Cardiac Sarcoidosis and Giant Cell Myocarditis in Finland

HELSINKI UNIVERSITY HOSPITALDIVISION OF CARDIOLOGYHEART AND LUNG CENTERFACULTY OF MEDICINEDOCTORAL PROGRAMME IN CLINICAL RESEARCHUNIVERSITY OF HELSINKI

RIINA KANDOLIN

DISSERTATIONES SCHOLAE DOCTORALIS AD SANITATEM INVESTIGANDAM UNIVERSITATIS HELSINKIENSIS 104/2015

104/2015

Helsinki 2015 ISSN 2342-3161 ISBN 978-951-51-1768-7

RIIN

A K

AN

DO

LIN

Card

iac Sarcoid

osis

Gian

t C

ell

Myocard

itis in

Fin

land

Recent Publications in this Series

85/2015 And DemirClinical Use of Urinary Gonadotropin Determinations in Children and Adolescents86/2015 Raisa HaavikkoSynthesis of Betulin Derivatives with New Bioactivities87/2015 Hongxin ChenHeterobasidion annosum sensu stricto Pathogenesis: Bioinformatic and Functional Study of Cerato-Platanin Family Proteins88/2015 Kristiina RajamäkiRole of the Infl ammasome Pathway in Atherosclerosis89/2015 Pirkko KriikkuToxicological Abuse Profi le of New Recreational Drugs in Driving-Under-The-Infl uence and Post-Mortem Cases in Finland90/2015 Eeva MustonenRed Clover Isofl avonoids in Feed, Plasma and Milk of Ruminants91/2015 Flavia RestituttiCentral, Cardiovascular and Metabolic Effects of Dexmedetomidine Associated with the Selective Peripheral Alpha2-adrenoceptor Antagonist MK-467 in Dogs92/2015 Katariina NurmiPrevention of Infl ammatory Cellular Responses by Ethanol and Hemin – Interplay Between Infl ammasomes and Processes Inhibiting Infl ammation93/2015 Liang HeRare Genetic Variants and Complex Diseases - a Bayesian Approach94/2015 Jani SaksiStroke-Associated Carotid Atherosclerosis - a Role for FABP4 in Atheroma Vulnerability and Cardiometabolic Risk95/2015 Otto HemminkiCancer Immunotherapy with a Gene Modifi ed Serotype 3 Oncolytic Adenovirus96/2015 Jukka KoskelaIndividual Trajectories in Asthma and COPD: A Longitudinal Perspective to Obstructive Lung Disease97/2015 Miira KlemettiTrends In Obstetric and Perinatal Outcomes of Women with Type 1 Diabetes During 1988-2011 – A Finnish Population-Based Observational Study98/2015 Ileana B. QuinteroPhysiological Functions of Prostatic Acid Phosphatase99/2015 Minna MatikainenLaunching New Products in the Finnish Pharmaceutical Industry: A Relationship Approach100/2015 Päivi Ylä-AnttilaPhagophore Membrane Connections and RAB24 in Autophagy101/2015 Kaisa KyöstiläIdentifi cation of Novel Mutations and Molecular Pathways for Canine Neurodegeneration and Chondrodysplasia102/2015 Emmi JoensuuEpigenetic Alterations in Sporadic and Familial Cancers103/2015 Elina ReponenPreoperative Risk-Assessment Methods, Short-Term Outcome, and Patient Satisfaction in Elective Cranial Neurosurgery

and

§

26

27

2.4.1ConductionabnormalitiesConductionabnormalitiesarethemostcommonmanifestationsofCSduetotheaffinityofgranulomas for theatrioventricularseptumand thevulnerabilityof theconductionsystemtogranulomatousinfiltration(111).Yetanotheroftenmentioned,yetprobablyraremechanismforconductiondisturbancesisgranulomatousinfiltrationofthenodalarterycausingischemiaoftheconductionsystem(112).

Virtually any kind of conduction abnormality is possible in CS from bundle branchblocks to AV-blocks of any degree (110). Different conduction abnormalities arereportedin42-65%ofpatients(5,16,22,77,93,97-99,101,113,114).CompleteAV-block,present in 15 to 45% in the larger studies, is the most frequent symptomatic firstmanifestationinCS(5,16,97,115).AV-blockduetosarcoidosisoccursatayoungeragecomparedtopatientswithAV-blockofothercauses(116).InaJapanesestudy,clinicallyorhistologicallyverifiedCSwasdiagnosedin11.2%ofpatientsmanifestingwithhigh-degreeAV-block(117).Bundlebranchblocksarepresentin27-57%,withrightbundlebranchblock(RBBB)beingthemorefrequentfinding(113,114).

2.4.2ArrhythmiasActive inflammation may lead to increased automaticity and fibrous granulomatousscars,promotingre-entrysubstratesbycreatingslowconductionzonesinmyocardium.Furthermore,thevariableinvolvementofLVand/orRVisacharacteristicfeatureofCS.The combination of these elements makes ventricular arrhythmias in CS not onlycommon, but also challenging to treat (98,118-120). Ventricular arrhythmias in CSrangefromasymptomaticventricularprematurebeats(VPBs)toventricularfibrillation(VF).Asinothertypesofstructurallyabnormalhearts,theVTsusuallyresultfromscar-induced re-entrymechanisms,with themost commonmechanism likely tobemacro-reentrant arrhythmias around granulomatous scars (121,122). The distinguishingfeatureinCSisahigherprevalenceofelectricalstorm,inwhichactiveinflammationcanplay a significant role (123). An electrical storm (defined as ≥ 3 appropriate ICDtherapiesin24hours)occurredin14%(16/112)ofCSpatientsina2.4-yearfollow-upafter an ICD implantation (98). Characteristically, CS-associated VT is monomorphicwith either LBBB or RBBB morphology, but polymorphic VTs are also possible(98,119,124,125). Electrophysiological studies have demonstrated that most VTsoriginate fromconfluentregionsofRVandLVscarringwithpredilectiontotheseptaland peritricuspid areas (122,126). One additional possible etiology for VTs in CS isPurkinje fiber-related VT, due to conduction disturbances (118). In the study byMcArdleetal.(127),sustainedVTswereassociatedwithsignsofactiveinflammationbyPET, while Banba et al. (124) did not find a link between disease activity andarrhythmias. In patients dying suddenly from CS, histologic lesions more oftencontainedwell-formedgranulomasinsteadofearlyinflammatorycelllesionsorhealedcollagen-richlesions(87).

28

29

Variable ventricular tachyarrhythmia incidence numbers of 11% to 31% have beenpreviously reported based on autopsy and clinical follow-up studies in CS(5,16,22,93,97,101).Recently,pacemaker follow-upstudieshaveyieldedmorepreciseincidence numbers. In a rather selected CS population from an academic centerwithICDsimplantedeitherforsecondaryprevention(withhistoryofVT,VF,suddencardiacarrest, or syncope of unknown etiology, n=16) or primary prevention (withoutaforementioned history, n=29), the incidence of appropriate ICD therapies (shockand/or anti-tachycardia pacing) was 15% per year and there was no significantdifference in the incidence rate between the two groups (128). Two larger studiespresentedcorrespondingICDtherapyratesof36.2%(85/234)and32.1%(36/112)inmeanfollow-upsof4.2years(129)and2.4years(98). In thesetwoseries,secondarypreventionpatientshadhigherratescomparedtoprimarypreventionpatients.Overall,the frequency of appropriate ICD therapies and hence ventricular arrhythmias wascalculated to be annually 9-15% (4), beingmuch higher than in landmark studies ofheartfailure(88)3.ThetrueincidenceofsuddencardiacdeathinCSisnotknown,butcurrently up to 25%of patientswith CS are estimated to presentwith sudden death(16,17,88,113,130).Fromanelectrophysiologist’spointofview,CSrevealed tobe theetiology in 5-28% of patients referred to electrophysiological study and/or catheterablation due to monomorphic VT of unknown cause (121,126,131). CS derived VTswereinducibleinelectrophysiologicalstudiesinvirtuallyallpatients(119,121,132).

Given their diverse pathophysiology, ventricular arrhythmias may manifestunexpectedly, though certain predisposing factors have been identified. Logically, LVdysfunction was the strongest predictor of arrhythmias in all studies, but still asignificant number of patients with appropriate therapies had LV ejection fraction(LVEF) of >35% (98,128,129). Besides reduced LVEF, male gender (129), history ofsyncope(129),longerfollow-up(128),higherprevalenceofcompleteheartblock(128),andRVdysfunction(RVEF≤45%)predictedfutureVTevents(98).

Afewstudieshaveaddressedtheusefulnessofprogrammedventricularstimulationinidentifying CS patients with a risk of VTs. Mehta et al. performed programmedventricular stimulation in 76 patients with biopsy-proven systemic sarcoidosis, withevidenceofCSinPET/CMRbutwithoutcardiacsymptoms(133).Eightoutof76(11%)were inducible for VTs and received an ICD. Over amedian follow-up period of fiveyears,twooutofeightdiedandfouroutofeighthadappropriateICDshocks,comparedtonoarrhythmiceventsinthenon-induciblepatients.Likewise,inastudyofAizeretal.,32CSpatientsunderwentprogrammedventricularstimulationandpatientswitheitherspontaneous (n=6),or inducible (n=6)VTreceivedan ICD(134). Ina follow-up,after2.7 years, five out if six patients with spontaneous VTs and four out of six patientswithoutspontaneousbutwithinducibleVTsreceivedappropriateICDtherapies.Thus,stimulationidentifiedpatientswithfuturearrhythmiceventsandnoarrhythmicdeathsoccurred in these patientswith either spontaneous or inducibleVTwhen all of themreceivedanICD.However,theelectrophysiologicalstudywasnotfullysensitive,astwo

30

(10%) patients without spontaneous or inducible sustained ventricular arrhythmiasuffered suddendeathor sustainedVTon follow-up. Interestingly, inanother smallerstudy of 16 patients, the inducibility of VT/VF did not predict future ICD therapies(134).

Due to overlapping clinical presentation, patients first presenting with ventriculararrhythmiasmaybemisdiagnosedasARVCandmayeven fulfill theTaskForceARVCcriteria(121,135-137).ThedistinguishingfeaturesinthisrespectarelackofconductionabnormalitiesandpositivefamilyhistoryinARVC,comparedwithmorefrequentseptalinvolvement and LV dysfunction in CS (138). The disruption of desmosomal proteinsmaybeapotentiallinkbetweenthesetwoarrhythmogenicdiseases(65).

Supraventricular arrhythmias occur in 9-36% of CS patients (4,22,101,139-141). Thecauses of atrial arrhythmias are diverse,with direct atrial granulomatous infiltrationcausingabnormalautomaticity,triggeredactivityandre-entrantcircuitsasspeculatedmechanisms (4,123,140). Another possible cause for atrial arrhythmias is elevatedatrial pressure due to ventricular dysfunction or pulmonary hypertension (140,142).Atrial fibrillation is themost commonsupraventriculararrhythmia inCS, followedbyatrial tachycardia and atrial flutter (123,140). Left atrial enlargement was the onlyvariable found to be associated with supraventricular arrhythmias in CS (risk ratio,6.12;95%CI,2.19-17.11)(137).

2.4.3HeartfailureGranulomatous infiltration of the myocardium with the resulting inflammation andfibrosiscanleadtoeithersystolicordiastolicdysfunctionoracombinationofthetwo.Diffuse cardiac involvement leads to global dysfunction, whereas focal involvementproduces segmental hypokinesia and aneurysms independent of coronary circulationzones(143,144).CongestiveheartfailurewithreducedLVsystolicfunctionisasignofawidely advanced disease state (71). Besides extensive myocardial infiltration, pumpfailuremayresultfromrightventricularoverloadcausedbypulmonaryinvolvementorfrom atrioventricular valve regurgitation caused by papillary muscle involvement(145).

Prior to current diagnostic methods, overt heart failure was a common firstmanifestationofCSwithreportedprevalencevarying from25to75%(22,71,97,116).Today,asdiagnosis ismadeatearlierstages,around13-18%ofpatientshavesystolicdysfunction at diagnosis (93,146). CS is frequently misdiagnosed as DCM. In twoJapanesestudieswithpatientsundergoing leftventriculoplasty forsuspectedDCM,CSwas found in3.6-7%(14/384and8/110, respectively) (100,147). Ina third Japanesestudy involving 533 patients with non-ischemic cardiomyopathy undergoing cardiacresynchronization therapy (CRT), 4.5% (25 patients) had CS (148). The evolution ofLVEFduringdiseasecourseishighlyvariableanddataonitsnaturalcourseislacking.

31

2.4.4OthermanifestationsDirect valve involvement inCS is relativelyuncommon, but case reports involving allfour valves have been reported (110,149,150). In the study by Fleming et al. valvelesionswerereportedin8.4%ofpatients(22).ThemostcommonvalvedysfunctioninCS is mitral regurgitation secondary to either papillary muscle infiltration or LVdilatation (143,151). Symptomatic pericardial involvement is very rare, but mildpericardial effusion has been reported in 19% patients (22,152-154). ConstrictivepericarditisasamanifestationofCShasalsobeenreported(155).

Clinical involvement of the coronary arteries is possible but rare in CS. Despite this,granulomas have been detectedwithin atherosclerotic plaques andwithin themediaandperi-arterial tissue(156).Thalliumperfusionscanshaveshownischemic findingsin theabsenceof stenoses incoronaryangiographies (157,158).Microvascular spasmhas been speculated as a possible explanation for the ischemic findings (156,159).Another mechanism could be sarcoid coronary vasculitis (160,161). ST-segmentelevation is a rare finding in CS and is considered to be due to the presence ofmyocardial inflammation/fibrosis and abnormal wall motion, including ventricularaneurysm. In some cases ST-elevations may be transient and may attenuate aftersteroidtreatment(162).Withsarcoidosisbeingasystemicinflammatorydisease,uptoover half of all CS patients are reported to experience constitutional symptoms likefatigue(101,163).Finally,initsearlystageswithminormyocardialinvolvement,CScanbeaclinicallyfullysilentdisease.

2.4.5IsolatedcardiacsarcoidosisSarcoidosis can manifest with solely cardiac symptoms. Previously, there was amisunderstandingthattheheartismostlyaffectedinwidespreadsystemicsarcoidosis.Yet,alreadyintheearlyautopsystudyfromRobertsetal., theauthorsconcludedthatmostpatientswithCShavelittleornoclinicalevidenceofdysfunctionofextracardiacorgansystems(109).Furthermore,itwasdemonstratedthattheextentofextracardiacsarcoidosisdidnotdifferamongpatientswithmild,severeornocardiacdisease(17).Inamore recent autopsy study from theU.S., 40%of25CSpatientsdidnothavegrossextracardiac involvement (87).On thecontrary, ina Japaneseautopsyseriesof42CSpatients, extracardiac sarcoidosis involvement was found in literally all patients,involving lymphnodes in 100%, lungs in 88%, liver in 64%and spleen in 43% (16).Thus, even in autopsy studies the reported frequency of extracardiac involvement ishighlyvariableandclinical studiesareevenmoreprone todifferences,dependingonthe precision of the examination. In a Finnish cohort of nine CS patients, eight hadclinically isolated CS and one developed lung manifestation later (119). Recently,Blanksteinetal.reportedextracardiacinvolvementdefinedasabnormalFDGuptakeinwhole-body 18F-FDG-PET in 26% of 71 CS patients (108). In contrast, Tezuka et al.found extracadiac involvement after work-up including chest-computed tomography,

32

whole-body 18F-FDG-PET, ophthalmologic and dermatologic examinations in 73%(30/41)ofCSpatients(164).Similarly, inthestudyfromKronetal.,94.5%of235CSpatients had clinically and/or histologically detected extracardiac sarcoidosis (99).Although the differences in the proportion of patientswith extracardiac involvementmightpartiallyresultfromdifferencesingeneticandenvironmentalfactors,thevaryingextent of extracardiac work-up is the most probable explanation for the divergence(16,49,99,164).

TherearethreepossibleclinicalscenariosofisolatedCS:1)sarcoidlesionsappearfirstin the heart and involve other organs later in the course of the disease, 2) a trulyisolatedcardiacinvolvementexists,and3)sarcoidlesionsarepresentinotherorgansbut are undetectable by standard examination methods (165). The critical issue inisolated CS is the difficulty of its diagnosis, given the limited sensitivity of EMB(2,21,165). As one example, isolated CS may present clinically as DCM anddistinguishingthemcanbeproblematic.Inastudyof15patientswithCSand30withDCM,bothwithLVEFaround30%,thedisjunctivefeatureswerefemalepredominance(73%vs20%),highincidenceofAV-block(67%vs0%),RBBB(57%vs17%),abnormalwallthickness(73%vs17%),andunevenwallmotionabnormalitiesinLVangiography(113). As an additional detail, sometimes the disease course of CS may mimicpostpartum cardiomyopathy, which has to be recalled in differential diagnostics inpregnancyandpostpartumreckoningthatCScommonlyaffectsyoungfemales(31).

2.4.6GCMGCMisknownasamyocardialdiseasecharacterizedbyacuteonset,progressiveheartfailure,andhasafrequentlyfataloutcomeifuntreated.Inthetwolargest,multicenterstudies of 63 and 73 GCM patients from overlapping groups, 64-75% of patientspresented with congestive heart failure (5,11). Other first manifestations in theselandmark studies were ventricular arrhythmias in 14-32%, signs and symptomsmimickingacutemyocardialinfarctionin6-19%andcompleteheartblockin5-15%.Ina later studyof the samegroup,presentationwithheart failurewas less common (in45%, five out of 11) andmean LVEF at baseline onlymoderately reduced (44±18%)(107). In two small cohorts of five and eight patients respectively, mean LVEF wasmarkedly reduced, 37% at presentation (9,63). In a study addressing long-termoutcome,duringthemeanfollow-upof5.5yearsstartingoneyearafterGCMdiagnosis,50%(13outof26)patientshadneworworseningepisodesofheartfailure,suggestingdisease recurrence (166). During the course of the disease, 23-60% of the patientsdevelop ventricular arrhythmias that are often pleiotropic and can be refractory totreatment (5,9,11,107,167). Follow-up ventricular arrhythmias occur primarily inpatientswhoinitiallypresentedwithventriculararrhythmias(166).AV-blockhasbeenreported in combination with heart failure and VTs (5,9,11), but also as the solemanifestation of GCM (8). As an additional detail, a distinct clinicopathologic entityinvolvingonly theatriaandentailingamore favorableprognosis compared to classic

33

ventricular GCMhas been described (168). Finally, GCM can present in two differentforms; either as an isolatedmyocardial disease ormore rarely togetherwith skeletalmuscleinvolvement,calledthymoma-associatedgiantcellpolymyositis(66,169-171).

