasrecorded.fulfilmentoftheassessmentofspondyloarthritisinternationalsociety (ASAS)axSpAcriteriawasassessedatbaselineandfollowup. Results At baseline, 68 patients (38% male; mean age 34.9±10.3 years) were included. BME was visible at baseline in 24 (35%) patients, all fulfilling the ASAS axspa criteria. Twentythreeof these 24 patients hada followup MRI.Not taking into account the baseline MRI, 3 (13%) of these 23 patients would no longer fulfill the ASAS criteria duringfollowupbecauseofsubsidingbme.fortyfour(65%)patientshadanegative baseline MRI, of whom 39 have a followup MRI available. New BME at followup meant that 3 (8%) of these 39 patients now fulfilled the ASAS criteria. At followup, baselinebmelesionssubsidedcompletelyinmean47%ofsijquadrants(range33% 71%),newBMElesionsweredetectedinmean8%ofSIJquadrants(range2%11%). Conclusion BME shows a fluctuating course in patients with early IBP suspected for axspa. This may have an impact on diagnosis making and the overall performance of the ASAS axspacriteria.

115 NaturalcourseofBMEonMRIofthesacroiliacjoints 115 INTRODUCTION Inflammation of the sacroiliac joints (SIJ) detected on magnetic resonance imaging (MRI)isacommonfindinginaxialspondyloarthritis(axSpA).[1]Indailypractice,MRIis frequently used as a diagnostic tool in those patients suspected for axspa but with normalpelvicradiographs.inadditiontopelvicradiographs,mriisincorporatedinthe imagingarmoftheassessmentinspondyloarthritisinternationalsociety(asas)classi ficationcriteriaforaxspa.[2]activeinflammatorylesionsthatcanbedetectedonmri arebonemarrowedema(bme),synovitis,capsulitisandenthesitis.ofthese,onlybme isconsideredmandatoryforfulfilmentoftheasas/outcomemeasuresinrheumatoid ArthritisClinicalTrials(OMERACT)workinggroupdefinitionofapositiveMRI.[1] To date, there is limited knowledge about the exact distribution, frequency of occurrence, and evolution over time of BME detected on MRI of the SIJ (MRISIJ), especially in patients with short duration of disease. Knowledge about the natural courseofbmeisimportantbothforresearchanddailypractice.mriisfrequentlyused asan objective measureofoutcometoexaminetheefficacyof(biological)therapyin clinical trials. However, as an example, subsiding lesions may give the impression of efficacyoftreatment,whereasthismayinfactbethenaturalcourse.also,fluctuating or subsiding BME may affect the sensitivity and specificity of MRI in detecting sacroiliitis,whichmayhamperthediagnosticprocess. The aims of this study were first to describe the distribution and frequency of occurrenceofbme,andsecondtoassesstheevolutionovertimeofbmeonmrisijin patientswithearlyinflammatorybackpain(ibp)followedfor2yearswithannualmri. METHODS Patients with IBP of less than 2 years duration were enrolled in the Early SpondyloArthritis Clinic (ESpAC) study. In this prospective cohort study, systematic clinical and radiological examinations were performed at baseline and after 1 and 2 years. A more detailed description of the study population has been reported previously.[3]foribptobepresent,patientshadtofulfilatleast4ofthefollowing5 Calin criteria: onset of symptoms before the age of 40 years, duration of back pain more than 3 months, insidious onset, morning stiffness and improvement with exercise.[4]patientswhofulfilledonly3outof5ofthecalincriteriabutreportednight pain,werealsoincluded.presenceofextraaxialmanifestationsofspawaspreferred butnotobligatory.patientswerenottreatedwithbiologicaltherapyduringtheentire

116 116 Chapter6 studyperiod.theuseofnonsteroidalantiinflammatorydrugs(nsaids)wasallowed. Fulfilment of the ASAS axspa criteria [2], modified New York (mny) criteria [5], EuropeanSpondyloarthropathyStudyGroup(ESSG)criteria[6],andAmorcriteria[7] wasassessed.thestudywasapprovedbytheethicscommitteefromthemaastricht UniversityMedicalCenter.Thestudywasconductedaccordingtotheprinciplesofthe DeclarationofHelsinki.Allpatientsgavewritteninformedconsent. The MR images were scored using a combination of the Spondyloarthritis Research Consortium of Canada (SPARCC) method and a modified version of the Aarhus MRI scoringmethod.[8,9]incontrasttotheoriginalsparccscoringmethod,therewasno maximumtothenumberofevaluatedslices,butthenumberofsliceswithinapatient was kept the same for all time points. MRIs were scored as series per patient with unknowntimesequencebyanexperiencedreaderforthepresenceofbonemarrow edema (BME) on the Short Tau Inversion Recovery (STIR) sequence. Each SIJ was dividedinto4quadrantsandthepresenceofbmewasrecordedforeachsijquadrant. The ASAS/OMERACT definition was followed for fulfilment of a positive MRI for sacroiliitis.[1] An MRISIJ is considered positive for active sacroiliitis when at least 1 activelesionthatischaracteristicofsacroiliitisispresentinatleast2successiveslices orwhen2ofsuchlesionsaredetectedin1slice.[1] AnteroposteriorpelvicradiographsoftheSIJwereobtainedandindependentlyscored according to the mny criteria in a random time order by 2 readers who were not involvedinthemrireading.[5]incaseofdisagreement,judgmentofathirdreaderwas conclusive. DescriptivestatisticswereusedtoanalysethepresenceofBMEdetectedonMRIona perpatientandpersijquadrantbasis,andtheconsequencesoffluctuationsinbmeon MRISIJ for fulfilment of the ASAS axspa criteria at baseline and at followup.[2] Generalizedestimatedequation(GEE)analysiswasusedtoinvestigateiftherewasa timetrendforthemribmelesionscoresinindividualpatients.spsssoftwareversion 18.0wasusedforallstatisticalanalyses. RESULTS Patientcharacteristics BaselineMRISIJandpelvicradiographswereavailableinall68patientsincludedinthe ESpACstudy.Table6.1showsthebaselinedemographicsandclinicalcharacteristicsof

117 NaturalcourseofBMEonMRIofthesacroiliacjoints 117 all included patients. Sixtytwo (91%) patients had at least 1 followup MRI and 44 (65%)patientscompletedbothfollowupMRIs.Sixtyfive(96%)patientshadatleast1 followup pelvic radiograph and 48 (71%) patients completed both followup pelvic radiographs.adjudicationofpelvicradiographswasconsiderednecessaryin10(15%) of68patientsbecauseofdisagreementbetweenthefirst2readers. Table6.1 Baselinecharacteristics. Characteristic Allpatients (n=68) PatientswithcompleteMRIfollow updata (n=44) Malesex 26(38) 15(34) Meanage(SD)[years] 34.9(10.3) 36.0(11.7) Mediansymptomduration(IQR)[months] 18(1224) 18(1224) HLAB27positive 31(46) 17(39) Historyofinflammatoryboweldisease 10(15) 7(16) Historyofanterioruveitis 10(15) 8(18) Historyofpsoriasis 16(24) 12(27) Historyofperipheralarthritis 19(28) 12(27) FamilyhistoryofSpA 37(54) 26(59) MeanCRP(SD)[mg/l] 9(11) 9(12) ElevatedCRP a 16(24) 10(22) MeanESR(SD)[mm] 13(15) 13(16) ElevatedESR a 24(36) 13(30) PresenceofBMEonMRI 24(35) 14(32) FulfillmentASASaxSpAcriteria 40(59) 22(50) FulfillmentmNYcriteria 15(22) 9(20) FulfillmentESSGcriteria 58(85) 39(89) FulfillmentAmorcriteria 48(71) 31(70) The values are expressed as number (percentage) of patients unless stated otherwise. ESpAC=Early Spondyloarthritis Clinic; SpA=spondyloarthritis; IQR=interquartile range; SD=standard deviation; CRP=C reactiveprotein;esr=erythrocytesedimentationrate;bme=bonemarrowedema;mri=magneticresonance imaging; ESSG=European Spondyloarthropathy Study Group; ASAS=Assessment in SpondyloArthritis internationalsociety;axspa=axialspondyloarthritis;mnycriteria=modifiednewyorkcriteria. a In66of68 patientsbaselinecrpandesrmeasurementswereavailable.esrnormalrange:7mmformales;12mm forfemales.crpcutoffvalue,normalrange:<10mg/l. Sixtysix (97%) patients fulfilled the Calin criteria at baseline. The remaining 2 (3%) patientsfulfilled3ofthe5calincriteriaandreported nightpain.atbaseline,40(59%) outof68patientsfulfilledtheasasaxspacriteria.twentytwopatientsfulfilledboth the imaging and clinical arm of the ASAS axspa criteria; 9 patients only fulfilled the imagingarmand9patientsonlyfulfilledtheclinicalarm.atbaseline,58(85%)patients fulfilledtheessgcriteria,48(71%)patientstheamorcriteria,and15(22%)patients themnycriteria.noneofthepatientsmetthemnycriteriaforthefirsttimeatfollow upvisits.

118 118 Chapter6 Theestimatedmarginal(EM)meanscoreforBMElesionsonMRIwas6.8atbaseline; 4.6at1yearfollowup;5.0at2yearsfollowup(p=0.21). BaselinedistributionandfrequencyofoccurrenceofBME.Intotal,24(35%)patientshadsignsofBMEonMRISIJatbaseline.. Table 2 shows that in all four quadrants of the SIJ, BME could be detectedonmri.therightcaudaliliacquadrantofthesijwasslightlymoreaffected than other SIJquadrants. BME was detected in the right caudal iliac quadrant in 16 (24%)outof68patientsatbaseline.IntheremainingSIJquadrants,BMEwasdetected in8to13(12%19%)outof68patientsatbaseline. EvolutionofBMEintheSIJsatfollowup Figure6.1showsthefulfilmentoftheASASaxSpAcriteriaatbaselineand at the last followup in 62 patients with at least one followup MRI and pelvic radiograph available. In 23 of 24 patients with BME at baseline, followup MRI was available.bmesubsidedcompletelyin7(30%)ofthese23patientsatthelastavailable followupmri.fourofthese7patientsfulfilledthemnycriteriaatbaselineandfollow up,andwouldthereforeremainintheimagingarmoftheasasaxspacriteriaatthe lastavailablefollowupexamination.theother3patientsdidnotfulfilthemnycriteria, were HLAB27 negative, and because BME on MRI also completely subsided, these patients wouldwithout prior knowledge of the baseline MRInot longer fulfil the ASAS axspa criteria at the last available followup examination. All these 3 patients, however, still fulfilled the ESSG and Amor classification criteria because of various combinations of other SpA features such as psoriasis, anterior uveitis, peripheral arthritisand/orapositivefamilyhistoryforspa,besidesthepresenceofibp. AtleastonefollowupMRIwasavailablein39ofthe44patientswithoutBMEonMRI atbaseline.in6(15%)ofthesepatients,bmewasdetectedonatleast1followupmri. ThreeofthesepatientsalreadyfulfilledtheimagingandclinicalarmoftheASASaxSpA criteria,becauseofradiographicsacroiliitisatbaselineandapositivehlab27status. Theother3patientsnewlyfulfilledtheASASaxSpAcriteriaatfollowup(Figure6.1). However,2ofthese3patientsalreadyfulfilledtheESSGandAmorcriteriaatbaseline.

119 NaturalcourseofBMEonMRIofthesacroiliacjoints 119 Figure6.1FulfilmentofASASaxSpA criteriaatbaselineandfollowup.

120 120 Chapter6. Table 6.2 shows that BME remained present in 53% of the SIJ quadrants(range29%to69%)atfollowupandwaspersistentmostfrequentlyinthe rightandleftcaudaliliacquadrants(table2).bmesubsidedin47%ofthesiquadrants (range 33% to 71%) and was newly detected at the last followup moment in approximately8%ofthesiquadrants(range2%to11%).mostoftheselesionswere detectedintherightandleftcaudalsacralquadrants(table6.2). Table6.2 PresenceofBMEonMRIperSIJquadrantatbaselineandfollowup. SIJquadrant Baseline LastfollowupMRI* Rightproximaliliacquadrant Rightproximalsacralquadrant Rightcaudaliliacquadrant Rightcaudalsacralquadrant Leftproximaliliacquadrant Leftproximalsacralquadrant Leftcaudaliliacquadrant Leftcaudalsacralquadrant PresenceofBME PresenceofBME AbsenceBME (yesvsno) Yes(n=10) 3(30) 7(70) No(n=52) 1(2) 51(98) Yes(n=8) 5(62) 3(38) No(n=54) 3(6) 51(94) Yes(n=16) 9(56) 7(44) No(n=46) 5(11) 41(89) Yes(n=8) 5(62) 3(38) No(n=54) 6(11) 48(89) Yes(n=11) 6(55) 5(45) No(n=51) 3(6) 48(94) Yes(n=14) 4(29) 10(71) No(n=48) 3(6) 45(94) Yes(n=13) 9(69) 4(31) No(n=49) 4(8) 45(92) Yes(n=9) 6(67) 3(33) No(n=53) 6(11) 47(89) n=62 patients.* Followup at 1 or 2 years, depending on last MRI. The values are expressed as number (percentage) of patients unless stated otherwise. SIJ=sacroiliac joint; BME=bone marrow edema, MRI=magneticresonanceimaging. DISCUSSION The present study, in a cohort of patients with early IBP suspected for axspa, demonstratedthatbmesubsidedcompletelyinapproximatelyhalfofthesiquadrants during followup and in 30% of the patients with a positive MRISIJ at baseline. On average,inlessthan10%ofthesiquadrants,bmewasnewlydetectedatsomepoint intimeduringfollowup.nottakingintoaccountthebaselinemri,13%ofthepatients wouldnotfulfiltheasasaxspacriteriaatafollowupmoment,asaconsequenceof fluctuations of BME on MRI. Another 8% of patients newly fulfilled the ASAS axspa criteriaatfollowup.

121 NaturalcourseofBMEonMRIofthesacroiliacjoints 121 Indailypractice,MRISIJisorderedwhenadiagnosisofnonradiographic(nr)axSpAis suspected on clinical grounds and conventional pelvic radiographs are normal or inconclusive.inthiscontext,subsidingorfluctuatingbmemayaffectthesensitivityof MRIindetectingsacroiliitisandsubsequentlycouldcauseadelayinthediagnosisof axspa.inaddition,forclinicaltrialsitisimportanttobeawareofthefluctuatingcourse of BME on MRI, which may in fact be the natural course instead of any treatment effect. This is illustrated by a study of Barkham et al. that evaluated the efficacy of infliximaboninflammationdetectedonmriinhlab27positivepatientswithrecent onsetibp.[10]inthetreatmentgroup,62.7%ofthebmelesionsdetectedonmrisijat baselinehadcompletelysubsidedatweek16.remarkably,intheplacebogroup,29.4% ofbmelesionsdetectedatbaselinehadalsocompletelysubsidedatweek16.[10]ina studybymarzoortegaetal.,thefrequencyofbmelesionsonmrisijwasevaluated overa1yearfollowupperiodinpatientswithearlyibp.[11]ofthe34patientswitha baselineandfollowupmri,themajority(73.5%)ofthepatientsstillhadbmeonmri SIJat1yearfollowup,buttheoveralldiseaseactivityscore,definedbytheextentof BME,haddecreased.[11]Thesefindingsareconsistentwiththeresultsofourpresent study.however,incontrasttothestudyofbarkhamandmarzoortegaetal.,ourstudy not only evaluated the presence of BME per SIJquadrant over a 2year followup period,butalsoincorporateditsexactlocalisation.inapreviousstudywithdatafrom theespacstudy,wehavedemonstratedthatbothhlab27status(oddsratio(or)8.1, 95%confidenceinterval(CI) ,p<0.001)andMRIstatusatbaseline(OR22.0, 95%CI ,p<0.001)werestronglyandindependentlyassociatedwithapositive MRISIJovertime.[12] Some MRI studies reported that the caudal parts of the SIJ are more frequently affected than the proximal parts in patients with nraxspa and ankylosing spondylitis.[13,14]thesefindingscontributetothenotionthatinflammationismerely presentinthelowersynovialdorsocaudalpartofthesij.thisassumptionissupported byhistologicstudies.[15,16]inourstudywefoundthatthepercentageofaffectedsi quadrants is rather equally distributed, although there is a slight preference for the caudaliliacsiquadrants,observedbothatbaselineandatfollowup.thiscorresponds withastudybyalthoffetal.,inwhichnodifferenceintheamountofbmebetweensi quadrants was detected in patients with axspa with a symptom duration of 5 years.[17] It is known that approximately one third of the patients with axspa according to expertsdonotfulfiltheimagingarmoftheasasaxspacriteria.[2,18]severalfactors mightcontributetothisrestrictedsensitivityoftheimagingarm.first,mriofthespine

122 122 Chapter6 isnotincludedintheimagingarmoftheasasaxspacriteria.however,upto70%of thepatientswithaxspamayhavespinalinflammation,evenintheabsenceofactive lesionsonmriofthesijs.[18]second,structuralchanges(i.e.erosions)onmriarenot considered for the definition of a positive MRISIJ.[2] Recent data, however, suggest thatassessment of erosionson MRISIJ may enhance the diagnostic utility of MRI in earlyaxspa.[19]finally,althoughmrimightbethemostsensitiveimagingmodalityfor detection of SIJ inflammation, not all inflammation is captured by MRI. A study has shownthatinbiopsiestakenfromthesijclearinflammationonhistologywasfound, whereasthiswasnotvisibleontheconcurrentlyperformedmri.[20] There are several limitations of this study that need to be addressed. First, the MR imageswerescoredbyonereaderonlyforreasonsofexperiencedreaderavailability, which may in theory influence the reliability of the data. However, this reader (a musculoskeletalradiologist,andmemberoftheasas/omeractmriworkinggroup) was highly experienced, the MRIs were scored with unknown time sequence, and blinding was preserved for clinical and laboratory findings.[1] The intraobserver agreementofourreaderwasverygoodinanotherstudythatevaluatedbmeonmri SIJs in patients with early axial spondyloarthritis, using the Danish (Aarhus) scoring method(weightedkappavalue0.96(95%confidenceinterval: )).[21]further, a high interobserver agreement among experienced readers for scoring of MRISIJs has been reported in previous studies.[8,22] These arguments reasonably justify an unbiasedacquisitionofmriscores. Second,theintervalbetween2MRIexaminationswas1year.SerialMRIexaminations atforinstance3or6monthsfollowupcouldprovidemoreinformationaboutshort term fluctuations of BME on MRISIJ. Third, the MRI scoring method used in the presentstudyisacombinationofthesparccandaarhusgradingmethodwhichisnot assuchformallyvalidated.[8,9]thedisadvantageofthesparccgradingmethodisthat amaximumof6slicesarescored.withourmodificationanunlimitednumberofslices couldbeevaluated,whileensuringthatthesame(numberof)sliceswerescoredper patientovertime.theadvantageofthismethodisthatallqualitativelyoptimalslices arescored,therebymaximisingthechanceofdetectingactivelesions.furthermore,a generalconcernwhenscoringmrisijisthepossibilityofmisalignmentbetweentwo successivemriexaminations,whichmayeventuallyresultinmeasurementerror.we haveensuredthatthescoringstartsandendsatthesameanatomicallevel,sothatthe riskofmisalignmentisminimized. Fourth,inESpAC,theuseofNSAIDswasallowed.Detailedinformationabouttheuseof NSAIDsperpatientisnotavailable,sothattheeffectsofNSAIDsonMRIscorescould

123 NaturalcourseofBMEonMRIofthesacroiliacjoints 123 not be studied. It is therefore unclear whether continuous or ondemand treatment with NSAIDs may ultimately have led to subsiding inflammation detected on MRI. Results from the second part of the Infliximab as First Line Therapy in Patients with EarlyActiveSpondyloarthritisTrial(INFAST)suggestthatitmaynotbedifferent.Inthis study, patients who continued naproxen were compared with patients who discontinuedtherapy.attheendofthefollowupperiod,completeabsenceoflesions onmrisijwasfoundinsimilarnumbersofpatientsinthenaproxenandnotreatment groups(7.5%vs.10.0%).[23]fifth,espacpatientswerenotgivenaclinicaldiagnosisof axspa. However, 40 (59%) patients fulfilled the ASAS axspa criteria at baseline. The overall sensitivity and specificity of the ASAS axspa criteria is 82.9% and 84.4%, respectively.thediagnosticperformanceoftheasasaxspacriteriaisgood,withan overall posttest probability of 89%.[2] Considering the characteristics of the ASAS axspacriteriaandthecaveatsrelatedtothediagnosticutilityofmriinaxspa,most patients included in the ESpAC that fulfil the ASAS axspa criteria, will probably also haveordevelopaclinicaldiagnosisofaxspa.unfortunately,patientswerenotactively followed after the study was terminated. Finally, ESpAC is a relatively small cohort including patients referred by (related) medical specialties (i.e. dermatology, gastroenterology)andrelativesofmembersofthelocalankylosingspondylitissociety. This recruitment strategy may therefore explain the high proportion of patients fulfillingatleastoneoftheclassificationcriteriaforspa.furthermore,theproportion ofmalepatientsinespacisrelativelylow(38%)andtheproportionofpatientswitha negative HLAB27 status is relatively high (54%). However, these percentages are in accordance with other cohorts that included patients with early IBP suspected for axspa(24,25).inthespondyloarthritiscaughtearly(space)cohort,forexample,the proportion of male patients among axspa patients is 48%, and the proportion of patients with a negative HLAB7 status is 20%.[24] In the Devenir des Spondylarthropathies Indifférenciées Récentes (DESIR)cohort these proportions are 46%and43%,respectively.[25] Inconclusion,inthiscohortof68patientspresentingwithrecentonsetIBP,BMEon MRIwasequallydistributedthroughouttheSIJs,withaslightpreferenceforthecaudal iliac quadrants. Nearly half of the BME lesions subsided during followup. These changes in MRI status are important and should be taken into account in clinical practice when evaluating patients with IBP suspected for axspa and may have an impactontheoverallperformanceoftheasasaxspacriteria.