The presentation of GCM is highly variable. Its progression is often rapid, with timefromsymptomsonsettohospitaladmissionrangingfromtwotosixweeksinthelargerseries(5,11,63).Evenso,themanifestationofGCMcanvarywidelyfromsuddendeath(27), toamoreindolentcoursewithmildDCMprogressingtoend-stageinnineyears(9,37,172).Renetal.describedthreecaseswithaprotracteddiseasecoursewithtimefromactivemyocarditisepisode to transplantationordeathbeing5-10yearswithoutimmunosuppression (8). Recently, Maleszewski et al. reported prolonged survival ofover 19 years beyond initial diagnosis with immunosuppression (166). In a studycomparing GCM with CS, presentation with heart failure predicted GCM andpresentationwithAV-blockormorethannineweeksofsymptomswereinfavorofCS(5).Takentogether,CSandGCMmaypresentwithsimilarsymptomsanddifferentiatingthem on the basis of clinical symptoms is not possible (5,9). The distinctions inhistologyandoutcomearediscussedelsewhereinthisthesis.

2.5DiagnosticsofCSandGCMAn unequivocal diagnosis of CS is based on a clinical presentation compatible withmyocardial disease associated with histology of non-caseating granulomatousinflammationinheartmuscle.Aprobablebutclinicallysufficientdiagnosiscomesfromextracardiachistologyofsarcoidosiscombinedwithclinicalmanifestationsandcardiacimaging indicative of myocardial involvement. In cases of clinically isolated cardiacinvolvement, the diagnosis can be extremely difficult and rest on a high index ofsuspicion. Still today, in 57-100%of patients undergoing transplantation due to end-stage heart failure, the diagnosis is missed until examination of the explanted heart(82,84,173). The current guidelines of the American Thoracic Society/EuropeanRespiratory Society recommend screening for asymptomatic CS in patients withsarcoidosisdiagnosisfromotherorgans(174).Severalgroupshavepublishedstepwisediagnostic algorithms to aid clinicians inCSdiagnostics (4,28,165,175-177).Figure 5presentsasimplerecentlypublishedscreeningstrategy(4).

34

Figure 5. Suggested study algorithm of patients with biopsy-verified extracardiacsarcoidosis.Adaptedfrom(4).

2.5.1DiagnosticguidelinesandrecommendationsinCSInJapan,CShaslongbeenatargetofresearchandtheJapaneseMinistryofHealthandWelfare(JMHW)publishedguidelinesfordiagnosingCS in1993(175),Table3.Sincethen, the JMHW criteria have been widely applied as the standard diagnostic toolagainst which other diagnostic modalities have been compared. However, theseguidelinesarenotproperlyvalidatedforeitherclinicalorresearchapplications(13).In2006theguidelineswererevised;histologicalconfirmationwasno longerrequired in“clinicaldiagnosis”groupandLGE-CMRwasincludedasaminorcriterion(178).Itisofnote that 18F-FDG-PETwas still not included. It has become evident that the JMHWguidelineshavetoolowasensitivityindetectingCS,especiallyitsearlymanifestations

Biopsyprovenextra-cardiacsarcoidosis

Cardiachistory,ECG,Echocardiogram

1)Symptom(s)positive(signimicantpalpitations*/pre-syncope/syncope)2)AbnormalECG¶3)AbnormalEchocardiogram§

Oneormoreof1-3

AdvancedcardiacimagingLGE-CMRand/

or18F-FDG-PET

Noneof1-3

Negative–>Lowprobabilityofcardiac

sarcoidosis

*Palpitationsweredefinedas“prominentpatientcomplaintlasting>2weeks”.¶Abnormalelectrocardiogramdefinedascompleteleftorrightbundlebranchblockand/orpresenceofunexplainedpathologicalQwavesintwoormoreleadsand/orsustained2ndor3rddegreeAV-blockand/orsustainedornon-sustainedVT.§AbnormalechocardiogramdefinedasLVEF<40%and/orbasalseptumthinningand/orregionalwallmotionabnormality.

35

(1,2). Merely the fact that JMHW guidelines emphasize abnormal electrocardiogram(ECG)results,yeteveninpatientswithextensivemyocardiallesionsinautopsy,thepre-mortem ECGmay be normal in 25%,makes these guidelines too insensitive (17,91).Recent advances in imagingmethods highlight this inaccuracy. As an example, in thestudybyPateletal.,theJMHWcriteriafoundonly52%ofsarcoidosispatientswithLGEinCMR(92). Insomerecentstudies,patientswithpulmonarysarcoidosisandcardiacabnormalities in LGE-CMR have been defined as having CS even without clear CSsymptoms (chest pain, palpitations or no cardiac symptoms at all), which is also asomewhatproblematicdefinition(179).

Table3.GuidelinesfordiagnosingCSfromtheJapaneseMinistryofHealthandWelfare1993(175).

1.Histologicdiagnosisgroup

CSisconfirmedwhenhistologicanalysisofoperativeorEMBspecimensdemonstratesepithelioidgranulomawithoutcaseatinggranuloma.

2.Clinicaldiagnosisgroup

Inpatientswithahistologicdiagnosisof extracardiac sarcoidosis,CSissuspectedwhenitem(a)andoneormoreofitems(b)through(e)arepresent:

a) CompleteRBBB,leftaxisdeviation,AV-block,VT,VPBs(≥Lown2*),orabnormalQorST-TsegmentchangeontheECGorambulatoryECG

b) Abnormalwallmotion,regionalwallthinningordilatationoftheLVc) Perfusion defect by Thallium-201 myocardial scintigraphy or abnormal

accumulationbygallium-67ortechnetium-99mmyocardialscintigraphyd) Abnormalintracardiacpressure,lowcardiacoutput,orabnormalwallmotionor

depressedLVEFe) Interstitial fibrosis or cellular infiltration over moderate grade, even if the

findingsarenonspecific

_____________________________________________________________________________________________________CS= cardiac sarcoidosis, EMB = endomyocardial biopsy, RBBB= right bundle branchblock,AV-block=atrioventricularblock,VT=ventriculartachycardia,VPB=ventricularprematurebeat,*30ormoreVPBsperhour,ECG=electrocardiogram,LV=leftventricle,EF=ejectionfraction.

In 2014, an expert group of the Heart Rhythm Society published consensusrecommendations for diagnostic criteria, diagnostic pathways in different clinicalscenarios and risk stratification in CS Table 4 (4). In the same year, the WorldAssociationofSarcoidosisandOtherGranulomatousDisordersupdatedaninstrumentbased on expert consensus for assessing different organ involvement in sarcoidosis(177). The tool classified organ involvement into three categories; highly probable(>90% likelihood), probable (50-90% likelihood), and possible (<50% likelihood). Inclinicalstudies,considerablyvariableinclusioncriteriaforCShavebeenapplied.

36

Table 4. Expert consensus recommendations on criteria for the diagnosis of CS.Adaptedfrom(4).

1. HistologicaldiagnosisfrommyocardialtissueCS is diagnosed in the presence of non-caseating granuloma on histologicalexaminationofmyocardialtissuewithnoalternativecauseidentified(includingnegativeorganismalstains,ifapplicable).

2. Clinicaldiagnosisfrominvasiveandnon-invasivestudies:Itisprobable*thatthereisCSif:a) Thereisahistologicaldiagnosisofextracardiacsarcoidosis

andb) Oneormoreoffollowingispresent

-Steroid+/-immunosuppressantresponsivecardiomyopathyorAV-block-UnexplainedreducedLVEF(<40%)-Unexplainedsustained(spontaneousorinduced)VT-MobitztypeII2ndor3rddegreeAV-block-PatchyuptakeondedicatedcardiacPET(inapatternconsistentwithCS)-LategadoliniumenhancementonCMR(inapatternconsistentwithCS)-Positivegalliumuptake(inapatternconsistentwithCS)and

c) Othercausesforthecardiacmanifestation(s)havebeenreasonablyexcluded

* In general, “probable involvement” is considered adequate to establish a clinicaldiagnosisofCS.

2.5.2Diagnosticexaminationsandprocedures2.5.2.1ECGandHoltermonitoringAccordingtoMehtaetal.,the12-leadECGhasahighspecificityofupto97%,butlowsensitivityindetectingCSinpatientswithsystemicsarcoidosis(93).Inacohortof112biopsy-provenpulmonarysarcoidosispatientswithsymptomssuggestiveofCS,RBBBwas found in23%ofpatientswithcardiac involvementbymodified JMHWcriteriavsonlyin7%ofnon-CSpatients(114).Likewise,therewasamuchhigherprevalenceoffragmented QRS complex with cardiac involvement (75% vs 34%) (146). Other ECGfindings associatedwithCS are anydegree ofAV-block, frequent atrial or ventricularectopic beats, and pathologic Q-wave or ST-T changes indicating local myocardialinfiltration (123). More delicate ECG tools, such as microvolt T wave alternans andsignal-averaged ECG, are more sensitive, with 56-75% and 79-96% positive andnegative predictive values in CS diagnosed by JMHW`s early or revised criteria(146,180).

37

In studies focusing on screening sarcoidosis patients for cardiac involvement, Holtermonitoringhasshownabnormalitiesin50-67%ofCSpatients(diagnosedbyearly-orrevised-, or modified to include LGE-CMR, JMHW criteria) with specificities rangingfrom21to97%(93,125,181).Nevertheless,theabnormalitiesobservedarenonspecificincludingVPBsandnon-sustainedVTs.Suzukietal.detected≥100VPBsin67%(8/12)ofCSpatients,8%(2/26)ofsarcoidosispatients,andin5%(3/58)ofhealthycontrols(181). The sensitivities and specificities of individual tests in three studies arepresentedinTable5.

Table5.SensitivitiesandspecificitiesofdiagnosticstudiesandfindingsinCS.Adaptedfrom(93,125,146)._____________________________________________________________________________________________Abnormalityatbaseline Sensitivity(%) Specificity(%)ECG 8-26 64-97Signal-averagedECG 28–52 80-82(anyabnormaldomain)Holter 50–89 21–97EPstudy 40 71_____________________________________________________________________________________________CS=cardiacsarcoidosis,ECG=electrocardiogram,EP=electrophysiologic2.5.2.2EchocardiographyGiven its wide availability and noninvasiveness, two-dimensional transthoracicechocardiography is theprimary imagingscreening tool forCS,but theresultscanbenormal in over two-thirds of patients at diagnosis (93,182). More specifically,echocardiography commonly fails to detect early, mild or localized disease (1).Echocardiogaphic abnormalities are nonspecific, with LV systolic and diastolicdysfunction, wall motion abnormalities in a non-coronary distribution, and septalabnormalities as themost common findings (93,113,183). Septal abnormalities, boththickeningduetoinfiltrationoredemaandthinningduetoscarring,arecharacteristicofCSandconstituteoneofthestrongestpredictorsofcardiacinvolvementinpatientswithsystemicsarcoidosis(90)(Figure6).

38

Figure 6. Echocardiographic image of CS with basal thinning. IVS=interventricularseptum, A= point located 10mm from the aortic annulus, i.e. basal IVS, B= IVS at thestandardlevelofthemitralvalveleaflettips,Ao=aorta,LA=leftatrium,LV=leftventricle.Adaptedfrom(184).

Severalstudiesexaminingpatientswithpulmonarysarcoidosishavedemonstratedthatdiastolic dysfunction and subnormal systolic function (evidenced by tissue dopplermeasurements, isovolumic acceleration etc.) are more common compared to healthycontrols (163,182,185-188). These findings can be related to subclinical myocardialinvolvement but the abnormalities are subtle and cannot usually be detected bytraditional echocardiographic methods (189). As the conventional two-dimensionalechocardiography lacks sensitivity to evaluate CS, new more sensitive methods areneeded (190). Recently, speckle-tracking imaging which assesses the intrinsicdeformation(strain)ofthemyocardium,hasshownpromiseinthisrespect(80). Inastudy of 39 patients, 3-D speckle-tracking radial strain showed good potential todistinguishCS(diagnosedbyrevisedJMHWcriteria)fromDCM(191).Furthermore,inastudy of 96 biopsy-proven systemic sarcoidosis patients without abnormalities inconventional echocardiography, the LV global longitudinal strains were lowercompared to healthy controls (192). Finally, in a study of 100 systemic sarcoidosispatientsand100healthycontrols,theaveragegloballongnitunidalstrainwaslowerinsarcoidosisandalsoaprognosticmarkerforfutureadversecardiacevents(80).

Echocardiographyhasimportantprognosticvalue,asLVdysfunctionanddilatationarestrong prognostic markers for future adverse cardiac events (80,97). Finally,echocardiography is the primary follow-up tool for assessing the development of LVdysfunction.

39

2.5.2.3LGE-CMRLGE-CMRhasbecomethepreferredimagingmodality inthediagnosisofCS(1,77,90).CMRoffersbothstructuralandfunctional informationwithoutradiationexposure(1).Evenminor areas of myocardial damage can cause LGE at CMR, which thus enablesearlydiagnosispriortoLVdeterioration(1).ThestructuralabnormalitiesencounteredinCSincludelocalwallmotionandthicknessalterationsthatpredominantlyinvolvetheseptumandmaysometimesresult inaneurysmformation(164,165)(Figure 7).CMRalsoaidsthepreciseassessmentofLVfunctionandchambersizes.InCSdiagnostics,thesensitivityandspecificityofLGE-CMRcomparedtomoderatedJMHWguidelineswere75-100%and77-92%,withpositiveandnegativepredictivevaluesof55%and100%,respectively (164,193-195). In a cohort of 115 CS patients (with cardiac biopsyconfirmationorextracardiacbiopsyconfirmationandelectrophysiologicaldisturbancescompatible with CS), LGE-CMR findings were normal in 14% and the septum wascommonlyaffected(129).

Figure7.ShortaxisandfourchamberCMRimageofCSwithbasalseptalthinningandmyocardial lategadoliniumenhancement.(WithpermissionfromDr.MiiaHolmström,HUSMedicalImagingCenter,Radiology,UniversityofHelsinkiandHelsinkiUniversityHospital).

The physiological mechanism of myocardial LGE is as follows. The contrast agent,gadoliniumchelate, isabiologically inert tracer thatdistributes freely toextracellularspace but does not cross intact cell membranes (196). In damagedmyocardium, thecombination of increased extracellular volume due to edema and slower washoutkineticsresultsinanaccumulationofthetracerseenasgadoliniumenhancement(196).LGE imaged 10-15 minutes after contrast injection is the cornerstone finding inmyocarditis such as CS, visualizing myocyte necrosis and fibrosis (197). The mosttypicalLGEpatterninCSconsistsofmultiple,patchymid-myocardiallesionsinanon-

40

coronary (i.e., sparing the endocardiumand localizing independently of any coronarycirculationarea)distributionwithseptalandRV involvement (91,96).Yet,nospecificpatternofLGEispathognomonicforCS(4,195).Evensubendocardialenhancementinacoronary distribution resembling ischaemic heart disease has rarely been observed(90).ItisrecommendedthattheCMRstudiesbeinterpretedbyaspecializedradiologistandintegratedwiththeclinicaldata(4).

WhileLGEimagingcanhelpindistinguishingpatternsofmyocytenecrosisandfibrosisfrom ischemic injury, T2-weighted and early gadolinium enhancement imaging(measured rapidly after contrast) indicate other inflammatory processes such asedema, capillary leakageandhyperemia (198).Active sarcoidosis lesions in theheartmuscle are characterized by focal wall thickening due to infiltration and edema,combinedwithwallmotion abnormalities, increased signal intensity on T2-weightedimages, and early gadolinium enhancement (94,193). In the chronic phase, wallthinninganddelayedgadoliniumenhancementareseenassignsofscarringandfibrosis(199,200). In reality, these two phases commonly overlap (145). Despite its highsensitivity, CMRmaymiss early inflammation anddifferentiating active inflammationfromfibrosiscanbeproblematic(1,198).

Inadditiontoitsdiagnosticutility,contrastCMRisapotentriskstratificationtool,withincreasingevidence fromseveralstudiesshowingthat thepresenceandmagnitudeofLGE predict future adverse cardiac events in patients with systemic sarcoidosis(91,95,96,201) and in patients with CS (202-204). Despite this, one recent studyinvolvingmostlylungsarcoidosispatients,allwithoutcardiacsymptomsandpreservedLVEF, suggested the contrary as there was no difference in cardiac events in LGEpositive and negative groups (90). Transmural LGE lesions and a higher number ofaffectedsegmentsareassociatedwithlargerventriclesandlowerLVejectionfractions(96,196,205).Furthermore,LGE-CMRhasbeenshowntobeusefulintheassessmentofsteroidresponseandtargetingEMB(1,95,203,206).

As CMR studies are booming and new, even more sensitive methods such asmulticontrast late-enhancement are developed (207), more sarcoidosis patientswithoutcardiacsymptomsbutwithpositiveCMRfindingsarelikelytobefound.Thereisnoconsensusonthetreatmentstrategyofthesepatients(197).

2.5.2.418F-FDG-PET18F-FDG-PETisanotheradvancedimagingmethodusedinthediagnosisandfollow-upof CS. 18F-FDG imaging is based on the fact that inflammatory cells, especiallymacrophages,useglucoseintheirmetabolism.18F-FDGisaglucoseanalog,whichafterbeing absorbed phosphorylates and thus cannot be released from the cells. Underaerobic circumstances, healthy myocardium utilizes mainly free fatty acids in itsmetabolism (145). As up-regulation of glucose metabolism occurs at sites of

41

macrophage-mediated inflammation, accumulation of 18F-FDG suggests an activeinflammatory process (208-210). 18F-FDG-PET study is usually combined withmyocardialperfusionscantodetectassociatedscarring(145,211). Intheory, the18F-FDG-PETfindingscanbedividedintothreephasesaccordingtothediseaseactivity.Inthe early inflammatory stage, focal orpatchyFDGaccumulationprevails. In themoreadvancedstagea typical finding isa “mismatch”or “hotspot”withFDGaccumulationsuperimposeduponaperfusiondefect.Ata latestageofCS,withpermanentscarringbutwithoutongoinginflammation,onlyaperfusiondefectisdetected(145).

Theadvantageof18F-FDG-PETindiagnosticsofCS is itshighsensitivity,being100%comparedtoJMHW1993guidelinesinmanystudies(209,212).Asystematicreviewofseven studies that evaluated the accuracy of 18F-FGD-PETwith JMHW 1993 criteriayielded 89% sensitivity (208). The specificity of 18F-FDG-PET is lower, ranging from38.5%to90%andbeing78% in thepooledanalysisdefinedby the JMHWguidelines(193,208,212,213). A speculated explanation for this might be that 18F-FDG depictsearly-stage sarcoid lesions before patients meet the diagnostic criteria (193). Manystudies have used qualitative visual analysis of FDG uptake in the heart muscle.Recently,theintroductionofquantitativeassessmentusingstandardizeduptakevaluesandFDGvolume-intensityanalysishasincreasedtheaccuracyof18F-FDG-PET,therebyimprovingitsdiagnosticandprognosticvalue(210,213,214).

In addition to its diagnostic utility in CS detection, 18F-FDG-PET provides functionalinformation of inflammatory activity and can thus help evaluate disease activity andtreatmentresponse(211-215).LikeCMR,18F-FDG-PET imagingalsopredictsadversecardiac events (108,213) with a mismatch pattern (see above) and right ventricularinvolvementbeingparticularlystrongsignalsofhigharrhythmiarisk(213).Yetanotherbenefitof18F-FDG-PET imaging is thepossibility todetectextracardiac inflammationand thereby targets for biopsies to verify sarcoidosis histology, most notablymediastinallymphnodebiopsies(1,213,216)(Figure8).

42

Figure8.18-F-FDG-PETwhole-bodyimages.18-F-FDGuptakeinmediastinalandhilarlymphnodesandlungs(Leftpanel)andnoabnormal18-F-FDGaccumulationafter10monthsoftreatment(Rightpanel).Adaptedfrom(216).

A proper technical imaging protocol is of primary importance in acquiringrepresentative diagnostic data from 18F-FDG-PET (193,210). Since normal myocytesalso utilize glucose in metabolism, it is necessary to suppress the physiological FDGuptake through a high-fat/low-carbohydrate diet, overnight fasting and/oradministration of unfractioned heparin (to stimulate lipoprotein lipase whichhydrolyzestriglyceridestofreefattyacidsandglycerol)(209,210,212,217).

Finally,18F-FDG-PETandLGE-CMRcanbeutilizedascomplementaryimagingmethods(193).Asdiscussed,18F-FDG-PETismoresensitiveandservesbetteratassessingearlyinflammatory phasewhereas CMR in detecting even small areas of scars (1,26). Thesensitivitiesandspecificitiesofdifferentimagingmodalitiesvary,partlydependingonwhichCSdiagnosticcriteriaareapplied(Table6).

43

Table6.ApproximatedsensitivitiesandspecificitiesofimagingmodalitiesforthedetectionofCS.Adaptedfrom(110)._____________________________________________________________________________________________Imagingmodality Sensitivity Specificity_____________________________________________________________________________________________Echocardiography Low LowTI-201orTc-99m Moderate Moderatescintigraphy Ga-67scintigraphy Low High18F-FDG-PET High ModeratetohighLGE-CMR Moderatetohigh High_____________________________________________________________________________________________CS=cardiacsarcoidosis,TI-201=thallium-201,Tc-99m=technetium-99m,Ga-67=gallium-67,18F-FDG-PET=18F-fluorodeoxyglucosepositronemissiontomography,LGE-CMR=lategadoliniumenhancementcardiacmagneticresonanceimaging.

2.5.2.5OtherimagingstudiesinCSVarious radionuclides, including thallium (TI-201), technetium-99m (Tc-99m) andgallium-67 (Ga-67) have been used in myocardial scintigraphies to evaluate CS(1,158,218,219). The typical abnormality in thallium-201 and technetium-99mmyocardial perfusion imaging is “reverse distribution”, i.e., the presence of a focalperfusion defect at rest that decreases or disappears under stress (220). Thehypothesized explanation for this is focal reversible vasoconstriction at rest inarterioles adjacent to granulomas (1). Gallium-67-citrate is a radionuclide used forimaginginflammationthatwaspreviouslywidelyutilizedinlungsarcoidosisandinthedetection of myocardial involvement and response to corticosteroids (221). Positivegallium-67 uptake is still included in the revised JMHW guidelines (178). Gallium-67locates inflammatory areas ofmyocardiumwith high specificity (221,222). However,the reported sensitivity is lower and gallium-67 may miss fibrosis and dormant CS(219,223).Where available, 18F-FDG-PET andLGE-CMRhave replaced theuse of theotherradionuclideimagingstudies(1).

Thoracic computed tomography can identify mediastinal lymphadenopathy and thushelp to differentiate CS fromDCM. Otsuka et al. reported that 88% (7/8) of patientslater diagnosed with CS, compared to 5% (1/20) of age-matched patients withidiopathic DCM, exhibited significant mediastinal lymph node swelling at computed

44

chest tomography (100). It is notable that, conventional chest X-rays did not showlymphadenopathy in any of the patients. Further, computed tomography can help torevealsarcoidpulmonaryandmediastinalinvolvementandtoexcludecoronaryarterydiseaseinyoungpatientswithlowpretestprobability(4).

2.5.2.6CardiacimaginginGCMIn acute GCM, a typical finding in echocardiography is reduced LVEF together withnormal LV dimensions, which progresses to further LVEF deterioration and LVdilatation, sometimes over a matter of days (224,225). Other echocardiographicabnormalities in GCM may include aneurysms and wall thickness abnormalities,particularlythickeningcausedbyextensiveinflammatorycellinfiltrate(226,227).DataonLGE-CMR inGCM is very limitedbecause the largestGCMstudieswere conductedprior to the golden age of CMR, andmany patients are too unstable to undergo theimagingintheacutephase(26).IncasereportsofsevereGCM,CMRhasshownmultiplemyocardial LGE areas in different myocardial layers, but the findings only partiallycorrelatedwithhistologicalexaminationsatautopsyorofexplantedheart(228,229).Ina population of 86 patients with suspected non-ischemic cardiomyopathy, moreextensive LGE predicted CS and GCM (230). Finally, CMR is a widely studied andestablishedimagingmethodinmyocarditis,involvinglargelythesamepathophysiologicmechanisms including edema, hyperemia, capillary leakage, necrosis and fibrosis asGCM(231-233).Thedataon18F-FDG-PETinGCMisextremelylimited(230).

2.5.2.7DiagnosticsbiopsiesAdefinite diagnosis of CS andGCM requires a histologic confirmation from the heart(2,4).Thehistologicalhallmarkofsarcoidosisisanon-caseatinggranuloma,asopposedtocaseatinggranulomasintuberculosis(5).Thesarcoidgranulomasarehistologicallydistinct from interstitial rheumatic granulomas and vascular lesions in Wegener`sgranulomatosis (27).Furthermore,other causesof granulomatous infiltrationneed tobe excluded on special stains (tuberculosis on Ziehl-Neelsen and fungal infections onGomorrimethenamine silver). Sarcoid granuloma is a nodus containingmononuclearphagocytes(macrophages,epithelioidcellsandgiantcells)inthecenter,surroundedbyT-lymphocytes, plasma cells andmast cells and circulated by a fibrous ring (3). Thehistologiccriteria forGCMontheotherhand,requiresthepresenceofmultinucleatedgiant cells, myocyte necrosis and lymphocytes in the absence of well-formedgranulomas, although sporadic, partial granulomas may be observed (10). Besidesnecrosis, eosinophils are more common in GCM, while fibrosis, especially whenwidespread and well-organized, strongly favors the diagnosis of CS (5,10,105).DisregardingthesometimesproblematicdifferentialdiagnosticsbetweenCSandGCM,EMBishighlyspecificinthesediseases(10).

In young or middle-aged patients with 2nd or 3rd degree AV-block, new-onset heartfailure or ventricular arrhythmias without ischemic heart disease, the current

45

guidelines recommend EMB (4,231,234). EMB is a class IB recommendation(recommended with intermediate evidence of usefulness) in new-onset (<2 weeks)heartfailurewithhemodynamiccompromise(234).Thesamerecommendationappliesfor heart failure of twoweeks to threemonths in durationwith dilated LV and newventriculararrhythmias,or2ndor3rddegreeAV-block,orfailuretorespondtonormalcarewithinone to twoweeks (234). Furthermore, EMB is a class II recommendation(may be considered with conflicting evidence) in heart failure of >3 months orunexplainedventriculararrhythmias(231,234).Theserecommendationsarebasedonincremental diagnostic, prognostic and therapeutic value, which histologicalconfirmationofCS,GCM,eosinophilicorothertreatablemyocarditissignals(86,234).

Nevertheless, due to the heterogeneous, patchy distribution of the microscopicgranulomas in myocardium, the diagnostic yield of EMB in CS is poor. Sarcoidosisclassically involves the base of the heart, leaving standardmid-ventricular and apicalbiopsiesunrevealing(235).DifferentstudieshavedemonstratedEMBsensitivityof15-30%inCS(2,21,86,91).Thehistologicdiagnosticrate isdependentonthestageofCSandthetimingofthebiopsies.Thesuccessratewas36.4%inaDCM-likeclinicalpicturecompared to 6.7% in cases with conduction disturbances and preserved systolicfunction (2). The study by Ardehali et al. suggested that a limited cardiac sarcoidinvolvement that was not detected by EMB, predicted better survival, though by nomeans freedom fromcardiacdeath, compared toEMBpositive cases (86). InGCM, atleastinthecaseoffulminantdisease,theyieldofEMBishigher,withasensitivityof80-85%(236).Agreaternumberofbiopsiesincreasesthediagnosticyieldinmyocarditisingeneral,butthevalueofmultipleand/orrepeatedEMBsinCS/GCMwaspreviouslyundetermined (237). Conventionally, EMBs are taken from theRVbut in some cases,e.g., inimagingfindingsconcentratingontheleftside,LVbiopsiesmaybeusefullwithreportedcomplicationratessimilartoright-sidedprocedures(237,238).

Pre-procedural CMR and/or 18F-FDG-PET imaging is recommended and usuallyperformed in clinical practice (231). Biopsies targetting areas with detectedinflammation has raised interest, but there are no prior reports on its usefulness inCS/GCM and the utility in myocarditis in general is poorly defined (231,237). NovelmethodstoimprovethediagnosticyieldofEMBsareneeded.Recently,electroanatomicmapping-guided EMB targeting low voltage and fragmented signal areas significantlyimprovedthediagnosticyieldinsuspectedmyocarditisorCS(137,235).Furthermore,arecent pilot study utilizing three-dimensional electroanatomic mapping with therecording electrode placed at the tip of the bioptome showed promise for betterdiagnosticgain(239).

Finally,duetothelimitedsensitivityofEMB,histologicconfirmationofsarcoidosiscanbeacquiredfromextracardiacorgansinthecaseofcompatiblecardiacsymptomsandsigns.The recentexpert consensus statement recommendedextracardiacbiopsy sitestobetargetedfirstbecauseofhigherdiagnosticyieldandlowerproceduralrisk(4,177).Onesmallpriorreportdemonstratedmediastinallymphadenopathyatchest-computed

46

tomographyin88%(7/8)ofCSpatientscomparedto5%(1/20)ofDCMpatients,butdidnotreportlymphnodesaspotentialbiopsytargets(100).

Gene expression profiling of EMB samples is a promising diagnostic tool fordifferentiating CS, GCM, active lymphocytic myocarditis and non-inflammatorymyocardialdiseases.Todate,twosmallstudieshavedemonstratedthatgenesencodingregulators of immune response, cellular receptors including Toll-like receptors, andproteinsof themitochondrialenergymetabolismaredifferentiallyexpressed in theseconditions(60,104).Thisisanimportanttargetforfutureresearch.

2.5.3LaboratorymarkersSarcoidosis presents with a weak systemic immunologic response despite theassociatedextensivelocalinflammation,andthusbiomarkersmeasuredfrombloodarenot of notable clinical use (33). Granulomas produce angiotensin-converting enzymeand their serum levels are elevated in 60% of patients with sarcoidosis. Still, thediagnostic value of angiotensin-converting enzyme is controversial in systemicsarcoidosis, let alone detection of cardiac involvement (33). Circulating lysozyme orcalcium concentrations are no more helpful. Elevated plasma N-terminal pro-B-typenatriureticpropeptide(NT-proBNP),amarkerofcardiacoverload,hasbeenobservedinsarcoidosispatientswithcardiaccomplications(240).

An early report found first-generation cardiac troponin-T measurements to beineffective in CS (240). Instead, high-sensitivity cardiac troponin-T (hs-cTnT)measurements were recently shown to correlate with 18F-FDG-PET findings in CS,beingabnormallyelevatedinmostPET-positivepatientsandnormalinthemajorityofPET-negative patients (241). One case report described a rapid decrease in high-sensitivity cardiac troponin-I (hs-cTnI) following intravenous methylprednisoloneadministration (242). The value of high-sensitivity troponins in the detection ofmyocyte injuryandevaluationofdiseaseactivityremains tobesolved.Galectin-3 isapromisingbiomarkerinvolvedinheartfailure,inflammationandfibrosisthathasbeenstudiedinasmallcohortofsixCSpatients(179,243).Galectin-3wasfoundtobehigherin pulmonary sarcoidosis patients compared to healthy controls, but did not differbetweenCSandnon-CSgroups(179,243).

2.6ManagementThe treatment of CS involves immunosuppressive medication for sarcoidosis andcardiac-specific treatment forheart failure, conductiondisturbancesandarrhythmias.NointernationalguidelinesexistonthemanagementofCSandthebestavailabledataisbasedonmeta-analysisandexpertconsensus(197,244).In2012aconsensusstatementon the management of CS, based on responses from 42 commendable sarcoidosis

47

experts to a Delphimethod questionnaire, was published in the U.S. (197). Althoughthere were significant differences of opinion among the experts highlighting thedifficulties inmanagingCS, sufficient agreementwas reached to formulateaproposalforbestpractice,Figure9.

Figure9.ProposedbestpracticeforthediagnosisandtreatmentofCSbasedonexpertpanelconsensus.Adaptedfrom(197).

CS=cardiacsarcoidosis,ECG=electrocardiogram,CMR=cardiacmagneticresonanceimaging,18F-FDG-PET=18F-fluorodeoxyglucosepositronemissiontomography,EP=electrophysiologic,LV=leftventricle,RV=rightventricle,SCD=suddencardiacdeath.

ScreeningforCS

-Clinicalsymptomsandphysicalexam

-12-leadECG-Echocardiography-(Holtermonitoring)

Immunosuppressiontherapy-Prednisone(30-40mg/day)-Methotrexate/azathioprine-Others

Startingimmunomodulatorytherapy

-Positive18F-FDG-PET-LVdysfunction-Conductiondefects-Ventriculararrhythmias-(LGEonCMR)

WorkupforCS-12-leadECG-Echocardiogram-CardiacCMR-Cardiac18F-FDG-PET-Holtermonitoring-(EPevalutation)

ICD-PositiveEPstudy-AbortedSCDevent

Evaluationoftherapyresponse

-Symptoms-Cardiac18F-FDG-PET-CardiacCMR

Electrophysiologicevaltuation

-EPstudy-RVmapping

Follow-up Positive

Positive

Negative

Negative

48

2.6.1Corticosteroids2.6.1.1ImplementationThegoalofimmunosuppressivetreatmentinCSistoreverseongoinginflammationandthereby potentially prevent progression to fibrosis and adjacent organ dysfunction(197). As in sarcoidosis affecting other organs, corticosteroids are the mainstay oftherapyforcardiacinvolvement.Corticosteroidsfunctionbysuppressingmultiplepro-inflammatory cytokines and chemokines (245).However, debate exists as towhethertreatment can change theoutcomeof sarcoidosis in general and the evidence is evenmoreincompleteinCS(29,43).

Astherearenorandomizedcontrolledtrials,thechoice,dosageanddurationoftherapystill remain to be judged by clinicians. In general, prednisone starting dosage variesfrom20-60mg/dayor50-60mgonalternatedays. In the studybyYazakiet al. therewas no difference in outcome in low (≤ 30mg) compared to high (≥40mg) initialprednisone daily dose (97). The dose of prednisone is gradually tapered down to amaintenancedoseof5-15mg/day (6,97,244).Controversy remainson thedurationofthetreatment(244).Somecentersrepeat18F-FDG-PETscansorothercardiacimagingmodalities to guide long-term steroid treatment(197,244). Finally, after stoppingimmunosuppressive therapy, close follow-up for possible future relapses isrecommendedbytheexperts(244).