124 124 Chapter6 REFERENCES RudwaleitM,JurikAG,HermannKG,etal.Definingactivesacroiliitisonmagneticresonanceimaging (MRI)forclassificationofaxialspondyloarthritis:aconsensualapproachbytheASAS/OMERACTMRI group.annrheumdis2009;68: RudwaleitM,vanderHeijdeD,LandewéR,etal.ThedevelopmentofAssessmentofSpondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: HeuftDorenboschL,LandewéR,WeijersR,etal.Performanceofvariouscriteriasetsinpatientswith inflammatorybackpainofshortduration;themaastrichtearlyspondyloarthritisclinic.annrheumdis 2007;66:928. Calin A, Porta J, Fries JF, et al. Clinical history as a screening test for ankylosing spondylitis. JAMA 1977;237: VanderLindenS,ValkenburgHA,CatsA.Evaluationofdiagnosticcriteriaforankylosingspondylitis.A proposalformodificationofthenewyorkcriteria.arthritisrheum1984;27:3618. Dougados M, van der Linden S, Juhlin R, et al. The European Spondylarthropathy Study Group preliminarycriteriafortheclassificationofspondylarthropathy.arthritisrheum1991;34: Amor B, Dougados M, Listrat V, et al. Are classification criteria for spondylarthropathy useful as diagnosticcriteria?revrhumengled1995;62:105. Maksymowych WP, Inman RD, Salonen D, et al. Spondyloarthritis research consortium of Canada magnetic resonance imaging index for assessment of sacroiliac joint inflammation in ankylosing spondylitis.arthritisrheum2005;53:7039. PuhakkaKB,JurikAG,EgundN,etal.Imagingofsacroiliitisinearlyseronegativespondyloarthropathy. Assessment of abnormalities by MR in comparison with radiography and CT. Acta Radiol 2003;44: Barkham N, Keen HI, Coates LC,et al. Clinical and imaging efficacy of infliximab in HLAB27Positive patientswithmagneticresonanceimagingdeterminedearlysacroiliitis.arthritisrheum2009;60: MarzoOrtegaH,McGonagleD,O'ConnorP,etal.Baselineand1yearmagneticresonanceimagingof the sacroiliac joint and lumbar spine in very early inflammatory back pain. Relationship between symptoms,hlab27anddiseaseextentandpersistence.annrheumdis2009;68: Van Onna M, Jurik AG, van der Heijde D, et al. HLAB27 and gender independently determine the likelihoodofapositivemriofthesacroiliacjointsinpatientswithearlyinflammatorybackpain:atwo yearmrifollowupstudy.annrheumdis2011;70: MucheB,BollowM,FrançoisRJ,etal.Anatomicstructuresinvolvedinearlyandlatestagesacroiliitis in spondylarthritis: a detailed analysis by contrastenhanced magnetic resonance imaging. Arthritis Rheum2003;48: Weber U, Pfirrmann CW, Kissling RO, et al. Whole body MR imaging in ankylosing spondylitis: a descriptive pilot study in patients with suspected early and active confirmed ankylosing spondylitis. BMCMusculoskeletDisord2007;8:20. François RJ, Gardner DL, Degrave EJ, et al. Histopathologic evidence that sacroiliitis in ankylosing spondylitisisnotmerelyenthesitis.arthritisrheum2000;43: Puhakka KB, Melsen F, Jurik AG, et al. MR imaging of the normal sacroiliac joint with correlation to histology.skeletalradiol2004;33:1528. Althoff CE, Sieper J, Song IH, et al. Active inflammation and structural change in early active axial spondyloarthritisasdetectedbywholebodymri.annrheumdis2013;72: SieperJ,vanderHeijdeD.Review:Nonradiographicaxialspondyloarthritis:newdefinitionofanold disease?arthritisrheum2013;65: WeberU,ZublerV,PedersenSJ,etal.DevelopmentandvalidationofanMRIreferencecriterionfor definingapositivesijmriinspondyloarthritis.arthritiscareres(hoboken)2013;65: Gong Y,ZhengN, ChenSB,etal.Tenyears'experiencewithneedle biopsyintheearlydiagnosisof sacroiliitis.arthritisrheum2012;64: ArnbakB,JensenTS,MannicheC,etal.SpondyloarthritisrelatedanddegenerativeMRIchangesinthe axialskeletonaninterandintraobserveragreementstudy.bmcmusculoskeletdisord2013;14:274.

125 NaturalcourseofBMEonMRIofthesacroiliacjoints eber U, Maksymowych WP, Jurik AG, et al. Validation of wholebody against conventional magnetic resonance imaging for scoring acute inflammatory lesions in the sacroiliac joints of patients with spondylarthritis.arthritisrheum2009;61:8939. SieperJ,LenaertsJ,WollenhauptJ,etal.Arandomised,openlabelstudytoexplorewhetherpartial remission van be maintained with naproxen or no treatment in patients with early, active axial spondyloarthritis:preliminaryresultsresultsofinfastpartii.annrheumdis2012;71suppl:248. van den Berg R, de Hooge M, Rudwaleit M, et al. ASAS modification of the Berlin algorithm for diagnosingaxialspondyloarthritis:resultsfromthespondyloarthritiscaughtearly(space)cohortand from the Assessment of SpondyloArthritis international Society (ASAS)cohort. Ann Rheum Dis 2013;72: Dougados M, d'agostino MA, Benessiano J, et al. The DESIR cohort: a 10year followup of early inflammatory back pain in France: study design and baseline characteristics of the 708 recruited patients.jointbonespine2011;78:

126

127 PARTIII Timelyidentificationofspondyloarthritisin primarycare

128

129 CHAPTER7 Generalpractitioners perceptionsoftheir abilitytoidentifyandreferpatientswithsuspected axialspondyloarthritis:aqualitativestudy MarloesvanOnna,SimoneGorter,AniekvanMeerendonk,AstridvanTubergen

130 130 Chapter7 ABSTRACT Objectives Toexploretheknowledge,beliefsandexperiencesofgeneralpractitioners(GPs)about inflammatorybackpain(ibp)andaxialspondyloarthritis(axspa)andpotentialbarriers forreferralofpatientssuspectedforaxspa. Methods A qualitative study involving semistructured interviews with GPs was conducted. Transcripts of the interviews were independently read and annotated by 2 readers. Illustrative themes were identified and a coding system to categorize the data was developed. Results TenGPs(allmen;meanage49years)wereinterviewed.Allcouldadequatelydescribe classic ankylosingspondylitis(as)andmentionedchronicbackpainand/orstiffness askeyfeatures.allgpsthoughtthatasisalmostexclusivelydiagnosedinmen.sixgps knewthatthereisadifferencebetweenmechanicalbackpain(mbp)andibp,butcould recallonlyalimitednumberofparametersindicativeofibp,suchas:awakeningnight pain(4gps),insidiousonsetofbackpain(1gp),improvementwithmovement(1gp) and(morning)stiffness(2gps).twogpsmentionedperipheralarthritisasotherspa features,nonementioneddactylitisorenthesitis.gps awarenessofassociatedextra articular manifestations was low. Most GPs expressed that (practical) referral parameterswouldbeuseful. Conclusion GPs are aware of classic, but longterm features of axspa. Knowledge about the diseasespectrumandearlydetectionis,however,limited.addressingtheseissuesin training programmes may improve better recognition of axspa in primary care. This may ultimately contribute to earlier referral, diagnosis, and initiation of effective treatmentinpatientswithaxspa.

131 GPrecognitionofSpA 131 INTRODUCTION Spondyloarthritis(SpA)comprisesagroupofinterrelatedinflammatorydisorderswith overlappingclinicalfeaturesandsharedgeneticmarkers.theestimatedprevalenceof SpA in Caucasian populations is approximately 1%, similarly to that of rheumatoid arthritis.[1] Symptom patterns and physical signs of SpA can be divided into predominantly axial involvement, with inflammatory back pain (IBP) as the most important clinical feature and predominantly peripheral involvement including peripheralarthritis,dactylitisandenthesitis.[2]extraarticularmanifestationsrelatedto axial and peripheral SpA include psoriasis, anterior uveitis and inflammatory bowel disease. Axial SpA (axspa) comprises a disease continuum, including both nonradiographic axspa and ankylosing spondylitis (AS).[3] Patients with nonradiographic axspa have similar clinical characteristics, disease activity and response to treatment as patients withestablishedas,emphasizingtheneedforearlyandcorrectdiagnosis.[4]however, thediagnosisofaxspaisoftendelayedduetotheinsidiousonset,theheterogeneous clinicalpicture,andalimitedknowledgeaboutthemanifestationsbelongingtoaxspa by general practitioners (GPs) or other referring physicians.[5] Offering tools for referral may be helpful in improving early diagnosis. Several initiatives have been performedtostudytheeffectofreferralstrategiesinprimarycare.theobjectivesof these referral programs were to identify patients with possible axspa early, make a correct diagnosis, and to provide the best possible care as early as possible.[6] However, limited knowledge of manifestations belonging to axspa might prevent successfulimplementationofthesereferralstrategiesintheprimarycaresetting. The aim of the present qualitative study is to explore by using semistructured interviewstheknowledge,beliefsandexperiencesofgpsaboutibpandaxspa,andthe potentialbarriersforreferralofpatientssuspectedforaxspa. MATERIALSANDMETHODS Studydesignandparticipants For this qualitative study, GPs, acquainted with the interviewers, without known specific interest or knowledge of musculoskeletal diseases and with various years of experience,wereinvitedforasemistructuredinterview.asemistructuredinterviewis atechniquetocollectqualitativedataaboutthetopicofinterestbycombiningopen

132 132 Chapter7 questionswiththeoptiontofurtherexploreparticularanswers.[7]thedurationofone interviewwasaboutonehour.theinterviewstookplacein2012andallinvitedgps wereworkingintheregionoflimburg,thenetherlands.thestudywasapprovedby theethicscommitteefromthemaastrichtuniversitymedicalcenter.allparticipants providedwritteninformedconsentandtherebyagreedtopresentthecollecteddata andquotesinanonymizedform. Datacollection An interview guide that consisted of both open and closedended questions was developed to secure uniform data quality and comparability. A pilot interview was conductedtoensurethatthequestionswereclearandaddressedallimportanttopics. Eachinterviewwasaudiotapedandafterwardsfullytranscribed.Eachtranscriptwas offeredtothematchinggptoreviewforvalidation. Thetopicsaddressedintheinterviewincluded: General questions: age, working experience in years as a GP, specific medical interests. Approach to patients presenting with chronic back pain, knowledge about symptomsindicativeofmechanicalbackpain(mbp)oribp,managementofback pain,motivatingfactorstoreferapatienttoarheumatologist. Perceptions and knowledge about axspa, including nonradiographic axspa and AS.Awarenessaboutdiagnosticdelay,knowledgeofextraarticularmanifestations ofaxspa. ApproachtopatientsalreadydiagnosedwithaxSpA,diseasemanagement. Awareness about treatment options and opinion of GPs about the current standardsofcareforpatientswithaxspa. Dataanalysis The transcripts were independently analyzed by 2 readers. All transcripts were repeatedly read and annotated. A coding system based on the grounded theory approach was developed by defining categories and developing a taxonomy of the data.[7]thereadersmetregularlytodiscusscodingandinterpretationofdata.incase ofdisagreement,consensuswasreachedbetweenthetworeadersafterrereadingthe specificpassageofthetranscript.whileanalyzingthedata,illustrativequotesmadeby GPswerecollected.

133 GPrecognitionofSpA 133 RESULTS Participantcharacteristics Ten of 16 invited GPs agreed to participate and were interviewed. All included GPs weremenandthemeanagewas49years(range3758years;standarddeviation(sd) 6.4years).ThemeannumberofyearsofexperienceasaGPwas20years(range1029 years; SD 6.0 years). Three GPs had a specific interest in musculoskeletal disorders. WhenGPswereaskedtoestimatethemeannumberofpatientswithASregisteredin their practice, the range of answers was between 0 and more than 10 patients (withoutfurtherspecification). Whenanalyzingthedata,anumberofthemesandpatternswereidentifiedacrossthe interviews.thesethemesandpatternsaredescribedbelowandexemplifiedinquotes (Table7.1). AbilitytodifferentiateMBPfromIBP FourGPswerenotfamiliarwiththetermsMBPandIBP(quote1).SixGPsknewthat thereisadifferencebetweenmbpandibp,butthesegpscouldrecallonlyalimited number of typical parameters to differentiate MBP from IBP. Four of these 6 GPs mentionedawakeningnightpainasatypicalfeatureofibpandconsidereditarelevant symptomthatneededattention(quote2).twogpsalsomentionedinsidiousonsetof back pain and improvement of back pain with movement as typical features of IBP. Morningstiffnesswasmentionedby2GPs.SevenGPsmentionedstiffnessoftheback astypicalforasbutdidnotelaborateonthecourseofstiffnessduringtheday. Knowledgeabouttheterms classic ASandaxSpAandawarenessabout diagnosticdelay AllGPswerefamiliarwiththetermASandmentionedbackpainand/orstiffnessofthe backasprominentfeaturesofas.threegpsalsoconsidered(severe)kyphosisasan importantfeature of AS.Noneof the GPscould give anadequatedescription ofthe termaxspa. Whenaskedabouttheageatonsetoffirstsymptoms,allGPsansweredthatsymptoms firstappearinearlyadulthood.allgpsthoughtthatasisalmostexclusivelydiagnosed in men. Two GPs thought that the delay in diagnosis was less than one year. The remainderofgpsansweredthatthedelayindiagnosiswasuptoseveralyears,without

134 134 Chapter7 furtherspecification.afewgpscommentedthatthisisprobablyduetoapatients and doctors delay(quote3). Table7.1 Illustrativequotesmadebygeneralpractitioners. Quote* 1 Ireallydonotknowthedifferencebetweenmechanicalandinflammatorybackpain.Idonotseea lotofpatientswithahistoryofinflammatorybackpain.( )Whenapatienthasbackpainforalong periodoftime,iusuallyreferthemtoarheumatologist.butitcertainlywouldnotsurprisemeif thereareseveralundiagnosedpatientswithasinmypractice. 2 Ifapatientpresentswithahistoryofbackpain,Iaskifheorshecanstillperformhousehold choresandworkrelatedduties.iaskifthepainiscontinuousornotandifthereisnightpainor painwhenwakingup.( )Whentherearesignsofawakeningnightpain,Itendtolookmore seriouslyatthesymptoms.duringthephysicalexaminationichecktherangeofmotionandthe stiffnessoftheback. 3 IthinkthatthetimebetweenfirstcomplaintsanddiagnosisofASvaries.Thereisapatientdelay, butalsoadoctordelay.whentherearefamilymemberswithas,youtendtolookmoreseriously andwillprobablyreferthispatienttoarheumatologistatanearlystage.butifthisisnotthecase HowlongitwilltakebeforeaGPwillreferapatientwithchronicbackpain?Idonotknow,months toyearsmaybe? 4 WhetherIcanmentionothersymptomsassociatedwithAS?Eyecomplaintsprobably,butIdonot thinkitisverytypical.conjunctivitismaybe?psoriasisalso,butthatisnotreallyinflammation,butit belongstoanothergroupofautoimmunedisorders.itisnotreallycleartome. 5 IftheHLAB27testispositiveornegative,itwillnotsolvethediagnosticproblem.Whenthetestis positive,youthink, OK,maybe...,butwhattodowhenthetestisnegative?Incaseofanegative testresult,thatdoesnotmeanthatthepatientdoesnothaveas.istillhavetoreferthepatientto therheumatologist. 6 WhenapatientpresentswithahistoryoflowbackpainandtherearenoabnormalitiesontheX ray,iwillreferthispatienttotheneurologist.itisveryunlikelythatireferthispatienttothe rheumatologist.providedthatlowbackpainistheonlysymptom. 7 IwanttoknowmoreabouthowtorecognizeAS.Aretherespecifictoolsordiagnostictestsyou canuseasagptomakeadiagnosisofasmoreorlesslikely?ifso,iwillperformthosetestsand consultarheumatologistoriwillreferthepatient.ialsowanttoknowmoreabouthowyoutreat patientswithas.whataretheresults? 8 IthinkthatImissthediagnosisfrequently.Yes,toomanytimes.Thereasonforthis?Probablydue tolackofknowledge. *QuotesweretranslatedfromDutch. KnowledgeaboutassociatedclinicalmanifestationsofaxSpA MostGPscoulddescribeonlyalimitednumberofclinicalfeaturesbelongingtoaxSpA. Two GPs considered peripheral arthritis as belonging to the spectrum of axspa; dactylitisandenthesitiswerenotmentionedatall.whenaskedaboutextraarticular manifestations of patients with axspa, 5 GPs mentioned anterior uveitis and 1 GP

135 GPrecognitionofSpA 135 mentioned eyecomplaints (quote4).inflammatoryboweldiseasewasmentionedby 2GPsandpsoriasisby3GPs. Useofdiagnostictestsintheprimarycaresetting None of the GPs would order a Human Leukocyte Antigen (HLA)B27 test when a patientpresentswithchronicbackpain.afewgpscommentedthatthistestshould only be ordered by the rheumatologist (quote 5). Most GPs specifically commented thattheywouldonlyorderaconventionalradiographincaseofchronicbackpain.one GPmentionedthatanormalpelvicradiographinapatientpresentingwithbackpain, would be a motivating factor to refer this patient to a neurologist and not a rheumatologist(quote6). PerceptionsaboutmanagementofaxSpA Adecreaseinpainandstiffnessofthebackandmaintainingfunctionwerejudgedas the most important treatment goals by the majority of the GPs. The use of non steroidalantiinflammatory drugs (NSAIDs) were considered an adequate treatment optionbyallgps.mostgpsalsomentionedphysicaltherapyorthatthepatientshould do homebased exercises. Five GPs indicated that antitnfalpha therapy can be prescribedtopatientswithaxspa.fourgpswereawareofthefactthatanincreased riskof(serious)infectionsisanimportantsideeffectofantitnfalphatherapy. PreferencesforeducationalprogrammesaboutaxSpA Most GPs expressed that (practical) referral parameters to decrease the delay in diagnosiswouldbeusefulinclinicalpractice(quote7).mostgpsalsowantedtoknow moreaboutthetreatmentoptions,includingantitnfalphatherapy.onegprevealed thatherecentlyvisitedaneducationaltrainingthatfocusedonaxspa.attheendof thistrainingherealizedthattherewereprobablyseveralundiagnosedpatientsinhis practice(quote8). DISCUSSION The present study demonstrated that there are several inconsistencies in the perceptionsofgpsaboutdiagnosisandmanagementofaxspa,includingas.mostgps couldprovideanadequatedescriptionabout classic ASandwereawareofthefact

136 136 Chapter7 thatthereisasubstantialdelayindiagnosis.gpsalsoknewthatthereisadifference betweenmbpandibp,butwereunabletoexplainhowtodifferentiatembpfromibp. Knowledge about the disease spectrum of axspa and associated extraarticular manifestationswaslimited.allgpswereawareofthebenefitsofphysiotherapyand NSAIDs, and half of the GPs knew that antitnfalpha therapy can be prescribed in patientswithaxspa. Chronicbackpainisacommonsymptominthegeneralpopulationanditisestimated thatin5%ofthesecasesaxspaistheunderlyingdisease.[8]inapproximately75%of thepatientswithaxspa,thechronicbackpainhasaninflammatorycharacter.several criteria sets to define IBP have been proposed, consisting of several parameters in ordertodifferentiateibpfrommbp.singleparameterswereinsufficientlypredictivein defining IBP, because they are also frequently present in patients without an inflammatory cause of their back pain.[9] Overall, the IBP criteria sets have a comparablesensitivityandspecificityofabout7580%.[911]ibphasbeentestedasa single referral parameter and as part of a composite referral strategy in several studies.[5,1214]whenpatientswerereferredbygpsbecauseofibpalone,axspawas diagnosedin1633%ofthereferredpatients.[5,1213]however,whenpatientswere referred because of IBP in combination with other parameters such as HLAB27 or sacroiliitis on imaging, axspa was diagnosed in 3556% of the referred patients.[5,12,14] KnowledgeofimportantfeaturesassociatedwithaxSpAisessentialbeforeareferral strategy can successfully be implemented in the primary care setting. Six GPs in our studycouldrecallonlyafewitemsindicativeofibpand4gpswerenotfamiliarwith thetermsmbpandibp.thiswasalsoobservedinastudybyjoisetal.[15]only5%of GPs in their study could identify all parameters indicative of IBP when a list of pre specifiedresponsechoiceswaspresentedtothem.furthermore,recentstudieshave shownthatthedegreeofagreementbetweenreferringphysicians(includinggps)and rheumatologistswhenevaluatingibpinpatientswithsuspectedaxspaispoor(kappa valuesbetween ).[5,16]EducatingGPsaboutthefullrangeofparameters indicativeofibpthereforeseemsthefirststepbeforeibpcansuccessfullybeusedina referral tool. The term axial spondyloarthritis will also increasingly be used in correspondence from rheumatologists to GPs. It istherefore importantto makegps alsofamiliarwiththisnewterminology. In the present study, GPs could recall only a limited number of extraarticular manifestations associated with axspa. In a some cases, GPs mentioned eye complaints or skinproblems.dactylitisandenthesitiswerenotatallmentionedby