2.6.1.2EfficacyDespite over 50 years of the use of corticosteroids in CS, there still is no proof ofsurvivalbenefit fromimmunosuppressivetreatment(43,244). Inthe largeststudyonlong-termsurvival, the five-year survival ratewasonly10% inpatientsnot receivingsteroidscomparedto75%insteroid-treatedpatients,butsinceall the20non-treatedpatientswerediagnosedatautopsy,noconclusionontheefficacyofsteroidtreatmentcould be drawn (113). The same dilemma applies to other studies as there are noplacebo-controlledtreatmenttrials.Katoetal.retrospectivelyevaluatedthesurvivalofsevenpatientsreceivingcorticosteroidsand13patientsnotreceivingtheseagentsandfound that in the untreated group, two patients died compared to no deaths in thetreatmentgroup,butthedifferencewasstatisticallyinsignificant(246).Likewise,Nagaiet al. found no effect on cardiac deaths in the 7.6-year follow-up in 67 patientswithcorticosteroidtreatmentcomparedto16patientswithoutcorticosteroids(reasonsforwithholding corticosteroids; cardiac imaging without active inflammation in seven,patientrefusalinfive,activeinfectiousdiseaseintwoandothersintwo)(115).

ThedataconcerningtheefficacyofcorticosteroidsonLVfunctioniscontradictory.Inacohort of 30 patients in a study by Takaya et al., LVEF remainedmildly reduced yetstabileover12monthsofsteroidtreatment(215).Conversely,inacohortof10patientsfromKudohetal.,meanLVEFimprovedfrom35%to49%aftersixmonthsofsteroidtherapy (247). In a study of 20 CS patientswith AV-block and initially normal LVEF,sevencorticosteroid-treatedpatientsdemonstratednochangeinLVEF,comparedto11

49

out of 13 non-treated patients with marked decline in LVEF over a mean follow-upperiod of 79months (246). Similarly, in a recent study byNagai et al., patientswithcorticosteroid therapy (n=67, initialmeanLVEF36%),hadan increase inLVEF (+8%±36%)versusdecrease inLVEF (-17±35%) innon-steroid group (n=16, initialmeanLVEF35%)(115).Furthermore,thedifferenceinLVEFwasemphasizedinpatientswithbaseline LVEF ≥35% and steroid-treated patients also had fewer hospital admissionsdue to heart failure (115). Furthermore, Chiu et al. demonstrated no effect on LVEFeither in the total groupof 43CSpatients or in patientswith severely reducedLVEF(<30%, n=11), but a significant improvement in LVEF and LV end-diastolic diameter(LVEDD)inthegroupwithmildlytomoderatelydecreasedLVEF(30-54%,n=10)(6).Asystematicreviewsummarizingfourstudies(notincludingtherecentstudyfromNagaiet al.) concluded that steroid therapy seems tomaintain LV function in patientswithnormalLVEFatdiagnosis,improveLVEFinpatientswithmildtomoderatedysfunction,but have no effect in patients with severe dysfunction (244). Anyhow, the authorsunderline that it isunclearwhether the improvementsweredue tocorticosteroidsorweresimplyindicativeofthenaturalcourseofthedisease(244).

Several small patient series have shown that AV-conduction blocks mainly developduring the inflammatoryphaseofCSandthatrecoveryofconduction ispossiblewithcorticosteroidtreatment(124,246,248). InthestudybyYodogawaetal.(2013)(248),AV-blockresolvedwithinthefirstweekofsteroidtherapyinfourofsevenpatientsandlater(upto14months)intheremainingthreepatientsoutofatotalof15patients.TherestorationofconductionwasmorecommoninpatientswithpreservedLVEF(248).

In studies with CS patients having ventricular arrhythmias at presentation, initiatingsteroid treatment has shown no beneficial effect on the arrhythmia load (124,132).However, in a small study of patients without presenting ventricular arrhythmias,14.3% (1/7) of patients treated with steroids compared to 61.5% (8/13) of non-steroid-treatedpatientshadventricular arrhythmiasduring thedisease course (246).ThestudybyYodogawaetal. (2011)showedasignificantdecrease inVPBsandnon-sustained VTs, as well as an improvement in signal-averaged ECG parameters inpatientswithEF≥35%,butnotinpatientswithEF<35%overaseven-monthperiodofcorticosteroid therapy (249). In 45 CS patients with ICD, post-implantationimmunosuppressionhadnoeffectontheventriculararrhythmias(128).

Several 18F-FDG-PET studies have demonstrated the diminution or disappearance of18F-FDG uptake in CS after one to two months’ steroid therapy (211,214,215).Likewise,inascintigraphystudy,corticosteroidsabolishedgallium-67uptakewithinsixmonthsinallninepatientswithpositiveuptakeatpresentation(124).InaCMRstudyof12histologicallyverifiedpatientswithsystemicsarcoidoisisandsuspectedCSbasedonscintigraphy results, six (including three with cardiac symptoms) receivedcorticosteroidswithCMRshowingclearedorimprovedfindingsinallatthe12-monthfollow-up, compared to a worsening or stability of CMR findings (signal intensity,contractilityandmyocardial thickness) inthesixnon-treatedpatients(206).Whether

50

the reductionofmyocardial inflammationmeasuredby imaging findings is related toclinical improvement is controversial. Osborne et al. found a significant inverserelationshipwithLVEFandstandardizeduptakevaluesinPET(214),whereasTakayaetal.demonstratedresolutionof18F-FDGandgallium-67scintigraphyactivitywithoutany improvement in LVEF (215). Altogether, there is evidence from relatively smallstudiesthatthecorticosteroidtreatmentresponseinCSmightbepositiveintheearlyinflammatoryphaseofdiseasebutnolongerwhenmyocardial fibrosisandscarshaveformed(6,124,215,249).

2.6.2OtherimmunosuppressantsLong-term use of steroids is associated with adverse effects (diabetes, cataract,osteoporosis, weight gain, gastrointestinal bleeding, infections, etc.), and alternativeagents are also required for patients refractory to corticosteroids. No randomizedcomparativestudiesofsteroid-sparingimmunosuppressantsinCSexist.Inastudyof17patients,acombinationofmethotrexate(6mg/week)andalow-dose(5-15mg/day)ofprednisoloneshowedatendencytowardsbetteroutcomescomparedtocorticosteroidsalone (with no information ofmaintenance dose)measured by higher LVEF, smallerLVEDD and lower NT-proBNP over five years of follow-up (250). In a Delphi studycollecting the opinions of sarcoidosis experts, methotrexate was the second mostcommonly used drug, followed by azathioprine, mycophenolate mofetil,hydroxycholoquine, and anti-TNF-α agents (88,197). There are case reports showingthatanti-TNF-αagentinfliximabmaybeeffectiveinCS(251,252).Theuseofinfliximabis theoretically supported by the well-known role of TNF-α in granuloma formation(32,41,42). However, infliximab has been shown to worsen the outcome in severesystolic heart failure in general (253), and it should therefore be used with greatcaution,particularlywithadvancedCS.

2.6.3ManagementofheartfailureandarrhythmiasBesides immunosuppression, the treatmentofCS includesmanagementofventriculardysfunction, heart failure, and cardiac rhythm disturbances (88). Ventricuardysfunction ismanaged according to the established guidelines for heart failurewithbeta blockers, angiotensin-converting enzyme inhibitors, and mineralocorticoid-receptorantagonists(254).Thetreatmentofventriculararrhythmiasisbasedonbothdrugs and devices (see below). In a multicenter study of 235 patients, the mostfrequently used antiarrhythmic agents, in addition to beta blockers,were sotalol andamiodarone(129).ClassI(Vaughan-Williams)antiarrhythmicdrugsshouldbeavoidedin ventricular and atrial arrhythmias alike due to their potential harmfulness instructuralheartdisease(255).

51

2.6.4DevicetherapyandcatheterablationSince ventricular arrhythmias in CS usually recur, ICD therapy has long beenrecommended for patients with sustained VT or VF (132). More recently, ICDs havebeen increasingly implanted for primary prevention, especially in patientswith non-sustained arrhythmias on ambulatory ECG, unexplained or near syncope, or positiveelectrophysiological study in the setting of reduced EF (90,197). In large studiesrecently published in the U.S., 63-74% of ICDs in CS patients were implanted forprimary prevention in patients without prior significant ventricular arrhythmias(98,128,129). Besides a high rate of appropriate ICD therapies (see Arrhythmiassection), CS patients are also reported to experience inappropriate shocks (13.3-25%)(98,128,129) and device-related complications (15.6-18%) relatively commonly(99,128,129).InthegeneralACC/AHA/HRS2013guidelinesforcardiacpacing,CSwasclassifiedasIIA,levelofevidenceCindicationforICD,meaningthatimplantationshouldbe considered (256). In the 2013 ESC/EHRA guidelines CS was not separatelyaddressed (257). An expert consensus statement on CS recommends ICD for all CSpatientswithspontaneoussustainedVTsand/orLVEF≤35%,despiteoptimalmedicaltherapy and a period of immunosuppression (if there is active inflammation) (4).Independent of LVEF, ICD implantation can be useful in cases of indication forbradycardia pacemaker, unexplained syncope or near syncope or positiveelectrophysiological studies (4). In case of mildly to moderately reduced LVEF (36-49%) and/or RVEF <40%, ICD implantation may be considered (4). If none of theaforementioned symptoms or findings are present and there is no LGE on CMR, ICDimplantationisnotrecommended,butthepatientsshouldbecloselymonitored(4).

Asthedisease-specificdataonCRTinCSisminimal,generalrecommendationsshouldbeapplied (4,257,258). Inone Japanese study,CSpatients (n=25) seemed to respondlesstoCRTcomparedtoothernon-ischemiccardiomyopathypatients(148)butthishasnot been confirmed. CRT is recommended in patientswith sinus rhythm, LBBBwithQRS duration ≥ 120ms, LVEF ≤ 35% and NYHA class ≥ II despite adequate medicaltreatment and should/may be considered in non-LBBB patients with the otheraforementionedfeatures(257).Inpatientswithconventionalpacemakers,upgradingtoCRTshouldbedoneaccordingtotheaforementionedindications(257).Theevidenceinpatients with conventional indication for anti-bradycardia pacing is less clear, butsuggeststhatpatientswithmoderate-to-severeLVdysfunctionmightbenefitfromCRTcomparedtoRVpacingalone(257).

In the case of ventricular arrhythmias refractory to drug treatment or frequentadequateICDtherapies,radiofrequencycatheterablationisanoption.Inasurveystudyfrom 2012, 14.9% of 235 patients with ICD implantation for CS in majorelectrophysiologycenters intheU.S.,Canada,andIndia,underwentVTablation(129).In two cohorts of patients with multiple antiarrhythmic medications without properarrhythmia control, catheter ablation resulted in freedom of VTs in four out of eightpatients(121)andeitherreducedorcompletelyeliminatedVTsinfourandfiveoutof

52

nine patients, respectively (122). In a study of 21 patients who were refractory tomedical treatment, VT ablationwas successful in terminating ≥ 1 spontaneous VT in90%and ineliminatingelectrical storm in78%ofpatients (126).However,probablydue to multiple arrhythmia substrates, VT recurred in all but three patients after asingleprocedure,andrepeatedprocedureswereneededtoachievebetterarrhythmiacontrolwithfewerantiarrhythmicdrugs(126).Altogether,catheterablationappearstobe effective in treating VT storm in CS and may provide palliation for recurringarrhythmias, but due to diffuse myocardial involvement, achieving full arrhythmiacontrol is difficult (126).Data concerning ablation of atrial tachyarrhythmias in CS islimitedtoonestudythatshowedlastingresultsinsevenofninepatientsduringameanfollow-upperiodof1.8years(140).

2.6.5TransplantationCardiac transplantation is a viable treatment in end-stage heart failure oruncontrollable ventricular arrhythmias when other organs are not significantlydamaged.CSaccountsforupto2.5%ofallcardiactransplantations(82-84,173,259).

Sarcoidosis has been reported to recur in the transplanted heartwith a rate rangingfrom0%(0/19)to14.3%(2/14)inapproximatelyfiveyears(82,173).Arecurrencecanfollow the tapering of corticosteroids (260) and is often treatable with an elevatedcorticosteroiddose(82,173,261).

The data concerning prognosis after cardiac transplantation in CS is somewhatcontradictory. The largest multicenter study of 38,230 transplantations with 65 CSpatients demonstrated slightly better one-year survival in CS patients compared topatients transplanted for other cardiac conditions (259). Conversely, a single-centerstudyof825transplantationsincluding14CSpatientsshowedatrendtowardshighermortality in CS patients with a five-year post-transplantation survival rate of 58.9%comparedto76.2%(82).Inthelateststudynosurvivaldifferencewasobserved,withfive-yearpost-transplantationsurvivalbeing79%in19CSpatientscomparedto83%in1,050 other heart transplant recipients (173). A case of possible transmission ofsarcoidosisfromadonortoarecipienthasbeendescribed(262).

2.6.6ManagementofGCMThe only trial designed to evaluate immunosuppressive treatment response withplacebo in GCM, by Cooper et al., failed to recruit patients in the non-immunosuppression arm,making the estimationof survival benefit impossible (107).Nevertheless, the study design was modified to examine the effect ofimmunosuppressivetherapyassuchandrecruited11patientsfrom17medicalcentersaround the world. EMBs controlled at four weeks after the start of combinedimmunosuppressive treatment showed significant a reduction in the amount of

53

inflammatory cells and necrosis. Anyhow, LVEF did not change significantly after amonthof treatment,presumablybecauseof relativelymild cardiacdysfunctionat thestart of the study (mean 44±18%). At the follow-up after one year, two patientsrequiredhearttransplantationandonepatientdiedofrecurrentGCMafterstoppingallimmunosuppressivemedication.InadditiontothestudybyCooperetal.,asmallstudyof fiveGCMpatients also found improvement inhistological findingsbutno effect onLVEF after six months of immunosuppressive therapy (9). Recently, a stepwisehistopathologic improvement with multiple immunosuppressants was documented,from fulminant myocarditis with giant cells to a myocarditis characterized by alymphocytic-eosinophilic infiltrate, further to a smoldering lymphocytic myocarditisand lastly to interstitial (mostly replacement-type) fibrosis in repeated EMBs (166).However, in the same study the degree of immunosuppression during the first yearafter diagnosis did not predict the subsequent composite endpoint event of death,transplantation, first ventricular arrhythmia or GCM recurrence (166). Furthermore,several case reports have demonstrated dramatic clinical and histological recoveriesfrom GCM cardiogenic shock treated by immunosuppression (69,263-265). Finally,experimental murine autoimmune myocarditis studies indicate a benefit from T-cellsuppressiontherapyinGCM(62).

No international guidelines on therapy of GCM exist, and the studies addressing itstreatmentaresmall.TheimmunosuppressionprotocolintheGCMtreatmentlandmarkstudy involving 11 patients was as follows (107). Intravenous methylprednisolone10mg/kgwasadministeredforthreedays,followedbyprednisone1mg/kgwithrapidtapering up to a year. Additionally, cyclosporine with serum target levels of 150-300ng/ml was used. Eighty-two percent of the patients received muromonab-CD3(OKT-3),amonoclonalantibodyagainstCD3-T-cells.Intheup-to-datereportof26GCMpatients from Germany and the U.S., the majority of patients received cyclosporine-basedmultiple immunosuppression after diagnosis, including muromonab-CD 3 or acomparable antilymphocytic agent in 42% (166). Of them, 27% were weaned ontocorticosteroid-onlytherapyduringthefirstyearand73%continuedwithcombinationtreatment(166).Atanypointintime,26patients(100%)receivedcorticosteroids,21(81%) cyclosporine, 10 (38%)mycophenolatemofetil, eight (31%) azathioprine, andeight (31%) sirolimus. Furthermore, there are case reports of tacrolimus, anti-thymocyte globulin and alemtuzumab (a monoclonal CD52 T-cell surface proteinantagonist)beingadministeredwithsuccessinGCM(26,69,172,265,266).Inadditiontoimmunosuppressants,thetreatmentofGCMencompassesheartfailureandarrhythmia-targetedtherapiesand,inpatientswithsevereLVdysfunction,inotropicagents.

Despitethemoderatesuccessofintensiveimmunosuppression,asignificantproportionof GCMpatients require a heart transplantation, sometimes urgently (7,37,267). As abridgetocardiactransplantation,mechanicalcirculatorysupportwithventricularassistdevices, intra-aortic balloon pumps and extracorporeal membrane oxygenation havebeenusedsuccessfullyinGCM(26,37,268).IntheGCMlandmarkstudy,54%(34/63)of

54

patientsunderwenttransplantationwithamedianofsixmonthsaftersymptomonset.On follow-up after 3.7 years, 26% (9/34) of transplanted patients died (11). Post-transplantrecurrenceofGCMismorecommoncomparedtoCS,with26-43%(9/34–3/7)oftransplantedpatientshavingGCMfindingsinsurveillancebiopsiesanaverageofthree years (ranging from three weeks to nine years) after transplantation (11,37).Asymptomatic recurrence does not necessarily carry a poor prognosis and usuallyresolveswithtemporarilyaugmentedimmunosuppression(37).InthestudybyCooperet al. (1997) six of the nine recurrences were asymptomatic and one died of heartfailure(11).

2.7Prognosis2.7.1PrognosisofCSThesurvivaldataconcerningCSisheterogeneous,dependingonthediagnosticcriteriaused,theinclusionorexclusionofautopsyandpost-transplantationdiagnoses,andthereported survival definition reported (overall-, cardiac- or transplant-free cardiacsurvival)(Table7).Somekindoftreatmentwithapossibleeffectonthediseasecoursehasbeenusedsincetheearlystudies,thustherearenoreportsstrictlyonthenaturalhistory of CS. In the first large series of 250CS patients, inwhich the treatmentwasvariablebutsteroidswerecommonlyused,thefive-yearsurvivalratewas40%(22).Inan early Japanese study, 55% (23/42) of CS patients died within one year from theonsetofthecardiacsymptoms(16).Themostwidelycitedsurvivalratesarefromtwoseries from thebeginningof the21st century,with five-year survival ratesof60-75%(5,97).Regarding the consequence of cardiac involvement in systemic sarcoidosis, CSwas reported to cause two-thirds of deaths in Japanese sarcoidosis patients, but lessthan20%ofdeathsinCaucasians(49,269).

Table7.PrognosisofCSindifferentstudies._____________________________________________________________________________________________Study,year N Survival Comments_____________________________________________________________________________________________Chiuetal.2005(6) 43 Overall Nocardiacdeathsin patientswithnormal Lifetimediagnosis LVEFatdiagnosis

(N=43)

1year98% 5years90% 10years84%_____________________________________________________________________________________________

55

Okuraetal.2003(5)42 Transplant-free Survivalsignificantly bettercomparedtoGCM Lifetime/autopsy/

explantdiagnosis(N=42)5years61%

Lifetimediagnosis(N=30)5years70%

_____________________________________________________________________________________________Yazakietal.2001(97) 95 Overall Overallsurvivalin patientswithlifetime Lifetime/autopsy diagnosisandLVEF≥ diagnosis(N=95) 50%;1year95%,5

1year85% years89%and10years5years60% 89%10years44%

Lifetimediagnosis(N=75)1year92%5years75%10years61%

_____________________________________________________________________________________________Flemingetal1986(22) 250 Overall Steroidusedcommonly, otherwisetreatment Autopsydiagnosis datanonexistent 5years40% 10years14%_____________________________________________________________________________________________Robertsetal.1977(109) 113 Overall Autopsydiagnosis 1year27%_____________________________________________________________________________________________Matsuietal.1976(16) 42 Overall *Only5(12%)lifetime diagnoses Mostlyautopsy

diagnoses* 1year45% 2years26%CS=cardiacsarcoidosis,LVEF=leftventricleejectionfraction,GCM=giantcellmyocarditis.

56

Alltheaforementionedstudiesareretrospective.Subsequently,severalsmallstudiesonpatients with systemic sarcoidosis, diagnosed with CS and monitored prospectivelysince, have shown comparable survival rates. Among 101 pulmonary sarcoidosispatients from the Netherlands, 16 patients were diagnosed with symptomatic CSaccording to themodified JMWH criteria, and 25% (4/16) of themdied over ameanfollow-up of 1.7 years (77). Similar results were reported from the U.S., with 19%(4/21)ofpatientswithbiopsy-provenextracardiacsarcoidosisandLGE-CMRfindingscompatiblewithCSsufferingcardiacdeathinthemeanfollow-upof1.8years(91). Incontrast, in the study by Mehta et al., none of the 24 patients with pulmonarysarcoidosis andmostlyasymptomaticCSdiagnosedbasedonLGE-CMRand18F-FDG-PETimagingdiedduringameanfollow-upperiodof1.8years(93).Onemightspeculatethat the outcome with modern heart failure treatment and prophylactic use of ICDsshould be better. Unfortunately, the recent large ICD follow-up studies have notreportedsurvivalnumbers(98,129).Inthelateststudywithsurvivaldata,15%(7/47)of patients with CS diagnosed according to the JMHW 2006 guidelines died over amedianfollow-upperiodof15months(270).LVsystolicfunctionisthemostpowerfulpredictorofoutcomeinCS. Inastudyof95patients,10-yearsurvivalwas89%inpatientswithpreservedEF(≥50%)comparedto27%inpatientswithreducedEF(<50%)(97).Inanotherstudyof43patients,10-yearcardiacsurvivalwas100%inpatientswithnormalEF(≥55%)butonly19%inthosewithseverelyreducedEF(<30%)atthestartof thefollow-up(6).Otherpredictorsofsurvival inCS includeLVend-diastolicdiameter,NewYorkHeartAssociation (NYHA)functional class, and sustained VT (97). Interestingly, the presence of pulmonaryinvolvementhasbeenassociatedwithbettersurvival,probablyduetoearlierdetectionof cardiac involvementduring the follow-upofpulmonary lesions (97).Eventually, inrecent years LGE in CMR has been suggested as an independent predictor of futureadversecardiaceventsinCS(202-204)(see2.5.2.3).

2.7.2PrognosisofGCMThecourseofGCM isoften fulminantand itsprognosis is therefore ingeneralpoorerthan the prognosis of CS. In an early study of five EMB-verified patients, 60% (threepatients)haddiedorundergonetransplantationduringafollow-upperiodof1.7years(9). In themulticenter landmark study of 63 patientswithGCM, the rate of death ortransplantationwas89%andmedian transplant-freesurvival fromsymptomonset inallpatientswas5.5months(11).Notably,atotalof25patients(39%)wereidentifiedand included only after autopsy or from explanted hearts. Combinedimmunosuppressive treatment including cyclosporine together with anotherimmunosuppressant,butnotcorticosteroidsalone,prolongedthemediansurvivalfrom3.0 (without any immunosuppressants) to 12.3 months. However, the finding wasbiased resulting from the selection of survivors, i.e., thosewhowould have survivedlongerevenwithoutimmunosuppressivetreatmentweremorelikelytobetreatedwithimmunosuppressivetherapy.Alaterstudywithalongerfollow-upfromthesamegroup

57

demonstrateda10%transplant-freesurvivalrateafterfiveyearsfromsymptomonsetin all 73 patients and 22% five-year transplant-free survival in the 38 patientsdiagnosedbeforedeathortransplantationandtreatedwithatleastcorticosteroids(5).Recently,Maleszewskietal.examinedthelong-termoutcomein26GCMpatientswhosurvivedthefirstyearfollowingdiagnosiswithouttransplant(166).Overthefollow-upperiodof5.5yearsstartingoneyearafterdiagnosis,31%ofpatientseitherdied(n=3)orunderwenttransplantation(n=5).ThefactorsleadingtofulminantprogressandpoorprognosisorindolentcourseinGCMareunknown.

58

3AIMSOFTHESTUDY

This study was designed to assess the epidemiology, characteristics, diagnostics andoutcomeofCSandGCMinFinland.Morespecificallythegoalswerethefollowing:

1.ToevaluatethedetectionrateandprevalenceofCSsincetheturnofthe1990s(I).

2.Toelucidate theroleofCSandGCMintheetiologicspectrumofAV-block inyoungandmiddle-agedadults(III).

3.ToassessthebiopsydiagnosticsofCSclinicallyisolatedintheheart(II).

4.ToexaminetheclinicalmanifestationsofCS(I).

5.TodeterminetheprognosisofCSwithmoderntreatment(I).

6.Toassesstheclinicalutilityofhs-cTnT/IinCS(IV).

7.ToevaluatethecharacteristicsandoutcomesofGCMwithmoderntreatment(V)

59

4MATERIALSANDMETHODS

Thisworkisessentiallyaretrospectivestudywherethedatageneratedduringroutineexaminations, treatment, and follow-up of patients with CS and GCM were collectedafterwards for research purposes. The project was launched by the real life clinicalobservationthattherewasaremarkableincreaseintheamountofpatientsdiagnosedwithCS,formerlyconsideredanextremelyraremyocardialdisease,inourinstitute.Thestudy was planned and initiated at Helsinki University Hospital (HUH) and furtherconducted in collaboration with the other Finnish university and central hospitalstreating patients with CS and GCM. During this project, the Myocardial Diseases inFinland(MIDFIN)-studygroupwasestablished,consistingofacademiccardiologistsatthe five Finnish university hospitals. We also founded an internet-based nationalregistry of CS and GCM to help with the collection and storage of data for futureresearch.

4.1StudypopulationPatientsdiagnosedwithbiopsy-provenCS(n=229)andGCM(n=49)inFinlandbetweenOctober1988and July2015wereenlisted through the followingprocess. In2009,allfiveFinnishuniversityhospitalsand17centralhospitalscoveringthewholeofFinland,werecontactedbyemailandifnecessarybytelephone.PossiblepatientswerescreenedfromhospitaldischargedataregistriesusingtheICD-10codeI41.8*D86.8(sarcoidosisof heart) and the code D86 (sarcoidosis) combined with one of the codes I42(cardiomyopathy), I44.1-2(2ndor3rddegreeAV-block), I45.3 (trifascicularblock), I46(cardiac arrest), I47.2 (VT), I49.0 (VF), I49.3 (premature depolarization), I50 (heartfailure) or R00.1 (bradycardia). In addition, colleagues in different hospitals treatingcardiomyopathieswere asked to gather patientswith CS and GCM in their care. Theinvestigator (RK)personally visited eachhospitalwithpotential adult (>18years)CSpatientsandscrutinizedthescreeningdata.TheGCMpatientswereidentifiedfromHUHhospitalregistries.

4.1.1DiagnosticcriteriaFor inclusion, ahistologic confirmationof sarcoidosiswas requiredand thuspatientswith clinically diagnosed CS were excluded. Histological analysis of the heart (EMB,cardiac tissue sample taken during open heart surgery or heart transplantation or atautopsy) was the golden standard. The histological criteria for sarcoidosis were thepresenceofat leastonenon-necrotizingepithelioidcellgranulomawithnomorethansolitary giant cells and eosinophils and without myocardial necrosis and specialstainings to rule out other causes of granulomatous inflammation if consideredpertinent. The histological diagnosis of GCM required widespread inflammatory

60

infiltrateincludingmultinucleatedgiantcells,eosinophlis,lymphocytesandhistiocytestogetherwith a variabledegreeofmyocardialnecrosis in the absenceofwell-formedgranulomas. The histological tissue sampleswere originally analyzed by pathologistsfrom different university hospitals. For study V, the GCM biopsy samples werereanalyzedbytwoexperiencedcardiacpathologistsinHUH.Ifwell-formedgranulomasweredetected, thepatientswereexcluded forhavingCSaccording to theMulticenterGCMStudyGroupcriteria(11).

In case of missing cardiac tissue verification, extracardiac organ histologic proof ofsarcoidosis in combination with both clinical manifestation indicative of myocardialdisease and abnormalities compatible with CS in either LGE-CMR, 18F-FDG-PET orechocardiographywasrequired.ThetypicalimagingfindingscompatiblewithCSwereLGEinanon-coronaryarterydiseasepatterninLGE-CMR,focal18F-FDGuptakewithorwithout a concomitant perfusion defect in PET, and LV dysfunction and/or septalabnormalitiesatechocardiography.Intotal,ourinclusioncriteriaadheredtotherecentexpertconsensusguidelines(4).

Isolated CS was defined as sarcoidosis involving the heart without prior history orevidentsignsorsymptomsofsarcoidosisoutsidetheheart.Thediagnosticwork-upforscreening extracardiac organs included chest X-ray, clinical examination, basiclaboratory exams, and a detailed medical history, but further studies such as chest-computedtomographyorophthalmologicexaminationswerenotroutinelyconducted.Patientswhohad18F-FGDuptake inmediastinal lymphnodes inPETbutnosignsofsarcoidosiselsewhereoutsidetheheartwereincludedintheisolatedCSgroup.

4.1.2StudiesI-V-The cohort in study I constituted of the 110 patients diagnosed with histologicallyproven CS across the whole of Finland between October 1988 and February 2012detectedthroughtheaforementionedconduct.AllthepatientsinstudiesII-IIIandsomeofthepatientsinstudyIV,weresubgroupsofthiscohort.

-The cohort in study II (conductedprior to study I)was screened fromHUHmedicalrecordsandtheEMBregistry.BetweenJanuary2000andDecember2010,576patientsunderwentdiagnosticEMBs,excludingtransplantsurveillancebiopsies.Biopsy-provenCSwasdiagnosed in 52patients and33 of themwithout extracardiacmanifestationswereincludedinstudyII.

-The cohort in study III (conducted prior to study I) was identified from the HUHpacemaker registry. Between January1999 andApril 2009 therewas a total of 6420pacemakerimplantations,ofwhich133werefor2ndor3rdAV-blockinpatientsaged18-55years,andthesepatientsconstitutedthepopulationinstudyIII.

61

-The cohort in study IV consistedof62patients fromsixFinnishhospitalswithnew-onsettreatmentnaïveCSdiagnosedbetweenDecember2010andDecember2014withhs-cTnT/Imeasuredbeforeandfollowingtheinitiationofcorticosteroidtreatment.Tobe included, the patients had to have estimated glomerular filtration >60ml/min/1.73m2bytheModificationofDietinRenalDisease(MDRD)formula.

-The cohort in study V consisted of the 32 patients diagnosed with histologicallyconfirmed GCM and seen in HUH between January 1991 and December 2011. ThepatientswereidentifiedfromHUHmedicalrecordsandHUHpathologyregistry.

4.2DatacollectionAfterpatient inclusion,RK reviewed the relevanthospital charts fordataonpatients’demographics, histologic findings, symptoms, initial and later clinical manifestations,results of imaging and laboratory studies, evolution of LVEF, invasive procedures,details of treatmentwith drugs and devices, and the occurrence of adverse outcomeevents. Since the initial data collection on-site in different hospitals, the follow-upinformationondiseaseprogressionandadversecardiaceventswascollected throughtheMIDFINnetwork.ThelastCSpatientswereincludedinFebruary2012andadverseeventswererecordeduptotheendofDecember2013(I).Forthisthesis,thenumberofnew CS and GCM caseswas updated until July 2015 and the updated data is in partunpublished. The work started in 2008 and data collection was fully retrospectivebetween1988 and 2008 andpartly prospective thereafter. The causes of deathweredeterminedbyhospitalchartreview.Themortalitydatawasdouble-checkedfromtheFinnish Population Register Centre in January 2014. All imaging studies wereperformedaspartofroutineclinicalcare.Themeasurementsofbiomarkersweretakenandanalyzedaspartofclinicalroutineexceptsomeofthehs-cTnTmeasurementstakenduring outpatient visits at HUH (Study IV). Detailed imaging and biomarker analysisprotocolsaredescribedintheoriginalarticles.

4.3EthicalaspectsThe nationwide study (I) had the approval of the national ethical review board(STM/1219/2009).Theotherstudies(II-V)wereapprovedbytheinstitutionalreviewboard (Ethics Committee, Department ofMedicine, Helsinki UniversityHospital). ThestudieswereconductedaccordingtotheDeclarationofHelsinki.Thepatientsgavetheirinformed,writtenconsenttodatacollectionforthenationalMIDFINregistry.

62

4.4StatisticalanalysesContinuousvariableswerepresentedasmeanvalue±SDwhensampledistributionwassymmetric,andmedian(min-max)whensampledistributionwasskewed.Categoricalvariableswerepresentedasabsolutenumbersandpercentages.Comparisonsbetweengroupswere analyzedusing theChi-square test, or Fisher’s exact testwhenexpectedvalueswerelessthan5,forcategoricaldata.Forcontinuousvariables,theStudent’st-test, Kruskall-Wallis, analysis of variance or Mann-Whitney test were applied asappropriate.Within-groupcomparisonsweremadeusingWilcoxon’ssignedrankstestand McNemar’s test. In all tests, a 2-tailed P<0.05 was considered statisticallysignificant.

Survivalanalyzeswerecalculated,first,fromtheonsetofsymptomsand,second,fromthe date of CS orGCMdiagnosis. Thus, analyseswere performed for total populationand separately after excluding the cases diagnosed post-mortem or from explantedhearts.InstudyItheoutcomeeventswerepresentedseparatelyfor1)cardiacdeath,2)a composite of cardiac death and transplantation, whichever came first, and 3) acomposite of cardiac death, transplantation, and aborted sudden cardiac death. Instudies II-V the time-dependent outcome events included 1) a composite of cardiacdeathandcardiactransplantation(V),2)abortedsuddencardiacdeath(VFtreatedbyexternalorinternaldefibrillation),andsustainedVT(II,III),and3)newhigh-gradeAV-block,whichevercamefirst(IV).SurvivalcurvesfreeofthecompositeendpointwereplottedbytheKaplan-Meiermethod,andfactorsinfluencingsurvivalwereanalyzedbythe log rank test andbyCox regression analysis. The analyseswereperformedusingSPSSversions17.0-22.0forWindows(SPSSInc.;Chicago,IL,USA).

63

5RESULTS

5.1IncidenceandprevalenceofclinicalCSandGCMinFinland(I,V)Atotalof229CSpatientswerediagnosed inFinland fromOctober1988through July2015(studiesI-IVandKandolinetal.unpublishedresults).TheannualdetectionrateofCSwas0.6per100000adults(>18years)inthelastfull2-yearstudyperiodbetween2013and2014.ThegeographicaldistributionofCSpatientsinFinlandispresentedinFigure10.TheprevalenceofhistologicallydiagnosedCScasesin2012was2.2per100000(I).

Figure10.DistributionofhistologicallydiagnosedCSpatientsinFinlandbetween1988and2015byhealthcaredistricts.

SHP=sairaanhoitopiiri(Finnishforhealthcaredistrict),HUS=Helsinginjauudenmaan,PP=Pohjois-Pohjanmaan,PS=Pohjois-Savon,P=Pirkanmaan,PK=Pohjois-Karjalan,VS=Varsinais-Suomen,KS=Keski-Suomen,EP=Etelä-Pohjanmaan,IS=Itä-Savon,KP=Keski-Pohjanmaan,SAT=Satakunnan,EK=Etelä-Karjalan,KH=Kanta-Hämeen,L=Lapin,PH=Päijät-Hämeen.

HUS126

PPSHP34PSSHP19

PSHP17

PKSHP10VSSHP8

KSSHP3EPSHP2

ISSHP2

KPSHP2SAT(=S)SHP2

EKSHP1

KHSHP1LSHP1

PHSHP1___________

=229

64

Overthestudyperiodof26years,thedetectionrateofCSmultiplied.Morespecifically,theannualnationwidedetection rateof0.5patientsperyear in theearliest two-yearperiod between 1988-1989 increased to 27 per year in the last follow-up two-yearperiod from2013-2014 (Figure 11). Thus thenumberofpatientsdiagnosedwithCSincreasedbymorethan50-foldduringthestudyperiod.

Figure 11.ThenumberofnewCScasesdiagnosed in two-yearperiods from1988 to2014inFinland.

65

Atotalof49patientswithhistologicallyconfirmedGCMwereseenatHUHduringtheGCMstudyperiodbetweenJune1991andDecember2014(studyVandKandolinetal.unpublished results). HUH is a nationwide referral center for cardiac transplantationandthustheseGCMcasesarelikelytopresentthemajorityofGCMcasesdiagnosedinFinland over the 23-year range. Therefore, the annual detection rate of histologicallyverifiedGCMinFinlandbetween2013and2014was0.13per100000.SimilarlytoCS,thediagnosticratemarkedlyincreasedoverthestudyperiod,especiallyoverthelatterhalfofthestudy(Figure12).