137 GPrecognitionofSpA 137 the GPs in our study. Jois et al. also investigated the recognition of extraarticular manifestations of SpA by GPs.[15] Psoriasis, inflammatory bowel disease and uveitis wererecognizedasanextraarticularmanifestationbyrespectively96%,68%and60% ofgps,whichisahigherproportionthaninoursmallsizedstudy.[15]however,inour study openended questions were used, which probably resulted in lower response ratesthanthesurveyusedinthestudyofjoisetal.allgpsinthepresentstudyalso indicated thatas isalmostexclusivelydiagnosedin men. Severalrecent studies that included patients with undifferentiated and nonradiographic axspa, however, demonstrated that the gender ratio is more equally distributed.[1719] Male gender has however been found to be a risk factor for developing radiographic sacroiliitis.[20,21] Furthermore, patients with radiographic sacroiliitis have in general higher inflammatory markers than patients with nonradiographic axspa.[4,22] IncreasingawarenessamongGPsthataxSpAisequallypresentinfemalesandmales, andmakingthemawareofthe SpAconcept,whichincludesaxial,butalsoperipheral andextraarticularmanifestations,willlikelyfacilitatereferralandtimelydiagnosis. HalfofGPsinthepresentstudywereawarethatthetherapeuticarmamentariumin patients with axspa is broadened with the introduction of antitnfalpha therapy. WhenGPswereaskedaboutthesideeffectsofantiTNFalphatherapy,6GPswerenot aware of the higher risk of (serious) infections. Collaboration and comanagement together with the rheumatologist is essential in managing patients with axspa. Educationabout(thesideeffectsof)antiTNFtherapyisthereforeanimportantstepto maintainandimprovethegeneralhealthstatusofapatientwithaxspa. Ingeneral,thelevelofknowledgeaboutaxSpAwaslow.NoneoftheGPscouldprovide a specific reason for this lack of knowledge. Possible explanations are relatively low attentiontothistopicinmedicalschooloratcontinuousmedicaleducation,andthe largeemphasisonanonspecificcauseofchronicbackpain.[23] Therearelimitationsinthispresentstudythatneedtobeaddressed.Thedesignofthe studyisqualitativeandthenumberofgpsincludedissmall.furthermore,onlymale andexperiencedgpswereincludedinthisstudy.severalfemalegpswereasked,but they unfortunately declined to participate. Logistically, it was extremely difficult to includerecentlyqualifiedgps,becauseinthenetherlandsalmostnoneofthemhave ownpractices.wecannotruleoutthatselectionbiasorknowledgebiashasoccurred. Thismaylimitreproducibilityofresultsandtheabilitytogeneralizethemtoawider population.however,themaingoalofthisstudywasnottoextrapolatethecurrent findingstoallgps,buttoexplorethelevelofknowledgeandawarenessthatprobably

138 138 Chapter7 needattentioninfutureeducationalprogrammes.furthermore,theoreticalsaturation wasreachedwiththisnumberofgps. Inconclusion,mostGPswerefamiliarwith classic butlongtermfeaturesofaxspa. KnowledgeaboutparametersindicativeofIBPandawarenessaboutthefullrangeof SpAfeatures,includingtheassociatedextraarticularmanifestations,waslimited.The diseasespectrumandmanagementofaxspahavesubstantiallychangedoverthelast few years. Educating GPs about the leading presenting symptoms of axspa and providinginformationaboutextraarticulardiseasemanifestationsandmanagementof axspa,willplayapivotalroleinthesuccessfulreferralofpatientswithsuspectedaxspa bygps.thismayultimatelycontributetoearlierinitiationofeffectivetreatmentand theimprovementofqualityoflife.

139 GPrecognitionofSpA 139 REFERENCES 1. StolwijkC,BoonenA,vanTubergenA,etal.Epidemiologyofspondyloarthritis.RheumDisClinNorth Am2012;38: Rudwaleit M, Khan MA, Sieper J. The challenge of diagnosis and classification in early ankylosing spondylitis:doweneednewcriteria?arthritisrheum2005;52: RudwaleitM,vanderHeijdeD,LandewéR,etal.ThedevelopmentofAssessmentofSpondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: RudwaleitM,HaibelH,BaraliakosX,etal.Theearlydiseasestageinaxialspondylarthritis:resultsfrom thegermanspondyloarthritisinceptioncohort.arthritisrheum2009;60: PoddubnyyD,VahldiekJ,SpillerI,etal.Evaluationof2screeningstrategiesforearlyidentificationof patientswithaxialspondyloarthritisinprimarycare.jrheumatol2011;38: Rudwaleit M, Sieper J. Referral strategies for early diagnosis of axial spondyloarthritis. Nat Rev Rheumatol2012;8: PopeC,MaysN.Qualitativeresearchinhealthcare.BlackwellPublishing, UnderwoodMR,DawesP.Inflammatorybackpaininprimarycare.BrJRheumatol1995;34: SieperJ,vanderHeijdeD,LandewéR,etal.Newcriteriaforinflammatorybackpaininpatientswith chronic back pain: a real patient exercise by experts from the Assessment of SpondyloArthritis internationalsociety(asas).annrheumdis2009;68: Rudwaleit M, Metter A, Listing J, et al. Inflammatory back pain in ankylosing spondylitis: a reassessment of the clinical history for application as classification and diagnostic criteria. Arthritis Rheum2006;54: Calin A, Porta J, Fries JF, et al. Clinical history as a screening test for ankylosing spondylitis. JAMA 1977;237: BrandtHC,SpillerI,SongIH,etal.Performanceofreferralrecommendationsinpatientswithchronic backpainandsuspectedaxialspondyloarthritis.annrheumdis2007;66: Hermann J, Giessauf H, Schaffler G, et al. Early spondyloarthritis: usefulness of clinical screening. Rheumatology(Oxford)2009;48: Sieper J, Srinivasan S, Zamani O, et al. Comparison of two referral strategies for diagnosis of axial spondyloarthritis:therecognisinganddiagnosingankylosingspondylitisreliably(radar)study.ann RheumDis2013;72: JoisRN,MacgregorAJ,GaffneyK.Recognitionofinflammatorybackpainandankylosingspondylitisin primarycare.rheumatology(oxford)2008;47: LópezGonzález R, HernándezSanz A, AlmodóvarGonzález R, et al. Are spondyloarthropathies adequatelyreferredfromprimarycaretospecializedcare?reumatolclin2013;9: HeuftDorenboschLH,LandewéR,WeijersR,etal.Performanceofvariouscriteriasetsinpatientswith inflammatorybackpainofshortduration;themaastrichtearlyspondyloarthritisclinic.annrheumdis 2007;66: van den Berg R, de Hooge M, Rudwaleit M, et al. ASAS modification of the Berlin algorithm for diagnosingaxialspondyloarthritis:resultsfromthespondyloarthritiscaughtearly(space)cohortand from the Assessment of SpondyloArthritis international Society (ASAS)cohort. Ann Rheum Dis 2013;72: Dougados M, d'agostino MA, Benessiano J, et al. The DESIR cohort: a 10year followup of early inflammatory back pain in France: study design and baseline characteristics of the 708 recruited patients.jointbonespine2011;78: VanTubergenA,WeberU.Diagnosisandclassificationinspondyloarthritis:identifyingachameleon. NatRevRheumatol2012;8: Feldtkeller E, Bruckel J, Khan MA. Scientific contributions of ankylosing spondylitis patient advocacy groups.curropinrheumatol2000;12: KiltzU,BaraliakosX,KarakostasP,etal.Dopatientswithnonradiographicaxialspondylarthritisdiffer frompatientswithankylosingspondylitis?arthritiscareres(hoboken)2012;64: BalaguéF,MannionAF,PelliséF,etal.Nonspecificlowbackpain.Lancet2012;379:48291.

140

141 CHAPTER8 Educationimprovesreferralofpatients suspectedofhavingspondyloarthritisbygeneral practitioners.astudywithunannounced standardizedpatientsindailypractice MarloesvanOnna,SimoneGorter,BasMaiburg,GerrieWaagenaar, AstridvanTubergen

142 142 Chapter8 ABSTRACT Objectives Toevaluatethepracticeperformanceofgeneralpractitioners(GPs)andGPresidentsin recognising and referring patients suspected for having axial or peripheral spondyloarthritis (SpA), and to investigate the influence of education on this performance. Methods GP(residents)werevisitedin2roundsbystandardisedpatients(SPs)simulatingaxial SpA,peripheralSpAorcarpaltunnelsyndrome(CTS)withinbetweenaneducational intervention on SpA for part of the participants. Participants were unaware of the nature of themedicalproblem andstudypurpose. CTS was included as diversionary tactic.theprimaryoutcomewas40%improvementin(considering)referralofthesps withspatotherheumatologistaftereducation.secondaryoutcomesincludedordering additional diagnostic tests, correct recognition of SpA and identification of variables contributingtothis. Results Sixtyeight participants (30 GPs and 38 GP residents) were included, of which 19 receivededucation.theprimaryoutcomewasmet.asignificantlyhigherproportionof GP (residents) from the intervention group referred patients to the rheumatologist comparedwiththecontrolgroupaftereducation(changescores,axialspa+71%vs. +15% (p<0.01); peripheral SpA +48% vs. 0% (p<0.001)). Participants who received education,morefrequentlycorrectlyrecognisedspacomparedwithcontrols(change scores,axialspa+50%vs.5%(p<0.001);peripheralspa+21%vs.0%(p=0.01). Conclusions RecognitionandreferralofpatientssuspectedforhavingSpAbyGP(residents)islow, but targeted education markedly improved this. This supports the development of educational initiatives to improve recognition of SpA and hence referral to a rheumatologist.

143 RecognitionandreferralofpatientssuspectedofhavingSpA 143 INTRODUCTION Musculoskeletalcomplaintsaccountfor20%ofallconsultationsinprimarycare.[13] Thechallengeforgeneralpractitioners(GPs)istofilterpatientswithahighsuspicionof an inflammatory rheumatic disorder.[4] Insufficient knowledge might result in a diagnostic delay, which subsequently may have a negative impact on physical functioning,socialparticipationandqualityoflifeinanindividualpatient.[5,6]among allrheumaticdiseases,spondyloarthritis(spa)hasthelongestdiagnosticdelay,which maybeupto10yearsorevenlonger.[7,8] ManySpApatientsarenotadequatelyrecognised,asillustratedbyastudyinprimary carewhere24%ofthepatientswithchroniclowbackpainthatstartedbeforetheage of 45 years, were classified as having axial SpA after careful evaluation.[9] It is important to obtain an early diagnosis in order to tailor treatment to the individual needs of a patient and to prevent a debilitating disease course.[10] A short disease duration before initiation of treatment is also a good predictor of achieving a major clinicalresponseontreatment.[1113] Inordertoimprovetimelydiagnosis,severalreferraltools,whichincludecharacteristic featuresofspa,havebeenproposedforaxialspa.[1418]applicationofareferraltool increases the probability of a disease in referred patients from 5% to 3345%.[18] However, for successful implementation of such a tool, knowledge about SpA in primary care is essential.[19,20] A qualitative study involving GPs showed that GPs wereawareof classic,butlongtermfeaturesofaxialspa,i.e.hyperkyphosisanda bamboospine.knowledgeabouttheentirediseasespectrum,includingearlydisease, extraarticularmanifestationsandothercharacteristicspafeatures,waslimited.[21] We hypothesised that education with special focus on SpA might improve the recognitionofspafeaturesandearlyreferralofpatientssuspectedforhavingspa.the objectivesofthisstudyweretoevaluate(1)thecurrentpracticeperformanceofgps and GP residents in recognizing and (considering) referral of patients suspected for having axial or peripheral SpA, and (2) to assess the influence of education on this performance, by using unannounced standardised patients (SPs) who visit GP (residents)intheirownpractices.

144 144 Chapter8 METHODS Studydesignandparticipants This study is a prospective controlled multicenter educational intervention study in primary care. GP residents and their supervising GPs were recruited through the department of General Practice from the Maastricht University Medical Centre (MUMC). Every trimester, 12 group(s) of 1012 GP residents enter the final year of their residency. Once a week, they meetat the MUMC for training. We used this structure for providing an educational programme, the intervention, to half of the groups. Each group was alternately assigned to either the intervention group or the controlgroup.thisallocationstrategywasonlyappliedtothegpresidents.thegroups didnothavedirectcontacthourswitheachother.asimilartrainingstructurewasnot available for GPs. It was logistically not possible to organize an extra training conference for GPs without revealing the topic of interest in advance. Therefore, despitegpsmaybesupervisingaresidentwhoreceivedintervention,wedecidedto assignallgpstothecontrolgroup. The ethics committee of the academic hospital Maastricht considered this study as evaluation and improvement of daily clinical practice. No further approval was required. The GP (residents) were informed about SPs visiting their practice for evaluation of using SPs in daily practice and education, and were asked to sign informedconsent.nofurtherspecificationwasprovidedonthepurposeofthestudy andthenatureofthemedicalproblems,norweretheyinformedthattheeducation wasrelatedtothestudywithunannouncedsps. TheSPencounterstookplace3monthsbeforeand3monthsaftertheintervention. ThestudystartedinSeptember2012,andwasendedprematurelyinMay2014. Standardisedpatients SPs,recruitedfromapoolofSPsworkingattheMaastrichtUniversitymedicalschool andamongclerkships,hadtomeetthefollowingcriteria:stablehealth,abilitytoplay theroleandtofilloutthecasespecificchecklists,noconfoundingphysicalsymptoms andsufficienttimeavailableforthevisits.[22] AllSPsweretrainedtosimulateonecase.Two2hourtrainingsessionswereorganised andguidedbyseveralgpsandrheumatologists,duringwhichthespsweretrainedin playingtheirrole,andhowtobehaveduringthephysicalexamination,inavalidand reliableway.gpsandarheumatologistnotinvolvedinthedevelopmentofthecases

145 RecognitionandreferralofpatientssuspectedofhavingSpA 145 judgedwhetherthespsimulatedrealistically.closeattentionwaspaidoncompletion of the checklist to secure uniform data quality and comparability. Discrepancies in checklist rating scores were discussed.[22] Based on good reproducibility demonstratedinpreviousstudies,andafterthisthoroughtraining,weassumedgood representationofthecasesbythesps.[2325] AllparticipantsfacedtwocasesofaxialSpA,twocasesofperipheralSpA,andonecase ofcarpaltunnelsyndrome(cts).figure8.1showsanexampleofapredefinedschedule for two participants. The CTS case was included as a diversionary tactic, preventing premature identification of the objectives of the study, but was also considered as common knowledge. Recognition of CTS by the far majority of participants was expectedandthereforeonlysimulatedduringround1.eachspacasewassimulatedby amaleandafemalesp,inrandomorder,accordingtoapredefinedschedule(figure 8.1showsanexample).AshortdescriptionoftheincludedcasesisprovidedinBox8.1. TheSPswereunawarewhichparticipantsreceivededucation.TheSPsreceivedasmall allowanceforeveryvisit. Figure8.1 Exampleofapredefinedschedulefor2participants.SpA=spondyloarthritis. Practicevisitsandchecklist Atthepracticevisit,theSPsidentifiedthemselvesasanSP,withoutprovidingfurther information. The duration of one consultation was 1015 minutes, corresponding to thestandardconsultationtimebyagp(resident). Afterthevisit,theSPimmediatelycompletedthecasespecificchecklistreportingthe activitiesofparticipantsduringthevisit,whichconsistedof:

146 146 Chapter8 Disease related items (e.g. onset of symptoms, presence of low back pain, family medicalhistory); Itemsonphysicalexamination(e.g.ofthejointsand/orback). The GP (residents) indicated for this specific case which additional diagnostic investigationstheywouldhaveordered,whichmedicationtheywouldhaveprescribed (ifany),andwhetherreferraltoanotherhealthcareprofessional(andwhich)wouldbe advised.participantsalsorankedtheirdifferentialdiagnosis,from1(mostlikely)to3 (lesslikely).thespswereresponsibleforcollectingandreturningallformstothestudy coordinator. Box8.1 Summaryofincludedcases,simulatedbystandardizedpatients. EarlyaxialSpA: Case1a:a27yearoldmale/female,sufferingfromchronicbackpainwithaninflammatorycharacter sinceoneyear.he/shehasahistoryofachillestendinitis.physicaltherapyhasalimitedeffectinbackpain relief.thepatientvisitsthegp(resident)becausethebackpainisnowalsopresentinthethoracicspine. AnaunthasCrohn sdisease. Case1b:a26yearoldmale/femalewithchronicbackpainwithaninflammatorycharactersince 1.5years.Therearealsosymptomsofanteriorchestwallpain.Physicaltherapyhasalimitedeffecton backpainrelief.thepatientvisitsthegp(resident)becauseofprogressiveworkdisability.abrotherhas psoriasis. EarlyperipheralSpA: Case2a:a27yearoldmale/female,whopresentswithapainfulandswollenmiddlefingeroftheright handwithmorningstiffnesssinceafewweeks.thesphandsoveraphotographtothegp(resident), showingdactylitisoftheaffectedfinger.he/shehasahistoryofkneearthritisthreeyearsagothat resolvedwithnonsteroidalantiinflammatorydrugs(nsaids).themotherhaspsoriasis. Case2b:a26yearoldmale/female,whopresentswithandapainfulandswollensecondtoeoftheright footwithmorningstiffnesssinceafewweeks.thesphandsoveraphotographtothegp(resident), showingdactylitisoftheaffectedtoe.thepatientexperiencedsimilarcomplaintsofafingeraboutone yearago.thebrotherhaspsoriasis. CTS: Case3:a50yearoldmale/female,withatinglingandburningsensationoftheindex,middleandring fingersincethreemonths.thesymptomsareworstatnight.flickingthewristgivessymptomrelief. Educationalintervention The interactive 3hour casebased educational programme took place at the department of General Practice of the MUMC. Three topics, were presented and discussedbytworheumatologists: Diagnosisandmanagementofgout(duration45minutes);

147 RecognitionandreferralofpatientssuspectedofhavingSpA 147 Axial and peripheral SpA (duration 90 minutes), i.e. concept and epidemiology of SpA,historytakingandphysicalexaminationofpatientssuspectedforhavingSpA, andcriteriaforreferralofthesepatientstotherheumatologist. Safety considerations for biologic therapy (i.e. antitumor Necrosis Factor (TNF) alphatherapy;duration45minutes). Printedmaterials,includingSpAfeatures,weresuppliedtotheparticipantssupporting selfdirectedlearningafterthetraining. Studyoutcomes OurprimaryoutcomewasreferralorconsideringreferraloftheSPtoarheumatologist by the GP (resident). We decided to combine both referral and considering referral, because GPs may spread diagnostic interventions over several consultations or only refertosecondarycarewhencomplaintsfailtoresolvewithinafewweeksafterthe first consultation. Secondary outcomes included (1) correct recognition of axial SpA, peripheral SpA and CTS respectively by the GP (resident), (2) ordering of additional diagnostictests,(3)identificationofvariablescontributingtocorrectrecognitionofspa orcts(gpversusgpresident,andgenderofthesps). Statisticalanalysis Thesamplesizecalculationwasbasedontheprimaryoutcome,(considering)referral ofthesptotherheumatologist.weestimatedthat20%ofthespswouldbereferred without education and aimed at increasing this by 40%. In order to detect a 40% differenceinthechangescoresbetweentheinterventiongroupandcontrolgroupin the proportion of SPs referred to the rheumatologist, 23 complete pre and post educationspencounterswereneededpergroup(80%power,alphaof0.05). Descriptiveanalyseswereusedforthedemographicdata.Chisquaretestsandfisher s exacttests,asappropriate,wereusedtoanalysetheprimaryandsecondaryendpoints. The difference in change scores between the intervention group and control group with regard to (considering) referral of the SP and correct recognition of SpA was compared with the MannWhitney U test. Withingroup changes in referral and recognitionofspabeforeandaftereducationwereanalysedwithmcnemartests.only participants that completed both rounds of SP encounters were included in these analyses. Descriptive analysis was used for investigating which diagnoses were mentionedbygp(residents)andthefrequencyoforderingadditionaldiagnostictests bygp(residents).spsssoftware20.0wasusedforallanalyses.

148 148 Chapter8 RESULTS Participantcharacteristics Intotal,117GPsresidentsandtheirsupervisingGPswereinvited,ofwhich68(38GP residentsand30gps)participatedinthestudy.reasonsfornonparticipationwerenot collected. The study was ended prematurely, because many GP (residents) declined participation,andthechancethatnewlyenrolledgpresidentscameintodirectcontact withgp(residents)thatalreadyparticipatedwasconsideredhigh.theapriorisample sizewasthereforenotpursued. Baseline characteristics of the participants are shown in Table 8.1. Three (4%) participantsmentionedaninterestinmusculoskeletaldisorders. In total, 256 SP encounters took place, excluding the CTS cases. Both rounds were completedby61(90%)and59(87%)participantsfortheaxialspaandperipheralspa case, respectively. Reasons for incomplete SP visits were: illness (n=6), unable to scheduleanappointmentwithinthegiventimeframe(n=4),latearrivalbyspdueto traffic problems (n=3), GP left for a medical emergency (n=2), and maternity leave (n=1). Table8.1 Characteristicsoftheparticipatinggeneralpractitionersandresidents. GPresidents (n=38) GPs (n=30) Allparticipants (n=68) Age[years] 28(1.6) 52(5.9) 39(12.9) Male 12(32%) 24(80%) 36(53%) Workingexperience,includingtraining[years] 2(0.4) 22(7.2) 11(10.9) Weeklyconsultations[number] 62(15.6) 107(26.5) 82(30.6) Specificinterestinmusculoskeletaldisorders 2(7%) 1(3%) 3(4%) Thevaluesareexpressedasmean(standarddeviation)orasnumber(%).GP=generalpractitioner. Axialspondyloarthritis. In the first round of SP encounters, 6% of the participants in the intervention group (n=18) and 10% of participants in the control group (n=43) referred or considered referral of thesp totherheumatologist (Figure 8.2).Participantswhoreceivedtheeducationalprogrammeclearlymoreoftenreferred

149 RecognitionandreferralofpatientssuspectedofhavingSpA 149 or considered referral in the second round of SP encounters than controls (changes scores:+71%vs.+15%(p<0.001);figure8.2). Figure8.2 Referral of standardized patients simulating axial and peripheral spondyloarthritis to a rheumatologist.. In the first round of SP encounters, 4 (22%) out of 18participants in the intervention group and 8 (19%) out of 43 participants in the control group ranked axial SpA as their no. 1 diagnosis. Nonspecific back pain was most frequently ranked as no. 1 diagnosis by 10 (56%) out of 18 participants in the interventiongroupand31(72%)outof43participantsinthecontrolgroup(table82). In total, 34 (56%) out of 68 participants ranked axial SpA as no. 1, 2 or 3 in their differentialdiagnosisbeforeeducation. Inthesecondround,theinterventiongroupmorefrequentlyrankedaxialSpAastheir no. 1 diagnosis (round 1: 22% vs. round 2: 72% (p=0.01); Table 8.2), which was statistically significantly different from the control group (changes scores: +50% interventiongroupvs.5%controlgroup,(p<0.001);table8.2).inthesecondround, nonspecificbackpainremainedtheno.1diagnosisin74%oftheparticipantsfromthe controlgroup(table8.2). IntheinterventiongrouplessHLAB27testswereordered in the second round, whereas in the control group the opposite was seen (changes scores:22%interventiongroupvs.+12%controlgroup,(p=0.01);table8.2).

150 150 Chapter8 Table8.2 Diagnosisandmanagementinstandardizedpatientssimulatingaxialspondyloarthritis. Educationalgroup(n=18) Controlgroup(n=43) PValue (change scores, between groups) PValue (within group) Change score(%) Round2 (%) Round1 (%) PValue (within group) Changescore (%) Round2 (%) Round1 (%) < NA 2(5) 1(2) 2(5) 1(2) 0(0) 6(14) 32(74) 4(9) 1(2) 0(0) 8(19) 31(72) 2(5) 2(5) 0(0) (50) 5(28) 2(11) 1(6) 1(6) 13(72) 5(28) 0(0) 0(0) 0(0) 4(22) 10(56) 2(11) 1(6) 1(6) Number1diagnosis AxialSpA Nonspecificbackpain Sacroiliacjointdysfunction Herniatednucleipulposi Poorposture Additionaldiagnostictests Laboratorytests(general) 5(28) 6(33) 1(5) (30) 18(42) 5(12) HLAB27test 6(33) 2(11) 4(22).13 4(9) 9(21) 5(12) Pelvicradiograph 3(17) 3(17) 0(0) (9) 8(19) 4(10) Management NSAIDsprescribed 4(22) 5(28) 1(5) (19) 8(19) 0(0) FollowupconsultationwithGP 14(78) 13(72) 1(6) (67) 30(70) 1(3) (resident)arranged SpA=spondyloarthritis;SPs=standardizedpatients;GP=generalpractitioner;HLAB27=HumanLeukocyteAntigenB27;NSAIDs=nonsteroidalantiinflammatorydrugs; NA=notassessed.Valuesareexpressedasnumber(percentage)ofparticipants.Numbersmaynotaddupduetorounding.NcNemartestswereusedwithingroups andmannwhitneyutestswereusedbetweengroups.

151 RecognitionandreferralofpatientssuspectedofhavingSpA 151 NodifferencebetweentheGPsand GPresidentcharacteristicswithregardtocorrectrecognitionofaxialSpAwasfound (datanotshown).inaddition,maleandfemalespswereequallyconsideredtohave axialspa(datanotshown). Peripheralspondyloarthritis In the first round, participants in both the intervention(n=19)andcontrolgroup(n=40)referredthespstotherheumatologistin 5%ofcases(Figure8.2).Theinterventiongroupreferredorconsideredreferralmore frequentlycomparedtothecontrolgroupinthesecondround(changescores:+48% vs.0%(p<0.01);figure8.2). In the first round of SP encounters, 2 (11%) out of 19participants in the intervention group and 4 (10%) out of 40 participants in the controlgrouprankedperipheralspaastheirno.1diagnosis(table8.3).allspecified this as reactive arthritis. Two participants ranked psoriatic arthritis in their differentialdiagnosisasno.2or3. Arthritisnototherwisespecified wasrankedby mostparticipantsasno.1diagnosis(5(26%)participantsintheinterventiongroupand 15(38%)participantsinthecontrolgroup;Table8.3). Four (21%) of the 19 participants from the intervention group correctly recognized spondyloarthritis or psoriaticarthritis aftereducation,comparedwithnoneofthe participantsinthecontrolgroup(changescores:+21%vs.0%(p=0.01);table8.3). Nodifferencesbetweentheinterventionandcontrolgroup regardingorderinglaboratoryanddiagnosticimagingtestswerefoundinbothrounds ofspencounters(table8.3). NodifferencebetweentheGPs andgpresidentswithregardtocorrectrecognitionofperipheralspawasfoundinthe first round of SP encounters. However, in general, GPs ordered more additional diagnostictestsinbothroundsofspencounters(table8.4). InthefirstroundofSPencounters,goutwasmoreoftenrankedasno.1diagnosisin malethaninfemalesps(malesps:8(26%)outof31diagnoses;femalesps:1(4%)out of28diagnoses(p=0.03)).

152 152 Chapter8 Table8.3 Diagnosisandmanagementofstandardizedpatientssimulatingperipheralspondyloarthritis. Educationalgroup(n=19) Controlgroup(n=40) PValue (changescores, betweengroups) PValue (withingroup) Change score(%) Round2 (%) Round1 (%) PValue (withingroup) Change score(%) Round2 (%) Round1 (%) NA NA (0) 1(2) 0(0) 0(0) 1(3) 4(10) 3(8) 0(0) 0(0) 1(3) 4(10) 4(10) 0(0) 0(0) 0(0) NA 3(16) 1(6) 3(16) 1(5) 0(0) 5(26) 1(5) 3(16) 1(5) 0(0) 2(11) 2(11) 0(0) 0(0) 0(0) Number1diagnosis PeripheralSpA Reactivearthritis Spondyloarthritis Psoriaticarthritis Tenosynovitis Arthritis,nototherwisespecified 5(26) 8(42) 3(16).45 15(38) 16(40) 1(2) Rheumatoidarthritis 5(26) 3(16) 2(10).69 2(5) 5(13) 3(8) Gout 1(5) 1(5) 0(0) (20) 6(15) 2(5) Sprain 4(21) 1(5) 3(16).25 6(15) 5(13) 1(2) Skinornailinfection/insectbite 2(11) 0(0) 2(11).50 3(8) 1(3) 2(5) Trauma 0(0) 1(5) 1(5) (3) 1(3) 0(0) Otherornodifferentialdiagnosis 0(0) 1(5) 1(5) (3) 2(5) 1(2) Additionaldiagnostictests Laboratorytests(general) 8(42) 5(26) 3(16).38 17(43) 18(45) 1(3) HLAB27test 0(0) 0(0) 0(0) NA 0(0) 0(0) 0(0) NA 1.00 IgMRFand/orACPAtest 5(26) 2(11) 3(16).38 11(28) 7(18) 4(10) Radiographofthehandorfoot 0(0) 1(5) 1(5) (10) 8(20) 4(10) Management NSAIDsprescribed 13(68) 13(68) 0(0) (53) 22(55) 1(3) (53) 10(53) 0(0) (58) 25(63) 2(5) Arrangedfollowupconsultationwith GP(resident) SpA=spondyloarthritis;SPs=standardizedpatients;GP=generalpractitioner;HLAB27=HumanLeukocyteAntigenB27;NSAIDs=nonsteroidalantiinflammatorydrugs; RF=rheumatoidfactor;ACPA=anticitrullinatedproteinantibody;NA=notassessed.Valuesareexpressedasnumber(percentage)ofparticipants.Numbersmaynotadd upduetorounding.ncnemartestswereused withingroupsandmannwhitneyutestswereusedbetweengroups.

153 RecognitionandreferralofpatientssuspectedofhavingSpA 153 Table8.4 Listofdiagnosesandmanagementinpatientssuspectedforcarpaltunnelsyndrome. GPresident (n=37) GP (n=28) PValue Allparticipants (n=65) Number1diagnosis CTS Osteoarthritis Sprain 30(81%) 1(3%) 6(16%) 24(86%) 0(0%) 4(14%) (83%) 1(2%) 6(16%) Additionaldiagnostictests Radiographyofthehand 0(0%) 2(5%).50 2(3%) EMG 3(11%) 3(8%) (9%) Management NSAIDsprescribed 3(8%) 3(11%).52 6(9%) Localinjectionwithcorticosteroids 2(7%) 8(22%).17 10(15%) Splint 7(25%) 15(41%).29 22(34%) FollowupconsultationwithGP 10(35%) 21(57%).13 31(47%) (resident)arranged Referraltoneurologist 1(4%) 2(5%).57 3(5%) CTS=carpal tunnel syndrome; EMG=electromyography; NSAIDs=nonsteroidal antiinflammatory drugs; GP=generalpractitioner. Recognitionandmanagementofcarpaltunnelsyndrome Asexpected,CTSwasrankedastheno.1diagnosisby54(83%)outof65participants, and by 61 (90%) in their top 3 (Table 8.5). No differences between the GP and GP residentorgenderofthespregardingrankingctsasno.1diagnosiswerefound.also the management and followup of the CTS case were similar for the GP and GP resident. Table8.5 Listofdiagnosesandmanagementinpatientssuspectedforcarpaltunnelsyndrome. GPresident (n=37) GP (n=28) pvalue Allparticipants (n=65) Number1diagnosis CTS Osteoarthritis Sprain 30(81%) 1(3%) 6(16%) 24(86%) 0(0%) 4(14%) (83%) 1(2%) 6(16%) Additionaldiagnostictests Radiographyofthehand 0(0%) 2(5%) (3%) EMG 3(11%) 3(8%) (9%) Management NSAIDsprescribed 3(8%) 3(11%) (9%) Localinjectionwithcorticosteroids 2(7%) 8(22%) (15%) Splint 7(25%) 15(41%) (34%) Followupconsultationwith 10(35%) 21(57%) (47%) GP(resident)arranged Referraltoneurologist 1(4%) 2(5%) (5%) CTS=carpaltunnelsyndrome,EMG=electromyography,NSAIDs=nonsteroidalantiinflammatorydrugs,GP =generalpractitioner.

154 154 Chapter8 DISCUSSION In this study we showed that education is an important means to change clinicians practice behaviour regarding recognition and referral of patients with SpA. While medicalhistoryandsymptomssimulatedbyspswouldhaveacknowledgedareferralto a rheumatologist, such a policy was executed by only 10% of the GPs. Specific SpA aimededucationimprovedthispolicydramatically.theprimaryoutcome,morethan 40% improvement in (considering) referral for both axial and peripheral SpA after education,wasmet. Approximately 20% of the adult population consult their GP because of musculoskeletal complaints, among which chronic low back pain is the most prevalent.[13,26]althoughitisimpossibleforgpstohaveconsiderableexpertiseinall areas,thehighexposuretomusculoskeletaldisorders(msd)warrantsdevelopmentof highqualitytrainingprogramsaimingatgainingandmaintainingsufficientexpertiseon MSD.Severalstudies,however,suggestthatgraduatedmedicalstudentsandresidents lack knowledge and confidence in this respect.[2730] Multifaceted education interventions,includingmixedinteractiveanddidacticlearningactivitiesfocussingon pertinentoutcomeshaveshowntosustainablychangephysicians behaviour.[31,32]in thepresentstudy,wealsoappliedmultifacetededucationaltoolsincludinginteractive powerpoint presentations, case vignettes and printed materials, which may have added to a better recognition of SpA three months after the intervention. Future studiesmayshedlightonthesustainabilityofeducationinthiscontext. Strengths of our study are that we have used a prospective, multicenter design and thatwehaveincludedacontrolgroupforanevaluationoftheeffectofeducation.in addition,ourstudywasconductedinprimarycarewhichisthesourceofmostreferrals ofspatotherheumatologist.furthermore,thespmodelhasprovenreliabilityforthe assessment of physicians knowledge and skills in a genuine clinical setting.[23 25,33,34] Severallimitationsofthisstudyrequirediscussion.First,SPsdidnottrulyhavesignsof their disease detectable at physical examination, which may have jeopardised recognition. SPs performing a role of peripheral SpA, for example, showed a photograph with dactylitis to the GP. Dactylitis is a relatively uncommon (albeit specific) manifestation of peripheral SpA. Nevertheless, one in two GPs (residents) ranked an inflammatory rheumatic disease as the diagnosis of highest likelihood in both SP encounters. This observation suggests that a knowledge deficit about peripheralspapreventedanadequatediagnosisbutnottherecognitionofarheumatic

155 RecognitionandreferralofpatientssuspectedofhavingSpA 155 disease.whilereferraltoamedicalspecialistwouldhavebeenthebestoptionhere, onlyaminorityofthepatientswasindeedreferred. Second,informationtoGPsaboutanSPvisitingtheirpracticemayhaveraisedarousal leadingtodifferentdiagnosticbehaviour.however,gp(residents)wereneitheraware ofthespecificcasepresentationorthepurposeofthestudynorweretheyinformed abouteducationbeingpartofthisstudy.inaddition,participantswerevisitedintheir ownpracticesbyspsduringregularworkinghours.sinceapreviousstudyhasfailedto demonstrateadifferenceinperformancebyresidentsevaluatingrealpatientsandsps [35], we believe the precautions we have taken have assured a most truthful performanceofgp(residents)indailypractice. Third,onemayarguethattheGP(residents)suspectedtheSPwassimulatingSpA,but thattheywereunawareofthefactthatreferralwouldhavebeenindicatedinthiscase. Makingacorrectdiagnosiswasasecondaryoutcomeinourstudy.Beforeeducation, onlyaminorityofgpscorrectlydiagnosedaxialspa(20%)andperipheralspa(10%).in contrast,ctswasrecognisedbythelargemajorityofparticipants(83%),indicatingthat GPs have sufficient knowledge of common disorders. However, they fall short regardingspa,whichismoreunfamiliarthancts. Fourth,wewereunabletoincludetheprojectednumberofparticipants.Nevertheless, theprimaryoutcomewasmet.asmallsamplesize,however,maylimitgeneralizability ofresultstoalargerpopulation. In conclusion, recognition and referral of patients suspected for having SpA by GP (residents) is in general low, but targeted education can markedly improve this. Increased awareness of a potential underlying inflammatory condition in patients presenting with musculoskeletal complaints and timely referral may prevent a debilitating disease course in patients with SpA. Therefore, we recommend the combinationofareferraltooltargetedatspaandeducationalactivitiesthatmaximize practitionerengagementandsupportforpracticechange.

156 156 Chapter8 REFERENCES 1. Jordan KP, Kadam UT, Hayward R, et al. Annual consultation prevalence of regional musculoskeletal problemsinprimarycare:anobservationalstudy.bmcmusculoskeletdisord2010;11: Jordan K, Clarke AM, Symmons DP, et al. Measuring disease prevalence: a comparison of musculoskeletal disease using four general practice consultation databases. Br J Gen Pract 2007;57: HoyD,MarchL,BrooksP,etal.Theglobalburdenoflowbackpain:estimatesfromtheGlobalBurden ofdisease2010study.annrheumdis2014;73: VilleneuveE,NamJL,BellMJ,etal.Asystematicliteraturereviewofstrategiespromotingearlyreferral and reducing delays in the diagnosis and management of inflammatory arthritis. Ann Rheum Dis 2013;72: Sieper J, Rudwaleit M. Early referral recommendations for ankylosing spondylitis (including pre radiographicandradiographicforms)inprimarycare.annrheumdis2005;64: Boonen A. A review of workparticipation, costofillness and costeffectiveness studies in ankylosing spondylitis.natclinpractrheumatol2006;2: FeldtkellerE,KhanMA,vanderHeijdeD,vanderLindenS,BraunJ.Ageatdiseaseonsetanddiagnosis delayinhlab27negativevs.positivepatientswithankylosingspondylitis.rheumatolint2003;23: Feldtkeller E, Bruckel J, Khan MA. Scientific contributions of ankylosing spondylitis patient advocacy groups.curropinrheumatol2000;12: VanHoevenL,LuimeJ,HanH,etal.Identifyingaxialspondyloarthritisindutchprimarycarepatients, ages2045years,withchroniclowbackpain.arthritiscareres(hoboken)2014;66: Van Tubergen A. The changing clinical picture and epidemiology of spondyloarthritis. Nat Rev Rheumatol2015;11: VandenBergR,BaraliakosX,BraunJ,etal.Firstupdateofthecurrentevidenceforthemanagementof ankylosing spondylitis with nonpharmacological treatment and nonbiologic drugs: a systematic literature review for the ASAS/EULAR management recommendations in ankylosing spondylitis. Rheumatology(Oxford)2012;51: RudwaleitM,ClaudepierreP,WordsworthP,etal.Effectiveness,safety,andpredictorsofgoodclinical response in 1250 patients treated with adalimumab for active ankylosing spondylitis. J Rheumatol 2009;36: Rudwaleit M,ListingJ,BrandtJ,etal.Predictionofamajorclinicalresponse(BASDAI50)totumour necrosisfactoralphablockersinankylosingspondylitis.annrheumdis2004;63: PoddubnyyD,VahldiekJ,SpillerI,etal.Evaluationof2screeningstrategiesforearlyidentificationof patientswithaxialspondyloarthritisinprimarycare.jrheumatol2011;38: BrandtHC,SpillerI,SongIH,etal.Performanceofreferralrecommendationsinpatientswithchronic backpainandsuspectedaxialspondyloarthritis.annrheumdis2007;66: Hermann J, Giessauf H, Schaffler G, et al. Early spondyloarthritis: usefulness of clinical screening. Rheumatology(Oxford)2009;48: Sieper J, Srinivasan S, Zamani O, et al. Comparison of two referral strategies for diagnosis of axial spondyloarthritis:therecognisinganddiagnosingankylosing SpondylitisReliably(RADAR)study.Ann RheumDis2013;72: Rudwaleit M, Sieper J. Referral strategies for early diagnosis of axial spondyloarthritis. Nat Rev Rheumatol2012;8: Van Tubergen A, Weber U. Diagnosis and classification in spondyloarthritis: identifying a chameleon. NatRevRheumatol2012;8: JoisRN,MacgregorAJ,GaffneyK.Recognitionofinflammatorybackpainandankylosingspondylitisin primarycare.rheumatology(oxford)2008;47; VanOnnaM,GorterS,vanMeerendonkA,etal.Generalpractitioners'perceptionsoftheirabilityto identify and refer patients with suspected axial spondyloarthritis: a qualitative study. J Rheumatol 2014;41: Gorter S, Rethans JJ, Scherpbier A, et al. How to introduce incognito standardized patients into outpatientclinicsofspecialistsinrheumatology.medicalteacher2001;23: UnderwoodMR,DawesP.Inflammatorybackpaininprimarycare.BrJRheumatol1995;34:10747.