Figure12.ThenumberofnewGCMcasesdiagnosedintwo-yearperiodsfrom1991to2014inFinland.

66

5.2Patientcharacteristicsandassociateddisorders(I)5.2.1CSThemean age of the 110 CS patientswas 51± 9 years, ranging from 27 to 69 years.Therewasafemalepredominance:65%(71patients)werewomen.Table8presentsthe associated illnesses in CS patients (Kandolin et al., unpublished results). CS wasassociated with other autoimmune diseases in 16% of cases (18 patients). Theautoimmune disorders preceded the onset of CS in all cases. Thyroid disordersweremostcommonin12%(13patients)andwereoverrepresentedcomparedtothegeneralpopulation(271).

Seventy-one (65%) of the 110 patients with CS underwent coronary angiography orcoronarycomputedtomography.Coronaryarterydiseasewasdiagnosedinfive(4.5%)patients. Two patients had mild coronary stenoses and were treated withpharmacotherapyalone.Onepatientunderwentapercutaneouscoronaryintervention.Furthermore,onepatientwhounderwentseveralpercutaneouscoronaryinterventionswas later referred to cardiac transplantation due to terminal heart failure and thediagnosisofCSwas revealed from theexplantedheart.Additionally, onepatientwithpulmonarysarcoidosisandVTsunderwentcoronarybypasssurgeryandthediagnosisof cardiac sarcoid involvement was found in cardiac tissue samples taken duringsurgery.

67

Table8.Associateddisordersin110patientswithCS(Kandolinetal.,unpublishedresults)._______________________________________________________AutoimmunedisordersHypothyroidism 11(10%)TypeIdiabetesmellitus 3(3%)Rheumatoidarthritis 1(1%)Inflammatoryboweldisease 1(1%)Coeliacdisease1(1%)Multiplesclerosis1(1%)________________________________________________________RespiratorydiseaseAsthma 10(9%)COPD 1(1%)________________________________________________________MalignancyThyroidcancer 2(2%)Breastcancer 2(2%)Coloncancer 1(1%)Lymphoma 1(1%)Thymoma 1(1%)Melanoma 1(1%)Myeloma 1(1%)________________________________________________________OtherHypertension 26(24%)Dyslipidemia 17(15%)TypeIIdiabetesmellitus 13(12%)Coronaryarterydisease 5(5%)(2withmedicaltreatmentalone)Kidneyinsufficiency 1(1%)______________________________________________________

5.2.2GCM(V)Themean age of the 32 GCM patientswas 52.5 ± 12.7 years, ranging from 35 to 69years. Themajority of them, 69% (22patients)were female.Among the32patientswithGCM,19%(6patients)hadassociatedautoimmunedisorders, including reactivearthritis, iritis, and thyreoiditis (n=1), coeliac disease (n=2), psoriasis (n=1),rheumatoid arthritis (n=1), and hypothyroidism (n=1). One patient with vitiligo hadthymoma-associated orbital myositis, i.e., symptoms compatible with giant cellpolymyositis. Furthermore, in accordance with genetic susceptibility, among the 32patients, two women were sisters and they both had a medical history of coeliacdisease. These twopatients had ahealthybrother and their first-degree relativesdidnothaveevidenceofautoimmunediseaseorcardiomyopathy.

68

5.3Presentingmanifestations(I)ThefirstclinicalmanifestationsinCSandGCMarepresentedinTable9.AV-blockwasthemostfrequentfirstmanifestationinCS,presentingin48(44%)patients.FollowingAV-block in frequency, ventricular arrhythmias were the second most commonmanifestation,presentingin33%patients.Thethirdmostcommonmanifestationwassymptomaticheartfailure,presentingin18%.ThemedianLVEFatechocardiographyinpatientspresentingwithheartfailurewas30%(range:15-47%).

In patients with GCM, AV-block and heart failure were equally common firstmanifestations, in 31%, followed by ventricular arrhythmias in 25%. FirstmanifestationsinGCMaredescribedinTable9.

Table9.Modesofpresentationin110CSpatientsand32GCMpatients._____________________________________________________________________________________________Firstclinical AllCS Clinically CSwith AllGCMManifestation patients, isolatedextracardiac patients, N=110 CS,n=71*disease, n=32 n=39*_____________________________________________________________________________________________2ndor3rd 48(44%)34(48%)14(36%) 10(31%)degreeAV-block_____________________________________________________________________________________________VTorVF 36(33%) 27(38%) 9(23%) 8(25%) VT31(28%) VT7(22%) VF5(5%) VF1(3%)¶_____________________________________________________________________________________________Heartfailure 20(18%) 8(11%) 12(31%) 10(31%)_____________________________________________________________________________________________Others 6(6%)† 2(3%) 4(10%) 4(13%)§_____________________________________________________________________________________________Thedatapresentsnumbersofpatientsand(%).CS=cardiacsarcoidosis,GCM=giantcellmyocarditis,AV-block=atrioventricularblock,VT=ventriculartachycardia,VF=ventricularfibrillation.*For comparison between clinically isolated CS and CS with extracardiac disease,p=0.016.¶Out-of-hospitalcardiacarrest.†Multipleventricularprematurebeats(n=4),mitralregurgitation(n=1)andpericardialeffusion(n=1).§ Chest pain and ST-segment changes in ECG,mimicking acutemyocardial infarction(n=3)andperimyocarditis(n=1).

69

Manifestations from outside the heart were observed in 39 (35%) CS patients. Thelungswerethemostcommonlyaffectedorganin23(21%)patients,followedbylymphnodes(n=8),skin(n=7),centralnervoussystem(n=5),eyes(n=4),andliver(n=3).

Thediagnosticdelay,ie.,themediantimefromfirstcardiacmanifestationtohistologicCS diagnosis, was 9.5 months (range 0.3-168) in the whole group of CS patients(n=110).IncasewheretherewasAV-blockasthefirstpresentation(n=48),themediandiagnostic delaywas 14months, compared to sixmonths in cases of heart failure orVT/VFaspresentingmanifestations(p=0.182).

5.3.1ImportanceofCSandGCMinidiopathicAV-block(III)AkeydiscoveryofthisworkwasthesignificanceofCSandGCMinthecausalspectrumof AV-block in young andmiddle-aged adults. From theHUHpacemaker registry,weidentifiedatotalof6420patientswhohadundergonepacemakerimplantationfor2ndor 3rd degree AV-block between 1999 and 2009. Patients <18 years were excludedbecause of the high proportion of congenital heart diseases and the low incidence ofsarcoidosisinthisagegroup.Respectively,patientsaged>55wereexcludedduetothehigh prevalence of conduction tissue degeneration. Consequently, there were 113patientsaged18-55years,and61ofthemhadaknownetiologyofAV-blockatthetimeofpacemakerimplantation.TheotheretiologiesarepresentedinFigure13(Kandolinetal.,unpublishedresults).Theremaining72patients(withamedianageof47,61%female)hadunexplainedAV-blockatthebaseline.Afterdiagnosticstudies,18ofthe72(25%, 95% CI 15-30%) patients with initially unexplained AV-block were diagnosedwitheitherCS(n=14)orGCM(n=4).Additionally,fourpatientshadechocardiographicand 18F-FDG-PET findings compatible with CS, but due to the lack of histologicverification, theywereclassifiedashaving idiopathicAV-blockandclinicallyprobableCS.Consideringall2ndto3rddegreeAV-blocksinthisagegroup,CSandGCMexplained14% (18 out of 133) and including the non-histologically proven CS cases, thepercentageis17%(22outof133).

70

Figure 13. Etiologies for AV-block with a known cause at the time of pacemakerimplantation(n=61)(Kandolinetal.,unpublishedresults).

TherewerecharacteristicdifferencesbetweenthepatientswithCSorGCMcomparedtothepatientsinwhomAV-blockremainedidiopathicafterthediagnosticwork-up.First,89%ofthepatientswithCSorGCMwerefemale,comparedtothe52%inpatientswithidiopathic AV-block (p<0.05). Other characteristic differences were lower LVEF atpresentation (median 50% vs 61%, p<0.05), a higher percentage of distal AV-block(88% vs 54%, p<0.01) and a higher proportion of septal abnormalities atechocardiography (56% vs 28%, p<0.01) in CS and GCM patients compared toidiopathic AV-block patients. None of these characteristics differed between CS andGCMpatients.

5.3.2ClinicallyisolatedCS(I)InthetotalCSpopulationofstudyI,71outof110(65%)hadclinicallyisolatedCStype(see 4.1.1). The remaining 39 patients (35%) had extracardiacmanifestations. Therewere marked differences in patients with isolated and extracardiac disease types(Table 10). First, isolated CSwas associatedwith female predominance. Second, thepatientswith isolated disease type had a higher frequency of LV dysfunction, LGE atCMR,andseptalabnormalitiesindicatingmorewidespreadcardiacinvolvement.Inlinewith this, the laboratory markers of granulomatous inflammation were higher inpatientswith extracardiac disease (see 5.6). Furthermore, isolated CSwas related toworse survival free of cardiac death, transplantation, and aborted sudden death (see5.8.2.1).

Postprocedural, 18

Ischemic heart disease, 8

Congenital heart disease, 8

Autoimmune disease, 5

Sceletal myopathy, 5

Dilated cardiomyopathy,

4

Heart transplant, 4

Endocarditis, 2

Radiation therapy, 3 Others, 3

Postprocedural Ischemic heart disease Congenital heart disease Autoimmune disease Sceletal myopathy Dilated cardiomyopathy Heart transplant Endocarditis Radiation therapy Others

71

Table 10. Differences in patient characteristics in clinically isolated CS vs CS withextracardiacmanifestations(I)._____________________________________________________________________________________________Characteristic Clinically CSwithknown P* isolatedCS, extracardiac n=71 disease,n=39_____________________________________________________________________________________________Age,years 51±9 50±9 0.463(mean±SD)_____________________________________________________________________________________________Gender,nof 53(75%) 18(46%) 0.003females_____________________________________________________________________________________________Echocardiographicfindings LVEF<50% 49/71(69%) 16/39(41%) 0.004LVdilatation 27/64(42%) 16/38(42%) 1.000Septalthinningor 48/70(69%) 15/39(38%) 0.002thickening_____________________________________________________________________________________________CMRfindingsLategadoliniumenhancement 36/38(95%) 13/21(62%) 0.002_____________________________________________________________________________________________18F-FDG-PETfindingsFocalFDGuptake 34/46(74%) 14/20(70%) 0.743Mediastinallymph 22/31(71%) 8/11(73%) 1.000nodeFDGuptakeFDGuptake 12/30(40%) 4/9(44%) 1.000outsideheartandmediastinum_____________________________________________________________________________________________CS= cardiac sarcoidosis, LVEF= left ventricle ejection fraction, LV= left ventricle,CMR=cardiacmagneticresonanceimaging,18-FDG-PET=18-fluorodexyglucosepositronemissiontomography.

72

5.4ImagingandECGfindings(I)5.4.1EchocardiographyIn thecohortof110CSpatients,echocardiographywasperformedat leastonce inallpatients, either at diagnosis or in case of post-mortem and post-transplantationdiagnoses, before diagnosis. At baseline, 65 out of 110 (59%) patients had reducedLVEF(<50%)withmedianLVEF45%±14%.LVwasdilated(LVEDD>60mminmenor> 55 mm in women) in 43 out of 102 (42%). Interventricular septal thinning orthickening was discovered in 63 out of 109 (58%). It is notable, that not all thesefindingswereevidentat the firstperformedechocardiography,but insteaddevelopedby the time of CS diagnosis. Other abnormalities detected were aneurysmatic wallmotion,pericardialeffusion(usuallymildbutrequiringfenestrationinonepatient)andmildtomoderatemitralinsufficiency.

In thecohortof32patientswithGCM,echocardiographywasperformed inallexceptone patient with sudden cardiac death as the first manifestation. Initially, LVEF wasreduced(<50%)in23outof32(74%)withamedianLVEFof38±13%.LVdilatationwasobserved innine(28%)patients.Septal thinningor thickeningwasdiscovered in21(68%)andLVaneurysmsinthree(10%)patients.

5.4.218F-FDG-PETOf the 110 CS patients, 66 (60%) underwent 18F-FDG-PET as a part of diagnosticevaluation.Some48(73%)ofthemhadfocallyincreasedmyocardialFDGuptakeandin44 (67%), FDGuptakewas superimposableuponaperfusiondefect.AccumulationofFDG in mediastinal lymph nodes was detected in 30 of the 42 (71%) patientsundergoing thoracic 18F-FDG-PET. In comparison, hilar or mediastinallymphadenopathy was observed in 15 (14%) patients in native chest X-rays. In 18patients, FDG-positive mediastinal lymph nodes offered a biopsy target in case ofnegativeEMBandhighsuspicionofCS.

Ofthe32GCMpatients,12(38%)underwent18F-FDG-PETrevealingfocalFDGuptakein 10 (83%) and involving the septum in all. A concomitant perfusion defect wasobservedinnineandtwopatientshadaperfusiondefectwithoutFDGaccumulation.

5.4.3LGE-CMROf the 110 CS patients, 59 (54%) underwent LGE-CMR as part of diagnostic process.Overall,LGEwasobservedin49(83%).Localwallthinningorthickeningwasdetectedin29outof50(58%).Otherfindingsperceivedwereventriculardysfunction,dilatationof ventricles, and aneurysmatic wall abnormalities. In the 32 GCM patients, CMRshowedareasofLGEineachoftheninepatientswhounderwentthestudy.

73

5.4.4ElectrocardiogramAn electrocardiogram was available in 107 out of 110 CS patients. At the time ofdiagnosis some abnormalities, including bundle branch blocks, any level ofatrioventricularblocks,pathologicQ-wavesorST-Tchanges,voltageabnormalities,orfrequentVPBswereobserved in92 (86%)patients. AV-blockwas themost commonfinding,presentin48(45%).RBBB,observedin40(37%)patients,wasalmosttwiceasfrequentasLBBB,foundin22(21%)patients.

5.5Biopsyconfirmation(I,II,V)Inthenationalcohortof110CSpatients,92patientshadundergoneoneormoreEMBsoftheRV,orofbothventricles(I).Ultimately,thehistologicconfirmationofsarcoidosiswas obtained from EMB in 55 (50%) patients. In 47 cases, biopsy verification wasacquired from extracardiac biopsy sites, including PET-positive mediastinal lymphnodes(n=18),lungs(n=11),peripherallymphnodes(n=8),skin(n=5),liver(n=3),andcentralnervoussystem(n=2).Intheremainingeightpatientsthehistologicverificationwasobtainedfromexplantedheartsattransplantation(n=6)oratautopsy(n=2).

ThevalueofrepeatedEMBsandmediastinallymphnodebiopsiesindiagnosingCSwasanalyzedinthecohortof55patientswithsuspectedCSundergoingdiagnosticbiopsiesatHUHbetween2000and2010(II).In33patients(60%),theCSdiagnosiswasfinallyverifiedbybiopsy,andin22(40%)thehistologicconfirmationremainedunobtainabledespite cardiac symptoms and imaging results compatible with CS. Considering thepatientswitheventualhistologicalCSconfirmation,thefirstEMBsessionrevealedCSin10outof31patients,resultinginasensitivityof32%.AsecondandthirdEMBsessioncontributedsignificantlytothediagnoses,asthecumulativeyieldofrepeatedEMBswas55% (17 out of 31). In one patient who had not undergone diagnostic biopsies, thehistologicdiagnosiswasmadefromthenativeheartremovedattransplantation.

Due to the poor sensitivity of EMB, the diagnostic strategy at HUHwas upgraded in2006totarget18F-FDG-PET-positive,i.e.,“hot”mediastinallymphnodesincaseofhighsuspicionofCSdespitenegativeEMBs.Twelvepatientswith “hot”mediastinal lymphnodesunderwentsamplingoftheselymphnodesatmediastinoscopyandin11(92%),the samples identified sarcoidosis. Nevertheless, the 22 patients without finalhistologicalCSconfirmationunderwentatotalof33EMBsandonemediastinallymphnodebiopsywithnegativeresults.IfthesepatientswereclassifiedashavingCS,inthesameway asmany recent studies have done, the sensitivity ofmultiple EMBs in thiscohortwouldbe31.5%(17outof54).Inthetotalpopulationofbiopsiedpatients,thesensitivityofEMBinhistologicallyorclinicallydiagnosedCSwas19%(10outof53)inthefirstEMB,22%(5outof23)inasecondEMBand50%(2outof4)inathirdEMB.Overthestudyperiod,themeannumberofrightventricularsamplesperbiopsysessionincreasedfrom5.5(range4-7)duringtheearlyhalf(from2000to2005)to6.2(range3-10)overthelatterhalf(years2006-2010).ThemeannumberofLVsamples,acquired

74

onlyafter2006butwith increasing frequency,was5.3 (range2-9)per session.Openchestcoreneedlebiopsieswerenotperformedinthiscohort.

ThediagnosticyieldofEMBwasbetterinGCM.Ofthe32GCMpatients,28(88%)hadundergone one or several EMBs (n=28) and surgical biopsies (n=3). The first cardiacbiopsy session revealed GCM in 19 out of 28 cases, resulting in a diagnostic yield of68%. The second cardiac biopsy gave diagnosis in five of seven patients, and a thirdcardiac biopsy in two of two patients. Therefore, the cumulative yield of repeatedcardiacbiopsieswas93%(26outof28).Thethreesurgicalmyocardialbiopsiesweretakeninassociationwithsurgicalprocedures(pericardialdrainage,ventricleresectionandventricularassistdeviceimplantation),andthesampleswerepositiveinall.Lastly,of the32GCMpatients, six ofwhom twohadundergoneEMBswithnegative results,werediagnosedatautopsy(n=4)orfromexplantedhearts(n=2).

5.6Laboratoryfindings(I,IV)Hs-cTnT/I were repeatedly measured in a cohort of 62 CS patients with new-onset,treatment-naïveCS(IV).Atenrollment,hs-cTnTwaselevated(>13ng/l)in26outof50(52%) and hs-cTnI (>0.04 ng/ml) in seven out of 12 (58%) of the newly diagnosedpatients.Elevatedhs-cTnT/IwasassociatedwithLVdysfunction(p=0.001),atendencyto higher NT-proBNP levels (p=0.085), and a tendency towards more heart failure(p=0.064)andlessAV-blockasafirstmanifestation(p=0.069).Morespecifically,inthenew-onset CS patients, LVEF averaged 43± 14% in patients with elevated hs-cTnT/Icomparedto53±10%inpatientswithnormalhs-cTnT/I.

The laboratorymarkersofgranulomatous inflammation, includingACE, lysozymeanddailyurinarycalciumexcretioninthe110CSpatientsarepresentedinTable11.Thelevels of these markers were higher in CS patients with extracardiac sarcoidosiscomparedtoclinicallyisolatedCS.

Table11.LaboratoryfindingsinCSpatients(I)._____________________________________________________________________________________________Laboratory AllCSpatients, Clinicallyisolated CSwithknownP*finding n=110 CSpatients, extracardiac n=71 disease,n=39_____________________________________________________________________________________________ElevatedACE 22/90(24%) 9/59(15%) 13/31(42%) 0.005_____________________________________________________________________________________________ElevatedLZM40/74(54%) 21/48(44%) 19/26(73%) 0.016_____________________________________________________________________________________________ElevateddU-Ca22/45(49%) 9/24(38%) 13/21(62%) 0.029_____________________________________________________________________________________________The data presents numbers of patients (%). CS= cardiac sarcoidosis, ACE= serumangiotensin-convertingenzyme,LZM=serumlysozyme,dU-Ca=dailyurinarycalcium.*p-valuesforcomparisonbetweenclinicallyisolatedandextracardiacdisease.

75

5.7Descriptionofmanagement(I,V)5.7.1ImmunosuppressivetreatmentAll the 102 CS patients diagnosed clinically before death or cardiac transplantationreceived disease-modifying immunosuppressive therapy (I). There were slightdifferencesbetweenthehospitalsinvolvedinthestudyinimplementingthetreatment,but the therapy followed same main principles. Corticosteroids were started atdiagnosis,withtheinitialdailyprednisone-equivalentdoserangingfrom30mgto80mg.Thestartingdosewas<60mgin42and≥60mgin60patients.Corticosteroiddosewasdecreasedslowlytoaprednisone-equivalentdailydoseofapproximately10mgwithinsix months from starting treatment. In 48 patients, steroids were used continuouslyuntil the end of follow-up. In the remaining 54 patients, steroids were eitherdiscontinuedor used intermittently due to fluctuatingdisease activity or steroid sideeffects.Inpatientswithactivediseasefollowingcorticosteroidtapering,steroid-sparingimmunosuppressants were introduced. Azathioprine was utilized in 50 patients,methotrexate in sixpatients,mycophenolatemofetil in threepatients, cyclosporine intwopatientsandinfliximabinonepatientduringthestudyperiod.

All the 26 patients diagnosed with GCM before death or transplantation receivedcombined immunosuppressionwith2-4drugs(V).The treatment includedsteroids in26 patients, azathioprine in 24 patients, cyclosporine in 20 patients, mycophenolatemofetil in three patients, muromonab in one patient, gammaglobulin in one patient,methotrexateinonepatientandintravenousmethylprednisoloneinsevenpatients.Themost common drug combination therapy was triple immunosuppression withcorticosteroid,azathioprine,andcyclosporineutilizedin17(65%)patients.Prednisonewasstartedatapproximately60mg(or0.75-1mg/kg)perdayandtaperedtoa10mgdaily dose within six months. Prednisone was discontinued if the patient had beenstabile for 6-12months or in the case of intolerable side effects. The target dose ofazathioprinewas1.5-2mg/kg/day.Thetargetconcentrationofcyclosporinewasatthelowertherapeuticrangeforpost-cardiac-transplantation(80-120μg/L).

5.7.2DeviceandmedicaltherapyDuringthestudyperiod,apermanentpacemakerwasimplantedin87outof110(79%)ofCSpatients (I).Toprevent suddencardiacdeathand tomanagemalignantVTs, anICDwasimplantedin59(54%)patients.TheindicationforICDwasprimaryprevention(LVEF≤35%) innine (15%)patients, secondaryprevention(historyof sustainedVT,including prior cardiac arrest) in 41 (69%), and other indication (treating clinician`sjudgment)innine(15%)patients(Kandolinetal.,unpublishedresults).Abiventricularpacemakerwasimplantedin17(15%)patients.ThedrugtreatmentinCSincludedbetablockers in 104 (95%) patients, angiotensin-converting enzyme inhibitors orangiotensin receptor blockers in 89 (81%) patients, diuretics in 48 (44%), andamiodaronein25(23%)patients.

76

Of the 32 GCM patients, 11 (34%) received inotropic agents including levosimendan(n=6),dopamine,ordobutamine(n=5),norepinephrine(n=1),epinephrine(n=1),andmilrinone (n=2) (Kandolin et al., unpublished results). Of the 26 GCM patientsdiagnosed before death or transplantation, 25 (96%) received beta blockers and 19(73%)antiarrhythmicdrugs (mainly amiodarone).Apacemakerwas implanted in21(81%)andICDin18(69%)patients.

5.8Outcome(I,IV,V)5.8.1Observationsduringtreatmentandfollow-upinCS

5.8.1.1ClinicalmanifestationsThecohortof110CSpatientswas followed foramedianof79months, i.e.,6.6years(range 12-303 months) from first symptom onset. In addition to the 48 patientsreceivingapacemakerfor2ndor3rddegreeAV-blockaspartoftheearlywork-up,nine(8%) patients developed a new 2nd or 3rd degree conduction block necessitatingpacemakerimplantationlaterduringthefollow-up.Thus,atotalof52%ofthewholeCScohort had significant atrioventricular conduction abnormalities during the course ofthedisease.

Furthermore,15(14%)patientsexperiencedVFand41(37%)patientshadsustainedVT,eitheras firstmanifestationor laterduringthediseasecourse.Duringthemedianfollow-upof53months,i.e.,4.4years(range0-183months)fromICDimplantationuntilthe close of the study period, 24 (41%) of the 59 patients with ICD received anappropriate ICD therapy (shock and/or anti-tachycardia pacing) (Kandolin et al.,unpublishedresults).

Regarding systolic dysfunction, 74 (67%) of the patients either had systolic LVdysfunction(EF<50%)atdiagnosis(n=65),ordevelopedit laterduringthefollow-up(n=9).

5.8.1.2TreatmenteffectevaluationTheclinicalresponsetocorticosteroidtherapywasevaluatedfromthreeaspects.First,initiationof corticosteroids, resulted in the recoveryof atrioventricular conduction in20% (7 of the 35 patients with accurate pacemaker follow-up data concerning thepercentage of pacing) defined as <10% ventricular pacing on follow-up. Second, LVfunctionatechocardiographywasanalyzedbeforeand12monthsaftertheinitiationofsteroids.Inthe102patientsreceivingimmunosuppressants,therewasnodifferenceinLVEF at diagnosis and after treatment (LVEF 44.9 ± 12% vs 45.4± 11%) (p=0.532).Nevertheless, in the subgroup of patients with severe LV impairment (EF <35%) atdiagnosis, mild yet significant recovery of LV function was observed (LVEF beforetreatment 27.9± 4.1% vs LVEF 12 months after treatment 34.1± 8.3%) (p=0.005).

77

However,noameliorationofLVEFwasperceivedinthesubgroupsofnormalLVEF(≥50%)(p=0.145)ormoderatelyimpairedLVEF(35-49%)(p=0.979).

Duringfollow-up,18F-FDG-PETwasrepeatedin22(33%)patientstoevaluatesteroidresponseanddiseaseactivity(Kandolinetal.,unpublishedresults).The18F-FDG-PETimageswereanalyzedmostlyqualitativelyandtheirdiagnosticandprognosticvalueatfollow-upwasnot consideredbeneficial for repeated studieswith theprotocol of thetime. Therefore, the use of repeated 18F-FDG-PET studies decreased until the end ofstudyperiod.

Theearlybiochemicalresponsetotheinitiationofcorticosteroidtherapywasassessedin the62patientswithnew-onsetCSandmultiplehs-cTnT/Imeasurements (IV).Theinitiationoftreatmenteitherdecreased(n=8,33%)ornormalised(n=16,67%)thehs-cTnT/I levels in four weeks in all the 24 patients with elevated pre-treatmentconcentrations who underwent repeated measurements after four (±1) weeks ofsteroid therapy. Hs-cTnT/I remained normal in all the 14 patients with normal pre-treatment concentrations (p=0.00003). Figure 14 shows the hs-cTnT response toinitiation of corticosteroid therapy in all 32 patients who underwent the one-monthrepeatmeasurement.

Figure14.Theearlyresponseofhs-cTnTtoinitiationofsteroidtherapy(IV).The p-value is from aWilcoxon signed rank test for paired comparisons. The dashedhorizontallinesignifiestheupperlimitofthenormalreferencerange(≤13ng/l).

78

During follow-up, three different profiles of serial hs-cTnT/Imeasurements could bedistinguished (Kandolin et al., unpublished results). First, persistent hs-cTnT/Ielevationwasfoundinsevenpatients(11.3%)ofwhomfourhadsevereLVdysfunction(EF ≤35%). Second, 48 patients (77.4%) had either normal hs-cTnT/I in allmeasurements or initially elevated hs-cTnT/I that normalised and remained normalthroughout the follow-up. The third group comprised of seven patients (11.3%) inwhom hs-cTnT/I was initially normal, or normalised with treatment, but becameelevated later during follow-up. In the group with late increases in hs-cTnT/I, theelevations were frequently preceded by a reduction in the intensity ofimmunosuppressionandoftenrespondedtoanincreaseinthedoseorreinstitutionofprednisone. Figure 15 illustrates a case example of hs-cTnT evolution duringcorticosteroidtreatment(Kandolinetal.,unpublishedresults).

Figure 15. Serial hs-cTnT concentrations (lower panel) and daily prednisone dose(upperpanel)duringthethree-yearfollow-upofa43-year-oldfemalewithchronicCS.Hs-cTnT was elevated (>13 ng/l) at enrollment, and there were two additional lateincreases following reduction of the daily prednisone dose from 5mg to 2.5mg thatresponded each time to restoration of a higher dose (Kandolin et al., unpublishedresults).

Mildimmunosuppressivetreatmentadverseeffects,mostcommonlyinsomnia,edema,hypokalemia, impaired glucose tolerance due to corticosteroids and leukopenia,hepatic/pancreatic irritation, and nausea due to azathioprine, were commonlyobserved. Intolerable side effects (severemyopathy, psychosis, aseptic joint necrosis,and severe insomnia with diarrhea and hypokalemia) leading to permanent

79

discontinuation of corticosteroidswere observed in four patients. No life-threateningadverseeventswereobserved.

5.8.2SeriousoutcomeeventsandsurvivalinCSFigure16presentstheseriousoutcomeeventsinthecohortof110CSpatients.Atotalof 10 (9%) patients suffered a cardiac death and six (5%) died of non-cardiac causeduringthemedianfollow-upof79months, i.e.,6.6years(range12-303months)fromsymptomonsettodeath,transplantationorendoffollow-up.Ofthe10cardiacdeaths,nineweresuddendeaths(intwocasesdiagnosiswasnotmadeuntilautopsy)andonedue to terminal heart failure. The cardiac 1-, 5-, and 10-year Kaplan-Meier survivalprobabilitiesarepresentedinTable12forthetotalpopulationandforthe102patientsdiagnosed clinically before death or transplantation and thus receivingimmunosuppressivetreatment.

During the follow-up11 (10%)patients underwent a cardiac transplantation. Twoofthe transplanted patients died within one month of postoperative complications(multiple organ failure and intra-cerebral hemorrhage). Another two patients died ofgraft failure, without signs of sarcoidosis recurrence, three and five years post-transplantation. The remaining seven transplanted CS patients were alive withoutrecurrence at the end of follow-up, median 60 months (range 22-249) fromtransplantation.

Consequently,attheendofmedian6.6-yeartotalfollow-upfromsymptomonset,atotalof 20 out of 110 (18%)patients haddied and seven (6%)patientswere alivewith atransplantedheart.The cardiac survival freeof transplantation ispresented inTable12. Additionally, two patients were awaiting transplantation, one with a ventricularassist device, at the close of follow-up. The main indication for transplantation wasterminal heart failure in 11 listed and transplanted patients, and incontrollableventricular arrhythmias in two, both ofwhomboth also had severe heart failure (EF<35%). The time from symptom onset to transplantation ranged from five to 89months,withamedianof59months.

80

Furthermore,11(10%)otherpatientsexperiencedanabortedsuddendeathasa firstoutcome event. Eight of them were resuscitated from VF and three received an ICDshock for VF. Thus, a total of 32 out of 110 (29%) patients suffered cardiac death,transplantation or aborted sudden death. Table 12 also presents the survival rateswithout the composite event of cardiac death, transplantation and aborted suddendeath.

Figure16.Seriousoutcomeeventsin110CSpatients(I).

110#biopsy*verified#CS#pa3ents#

11#cardiac#transplanta3ons#**>2#post*opera3ve#deaths###**>2#allogra<#failure#deaths#

10#cardiac#deaths;#9#sudden#deaths,##

1#heart#failure#death#

11#aborted#sudden#deaths;#8#resuscita3ons#from#VF,##3#ICD#defibrilla3ons#to#VF#

6#non*cardiac#deaths#

96#cardiac#survivors,#90#survivors,#

89#transplanta3on*free#cardiac#survivors#

78#survivors#without#cardiac#death,#

transplanta3on#or#aborted#sudden#death#

81

Table12.Survivalfreeofmajorcardiacoutcomeeventsinallthe110CSpatientsandinthe102patientsdiagnosedpriortoautopsyandtransplantation(I).

_____________________________________________________________________________________________ 1-yearsurvival,%5-yearsurvival,%10-yearsurvival,%_____________________________________________________________________________________________Cardiacsurvivaln=110 99(94-100) 94(87-97) 89(82-94)n=102 100(96-100) 97(91-99) 93(85-97)_____________________________________________________________________________________________Cardiactransplantation-freesurvivaln=110 97(92-99) 90(82-95) 83(75-89)n=102 99(94-100) 95(88-98) 91(83-95)Patientswithheart failureasfirstmanifestation,n=20 90(67-98) 75(51-90) 53(30-74)_____________________________________________________________________________________________Cardiac,transplantation-freeandabortedsuddendeath-freesurvivaln=110 89(81-94) 78(69-85) 70(61-79)n=102 89(81-94) 82(73-89) 77(68-85)_____________________________________________________________________________________________CS=cardiacsarcoidosis.

5.8.2.1PredictorsofoutcomeHeart failure as a primary manifestation was associated with worse transplant-freecardiacsurvival,andthisappliedacrossthetotalpopulation(n=110)aswellasintheclinicallydiagnosed(n=102)patients(logrankp=0.0001andp=0.023,respectively).Inthepatientspresentingwithheartfailuresignsandsymptoms,the5-yearKaplan-Meiersurvival estimatewas 75% (95%CI, 50.6- 90.4%) compared to 90% (95%CI, 82.4-94.6%) in the total population (Table 12). On the other hand, AV-block as a firstmanifestation was associated with better outcome compared to other firstmanifestations (in the 102 patients, log rank p=0.037). Other predictors of cardiacsurvivalfreeoftransplantationinthetotalpopulationof110patientswereinitialLVEF(p=0.006byCoxregressionandp=0.011bylogrank)andNYHAclassatdiagnosis(logrank p=0.023). On the contrary, the outcomewas not related to age (Cox regression

82

p=0.469),sex(logrankp=0.163),orthetypeofCS(isolatedvsextracardiac)(logrankp=0.132).

We also analyzed characteristics associatedwith survival free of transplantation andabortedsuddendeath.InthetotalCScohortof110patients,LVEFwasassociatedwithsurvival free of this composite outcome event (by log rank p=0.046 and by Coxregressionp=0.017).Instead,survivalfreeoftransplantationandabortedsuddendeathwasindependentofage,sex,andheart failureasthefirstmanifestation.Furthermore,isolatedCStypewasrelatedtoworsesurvival freeof thiscompositeadverseevent inthetotalpopulation(n=110)andintheclinicallydiagnosed(n=102)patients(logrankp=0.005 and 0.015, respectively). This could be explained by the higher frequency ofpresentation with ventricular arrhythmias in isolated CS (see Table 9) and thusclusteringofearlyabortedsuddendeathsinthispatientgroup.

Implementing prednisone treatment with a high (≥ 60mg/day) versus low(<60mg/day) initial prednisone dose did not effect survival free of cardiac death ortransplantation (log rank p=0.561). Neither did the treatment delay from symptomonset to starting corticosteroids (<6 months vs ≥6 months) (log rank p=0.867).Furthermore, to evaluate the effect of corticosteroid in the subgroupof patientswithAV-block as a firstmanifestation, we scrutinized the adverse cardiac events (cardiacdeath, aborted sudden death, sustained VT, new systolic LV dysfunction andtransplantation)onfollow-upinpatientswithearlyCSdiagnosisandthusinitiationofsteroid treatment (<3 months from pacemaker implantation) compared to latediagnosisandinitiationofsteroidtreatment(≥3monthsfrompacemakerimplantation).Duringthefollow-upofamedianof6.6years,50%(9outof18)intheearlydiagnosisgroup compared to 63% (17 out of 27) in the late diagnosis group had an adversecardiac event, but the difference was not statistically significant (log rank p=0.821)(Kandolinetal.,ESCCongress2015abstract83429).