157 RecognitionandreferralofpatientssuspectedofhavingSpA FreedmanKB,BernsteinJ.Educationaldeficienciesinmusculoskeletalmedicine.JBoneJointSurgAm 2002;84: SkelleyNW,TanakaMJ,SkelleyLM,etal.Medicalstudentmusculoskeletaleducation:aninstitutional survey.jbonejointsurgam2012;94: TruntzerJ,LynchA,KruseD,etal.Musculoskeletaleducation:anassessmentoftheclinicalconfidence ofmedicalstudents.perspectmededuc2014;3: HosieGA.Teachingrheumatologyinprimarycare.AnnRheumDis2000;59: Cantillon P, Jones R. Does continuing medical education in general practice make a difference? BMJ 1999;318: Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients'care.lancet2003;362: SiminoffLA,RogersHL,WallerAC,etal.Theadvantagesandchallengesofunannouncedstandardized patientmethodologytoassesshealthcarecommunication.patienteduccouns2011;82: ShiraziM,LabafA,MonjazebiF,etal.Assessingmedicalstudents'communicationskillsbytheuseof standardizedpatients:emphasizingstandardizedpatients'qualityassurance.acadpsychiatry2014;38: Beullens J, Rethans JJ, Goedhuys J, et al. The use of standardized patients in research in general practice.fampract1997;14: Woodward CA, McConvey GA, Neufeld V, et al. Measurement of physician performance by standardizedpatients.refiningtechniquesforundetectedentryinphysicians'offices.medcare1985; 23: Erby LA, Roter DL, Biesecker BB. Examination of standardized patient performance: accuracy and consistencyofsixstandardizedpatientsovertime.patienteduccouns2011;85: NormanGR,TugwellP,FeightnerJW.Acomparisonofresidenceperformanceonrealandsimulated patients.jmededuc1982;57:70815.

158

159 CHAPTER9 Summaryandgeneraldiscussion

160

161 Summaryandgeneraldiscussion 161 SUMMARYANDDISCUSSIONOFMAINFINDINGS Thisthesisdescribesseveralstudiesthatfocusontheearlyidentificationofpatients withspondyloarthritis(spa). As described in Chapter 1, major advances have been made over the last years to facilitateearlydiagnosisofspa.oneoftheimportantkeychangeswastheproposalof the Assessment of SpondyloArthritis international Society (ASAS) to revise disease nomenclature. Consistent with this proposal, SpA is now categorised into predominantly axial SpA (axspa) and predominantly peripheral SpA.[14] In addition, axspahasfurtherbeendifferentiatedas radiographicaxialspa whichissynonymous to ankylosing spondylitis (AS) and nonradiographic (nr)axspa, in which pelvic radiographs are normal or equivocal.[1] By recognizing nonradiographic axspa and incorporating this concept in the ASASclassification criteria for axspa, the focushas shiftedtowardsanearlierdetectionofaxspa.[1,3] However,toachievetimelyandaccuratediagnosisofSpA,furtherresearchinseveral areasisrequired. Oneofthechallengeswasaggregatingtheavailableepidemiologicaldatainareliable manner so that it could be used to better understand disease patterns. This information may also help to provide insight about which health care resources are neededtodetectandmanagespaatanearlystage. Chapter2presentsthereforeasystematicliteraturereviewandmetaanalysisonthe prevalenceofspaanditssubtypes.severaldemographical(includinggeographical)and methodological variables were studied to explain variation in the prevalence of (subtypesof)spainthe84studiesthatwereincluded.thepooledprevalenceofspa was0.55%(95%confidenceinterval(ci): )andthepooledprevalenceofas was0.18%(95%ci: ).therewashoweverahighlevelofheterogeneityacross studies.substantialvariationbetweengeographicareaswasfound,whichcanpartlybe explained by the prevalence of HLAB27; in populations known to have a high prevalence of HLAB27, a higher prevalence of SpA was reported. The highest prevalence of SpA (including AS) in our review was reported in decreasing order in Northern Artic indigenous communities, North America, East Asia and Europe. Interestingly,prevalenceestimatesofSpAwerepositivelyrelatedtotheyearofdata collection.thisfindingmightbetheresultofincreasedawarenessandrecognitionof SpA as a separate disease entity. In addition, our study showed that the reported prevalenceofspawashighlydependentonthetypeofcasedefinitionthatwasused.

162 162 Chapter9 ThisisillustratedbyapopulationbasedstudyintheUS.[5]TheprevalenceofSpAwas 1.4% according to the ESSG criteria, but 0.9% according to the Amor criteria. This findingmakesclearthatthetypeofclassificationcriteriamayhaveanimpactonthe reportedprevalenceof(subtypesof)spa.consequently,theprevalenceofspausing thenewclassificationcriteriaforaxialandperipheralspamaybedifferent. In axspa, the development of new classification criteria was preceded by the introductionofmagneticresonanceimaging(mri)ofthesij,withtheaimtofacilitate an early diagnosis of axspa.[6] The Early SpondyloArthritis Clinic (ESpAC) was established by our group in order to better understand the role of MRI in patients suspected for axspa. The followup study included three full clinical and imaging examinations performed with one year time intervals. MRIs of the sacroiliac joints (MRISIJ)wereanintegralpartofthisstudy.Chapter3describestheevolutionofactive lesionsonmrisij,suggestiveforsacroiliitis,overa2yearfollowupperiod.twenty four(35%)outof68patientshadapositivemrisijatbaseline.wefoundthatbotha positivehlab27statusandpositivemrisijatbaselinewereindependentlyassociated with a persistently positive MRISIJ over time. A combination of a negative HLAB27 statusandanegativemrisijatbaselineessentiallyruledoutapositivemriovertime. EspeciallyinmaleHLAB27positivepatientswithanegativebaselineMRISIJ,afollow upmriseemsrationalsinceaconsiderableproportionofthesepatientsmaydevelopa positivemriovertime.ourfindingsnotonlycontributetoefficientlydiagnosepatients withaxspabutalsogivecredittotheprominentplaceofmrisijintheasasaxialspa classificationcriteria.[1,7,8] Todate,activelesions(bonemarrowedema(BME))highlysuggestiveforsacroiliitisare required for a positive MRISIJ.[9] There is increasing debate, though, if structural lesions(erosionsorfattylesions)mayalsohaveanindependentdiagnosticvalue.[10]in Chapter4weinvestigatedthepresenceandevolutionofstructuralchangesonMRISIJ in patients included in the ESpAC. Subsequently, we analysed whether structural changes on either MRI or pelvic radiographs were preceded by BME on MRI at the sameanatomiclocation.thenumberoferosionsonmriincreasedsignificantlyduring followup,butthesenewerosionswerenotassociatedwithprecedingbme.whilethe numberoffattylesionsonlyshowedaslightincrease,thefewnewfattylesionswere associated with preceding BME. Interestingly, they occurred preferably in locations that showed resolution of BME. The relation between inflammation and structural damage on MRI in patients with axspa had been extensively discussed over the last years.[1118]thedebateismainlyconfinedtothequestionwhetherinflammationand

163 Summaryandgeneraldiscussion 163 structuraldamageare coupled or uncoupled processes.earlierreportshaveshown conflicting results.[1417] Our current findings support the view that fatty lesions should be seen as a repair reaction in response to inflammatory triggers.[19] Other datahavesuggestedthatfattylesionsonmriarelinkedtosubsequentdevelopmentof syndesmophytes in the spine.[20] We did not find an association between active lesionsanderosionsonmriandsubsequentprogressionofsacroiliitisonconventional radiographs.largerstudiesthanourswithsequentialmriexaminationsoveralonger followupperiodwillbeneededinordertobeconclusive. DifferentMRItechniquescanbeusedtodetectactivelesionsonMRI;theseareshort tau inversion recovery (STIR) sequence, and fatsaturated T1weighted spinecho sequence after administration of the contrast agent gadolinium diethylene triaminepentaaceticacid(gddtpa).[9]inchapter5weshowedthatthepostgddtpa sequence did not have a surplus value compared with the STIR sequence in combinationwithpelvicradiographs.ourresultswerecomparabletotheresultsofthe SPondyloArthritisCaughtEarly(SPACE)cohortthatfounda100%agreementbetween thestirandpostgddtpamrisequencebothatbaselineand3monthfollowup.[21] All together, it seems as if the postgddtpa can safely be omitted. This finding increases feasibility of MRI and reduces costs regarding the use of MRI during the diagnosticworkupofpatientssuspectedforaxspa. Atthestartofthisproject,essentialinformationabouttheuseofMRIwaslackingin order to recommend it in all patients with a suspicion of axspa and a normal or equivocalpelvicradiograph.oneofthosequestionspertainedtohowstablealesion found on MRI is over time. Chapter 6 describes the exact distribution, frequency of occurrence,andevolutionovertimeofbmedetectedonmrisijinpatientsincludedin theespac.almosthalfofthebmelesionsdetectedatbaseline,hadresolvedduring followup. Vice versa, some negative baseline MRIs had become positive over time. BothhadconsequencesforthefulfilmentoftheASASaxSpAcriteriaatanyfollowup moment. This finding reflects a general concern inherent to diagnostic imaging in rheumaticconditions,beingthatimagingfindings(whichareinherentlyunreliabledue to measurement error and judgement variation) become dominant over clinical findingsintheprocessofdiagnosticpatternrecognition.myopicallyfocusingonmri reports,forinstance,couldleadtoadelayinthediagnosisofaxspa,ifsignssuggestive forsacroiliitisonmriareabsentinanotherwiseclinicalundisputablespapatient.on theotherhand,afocusonmrifindingscouldleadtoafalsediagnosisofspaiftoo subtle and not highly suggestive lesions on MRI (the innocent white spots ) are interpretedwithoutthecontextoftheclinicalsymptomsofapatientthatactuallydo

164 164 Chapter9 notraisesuspicionofspa.theentirepatternofaxspa,includingmanifestationssuch as psoriasis, enthesitis or uveitis should therefore be taken into account. This has actuallybeendoneinthe diagnosticalgorithm inwhichathoroughmedicalhistory andphysicalexaminationareprominentstepsintheworkupofpatientssuspectedfor axspa, in which imaging also plays a role.[8] MRI should only be ordered in those patientswithareasonablesuspicionforspa. OptimizingthemanagementofpatientswithSpAalwaysstartswith casefinding.in thedutchhealthcaresetting,thisusuallystartsinprimarycare.thelevelofknowledge andtheperceptionsofgeneralpractitioners(gps)aboutinflammatorybackpain(ibp) and axspa, and potential barriers for referral of patients suspected for axspa were studiedinchapter7.wefoundthataxspaisoftenunrecognizedinprimarycare,and even when recognized, it is often suboptimally managed. In general, the level of knowledge about IBP, axspa and the associated extraarticular manifestations is limited.gpsareawareof classic,butlatefeaturesofaxspa,suchashyperkyphosisas seeninas.onlyhalfofthegpsknew,forexample,thattumornecrosisfactor(tnf) alphainhibitingbiologicalsareatreatmentoptioninaxspa.improvingtheknowledge of GPs about the leading presenting symptoms and current treatment options of axspa,willlikelyfacilitatethereferralofpatientssuspectedofhavingaxspa. TheprimarycaresettingitselfdoesnotfacilitatetheidentificationofpatientswithSpA: GPshavetodealwithabroadvarietyofphysicalandpsychologicalconditions.Chronic backpainisacommonsymptominthegeneralpopulationanditisestimatedthatin only5%ofcases,axspaistheunderlyingcondition.[22]gpsmaythinkthatadequate recognitionofaxspaisofsecondaryconcern.however,finetuningtheprocedurein ordertoselectthosepatientsthatmayneedfurtherevaluationinsecondarycareisstill an important step to make. A recent study has shown that 24% of primary care patients with chronic low back pain starting before the age of 45 years could be classifiedashavingaxialspa.[23]gpscanprovide longitudinalcontinuity ofcareto thepatientthatisconsistentwiththepatient sotherneeds.thismeansthatgpsarein auniquepositiontorecognizethepatternofsymptomssuggestiveforspa(including theoccurrenceofspafeaturesovertime)whenapatientfirstseeksmedicalattention. In Chapter 8 we describe the current practice performance of GPs and GP residents withregardtorecognizingearlyaxialorperipheralspa.weusedstandardizedpatients (SPs)todemonstrate whatisreallygoingon indailyclinicalpractice.inaddition,the influence of education on this performance was investigated. In this study, 68 GP residents and their supervising GPs were included and visited by trained SPs who

165 Summaryandgeneraldiscussion 165 simulatedacaseofearlyaxialofperipheralspa.weconcludedthatrecognitionand referral of patients suspected for SpA to the rheumatologist by GP (residents) is generallylow.providingeducationhowevermarkedlyimprovedbothrecognitionand referralofpatientssuspectedforspatotherheumatologist.educationmaytherefore bean interesting startingpoint to increaseawarenessabout SpA,paving the wayto successfulimplementationofareferralstrategy. METHODOLOGICALCONSIDERATIONS ThemostimportantlimitationsofthestudiesthatweredescribedinChapter2to8of thisthesishavealreadybeendescribedintherespectivechapters.twomethodological considerationsarehighlightedanddescribedinthissection:generalizabilityofresults (Chapter3to6)andtheuseofstandardizedpatientsinoutcomeresearch(Chapter8). GENERALIZABILTYOFRESULTS Sincethemaingoalofresearchistoincreaseourunderstandingoftheworldaround us, researchresultsshould rather berelevant whenapplied to other patientswitha similarclinicalprofile.althoughespacisarelativelysmallcohort,itclearlyrepresentsa studypopulationofhighactualinterest.patientsinespachadibp2yearsatmostand this cohort is therefore able to provide more insight in the early stages of axspa. Patients included in the ESpAC were recruited from the regular rheumatology outpatient clinic (60%) or by (related) medical specialties (i.e. dermatology, ophthalmology) and relatives of members of the local AS society (40%). This recruitmentstrategymayexplainthehighproportionofpatientsfulfillingatleastone of the classification criteria for SpA (94%). In the ESpAC study, 31 (46%) out of 68patients fulfilled the imaging arm of the ASAS criteria. Although rheumatologists oftenseepatientsreferredbyothermedicalspecialists,selectingpatientsinthisway mayhampergeneralizabilityofourresultstoawiderpopulationtosomeextent.our results may be restricted to patients with a clearer clinical presentation (such as psoriasisandibp )orthosewithmoreactivedisease.thismeansthatgeneralization toanotherpopulation,forexamplepatientswithbackpainreferredbygps,shouldbe done with caution. It ishowever important to note thatour mainresearch question was how BME and structural changes on MRISIJ, suggestive for axspa, evolve over time. To answer this question, subject restriction by selecting patients with a high

166 166 Chapter9 chanceofabnormalitiesonmrisijmayhaveimprovedefficiencyandinternalvalidity. Notwithstanding, the lessons learned from ESpAC may help rheumatologists how to assessmrisijinclinicalpractice. STANDARDIZEDPATIENTS,EDUCATIONANDOUTCOME MEASURES In 1968, Barrows was the first to introduce SPs in order to assess the clinical performance of medical residents.[24] Since then, SPs are most commonly used for teaching communication and clinical skills.[25] The study with SPs described in this thesis, however, focused on changes in recognition of SpA and the decision to refer patients to a rheumatologist before and after education. SPs recorded several outcomes of the consultation before as well as after the educational intervention. Whenevaluatingoutcomeinastudythatincludesaneducationalprogramme,several outcomelevelscanbedistinguished[26]:.towhatextentdidtheparticipantslikethelearningevent?likert scalescanbeusedtomeasuretheparticipants levelofsatisfaction..towhatdegreedidparticipantsacquiretheintendedknowledge, skills and attitudes based on their participation in the leaning event? Computer testsorobjectivestructuredclinicalexaminations(osce)canbeusedtomeasure changesinforinstanceknowledge..towhatdegreeapplyparticipantswhattheylearnedduringthe training when they are back on the job? SPs that visit the GP before and after educationcanrecordchangesinforinstancehistorytaking. :Towhatdegreedoand wellbeingofapatientoccurasaresultofthelearningevent?surveysandannual reportscanbeusedtomeasurethesechanges. The majority of studies that report on educational program innovations focus on changes in knowledge, skills and attitudes (level 2).[26] Our study also assessed the changes in practice behaviour (level 3). The advantage ofcollectingdata onpractice behaviouristhatitoffersinformationaboutwhatactuallyhappensinclinicalpractice. TheSPmethodisvalid,butthoroughselectionandtrainingoftheSPisaprerequisite. In addition, SP visits have proven to be accurate and reliable in collecting information.[2730]inareviewofrethansetal.itwasconcludedthatbothintrasp reliability and intersp reliability are 0.85 or more, which indicates an excellent agreement.[27]animportantlimitationregardingtheuseofspsinourstudyisthefact

167 Summaryandgeneraldiscussion 167 thatthespvisitwasrestrictedtoone(first)consultation.especiallyinthecaseofrare or chronic conditions, GPs may unfold diagnostic and therapeutic interventions in severalconsultations.toovercomethislimitation,weincludedinthemedicalhistory of the SP that complaints were present for more than one year, multiple visits with othergpshadtakenplaceandphysiotherapyorswimmingwereunsuccessful.overall, theuseofspsseemsanexcellentmethodtoassesstheeffectofeducationonpractice performance.whentakingtherightprecautions,itseemsthatspscanalsobeusedfor lessstraightforwardmedicalproblems,suchaswedidinthecaseofspa. IMPLICATIONSFORCLINICALPRACTICE ImprovingevidencebasedcareofpatientssufferingfromSpAistheultimategoalof researchinthespafield. NewinsightsabouttheepidemiologyofSpA,asdescribedinthisthesis,suggestthat spondyloarthritisisasprevalentasrheumatoidarthritis.[31]theactualprevalenceof SpAisimportantforimplementingstandardsofcareforpatientswithSpA(including efforts to diagnose the disease earlier), and for decision makers when allocating resources to research and healthcare. Knowledge about (variation in) prevalence contributes to plan optimal quality of care within the limits of available health care resources. TheintroductionofMRItodiagnosesacroiliitisatanearlystagehasalreadyshownto be highly useful to both rheumatologists and patients. For rheumatologists, the introduction of MRI helps to improve the diagnostic approach of a highly prevalent clinicalproblem.whenrheumatologistshavearealsuspicionofaninflammatoryback probleminapatientaftercarefulevaluation,conventionalpelvicradiographsoftenfail to show radiographic sacroiliitis. The findings described in this thesis may guide rheumatologiststocorrectlyassessthevalueofmriinthesedaytodayclinicalcases. Forinstance,inanHLAB27positivemalepatientwithanegativeMRISIJ,asecondMRI can be considered in case of persisting complaints since a reasonable percentage of thesepatientswilldevelopactivelesionsonmriovertime. For patients, early diagnosis is important since the pain and physical limitations associated with axspa may have important socioeconomic consequences, such as decreased work participation.[32] Substantial diagnostic delay can have an adverse impactonseveraloutcomes,includingphysicalfunctioningandqualityoflife.[33]last butnot least,patientswithnraxspawhohave objective measures ofinflammation (elevated Creactive protein and active lesions on MRI) have a similar response to

168 168 Chapter9 treatmentascomparedtopatientswithestablishedas,emphasizingtheneedforearly andcorrectdiagnosis.[34,35]earlydiagnosismightfacilitateearlyinitiationofeffective treatment,therebydiminishingtheburdenofdiseaseforpatientsandpossiblyavoid longtermcostsofworkdisabilityforsociety.[36] Optimization of the diagnostic and management process by the rheumatologist is important,butdecreasingthediagnosticdelayinprimarycareisequallyessential.the greater part of the diagnostic delay pertains to primary care.[37,38] It is therefore importanttocommunicatenewdiseaseconcepts,inparticularthosewithimplications forprimarycare,tothegp.ourdatahaveshownthateducatinggpresidentsabout SpAincreasesthenumberofreferralstotherheumatologist.Postgraduatetrainingof GPsinrheumatologyseemsafeasibleapproachtoimprovethecareforpatientswith SpAanddeservesgreaterpriority. FUTUREPERSPECTIVES A definite diagnosis of axspa remains a challenge in daily practice due to the heterogeneousclinicalpictureofthedisease.althoughmrihasundoubtedlyfacilitated anearlydiagnosisofaxspa,severalresearchquestionsremainunsolved.first,although MRIiswidelyacceptedasimagingtechniquetodetectsacroiliitisatanearlystage,the specificityandpredictivevalueoftheseactivemrilesionsonthelongtermneedsto be further validated in prospective studies.[39] Second, further clarification of the potential contribution of structural changes on MRI in relation to classification and diagnosisofaxspaisneeded.severalstudiessuggestthatstructuraldamageonmri, especiallyerosions,maybehelpfulinestablishinganearlydiagnosisofaxspa.[40,41] DetectionoferosionsonMRISIJismorechallengingthandetectionofBME,sincethe spectrum of appearance of erosions is more heterogeneous.[42,43] Further defining thecharacteristicsoferosionsandtheirroleinfurtherimprovingthediagnosticutility ofmriisneeded.third,thelinkbetweenactiveandstructurallesionsonmriandbone formationonconventionalradiographs,needstobefurtherelucidated.inflammation mightbethefirststepinthedominoeffectleadingtoboneformation.[44]however, inflammation and bone formation might also be (partly) uncoupled processes. New therapeuticoptionsshouldthenalsoneedtotargetpathwaysofboneformation.[11] Prospective (longitudinal) cohorts are necessary to provide more insight into the predictive potential of MRI and may ultimately guide the rheumatologist in clinical decisionmaking.

169 Summaryandgeneraldiscussion 169 WithregardtoimprovingreferralofpatientssuspectedforaxialandperipheralSpAto specialist services, it is important that GPs are both informed and motivated to incorporate the latest insights about SpA into their daily work practices.[43] Much attentionhasbeengiventoreferralstrategies,butlackofknowledgeaboutspaand absence of incentives for change may prevent successful implementation of any referral strategy. Therefore, a targeted educational activity should not only include increaseknowledgeaboutspa,butalsocreatecommitmentamonggpstochangethe quality of care for this relatively young patient group.[4547] More research is necessary in order to develop and test educational activities that can ultimately be usedtosuccessfullyimplementreferralstrategiesinprimarycare. OVERALLCONCLUSIONS ThisthesisshowsthatthepredictivepotentialofMRISIJinpatientswithearlyIBPis highwithregardtothepresenceorabsenceofactivelesions,especiallyincombination with other parameters such as HLAB27 and gender. Signs of BME on MRI seem a triggerforreparativeprocessessuchasfattylesions,whichstrengthenideasaboutthe needforearlyantiinflammatorytreatment.earlyinitiationoftreatmentcanonlystart aftertimelyreferralofpatientssuspectedofspabygps.toachievetimelyreferral,itis important that GPs and rheumatologists learn from each other. Both have complementary skills that are necessary to ensure highquality care for patients throughouttheirdiseasetrajectory.knowingishalfthebattle.

170 170 Chapter9 REFERENCES 1. RudwaleitM,vanderHeijdeD,LandewéR,etal.ThedevelopmentofAssessmentofSpondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: Rudwaleit M, van der Heijde D, Landewé R, et al. The Assessment of SpondyloArthritis International Societyclassificationcriteriaforperipheralspondyloarthritisandforspondyloarthritisingeneral.Ann RheumDis2011;70: RudwaleitM.Newapproachestodiagnosisandclassificationofaxialandperipheralspondyloarthritis. CurrOpinRheumatol2010;22: Rudwaleit M, Khan MA, Sieper J. The challenge of diagnosis and classification in early ankylosing spondylitis:doweneednewcriteria?arthritisrheum2005;52: Reveille JD, Witter JP, Weisman MH. Prevalence of axial spondyloarthritis in the United States: estimatesfromacrosssectionalsurvey.arthritiscareres(hoboken)2012;64: BraunJ,SieperJ,BollowM.Imagingofsacroiliitis.ClinRheumatol2000;19: Van den Berg R, de Hooge M, Rudwaleit M, et al. ASAS modification of the Berlin algorithm for diagnosingaxialspondyloarthritis:resultsfromthespondyloarthritiscaughtearly(space)cohortand from the Assessment of SpondyloArthritis international Society (ASAS)cohort. Ann Rheum Dis 2013; 72: Yu D, Van Tubergen A. Clinical manifestations of ankylosing spondylitis in adults. In: DS Basow (Ed.), UpToDate.WalthamMA, RudwaleitM,JurikAG,HermannKG,etal.Definingactivesacroiliitisonmagneticresonanceimaging (MRI) for classification of axial spondyloarthritis: a consensual approach by the ASAS/OMERACT MRI group.annrheumdis2009;68: Weber U, Ostergaard M, Lambert RG, et al. Candidate lesionbased criteria for defining a positive sacroiliacjointmriintwocohortsofpatientswithaxialspondyloarthritis.annrheumdis2014:doi: /annrheumdis MaksymowychWP,ElewautD,SchettG.Motionfordebate:thedevelopmentofankylosisinankylosing spondylitisislargelydependentoninflammation.arthritisrheum2012;64: Maksymowych WP, Chiowchanwisawakit P, Clare T, et al. Inflammatory lesions of the spine on magnetic resonance imaging predict the development of new syndesmophytes in ankylosing spondylitis:evidenceofarelationshipbetweeninflammationandnewboneformation.arthritisrheum 2009;60: Wanders A, van der Heijde D, Landewe R, et al. Nonsteroidal antiinflammatory drugs reduce radiographic progression in patients with ankylosing spondylitis: a randomized clinical trial. Arthritis Rheum2005;52: VanderHeijdeD,LandeweR,EinsteinS,etal.Radiographicprogressionofankylosingspondylitisafter uptotwoyearsoftreatmentwithetanercept.arthritisrheum2008;58: VanderHeijdeD,LandeweR,BaraliakosX,etal,andtheAnkylosingSpondylitisStudyfortheEvaluation ofrecombinantinfliximabtherapystudygroup.radiographicfindingsfollowingtwoyearsofinfliximab therapyinpatientswithankylosingspondylitis.arthritisrheum2008;58: VanderHeijdeD,SalonenD,WeissmanBN,etal.Assessmentofradiographicprogressioninthespines ofpatientswithankylosingspondylitistreatedwithadalimumabforupto2years.arthritisresther 2009;11:R VanderHeijdeD,MachadoP,BraunJ,etal.MRIinflammationatthevertebralunitonlymarginally predicts new syndesmophyte formation: a multilevel analysis in patients with ankylosing spondylitis. AnnRheumDis2012;71: LoriesRJ,DereseI,DeBariC,etal.Evidenceforuncouplingofinflammationandjointremodelingina mousemodelofspondylarthritis.arthritisrheum2007;56: SongIH,HermannKG,HaibelH,etal.Relationshipbetweenactiveinflammatorylesionsinthespine and sacroiliac joints and new development of chronic lesions on wholebody MRI in early axial spondyloarthritis:resultsoftheesthertrialatweek48.annrheumdis2011;70:

171 Summaryandgeneraldiscussion Chiowchanwisawakit P, Lambert RG, ConnerSpady B, et al. Focal fat lesions at vertebral corners on magnetic resonance imaging predict the development of new syndesmophytes in ankylosing spondylitis.arthritisrheum2011;63: DeHoogeM,vandenBergR,NavarroCompánV,etal.Magneticresonanceimagingofthesacroiliac jointsintheearlydetectionofspondyloarthritis:noaddedvalueofgadoliniumcomparedwithshorttau inversionrecoverysequence.rheumatology(oxford)2013;52: UnderwoodMR,DawesP.Inflammatorybackpaininprimarycare.BrJRheumatol1995;34: VanHoevenL,LuimeJ,HanH,etal.Identifyingaxialspondyloarthritisindutchprimarycarepatients, ages2045years,withchroniclowbackpain.arthritiscareres(hoboken)2014;66: BarrowsHS.Simulatedpatientsinmedicalteaching.CanMedAssocJ1968;98: May W, Park JH, Lee JP. A tenyear review of the literature on the use of standardized patients in teachingandlearning: medteach2009;31: KirkpatrickDL.Evaluatingtrainingprograms:thefourlevels.SanFrancisco:BerettKoehlerPublishers; Beullens J, Rethans JJ, Goedhuys J, et al. The use of standardized patients in research in general practice.fampract1997;14: Tamblyn RM, Abrahamowicz M, Berkson L, et al. Firstvisit bias in the measurement of clinical competencewithstandardizedpatients.acadmed1992;67:s CarneyPA,DietrichAJ,FreemanDHJr,etal.Theperiodichealthexaminationprovidedtoasymptomatic olderwomen:anassessmentusingstandardizedpatients.anninternmed1993;119: Rethans JJ, Gorter S, Bokken L, Morrison L. Unannounced standardised patients in real practice: a systematicliteraturereview.mededuc2007;41: BijlsmaJWJ.EULARtextbookonrheumaticdiseases.BMJPublishingGroup; Boonen A. A review of workparticipation, costofillness and costeffectiveness studies in ankylosing spondylitis.natclinpractrheumatol2006;2: Sieper J, Rudwaleit M. Early referral recommendations for ankylosing spondylitis (including pre radiographicandradiographicforms)inprimarycare.annrheumdis2005;64: SieperJ,LandewéR,RudwaleitM,etal.EffectofCertolizumabPegolover96WeeksinPatientswith Axial Spondyloarthritis: Results from a Phase 3 Randomized Trial. Arthritis Rheumatol 2014: doi: /art SieperJ,vanderHeijdeD,DougadosM,etal.Efficacyandsafetyofadalimumabinpatientswithnon radiographic axial spondyloarthritis: results of a randomised placebocontrolled trial (ABILITY1). Ann RheumDis2013;72: RudwaleitM,ClaudepierreP,WordsworthP,etal.Effectiveness,safety,andpredictorsofgoodclinical response in 1250 patients treated with adalimumab for active ankylosing spondylitis. J Rheumatol 2009;36: FeldtkellerE,KhanMA,vanderHeijdeD,etal.AgeatdiseaseonsetanddiagnosisdelayinHLAB27 negativevs.positivepatientswithankylosingspondylitis.rheumatolint2003;23: Feldtkeller E, Bruckel J, Khan MA. Scientific contributions of ankylosing spondylitis patient advocacy groups.curropinrheumatol2000;12: DeRyckeL,MaasM,TakPP,etal.'MRItis'intheearlydiagnosisofaxialSpA:issuesandlimitations.Nat RevRheumatol2010;6: WeberU,ZublerV,PedersenSJ,etal.DevelopmentandvalidationofanMRIreferencecriterionfor definingapositivesijmriinspondyloarthritis.arthritiscareres(hoboken)2013;65: Weber U, Ostergaard M, Lambert RG, et al. Candidate lesionbased criteria for defining a positive sacroiliacjointmriintwocohortsofpatientswithaxialspondyloarthritis.annrheumdis2014;doi: /annrheumdis WeberU,MaksymowychWP.Advancesandchallengesinspondyloarthritisimagingfordiagnosisand assessmentofdisease.currrheumatolrep2013;15: Weber U, Pedersen SJ, Østergaard M, et al. Can erosions on MRI of the sacroiliac joints be reliably detectedinpatientswithankylosingspondylitis?acrosssectionalstudy.arthritisresther2012may 24;14:R Ramiro S, van der Heijde D, van Tubergen A, et al. Higher disease activity leads to more structural damage in the spine in ankylosing spondylitis: 12year longitudinal data from the OASIS cohort. Ann RheumDis2014;73:

172 172 Chapter9 45. Davis DA, TaylorVaisey A. Translating guidelines into practice. A systematic review of theoretic concepts, practical experience and research evidence in the adoption of clinical practice guidelines. CMAJ1997;157: CoomarasamyA,KhanKS.Whatistheevidencethatpostgraduateteachinginevidencebasedmedicine changesanything?asystematicreview.bmj2004;329: HosieGA.Teachingrheumatologyinprimarycare.AnnRheumDis2000;59:5003.

173 SamenvattinginhetNederlands

174

175 Samenvatting 175 INLEIDING Spondyloartritis(SpA)iseenverzamelnaamvooreenaantalinflammatoirereumatische aandoeningen van het bewegingsapparaat met verwante klinische kenmerken, genetischeaanlegenfamiliairoptreden.totdegroepvanspabehorenonderandere ankyloserendespondylitis(as),arthritispsoriatica,reactieveartritisenenteropathische artritis.[1]typischvoorspaisdatookextraarticulairemanifestaties(aandoeningenin andereorganendanhetskelet)deelkunnenuitmakenvanhetziektebeeld.[2] TegenwoordigwordtdegroepvanSpAookwelingedeeldinaxialeSpAenperifereSpA. Bij axiale SpA staan symptomen van het axiale skelet op de voorgrond, met als overheersende symptoom inflammatoire rugpijn veroorzaakt door sacroiliitis (ontstekingvanhetheiligbeen)en/ofspondylitis(ontstekingvandewervellichamen). De bekendste vorm van axiale SpA is de aandoening AS, oftewel de ziekte van Bechterew.[2]BijperifereSpAstaanartritis,enthesitis(inflammatieopdeplaatswaar pees aan bot vastzit) en dactylitis (worstvormig gezwollen teen of vinger) op de voorgrond.[2]intotaalkrijgt4050%vandepatiëntenmetspatijdenshetbeloopvan de ziekte een extraarticulaire manifestatie, zoals acute uveitis anterior (regenboog vliesontsteking), inflammatoire darmziekten en psoriasis.[3] SpA begint meestal op jongvolwassen leeftijd. Aanvang van de eerste klachten na het 45e levensjaar is zeldzaam.[4] Bij axiale SpA en in het bijzonder type AS kan uiteindelijk een klassiek radiologisch beeldmetankyloseringvandesacroiliacalegewrichtenopeenbekkenfotozichtbaar worden.[1] In de wervelkolom kunnen syndesmofyten ontstaan door verbening van intervertebrale ligamenten.[1] Dit leidt uiteindelijk in het meest gevorderde stadium totvolledigeankylosevandewervelkolom,eenzogehetenbamboospine.wanneerer geenofslechtsdubieuzeafwijkingenzichtbaarzijnopeenconventionelebekkenfoto, kan men besluiten een Magnetic Resonance Imaging (MRI) van de sacroiliacale gewrichten te maken. Op de MRI kunnen voor sacroiliitis kenmerkende afwijkingen worden gezien, bijvoorbeeld subchondraal beenmergoedeem. Indien de patiënt een afwijkendemriheeftincombinatiemeteenanderspakenmerk,danwordtdit niet röntgenologische axiale SpA genoemd.[5,6] Bij een aantal patiënten met niet röntgenologische axiale SpA is sacroiliitis op een MRI (nog) niet aantoonbaar. Deze patiënten worden op basis van genetische aanleg (humaan leukocytenantigeen B27 (HLAB27)) en minimaal twee andere kenmerkende SpAmanifestaties ook gerekend totdegroepvannietröntgenologischeaxialespa.

176 176 BijpatiëntenmetaxialeSpAduurthetgemiddeldnegenjaarnahetontstaanvande eerste rugklachten voordat de diagnose wordt gesteld.[4] Een belangrijke reden hiervoor is dat afwijkingen op een conventionele bekkenfoto, wat voorheen als een vereistegoldvoorhetstellenvandediagnoseankyloserendespondylitis,slechtslaatin hetziekteproceszichtbaarworden.demrikanhieruitkomstbieden,omdatopdemri vaakwelafwijkingentypischvoorsacroiliitiswordengezien.[6]daarnaastheeftmen langgedachtdataxialespaeentypischemannenziektewas.indeliteratuurzijnman vrouwverhoudingenbeschrevenvan8:1à10:1.erblijktechterspraketezijngeweest van onderdiagnose bij vrouwen. Bij vrouwen ziet men minder structurele schade op een conventionele bekkenfoto.[7] Tegenwoordig wordt verondersteld dat de preva lentievanaxialespainmannenenvrouwengelijkis. Eerder onderzoek heeft laten zien dat een goede en tijdige behandeling van SpA uiteindelijkleidttoteenverbeteringvanverschillendeuitkomstmaten,waaronderde matevanziekteactiviteit,arbeidsproductiviteitenkwaliteitvanleven.[810]dehuisarts is een sleutelfiguur wanneer het gaat om de herkenning en vroege verwijzing van patiënten met klachten die zouden kunnen wijzen op SpA. De afgelopen jaren zijn verschillendestrategieënontwikkelddieeensnelleverwijzingvanpatiëntenmetspa mogelijk zouden moeten maken.[1115] Huisartsen hebben echter moeite met het herkennen van het ziektebeeld.[16] Slechte herkenning kan er toe leiden dat de verwijsstrategieën niet goed worden toegepast in de huisartsenpraktijk. Educatie specifiek gericht op SpA zou de kennis, herkenning en daardoor verwijzing van patiëntennaardereumatoloogmogelijkkunnenverbeteren. De onderzoeken zoals beschreven in dit proefschrift richten zich op de vroege herkenningvanspa. DEELI:DEEPIDEMIOLOGIEVANSPONDYLOARTRITIS In hoofdstuk 1 wordt beschreven dat er de afgelopen jaren veel is verbeterd ten aanzien van de mogelijkheden om SpA eerder te diagnosticeren. Een belangrijke verandering is de ontwikkeling van nieuwe naamgeving door de Assessment in SpondyloArthritis international Society (ASAS).[1720] Momenteel wordt onderscheid gemaakttussenaxialespaenperiferespa.axialespawordtverderonderverdeeldin röntgenologische axiale SpA ofwel AS en nietröntgenologische axiale SpA, een aandoening waarbij geen of onvoldoende afwijkingen op een röntgenfoto aanwezig

177 Samenvatting 177 zijn.[17]doornietröntgenologischeaxialespaalszodanigteerkennenenoptenemen indeasasaxialespaclassificatiecriteria,heeftdefocuszichdeafgelopenjarenverlegd richtingvroegeherkenningvanaxialespa.[17,19] Ondanks(ofdoor)dezeontwikkelingen,bleefverderonderzoekophetgebiedvanSpA nodig.zoishetbelangrijkombetrouwbareepidemiologischegegevensteverzamelen over het wereldwijd voorkomen van (de verschillende subtypes van) SpA. Deze gegevenskunnenwordengebruiktommeerinzichttekrijgeninwelkemaatregelenen middelen nodig zijn om patiënten met SpA tijdig op te kunnen sporen en te behandelen. In hoofdstuk 2 wordt een systematisch literatuuronderzoek en meta analysenaardeprevalentievan(subtypesvan)spagepresenteerd.intotaalwerden84 studiesgeïdentificeerd.verschillendedemografische(waaronderookgeografische)en methodologischevariabelenwerdenbestudeerdomteachterhalenofzijeendeelvan de variatie in prevalentie tussen de verschillende studies konden verklaren. De gepooldepopulatieprevalentievanspawas0.55%(95%betrouwbaarheidsinterval(bi) ); de gepoolde populatieprevalentie van AS was 0.18% (95% BI ). Tussendegeïncludeerdestudieswerdechtereenhogeheterogeniteitgevonden.De variatie in prevalentie tussen de verschillende wereldregio s kon deels worden verklaarddoordebekendevariatieindeprevalentievanhlab27tussendezeregio s. InpopulatieswaardeprevalentievanHLAB27hoogis,werdeenhogeprevalentievan SpAgevonden.DehoogsteprevalentiesvanSpA(inclusiefAS)werdeninafnemende volgordegevondenbijinheemsebevolkingsgroependieleveninhetarctischgebied, NoordAmerika, OostAzië en Europa. Een andere interessante bevinding was dat de prevalentie van SpA hoger was in studies van meer recente datum. Wellicht is deze stijginginprevalentiedoordejarenheenteverklarendoorhetfeitdatspainmiddels beterwordtherkend.daarnaastwasprevalentieafhankelijkvandemanierwaaropspa werd gedefinieerd. Dit wordt mooi geïllustreerd in een populatiestudie uit de Verenigde Staten.[21] De prevalentie van SpA was 1.4% wanneer de ESSG criteria werdengehanteerden0.9%wanneerdeamorcriteriawerdengehanteerd.ditverschil inprevalentiemaaktduidelijkdathettypeclassificatiecriteriauiteindelijkinvloedkan hebben op de gerapporteerde prevalentie van (subtypes van) SpA. Het is daarom te verwachten dat door het toepassen van de nieuwe ASAS axiale en perifere SpA classificatiecriteria, de prevalentie van SpA wederom beïnvloed kan worden en/of anderepatiëntengeïdentificeerdzullenworden.

178 178 DEELII:HETGEBRUIKVANMRIBIJHETOPSPORENVAN VROEGESPONDYLOARTRITIS BijaxialeSpAwerddeontwikkelingvannieuweclassificatiecriteriavoorafgegaanaan deintroductievandemrivandesacroiliacalegewrichten.doordemritegebruiken, wordthetmogelijkdediagnoseaxialespaeerdertestellen.[22]opeenmrikunnen immers, in tegenstelling met conventionele röntgenfoto s, ook kenmerken van ontstekingsactiviteit worden gezien, zoals beenmergoedeem. Verder kunnen ook structureleveranderingenwordengezienzoalserosies,vervettingvanhetbeenmerg enankylose.[6]eenmrivandesacroiliacalegewrichtenwordtpositiefbevondenvoor sacroiliitiswanneerminstensééninflammatoirelaesie(beenmergoedeem)aanwezigis op twee opeenvolgende MRIsneden, of wanneer twee laesies aanwezig zijn op één MRIsnede.[6] De Early SpondyloArthritis Clinic (ESpAC) werd door onze onderzoeksgroep opgezet om meer inzicht te krijgen inderol van MRI bij patiënten met symptomen suggestief voor axiale SpA. De twee jaar durende followup studie bestond uit drie evaluaties met tussenpozen van telkens één jaar. Elke evaluatie bestondondermeeruiteenuitgebreidklinischonderzoekenbeeldvorming.mriscans vandesacroiliacalegewrichtenmaakteneenbelangrijkonderdeeluitvandezestudie. Hoofdstuk 3 beschrijft de ontwikkeling van ontstekingsactiviteit op de MRI van de sacroiliacale gewrichten, suggestief voor sacroiliitis, gedurende een twee jaar durende followup periode. Op baseline hadden 24 (35%) van de 68 geïncludeerde patiënteneenpositievemrivandesacroiliacalegewrichten.zoweldeaanwezigheid vanhlab27alseenpositievemriopbaseline,warenonafhankelijkgeassocieerdmet een positieve MRI tijdens de followup periode. Bij afwezigheid van HLAB27 en een negatieve MRI op baseline, was de ontwikkeling een positieve MRI tijdens followup zeeronwaarschijnlijk.enkelbijhlab27positievemannenmeteennegatievemrikan hetzinvolzijndemriteherhalen.eenaanzienlijkdeelvandezepatiëntenontwikkelde namelijkalsnogeenpositievemritijdensdefollowupperiode.onzeresultatenlaten zien hoe patiënten met klachten verdacht voor axiale SpA op efficiënte wijze het diagnostischtrajectkunnendoorlopen.daarnaastondersteunenonzebevindingenhet feitdatdemriinmiddelseenprominenteplaatsheeftgekregenindeasasaxialespa classificatiecriteria.[17,23,24] Opditmomentwordtalleenontstekingsactiviteit(beenmergoedeem)meegenomenbij debeoordelingofeenmrivandesacroiliacalegewrichtenaldannietpositiefis.eris momenteel echter een discussie gaande of structurele afwijkingen (erosies of beenmergvervetting)ooknieteenonafhankelijkediagnostischewaardehebben.[25]in hoofdstuk 4 onderzochten we de aanwezigheid en ontwikkeling van structurele

179 Samenvatting 179 afwijkingen op een MRI van de sacroiliacale gewrichten bij patiënten die zijn geïncludeerdindeespac.daarnaastwerdgekekenofstructureleafwijkingenopzowel de MRI als conventionele bekkenfoto vooraf werden gegaan door beenmergoedeem op de MRI op dezelfde locatie. Hoewel het aantal erosies op MRI tijdens followup toenam,werdergeenassociatiemetbeenmergoedeemopdezelfdelocatieeenjaar daarvoorgevonden.ondanksdaterslechtseenlichtetoenamewerdgevondeninde gebiedenmetbeenmergvervetting,werderweleenassociatiemetbeenmergoedeem opdezelfdelocatieinhetvoorafgaandejaargevonden.eeninteressantebevindingwas dat beenmergvervetting met name werd gezien op locaties waar beenmergoedeem wasverdwenen.derelatietussenontstekingsactiviteitenstructureleschadeopmribij patiëntenmetaxialespaisdelaatstejarenvolopbediscussieerd.[2633]dediscussie richtzichmetnameopdevraagofontstekingsactiviteitenstructureleschade'inrelatie tot elkaar staan' of 'los van elkaar staan'. Eerder onderzoek laat tegenstrijdige resultaten zien.[2932] Onze bevindingen ondersteunen de theorie dat beenmerg vervetting mogelijk een herstelreactie is op eerder aanwezige ontsteking.[34] Ander onderzoekheeftaangetoonddatbeenmergvervettingisgeassocieerdmetontwikkeling van syndesmofieten (benige vergroeiing) in de wervelkolom.[35] Wij vonden geen associatietussenontstekingsactiviteitopmrienerzijdsenontwikkelingvanerosiesop MRI of ankylose op de bekkenfoto anderzijds. Grotere studies dan ESpAC met meerderemri senröntgenfoto sovereenlangerefollowupperiodekunnenhierover uitsluitselgeven. OmontstekingsactiviteitopdeMRItekunnenvaststellen,kunnenverschillendeMRI sequentieswordengebruikt.ditzijndeshorttauinversionrecovery(stir)sequentie en de T1gewogen spinecho sequentie met onderdrukking van het vetsignaal na toedieningvanhetcontrastmiddelgadoliniumdiethylenetriaminepentaaceticacid(gd DTPA).[6] In hoofdstuk 5 lieten we zien dat de postgddtpasequentie geen toegevoegde waarde heeft boven de STIRsequentie in combinatie met de conven tionelebekkenfoto.onzeresultatenkomenovereenmetdievanhetspondyloarthritis CaughtEarly(SPACE)cohort,waarzowelopbaselinealsnadriemaandeneenvolledige overeenkomstwerdgezientussendestirenpostgddtpasequentie.[36]hetlijkter dus op dat de postgddtpasequentie kan worden overgeslagen terwijl de diagnostische waarde van de MRI behouden blijft. Deze bevinding leidt tot een kostenbesparing en vergroot daarmee de toepasbaarheid van de MRI wanneer patiënten verdacht voor axiale SpA tijdens het diagnostisch traject een MRI moeten ondergaan.

180 180 Bij aanvang van dit promotieonderzoek was er nog veel onduidelijkheid omtrent de toepasbaarheidvanmri.deaanbevelingdatallepatiëntenmetklachtenverdachtvoor axialespaeneennegatievebekkenfotoeenmrizoudenmoetenondergaan,konnog niet hard worden gemaakt. Een belangrijke vraag was of beenmergoedeem op MRI stabiel aanwezig blijft over de tijd. In hoofdstuk 6 onderzochten we daarom de verdeling,frequentievanvoorkomenendeontwikkelingoverdetijdvanbeenmerg oedeemopeenmrivandesacroiliacalegewrichtenbijpatiëntengeïncludeerdinde ESpAC.Bijnadehelftvandelokalisatiesmetbeenmergoedeemdatopbaselinewerd vastgesteld,verdweentijdensfollowup.omgekeerdwerdenernieuwelokalisatiesmet beenmergoedeem op een aantal MRI s tijdens followup vastgesteld, terwijl de voorafgaandemriopbaselinenegatiefwas.inbeidegevallenhadditgevolgenvoor hetaldannietvoldoenaandeasasaxialespaclassificatiecriteria.ditlaatziendathet toepassen van beeldvormende technieken om een reumatische aandoening te diagnosticeren ook beperkingen met zich meebrengt. Bevindingen bij beeldvorming zijn immers altijd onderhevig aan meetfouten van beoordelaars en variatie tussen beoordelaars; beeldvorming moet daarom geen belangrijkere rol toebedeeld krijgen dandebevindingentijdensanamneseenklinischonderzoek. Blindstaren opeenmri uitslagkan bijvoorbeeld leidentoteen vertraging in hetstellen van de diagnose van axialespa,wanneercrucialebevindingenopmritijdelijkafwezigzijnbijeenpatiënt diewelpastinhetklinischprofiel.omgekeerd,eenmriuitslagkanookleidentoteen foutpositievediagnosevanspa,wanneeraansubtieleenweinigsuggestievelaesiesop MRI ( onschuldige witte plekjes ) veel waarde wordt gehecht terwijl de klinische symptomenvandepatiënthiereigenlijkgeenaanleidingtoegeven.allemanifestaties diewordengezienbijaxialespamoetendaarominoverwegingwordengenomenbij debeoordelingvaneenpatiënt,waaronderbijvoorbeeldpsoriasis,enthesitisofuveitis. Dit gegeven komt eveneens terug in een recent ontwikkeld diagnostisch algoritme waarineengedegenanamneseenlichamelijkonderzoekonmisbarestappenzijnbijde beoordeling van een patiënt met verdenking op axiale SpA.[24] Alleen wanneer er voldoende verdenking is op axiale SpA, kan aanvullende beeldvorming worden aangevraagd. DEELIII:HERKENNINGVANSPONDYLOARTRITISINDE HUISARTSENPRAKTIJK DereumatoloogismededankzijdeMRIinstaatgestelddediagnoseSpAineeneerder stadiumtestellen.patiëntenmeteenverdenkingopspamoetenhiervoorweleerst

181 Samenvatting 181 tijdignaardereumatoloogverwezenworden.hetoptimaliserenvandezorgrondom patiënten met SpA begint dus altijd met zogenaamde case finding. Dit proces start meestal in de eerste lijn. In hoofdstuk 7 werden zowel het kennisniveau als de percepties van huisartsen over inflammatoire rugpijn en axiale SpA bestudeerd. Daarnaast werd gekeken welke potentiële barrières een tijdige verwijzing van patiëntenmetmogelijkaxialespaverhinderen.wetoondenaandataxialespavaak nietgoedindeeerstelijnwordtherkend.zelfswanneerhetwelwordtherkend,danis de zorg rondom deze patiënten vaak suboptimaal. In het algemeen was de kennis omtrent inflammatoire rugpijn, axiale SpA en extraarticulaire manifestaties beperkt. Huisartsen herkenden veelal slechts de klassieke, maar langetermijngevolgen van axialespa,zoalsbijvoorbeeldeenkyfosediewordtgezienbijas.slechtsdehelftvan degeïnterviewdehuisartsenwistdattumornecrosisfactor(tnf)alphablokkerende biologicals, een erkende en effectieve behandeling bij een specifieke groep van patiënten met bij axiale SpA, een behandeloptie zijn bij deze patiëntengroep. Vergroten van kennis bij huisartsen aangaande de symptomen en behandeling van axialespa,zalwaarschijnlijkdeverwijzingvanpatiëntenmetklachtenverdachtvoor axialespabevorderen. In de dagelijkse praktijk moet de huisarts adequaat kunnen anticiperen op velerlei lichamelijkeenpsychischeklachtenwaarpatiëntenzichmeepresenteren.herkenning van patiënten met SpA in de huisartsenpraktijk kan lastig zijn. Chronische rugpijn is immerseenveelvoorkomendsymptoomindealgemenepopulatieenisvoorpatiënten ookvaakeenredenomdehuisartstebezoeken.slechtsin5%vandegevallenisaxiale SpAdeonderliggendeoorzaak.[37]Hetfinetunenvandeprocedureomdepatiëntente selecteren die wel dienen te worden verwezen naar de tweedelijn, is echter een belangrijkenogtemakenstap.eenrecentestudieheeftlatenziendatbijnaeenkwart van de patiënten met chronische rugpijn die ontstond voor de leeftijd van 45 jaar geclassificeerd kon worden met axiale SpA nadat de huisarts deze patiënten had verwezennaardereumatoloog.[38] DehuisartsisonmisbaarbijhetherkennenvanpatiëntenmetSpA.Huisartsenverkeren in een unieke positie: ze kunnen longitudinale continue zorg leveren in overeen stemmingmetoverigebehoeftenvaneenpatiënt.ditbetekentdatzeinstaatzijnom hetpatroonkenmerkendvooraxialespa,bijvoorbeeldachtereenvolgensoptredenvan psoriasisenrugpijn,alseersteteherkennenwanneereenpatiëntdepraktijkbezoekt. In hoofdstuk 8 onderzochten we in hoeverre huisartsen en huisartsen in opleiding (HAIO)vroegeaxialeenperifereSpAherkennenenverwijzen.Wemaaktengebruikvan simulatiepatiënten om te onderzoeken hoe het er in de praktijk echt aan toe gaat.

182 182 Daarbij keken we ook naar de invloed van educatie op de mate van herkenning en uiteindelijk verwijzing. In totaal werden 68 huisartsen en HAIO s bezocht door simulatiepatiëntendiewarengetraindomeencasusvanvroegeaxialedanwelperifere SpA te simuleren. We bevestigden dat zowel de herkenning als verwijzing van patiëntenmetklachtensuggestiefvoorspaslechtsbeperktwas.educatieverbeterde duidelijkdeherkenningenverwijzingvanpatiëntennaardereumatoloog.hetgeven van onderwijs lijkt daarmee een belangrijke kans om de kennis wat betreft SpA te verbeteren, waardoor een succesvolle implementatie van een verwijsstrategie dichterbijkomt. PERSPECTIEF In hoofdstuk 9 werden bevindingen uit dit proefschrift samengevat en enkele methodologische aspecten verder bediscussieerd. Ten eerste werd stilgestaan bij de generaliseerbaarheid van de resultaten die zijn verkregen uit het ESpAConderzoek. PatiëntendiewerdengeïncludeerdindeESpACwarenreedsverdachtophethebben van SpA enwaren verwezen door de reumatoloog of door medisch specialisten die extraarticulaire manifestaties van SpA behandelen (waaronder dermatologen, oogartsen).veelpatiënteninespacpresenteerdenzichdusmetduidelijkeklachten, zoalsbijvoorbeeld inflammatoirerugpijnincombinatiemetpsoriasisofuveitis.door patiëntenlangsdezerouteteincluderenishetwaarschijnlijkdatsommigebevindingen niet meteen te generaliseren zijn naar patiënten met een minder duidelijk klinisch beeld, bijvoorbeeld patiënten met uitsluitend inflammatoire rugpijn in de huisartsenpraktijk.daarentegen,onzeprimaireonderzoeksvraagwashoeafwijkingen opdemriverdachtvoorsacroiliitiszichoverdetijdontwikkelen.hetincluderenvan patiëntenmeteenhoge pakkans vergrootdandeefficiëntieeninternevaliditeit. Ten tweede werden de voor en nadelen van het gebruik van simulatiepatiënten in onderzoekbesproken.doorsimulatiepatiëntenintezettenwarenwijinstaattekijken of educatie uiteindelijk ook een gedragsverandering in gang zet. Dit gaat dus verder dan hetafnemen vaneentoets naeeneducatiemoment. Onzeresultatenlaten zien dat simulatiepatiënten ook ingezet kunnen worden bij meer ingewikkelde medische problemen zoals bijvoorbeeld SpA. Er zijn dan wel meer voorbereidingsmaatregelen nodig. Specifiek in ons onderzoek was het bijvoorbeeld belangrijk om een goede voorgeschiedenisoptestelleneneennauwkeurigebeschrijvingtegevenvandereeds ondernomen stappen, zoals het effect van eerder ontvangen fysiotherapie op de klachten.

183 Samenvatting 183 Ten slotte werden er aanbevelingen gedaan voor de klinische praktijk en verder onderzoek. Met de introductie van de MRI van de sacroiliacale gewrichten is de reumatoloog in staat gesteld om patiënten met SpA eerder een diagnose te kunnen gevenenhierdoor,indiennodig,tijdigtestartenmeteenbehandeling.hetonderzoek beschreveninditproefschriftkanreumatologenhelpenhoezedemriindedagelijkse praktijk kunnen inzetten. Meer onderzoek is echter nodig om onder andere de diagnostischewaardevanstructureleveranderingenopeenmrivasttestellen. Snel instellen van een behandeling leidt mogelijk tot minder werkverzuim en arbeidsongeschiktheid. Om het gunstige effect van een vroege diagnose op de verschillendeuitkomstmatenoptimaaltebenutten,ishetwelbelangrijkdatpatiënten ooksnelwordenverwezen.wijhebbenlatenziendateducatieaanhuisartsenhierbij kan helpen. Meer onderzoek is nodig om te kijken welke onderwijsactiviteiten bijdragenaaneensuccesvolleverwijsstrategieis.bijhetbewerkstelligenvanditdoelis hetbelangrijkterealiserendatreumatologenenhuisartsenhierbijvanelkaarkunnen leren. Beide medische disciplines vullen elkaar mooi aan wanneer het gaat om het leverenvankwalitatiefgoedezorgvoorpatiëntenmetspa.

184 184 LITERATUUR 1. BijlsmaJWJ,vanLaarJM.Reumatologieenklinischeimmunologie.Houten:BohnStafleuvanLoghum, RudwaleitM.Newapproachestodiagnosisandclassificationofaxialandperipheralspondyloarthritis. CurrOpinRheumatol2010;22: Stolwijk C, van Tubergen A, CastilloOrtiz JD, et al. Prevalence of extraarticular manifestations in patientswithankylosingspondylitis:asystematicreviewandmetaanalysis.annrheumdis2013.doi: /annrheumdis FeldtkellerE,KhanMA,HeijdeDvander,etal.AgeatonsetanddiagnosisdelayinHLAB27negativevs positivepatientswithankylosingspondylitis.rheumatolint2003;23: Rudwaleit M, Khan MA, Sieper J. The challenge of diagnosis and classification in early ankylosing spondylitis:doweneednewcriteria?arthritisrheum2005;52: RudwaleitM,JurikAG,HermannKG,etal.Definingactivesacroiliitisonmagneticresonanceimaging (MRI) for classification of axial spondyloarthritis: a consensual approach by the ASAS/OMERACT MRI group.annrheumdis2009;68: Feldtkeller E, Bruckel J, Khan MA. Scientific contributions of ankylosing spondylitis patient advocacy groups.curropinrheumatol2000;12: KeatAC,GaffneyK,GilbertAK,etal.Influenceofbiologictherapyonreturntoworkinpeoplewithwork disabilityduetoankylosingspondylitis.rheumatology(oxford)2008;47: PrinceDS,McGuiganLE,McGirrEE.Workinglifeandphysicalactivityinankylosingspondylitispreand postantitumornecrosisfactoralphatherapy.intjrheumdis2014;17: VanderHeijdeD,HanC,DeVlamK,etal.Infliximabimprovesproductivityandreducesworkdaylossin patients with ankylosing spondylitis: results from a randomized, placebocontrolled trial. Arthritis Rheum2006;55: Rudwaleit M, Sieper J. Referral strategies for early diagnosis of axial spondyloarthritis. Nat Rev Rheumatol2012;8: BrandtHC,SpillerI,SongIH,etal.Performanceofreferralrecommendationsinpatientswithchronic backpainandsuspectedaxialspondyloarthritis.annrheumdis2007;66: PoddubnyyD,VahldiekJ,SpillerI,etal.Evaluationof2screeningstrategiesforearlyidentificationof patientswithaxialspondyloarthritisinprimarycare.jrheumatol2011;38: Hermann J, Giessauf H, Schaffler G, et al. Early spondyloarthritis: usefulness of clinical screening. Rheumatology(Oxford)2009;48: Sieper J, Srinivasan S, Zamani O, et al. Comparison of two referral strategies for diagnosis of axial spondyloarthritis:therecognisinganddiagnosingankylosingspondylitisreliably(radar)study.ann RheumDis2013;72: JoisRN,MacgregorAJ,GaffneyK.Recognitionofinflammatorybackpainandankylosingspondylitisin primarycare.rheumatology(oxford)2008;47: RudwaleitM,vanderHeijdeD,LandewéR,etal.ThedevelopmentofAssessmentofSpondyloarthritis international society classification criteria for axial spondyloarthritis (part II): validation and final selection.annrheumdis2009;68: Rudwaleit M, van der Heijde D, Landewé R, et al. The Assessment of SpondyloArthritis International Societyclassificationcriteriaforperipheralspondyloarthritisandforspondyloarthritisingeneral.Ann RheumDis2011;70: RudwaleitM.Newapproachestodiagnosisandclassificationofaxialandperipheralspondyloarthritis. CurrOpinRheumatol2010;22: Rudwaleit M, Khan MA, Sieper J. The challenge of diagnosis and classification in early ankylosing spondylitis:doweneednewcriteria?arthritisrheum2005;52: Reveille JD, Witter JP, Weisman MH. Prevalence of axial spondyloarthritis in the United States: estimatesfromacrosssectionalsurvey.arthritiscareres(hoboken)2012;64: BraunJ,SieperJ,BollowM.Imagingofsacroiliitis.ClinRheumatol2000;19:517.

185 Samenvatting Van den Berg R, de Hooge M, Rudwaleit M, et al. ASAS modification of the Berlin algorithm for diagnosingaxialspondyloarthritis:resultsfromthespondyloarthritiscaughtearly(space)cohortand from the Assessment of SpondyloArthritis international Society (ASAS)cohort. Ann Rheum Dis 2013;72: Yu D, Van Tubergen A. Clinical manifestations of ankylosing spondylitis in adults. In: DS Basow (Ed.), UpToDate.WalthamMA, Weber U, Ostergaard M, Lambert RG, et al. Candidate lesionbased criteria for defining a positive sacroiliacjointmriintwocohortsofpatientswithaxialspondyloarthritis.annrheumdis2014:doi: /annrheumdis MaksymowychWP,ElewautD,SchettG.Motionfordebate:thedevelopmentofankylosisinankylosing spondylitisislargelydependentoninflammation.arthritisrheum2012;64: Maksymowych WP, Chiowchanwisawakit P, Clare T, et al. Inflammatory lesions of the spine on magnetic resonance imaging predict the development of new syndesmophytes in ankylosing spondylitis:evidenceofarelationshipbetweeninflammationandnewboneformation.arthritisrheum 2009;60: Wanders A, van der Heijde D, Landewe R, et al. Nonsteroidal antiinflammatory drugs reduce radiographic progression in patients with ankylosing spondylitis: a randomized clinical trial. Arthritis Rheum2005;52: VanderHeijdeD,LandeweR,EinsteinS,etal.Radiographicprogressionofankylosingspondylitisafter uptotwoyearsoftreatmentwithetanercept.arthritisrheum2008;58: VanderHeijdeD,LandeweR,BaraliakosX,etal,andtheAnkylosingSpondylitisStudyfortheEvaluation ofrecombinantinfliximabtherapystudygroup.radiographicfindingsfollowingtwoyearsofinfliximab therapyinpatientswithankylosingspondylitis.arthritisrheum2008;58: VanderHeijdeD,SalonenD,WeissmanBN,etal.Assessmentofradiographicprogressioninthespines ofpatientswithankylosingspondylitistreatedwithadalimumabforupto2years.arthritisresther 2009;11:R VanderHeijdeD,MachadoP,BraunJ,etal.MRIinflammationatthevertebralunitonlymarginally predicts new syndesmophyte formation: a multilevel analysis in patients with ankylosing spondylitis. AnnRheumDis2012;71: LoriesRJ,DereseI,DeBariC,etal.Evidenceforuncouplingofinflammationandjointremodelingina mousemodelofspondylarthritis.arthritisrheum2007;56: SongIH,HermannKG,HaibelH,etal.Relationshipbetweenactiveinflammatorylesionsinthespine and sacroiliac joints and new development of chronic lesions on wholebody MRI in early axial spondyloarthritis:resultsoftheesthertrialatweek48.annrheumdis2011;70: Chiowchanwisawakit P, Lambert RG, ConnerSpady B, et al. Focal fat lesions at vertebral corners on magnetic resonance imaging predict the development of new syndesmophytes in ankylosing spondylitis.arthritisrheum2011;63: DeHoogeM,vandenBergR,NavarroCompánV,etal.Magneticresonanceimagingofthesacroiliac jointsintheearlydetectionofspondyloarthritis:noaddedvalueofgadoliniumcomparedwithshorttau inversionrecoverysequence.rheumatology(oxford)2013;52: UnderwoodMR,DawesP.Inflammatorybackpaininprimarycare.BrJRheumatol1995;34: VanHoevenL,LuimeJ,HanH,etal.Identifyingaxialspondyloarthritisindutchprimarycarepatients, ages2045years,withchroniclowbackpain.arthritiscareres(hoboken)2014;66:44653.

186

187 Valorisationaddendum

188

189 Valorisationaddendum 189 VALORISATIONADDENDUM Valorisation refers to the process of how academic research can be utilized and translatedtoclinicalandsocietalbenefit.therelevanceoftheresearchpresentedin thisthesisisdescribedinthesectionbelow. PartI:epidemiologyofspondyloarthritis Quantificationoftheburdenofrheumaticconditionsisimportantforraisingawareness among health care professionals, setting research priorities and initiating a policy debate.[1]rheumaticconditionshavemajorimpactontheindividualpatient,butalso onsociety.alsointhecaseofspondyloarthritis(spa),importantdecreasesinalmostall aspectsofhealthrelatedqualityoflifearereported.theonsetatarelativelyyoung age,beforethefourthdecade,addstotheyearslivedindisabilityforanindividual.[2] Asaconsequenceofdecreasedfunctioning,SpAhasanadverseimpactonthepatient and family by reduced participation in social roles.[3,4] The indirect costs related to SpAarefourtimesashighasthedirectcosts,reflectingtheimportantimpactofthe disease on work participation in terms of sick leave, disability pensions and early retirement.[5] To extrapolate how this individual burden would affect society, appropriatedataontheepidemiologyarewarranted.inthisthesis,wefoundthatthe global prevalence of SpA was 0.55% (95% Confidence Interval (CI): ). This prevalenceiscomparabletothatofrheumatoidarthritis.substantialvariationacross geographicregionswashoweverfound.forinstance,theprevalenceofspaineastasia was 0.79% (CI: ). The study described in this thesis adds to the available evidence on epidemiology of rheumatic diseases that can contribute to prioritize research,butalsotoinformhealthcaresystemsatthecountrylevel,whenallocating budgetstoimprovediagnosis,treatmentandpreventionofworkdisability.[1,68] PartII:theuseofMRIinearlydetectionofaxialspondyloarthritis TheintroductionofMRIfordetectingactivesacroiliitishasrevolutionizedthediagnosis of axial SpA (axspa), making an early diagnosis possible. This thesis showed that a positivemriisareliablefinding:apositivemriatbaselinewasstronglyassociatedwith a positive MRI of the sacroiliac joints (MRISIJ) over time, particularly in HLAB27 positivepatients.thisisimportantforclinicians,becauseitsuggeststhatadiagnosisof axspa incorporating a positive MRI is robust and credible. Furthermore, the post gadoliniumdiethylenetriaminepentaaceticacid(gddtpa)mrisequencecanbesafely omitted,whichincreasesthefeasibilityofmri,sinceitsavestimeandreducescosts.

190 190 Further, in axspa it is a challenge to identify the appropriate target for treatment: shoulditbediseaseactivity,preventionordelayinprogressionofstructuraldamage, or both? In this thesis, we focused on the relation between inflammation and developmentofstructuraldamage(erosionsandfattylesions)onmri.wehaveshown thatfattylesionsonmrisijpreferablydevelopafterbonemarrowedema(bme)has subsided.otherstudieshavesuggestedthatfattylesionsatvertebraledgespredictthe developmentofnewsyndesmophytes.[9]bmemaythereforebethefirstdominothat setsoffthechainthatleadstodevelopmentoffattylesionsandeventuallynewbone formation (syndesmophytes).[10,11] The true relation between BME and new bone formationstillneedstobedisentangledinfurtherresearch.thisthesishascontributed inoneofthemanystepstounraveltherelationbetweeninflammationandstructural damageonmriinordertoenableidentificationofdifferenttreatmenttargets. PartIII:earlyidentificationofspondyloarthritisinprimarycare Musculoskeletaldisorders(MSD)areamongthemostcommonreasonsforconsultinga GPandhaveamajorimpactonhealthcareresources.[12]TheGlobalBurdenofDisease 2010 study found that MSD, including rheumatic disorders, were the second main contributor to the number of years lived with disability.[1] Part of this burden is avoidable.theimportanceofmsdasmajorcauseof(avoidable)disability,however, seems insufficiently acknowledged by GPs.[13,14] Training in rheumatology is rarely mandatoryingeneralpracticetrainingprograms,despitethelargenumberofpatients thatpresentthemselvesinprimarycarewithrheumaticdisorders.[15]nonetheless,it seemsthatmusculoskeletalconditionsarenotapriorityinprimarycare.[16,17] MakingadiagnosisofaxSpAisoftendelayedupto10yearsorlonger,suggestingthat opportunities for early recognition and referral have been missed in primary care.[18,19]severalreferralstrategiesthatpromoteearlyreferralofaxspahavebeen developed, but successful implementation may be hampered by ineffective referral patterns due to lack of knowledgeaboutaxspa.[2023]from thisthesis, we learned thatthereisroomforimprovementwithregardtothelevelofknowledgeofgpsabout their ability to identify and refer patients with suspected axspa. Changing clinical practice behavior and assessing such change is a real challenge. The evidence that educationalinterventionsmayactuallychangeanythingislimited.[24]lackoftimeand resources often contribute to failure of the education intervention.[24] This thesis shows that the use of standardized patients (SPs), is a feasible and informative approach to assess the impact of an educational intervention. More importantly, a multifacetededucationalprogramcanplayakeyroleinimprovingdiseaserecognition

191 Valorisationaddendum 191 andreferralofpatientssuspectedforspa.thisimportantfindingmayfurtherimprove timelydiagnosisandinitiationoftreatmentofpatientswithspa.

192 192 REFERENCES 1. VosT,FlaxmanAD,NaghaviM,etal.Yearslivedwithdisability(YLDs)for1160sequelaeof289diseases and injuries : a systematic analysis for the Global Burden of Disease Study Lancet 2012;380: FeldtkellerE,KhanMA,vanderHeijdeD,etal.AgeatdiseaseonsetanddiagnosisdelayinHLAB27 negativevs.positivepatientswithankylosingspondylitis.rheumatolint2003;23: BoonenA,vanderLindenSM.Theburdenofankylosingspondylitis.JRheumatol2006;78: Boonen A. A review of workparticipation, costofillness and costeffectiveness studies in ankylosing spondylitis.natclinpractrheumatol2006;2: AkessonK,DreinhöferKE,WoolfAD.Improvededucationinmusculoskeletalconditionsisnecessaryfor alldoctors.bullworldhealthorgan2003;81: Badley EM, Rasooly I, Webster GK. Relative importance of musculoskeletal disorders as a cause of chronichealthproblems,disability,andhealthcareutilization:findingsfromthe1990ontariohealth Survey.JRheumatol1994;21: March L, Smith EU, Hoy DG, et al. Burden of disability due to musculoskeletal (MSK) disorders. Best PractResClinRheumatol2014;28: HoyD,MarchL,BrooksP,etal.Theglobalburdenoflowbackpain:estimatesfromtheGlobalBurden ofdisease2010study.annrheumdis2014;73: Chiowchanwisawakit P, Lambert RG, ConnerSpady B, et al. Focal fat lesions at vertebral corners on magnetic resonance imaging predict the development of new syndesmophytes in ankylosing spondylitis.arthritisrheum2011;63: Maksymowych WP, Chiowchanwisawakit P, Clare T, et al. Inflammatory lesions of the spine on magnetic resonance imaging predict the development of new syndesmophytes in ankylosing spondylitis:evidenceofarelationshipbetweeninflammationandnewboneformation.arthritisrheum 2009;60: SongIH,HermannKG,HaibelH,etal.Relationshipbetweenactiveinflammatorylesionsinthespine and sacroiliac joints and new development of chronic lesions on wholebody MRI in early axial spondyloarthritis:resultsoftheesthertrialatweek48.annrheumdis2011;70: Jordan KP, Kadam UT, Hayward R, et al. Annual consultation prevalence of regional musculoskeletal problemsinprimarycare:anobservationalstudy.bmcmusculoskeletdisord2010;11: WoolfAD,AkessonK.Understandingtheburdenofmusculoskeletalconditions.Theburdenishugeand notreflectedinnationalhealthpriorities.bmj2001;322: WoolfAD,AkessonK.Canwereducetheburdenofmusculoskeletalconditions?TheEuropeanaction towardsbettermusculoskeletalhealth.bestpractresclinrheumatol2007;21: RaskerJJ.Rheumatologyingeneralpractice.BrJRheumatol1995;34: HosieGA.Teachingrheumatologyinprimarycare.AnnRheumDis2000;59: Lanyon P, Pope D, Croft P. Rheumatology education and management skills in general practice: a nationalstudyoftrainees.annrheumdis1995;54: JoisRN,MacgregorAJ,GaffneyK.Recognitionofinflammatorybackpainandankylosingspondylitisin primarycare.rheumatology(oxford)2008;47; FeldtkellerE,KhanMA,vanderHeijdeD,etal.AgeatdiseaseonsetanddiagnosisdelayinHLAB27 negativevs.positivepatientswithankylosingspondylitis.rheumatolint2003;23: PoddubnyyD,VahldiekJ,SpillerI,etal.Evaluationof2screeningstrategiesforearlyidentificationof patientswithaxialspondyloarthritisinprimarycare.jrheumatol2011;38: BrandtHC,SpillerI,SongIH,etal.Performanceofreferralrecommendationsinpatientswithchronic backpainandsuspectedaxialspondyloarthritis.annrheumdis2007;66: Hermann J, Giessauf H, Schaffler G, et al. Early spondyloarthritis: usefulness of clinical screening. Rheumatology(Oxford)2009;48: Sieper J, Srinivasan S, Zamani O, et al. Comparison of two referral strategies for diagnosis of axial spondyloarthritis:therecognisinganddiagnosingankylosingspondylitisreliably(radar)study.ann RheumDis2013;72: Grol R, Wensing M. What drives change? Barriers to and incentives for achieving evidencebased practice.medjaust2004;180:s5760.

193 Dankwoord

194

195 Dankwoord 195 DANKWOORD Hetisaf.Zondereenofficieelpromotietrajectisditproefschrifterdantochgekomen. Incombinatiemetmijnopleidingtotreumatoloogwasditnietaltijdgemakkelijk,maar methulpvanveelmensenishetuiteindelijktochgelukt.dankaaniedereendiemijmet zijn/haar bijdrage heeft geholpen. Ik wil deze pagina s gebruiken om een aantal menseninhetbijzondertebedanken. Ten eerste wil ik alle patiënten van het ESpAC cohort bedanken voor hun medewerking, zonder jullie zou er voor mij niet veel te onderzoeken zijn geweest. Daarnaast wil ik alle simulatiepatiënten, huisartsen en huisartseninopleiding bedanken. Dankzij jullie enthousiasme en inzet ben ik in staat gesteld om veel belangrijkegegevensvoormijnproefschriftteverzamelen. Prof.dr.Landewé,besteRobert,tijdensmijnWESPstageopdeafdelingreumatologie hebjemekennislatenmakenmethetdoenvanonderzoek.daarnaasthebjeervoor gezorgddatikhetonderzoekkonvoortzettentijdensmijnopleiding.ikhebhiergeen moment spijt van gehad. Ik hoop dat jij dat ook niet hebt gehad Zeker omdat het allemaalietsmeertijdheeftgekostdannormaal.ikhebveelvanjegeleerdenwilje bedankenvoorjevertrouwen,geduld,kritischeblikenhumor. Prof. dr. Boonen, beste Annelies, wat ben ik blij dat ik nu steeds meer met je kan samenwerken! Ondanks je eigen drukke programma ben je altijd bereid om mee te denkenenhelpjemeweeropweg.jebenteengroteinspiratiebronvoormij.ikben heelergblijentrotsdatikdekanshebgekregenommeinmaastrichtverdertemogen ontwikkelenalsreumatoloogenonderzoeker. Dr.vanTubergen,besteAstrid,jebentdeafgelopenperiodeechteensteunentoever laatvoormegeweest.ikhebveelbewonderingvoorjekennis,scherpeanalysesenje vermogenomditallesduidelijkuitteleggen.inmiddelsdeelikeenkamermetje.nuik weermeertijdheb,benikzekervanplanomdiewandmetboeken/formulieren/stof eensopteruimenendekameropnieuwinterichten.deplantmagblijvenstaan,. Prof. dr. van der Heijde, beste Désirée, eigenlijk heb jij het fundament voor dit proefschrift gelegd, onder andere als initiator van het ESpAC cohort. Ik bewonder je

196 196 driveeninzetalswetenschapperenwiljebedankenvoorjegeduldommijnartikelen telezenendezemetmetebespreken. Ikwilgraagdeledenvandebeoordelingscommissie,Prof.dr.J.E.Wildberger,Prof.dr. R.A. de Bie, Prof. dr. F. van den Bosch, Prof. dr. J.A. Knottnerus, Dr. A.E.A.M. Weel bedankenvoorhetinhoudelijkcommentaar,adviezenendegoedkeuring. Veel mensen hebben essentiële bijdrage geleverd aan de totstandkoming van verschillende artikelen. Dr. Jurik, dear Anne Grethe, I would like to thank you for helping me to score all those MRIs, in only four days. We finished scoring just ten minutesbeforeihadtoleavetocatchmyflightbackhome!ienjoyedworkingtogether withyou.dr.heuft,liesbeth,dankzijjouwinvesteringinhetespaccohortbenikin staat gesteld om verder aan de slag te gaan met de followup gegevens. Dank daarvoor! Dr. Gorter, Dr. Maiburg, Drs. Waagenaar, beste Simone, Bas en Gerrie, bedankt voor de prettige samenwerking en jullie enthousiasme bij het opzetten en uitvoeren van het simulatiepatiëntenproject. Aniek en Carmen, het was erg leuk en gezelligommetjullieonderzoektemogendoen. Een proefschrift kan natuurlijk alleen maar worden geschreven als je daarvoor de ruimte en mogelijkheden krijgt. Debby, Thea, Caroline, Sandrine, bedankt voor jullie concrete hulp of steuntje in de rug op het juiste moment. Debby, ik wil jou in het bijzonder ookbedanken voor de kans die je mij hebt geboden om me in Maastricht verdertemogenontwikkelen.jemoetnutocheensgaanwinkelenvooreenmooietas, je hebt hem inmiddels wel verdiend. Caroline en Sandrine, eerst collegaaios, nu collegareumatologenenmijnparanimfen.ikbenergblijdatjullie achtermestaan op dedagzelf.onzemaandelijkseetentjesmoetenwenuechtvolzientehouden.thea,ik wiljebedankenvooraljehulpophetgebiedvanpatiëntenzorg,ikhebdaaraltijdveel steunaangehad.els,janine,mehmet,andyenyvonne,dankvoorjullieondersteuning bijhetcombinerenvanonderzoekenpatiëntenzorg. Prof.dr.vanderLinden,besteSjef,ikwilubedankenvooruwnuchtereadviezenen vertrouwen in mij. Dankzij u kon ik starten met de opleiding tot reumatoloog, een prachtigcadeau! Mijn opleiding heb ik mogen voltooien in Heerlen. Wat heb ik daar een leuke tijd gehad! Ik wil graag alle reumatologen, AIOS reumatologie, reumaconsulenten en polimedewerkers bedanken. Mede dankzij jullie inzet heb ik dat jaar echt tempo

197 Dankwoord 197 kunnen maken en veel werk kunnen verzetten. Een bijzonder woord van dank aan Mirian en Ralph, dankzij jullie uitstekende begeleiding kunnen AIOS in Heerlen het beste uit zichzelf halen. Jullie hebben me in de eindfase van mijn opleiding klaargestoomdvoorhetvak!femke,dankvoorjehulpbijhetverwerkeneninvoeren vandata. TijdensmijnopleidinghebikmogensamenwerkenmeteenclubvansuperAIOS.We hebben heel veel lol gehad. Een aantal heb ik hierboven al genoemd. Kristof, we hebbenhetzelfdegevoelvoorhumorenikmisnunogweleensjouweigenvisieopde werelddiejealtijdbereidbentmetanderentedelen.merdan,wemoestenvoorhet eerstsamenwerkentoenweallebeiderdejaarsgeneeskundestudentwaren.wiehad gedachtdatwetienjaarlaterallebeireumatoloogzoudenzijn?ward,bedanktvoorje opbeurende woorden als iets weer niet helemaal ging zoals ik van te voren had bedacht. Inmiddels zijn we allemaal klaar met de opleiding en werken we in verschillendeziekenhuizen,maarikweetzekerdatwecontactblijvenhouden. DaarnaastookdankaanalleAIOSenpromovendivandeafdelingreumatologie,voor deprettigesamenwerkingenjullieenthousiasme! Peggy,MarianenYvonne,julliestaangarantvooreensoepellopendsecretariaat.Wat kan ik me nog meer wensen? Jullie helpen me altijd perfect met de administratie. Dankzijjulliehebikgeleerddathetsomsechtbeteriswanneerjedezakenuithanden geeft. Edith, Maddy en Marjos, jullie zijn voor mij het gezicht van de polikliniek reumatologieenzorgenervoordatmijnspreekuursoepelloopt.erverandertnuveel op de polikliniek, maar ik weet zeker dat we er iets moois van gaan maken. Tiny, bedanktvoorjehulpbijdelayoutvanhetproefschrift. De eerste jaren van mijn opleiding heb ik doorgebracht op de afdeling interne geneeskundevanhetmumc.ikwilprof.dr.c.stehouwerenprof.dr.r.koopmans bedanken voor het verzorgen van een gedegen vooropleiding en de vrijheid om daarnaast ook onderzoek te kunnen blijven doen. Een speciaal woord van dank aan Evelien,Wubbo,RonaldenPatricia.Overlegmetjulliehebikaltijdalszeernuttigen waardevolervaren. Vrienden,(inmiddels)verwegmaartochookdichtbij.Julliehebbenmealtijddenodige gezelligheid en afleiding geboden. Floor, ik ken je al vanaf de eerste klas van de middelbare school. We hebben samen zowat heel Europa gezien. Ik ben blij dat we

198 198 ondanksdereisafstandelkaarregelmatigblijvenzienomgezelligbijtekletsen.cécile, volgensmijhebbenweintotaalmaartweewekensamengewerktopb5,maarwezijn altijdcontactblijvenhouden.ikwiljebedankenvoorjevriendschap.ikkijkuitnaarde nog komende theeenkoffiemomentjes. Sofia, je toont altijd interesse voor mijn bezighedenenbentimmerbereidtehelpen.bedanktdaarvoor! Endan.familie.Tantes,ooms,nevenennichten,bedanktvoorjulliebetrokkenheid. Berdien en Corrie, nogmaals dank voor jullie hulp bij het opknappen van mijn huis. Dankzij jullie heb ik mijn proefschrift kunnen schrijven met een door jullie geverfde muur(schoonenwit)alsprettiguitzicht. RogierenRoma,dankvoordegesprekkendienietoverhetproefschriftgingen. Lievepapaenmama,hartelijkdankvoorjullieonvoorwaardelijkesteun,adviezenen vertrouwen.zonderjulliehadikditnooitbereikt.nuhetproefschriftafis,isdatvoor julliedenkikookeenhelezorgminder.uitdegrondvanmijnhart,dankjulliewel!

199 CurriculumVitae

200

201 CurriculumVitae 201 CURRICULUMVITAE MarloesvanOnnawerdop2augustus1982geboreninBeuningen.Nahetbehalenvan haar gymnasiumdiploma aan het Kandinsky College in Nijmegen, studeerde zij van 2000tot2002GezondheidswetenschappenaandeuniversiteitvanMaastricht.In2002 besloot zij Geneeskunde te gaan studeren, omdat zij bij nader inzien toch geen onderzoeker maar dokter wilde worden. Tijdens een keuzecoschap Reumatologie enmaag,darmenleverziekteningentenvervolgenseenwetenschappelijkestage opdeafdelingreumatologieinhetmaastrichtuniversitairmedischcentrum(mumc) werdhaarinteressevoordereumatologiegewekt.zebehaaldehaarartsenbulin2008 cumlaude. In2009begonzijaanhaarvooropleidingInterneGeneeskundeinhetMUMC(opleider: Prof.dr.C.Stehouwer).Inhetzelfdejaarstarttezijeveneensmethetonderzoekwelk uiteindelijkheeftgeresulteerdinditproefschrift.deresultatenzoalsbeschrevenindit proefschrift werden gepresenteerd op verschillende nationale en internationale congressen.devervolgopleidingtotreumatoloogwerdin2012voortgezetinzowelhet MUMCalshetZuyderlandZiekenhuisinHeerlen(opleiders:Prof.dr.S.vanderLinden, Dr.D.VosseenDr.R.Peeters). Sinds2015werktzijalsstaflidReumatologieinhetMUMC.

202