Inthecohortof62newlydiagnosedCSpatientsandserialhs-cTnT/Imeasurementsatdiagnosis and during follow-up (IV), there was a trend towards poorer prognosis inpatients with elevated hs-cTnT/I at baseline (log rank p=0.068). More specifically,duringthemedianfollow-upof17months(1-48months),atotalof16outof62(26%)patients suffered a cardiac adverse event, including sudden cardiac death (n=2),aborted sudden cardiac death (n=5), symptomatic sustained VT (n=8), and newcomplete AV-block (n=1). In the groupwith elevated hs-cTnT/I, 11 out of 33 (33%)patients had an adverse event, compared to five out of 29 (17%) in the group withnormalhs-cTnT/Iatbaseline.

5.8.3OutcomeinGCM(V)Thecohortof32GCMpatientswasfollowedforamedianof15months(range0.3-90.3months)fromsymptomonsettodeath,transplantationorendoffollow-up.Duringthefollow-up,atotalof15(47%)patientssufferedcardiacdeath(n=5,16%)orunderwentcardiactransplantation(n=10,31%).Allthefivefatalitieswerearrhythmicandfourof

83

five patients who died were diagnosed post-mortem. The Kaplan-Meier estimates oftransplant-free survival are presented in Table 13. In the subgroup of 26 patientsdiagnosed clinically before death or transplantation and thus receivingimmunosuppressive therapy, one suffered cardiac death and eight underwenttransplantation.Theirtransplant-freesurvivalfromdiagnosisisalsopresentedinTable13.

Table13.Survivalfreeofmajorcardiacoutcomeeventsinallthe32GCMpatientsandinthe26patientsdiagnosedpriortoautopsyandtransplantation(V)._____________________________________________________________________________________________Cardiac*transplantation-free 1-yearsurvival,% 5-yearsurvival,%survival n=32 69%(50-83%) 52%(34-70%)n=26 77%(56-90%) 63%(42-81%)_____________________________________________________________________________________________*Sameastransplantation-freesurvival.GCM=giantcellmyocarditis.

All the eight patients with known pre-transplant diagnosis of GCM, were listed fortransplantationwithinninemonths fromthediagnosisand the transplantationswereaccomplished within one year of diagnosis in seven of the eight patients. Of the 10transplantedGCMpatients,threediedwithinsixweeksofpostoperativecomplications(large intracardiac thrombus (n=1), bleeding (n=1) andmultipleorgan failure (n=1)).The remaining seven transplanted GCMpatientswere alive at the close of follow-up.One patient had a recurrence of GCM in the graft, presenting as symptomatic heartfailure4.8yearspost-transplantation,whichensuedfromdiscontinuingcorticosteroidandwasresolvedwithaugmentedimmunosuppression.

Themajority of GCM patients had ventricular arrhythmias. The aforementioned fourpatientssufferedsuddenarrhythmicdeathandwerediagnosedonlyatautopsy.Ofthetwo patients diagnosed at transplantation, one had sustained VTs. Of the 26 patientsdiagnosed before death and transplantation, 17 (65%) had sustained VTs, one wassuccessfullyresuscitatedfromVF,andonediedofVFtwoweeksafterdiagnosticEMB.Intotal,24outof32(75%)ofGCMpatientsexperiencedsuddencardiacdeath,abortedsuddencardiacdeathorsustainedVTduringthediseasecourse.Inthe18GCMpatientswith ICD, eight (44%) received an appropriate ICD therapy (shock and/or anti-tachycardia pacing). In the subgroup of 17 transplant-free survivors, 10 (59%) hadsustainedVTsduringfollow-upandthreereceivedICDshocks.

Considering systolic dysfunction in GCM, 28 out of 31 (90%) patients who hadundergoneechocardiographyeitherhadreducedLVEF(<50%)atdiagnosis(n=23)ordevelopedsystolicdysfunctionduringthediseasecourse(n=5).Symptomatic(NYHAII-IV)heartfailure,withLVEFrangingfrom20%to40%,wasobservedinsevenoutof26patients with pre-transplant and pre-mortem GCM diagnosis. Six of them required

84

transplantation and one stabilized with medical therapy. Out of the four autopsydiagnosed patients, two had severe heart failure (EF ≤35%), one had normalechocardiographyayearbeforesuddencardiacdeath,andonesufferedsuddendeathwithout previous echocardiography. Both patients diagnosed at transplantation hadsevere systolic heart failure (EF <20%). Second or 3rd degree AV-block was presentfromthesymptomonsetin11(34%)patients(ofwhominonethepresentingsymptomwas classified as VT according to the principal symptom). None of the GCM patientsdevelopedanewAV-blockduringthestudyperiod.

Age,sexorpresentingsymptomswerenotpredictiveofoutcomeinGCM.InitialLVEFwas not statistically significantly associated with survival free of transplantation.Instead, at follow-up LVEF was higher in the transplant-free survivors (45±11%),comparedtothepatientswhodiedorunderwenttransplantation(26±7%)(p=0.003).

6DISCUSSION

6.1Methodologicalconsiderations6.1.1StudypopulationThisstudyisaretrospectiveanalysisofCSandGCMpatientsinFinland.TheCSpatientswereenlistedbycontactingalltheFinnishuniversityandcentralhospitals.Allthefiveuniversity hospitals and six out of 17 central hospitals responded and screened thepossible CS patients from their registries by using the given ICD-10 codes. The vastmajority of diagnosed CS patients in Finland are treated and followed in universityhospitals or have at least undergone diagnostic procedures in them during thediagnostic work-up. Thus the screening should have been comprehensive, yet somepatientsmayhavebeenmissedduetodeficienciesinhospitalregistries.Moreover,theICD-10classificationwasintroducedinFinlandin1996,andthereforesporadicpatientsdiagnosed between 1988 and 1996 who were lost during follow-up may have beendismissed.

ThereweredifferencesintheprevalenceofdiagnosedCSbetweendifferenthealthcaredistricts(Figure10).Althoughgeographicaldivergencecannotbeexcluded,differencesindiagnosticpracticesarethemostlikelycause.ThediagnosticbiopsypolicyinpatientswithsuspectedinflammatoryorothercardiomyopathywasparticularlyactiveinHUH,whichisanationalcardiactransplantationcenter,comparedtootherMIDFINnetworkhospitals. Of the 110 patients, 50% were from the HUH area, whereas the referralpopulation of HUH represents roughly one-quarter of the total Finnish population.OutsideHUH, patientsweremore often diagnosed clinicallywith CS and also treatedwithout histopathological proof. Yet another limitation of the present study is that itwas hospital-based and therefore sudden death in the community as a firstmanifestation of CS remained outside its scope. Finally, as a result of the active

85

diagnosticstrategy,theproportionofautopsyandpost-transplantationdiagnoses(7%)was significantly lower than earlier studies (21-29%) (5,97) which results indifferencesindiseasecourse.

ThestrictinclusioncriteriarequiringhistologicverificationofCSentailsprosandcons.The inclusion policy applied results in diagnostic specificity as the patients includedmost definitely have CS. In contrast, a selection bias towards a more severe diseasecoursemightpursue, as thepatientswith amore severedisease aremore rigorouslyexaminedandsubsequentlyincludedinthestudy.

FourfactssuggestthatthetruefrequencyofCSisprobablyhigherthanreportedinthisstudy.First,wedidnot include clinicallydiagnosedCSpatients, thenumberofwhichmight be double the number of histologically verified cases (272). The sensitivity ofmultiple EMBs in the HUH cohort was 31.5% (II) which is low but representative(2,21,86,91). Thus a considerable number of patients with suspect CS with negativeEMBsortotallywithouthistologicbiopsiesduringthestudyperiodwereleftoutofthescopeofthisstudy.Second,previousdatashowthatsymptomaticCSisjustthetipoftheiceberg(16,17,90,91,96,109)andthat20-30%ofpatientswithextracardiacsarcoidosishaveasymptomaticcardiacinvolvement(17,91,96).Duringthestudyperiod,systematiccardiacscreeningofpatientswithextracardiacsarcoidosiswasnotpracticedinFinland.Therefore, intheepidemiologicsenseadeficiencyofthisstudyisthat itonly involvesthe patients with symptomatic CS. Third, as indicated above, patients with outside-hospitalsuddencardiacdeathasfirstamanifestationofCScouldnotbeincluded.Intheearlier autopsy series up to 35% of patients suffered cardiac death as a firstmanifestation of CS (16,17). However, with today’s modern health care system thefigureislikelytobelower.Fourth,astheresponserateofthecentralhospitalswasonly35%,someCScasesmayhavebeenmissed.Anyhow,itislikelythatthehospitalsthatfailed to reply did not treat known CS cases and diagnosed cases from thecorrespondingareaswerediscoveredfromtheuniversityhospitalrecords.

InrespecttoGCM,thedatacollectionmayhavebeenincompleteasnospecificICD-10codeexitsforGCM.Additionally,thefactthat50%(16outof32)ofGCMpatientswerereferredtoHUHfromotherhospitalsand50%werefromtheHUHarea,whichcoversapproximately one-quarter of the Finnish population, also suggests that the trueincidence of GCM in Finland is higher than presented in this study. Compared toprevious studies, a significantly higher portion of GCM patients were diagnosed atbiopsybeforedeathortransplantation(81%vs0-56%)(5,37).Thehigherpercentageoflifetime diagnoseswas probably a consequence of the active EMBpolicy atHUH andfurther reflects on differences in the clinical picture and prognosis compared topreviousworks.

6.1.2DiagnosticsDiagnosing CS is challenging and can be pursued in three ways. Firstly, the goldenstandard and only absolute proof of CS is histologic confirmation of granulomatous

86

inflammationfromtheheartwithothercausesofgranulomasexcluded.However,EMBmoreoftenmissesthanhitsthepatchygranulomatousinfiltrateswithasensitivitythatis no better than15-30% in previous studies (2,21,86,91) and31.5% in our study.AsecondapproachtoCSdiagnosisisbyconfirminghistologyfromanextracardiacorgan,e.g., lungs or lymph nodes togetherwith symptoms and imaging findings compatiblewithCS.Third,acombinationoftypicalimagingfindingsandsymptomstogetherwithclinically diagnosed extracardiac sarcoidois but without histologic confirmation hasbeen considered towarrant diagnosis of CS (178). Anyhow, in this study a histologicverification either from the heart or an extracardiac organwas required,which is inaccordancewith the recently released CS guidelines (4). There are strong argumentspointing towards the fact that the confirmation of CS should be undisputed; thediagnosis triggers long-term immunosuppression with its potential side effects,commonly leads to ICD implantation, and portends a potentially life-threateningdiagnosis in often otherwise healthy young individuals. Despite this, the initiation oftreatmentafterrepeatednegativebiopsiesbutstrongclinicalgroundscouldbejustifiedonthebasisthatnottreatingmightleadtopermanentmyocardialdamageandpotentialfatality.

In termsofdiagnosingCS fromclinicalperspective, thereare threemainscenariosofpresentation. The first scenario comprises of patients presenting with cardiacconditions(mainly2ndto3rddegreeAV-block,ventriculararrhythmiasorunexplainedheart failure usually togetherwith conduction disturbances or VTs)without signs orsymptoms of sarcoidosis from other organs. This used to be considered a rarepresentation, but cumulating evidence from this study and others (108) shows thatsarcoidosisofteninvolvestheheartwithoutclinicallyapparentextracardiacdisease.Inthis clinical setting,after rulingout coronaryarterydisease,agreatdealof caution isneeded to pursue the CS diagnosis, often with multiple EMBs and sometimes withtargetingFDG-positivemediastinal lymphnodes forbiopsy confirmation. It isnotablethatmarkersofgranulomatousinflammation,includingangiotensin-convertingenzyme,lysozyme,anddailyurinarycalciumexcretionwerenotofsignificanthelpindetectingCS.Asecondclinicalscenarioencompassespatientswithbiopsy-confirmedextracardiacsarcoidosiswith symptoms of possibly cardiac origin. Given the positive histology inthesecases,typicalabnormalities inechocardiography,LGE-CMR,orcardiac18F-FDG-PET confirm the CS diagnosis with sufficient certainty. A third scenario involvespatients with biopsy-confirmed extracardiac sarcoidosis without cardiac symptoms.Thesepatientsshouldundergomedicalhistorychecks,aphysicalexamination,andanECG. Echocardiographymay be considered but lacks sensitivity and is recommendedmainlyifoneoftheaforementionedscreeningtestsispositive(88,197).Measuringhs-cTnT/IandNT-proBNPmightbeusefulinguidingdiagnosticdecisions.

The diagnostic path in this study was different from the earlier reports in that theproportion of patients diagnosed at autopsy or from the transplantedheart (8 out of110)wassmaller(5,97).Asmentionedpreviously, theexplanationofthedifferenceis

87

probablytheactivescreeningofyoungpatientswithAV-blockandactiveEMBpolicyinunexplained cardiomyopathy with conduction abnormalities or arrhythmias at ourinstitution. Furthermore, thiswas, to our knowledge, the first study to systematicallyutilize mediastinal lymph node biopsy in patients with negative EMB(s) and a highsuspicionofCS.

CS biopsy findings were collected from hospital registries. Because histologicconfirmationistheonlyabsolutemeansofdifferentiatingbetweenGCMandCS(5),theGCM tissue samples were re-examined by cardiac pathologists for purposes of thisstudytoascertainaccurateGCMdiagnosis.Infourcasestheinitialhistologicdiagnosiswas converted from GCM to CS, which demonstrates the pitfalls in pathologicaldiagnostics. On the other hand, the original histologic CS samples all containedgranulomasandthustheywereconsideredspecificforCS.

6.1.3DataThe data including patient characteristics, imaging and laboratory results, outcomeevents, etc., was principally collected from hospital charts and registries with a fewexceptions.First,theECGswerereanalyzedbytheinvestigatorwhenavailable.Second,the results of hs-cTnT/I measurements from other hospitals outside HUH werecollected using questionnaires. The retrospective study design inherently causesdeficiencies, such as lack of systematic diagnostic work-up and incompleteness offollow-updata.Anyhow, thepotential deficiencies in clinical datawereminimizedbyusing complementary data collection systems (hospital discharge registries, the HUHpathologyregistry,theHUHpacemakerregistryandthe,Centre).Notably,wedouble-checkedthemostimportantoutcomedata(i.e.mortality)fromthePopulationRegisterCentre. Furthermore, a single investigator (RK) scrutinized the patient data, thusenablinguniformity indatahandling.Reporting theoutcome,we focusedon thehardend-points(death, transplantation,VF,VT,AV-block).Thus,weeludedthedifficult-to-defineend-pointssuchasheartfailure,whichareproblematic,especiallyinthecaseofretrospectivestudydesign.

Alltheechocardiographicdatainthisstudywasmeasuredasapartofclinicalwork-upby various cardiology physicians, and a standardized echocardiography protocolwasnot applied. LVEFwasdefinedby Simpson’smethod,M-mode and visual assessment.Echocardiographic analysis is known to be prone to inter-observer differences andincompletereproducibility.

As to the diagnosis of isolated CS, screening for extracardiac sarcoidosis, such asophthalmologicalanddermatologicalconsultations,werenotroutinelyconducted.Mostimportantly,chest(high-resolution)computedtomographywasrarelyperformed.Asaresult, although these patients have isolated CS from the cardiologist’s perspective,someofthemprobablyhaveclinicallysilentextracardiacinvolvement.

Althoughthenatureofthisstudywaspurelyobservational,theinformationgainedover

88

the years was acknowledged by the treating physicians and has thus influenced thepatient care. This particularly applies to the hs-cTnT/I study, where modulations ofimmunosuppressivetreatmentwereconductedbasedonhs-cTnT/Ialterations.

6.1.4AnalysesTheseriesof110CSpatientsisamongthelargestworldwidetodate,andtheseriesof32 GCM patients is the largest involving pre-mortem diagnoses. Anyhow, from astatistical perspective the patient numbers and the numbers of outcome events aresmall, limiting the statistical power. All the studies were uncontrolled, thus makingconclusions about treatment effect impossible. Furthermore, in respect to thesensitivity of diagnostic biopsies and imaging studies, the diagnoses of patientswithnegativeresultsandthusthenumberoffalse-negativestudiesremainedunknown.Theoutcomeofthesepatientsshouldbeevaluatedinfurtherinvestigations.

6.2Comparisonwithpreviousstudies6.2.1EpidemiologyofCSandGCMOur study presents the first, yet limited population-based nationwide data on theprevalenceandincidenceoflifetimesymptomaticCS.Untilnow,epidemiologicdataonthe prevalence of CS has been based on studies of lung sarcoidosis with cardiacinvolvement(31,77,79),andautopsystudies(16,17,49,109).Wefoundthatdespitethemarkedincreaseinthedetectionrateoverthe25-yearperiod,CSstillremainedaveryrare condition in Finland.The annual detection rate of symptomatic CS confirmedbybiopsy was 0.6 per 100 000 adults and the prevalence 2.2 per 100 000 in 2012.Nevertheless, as discussed in the Methodological considerations section, the trueprevalenceofCSincludingclinicallymilddiseaseislikelytobesignificantlyhigher.

Themoststrikingresulttoemergefromthisstudywastheover50-foldincreaseinthedetectionrateofCSover26yearsinFinland.ThisincreaseislikelytobeexplainedbygrowingawarenessofCSandadvancesindiagnosticmethods.Thefactthatthelargestincrease in detection ratewas dated over the yearswith the implementation of LGE-CMR and 18F-FDG-PET in clinical practice supports this theory. Despite this, a trueincrease in the incidenceofCS,possiblydue toanenvironmentalor infectious factor,cannot be ruled out. Previous epidemiologic data on lung sarcoidosis showed similarincidencenumbersinFinnscomparedtootherwhitepopulations(74-76).Henceitcanbeassumedthatourdataisrepresentableinotherwhitepopulations,yetcautionmustbeappliedingeneralizingthesefindingstootherethnicgroups.

GCM is an even rarer myocardial disease, demonstrated by the fact that largestmulticenter collaboration, collectingdata fromalmost40 centersworldwidebetween1982 and1997, could comeupwith nomore than73GCMpatients (5). The presentGCMcohortofstudyVconsistedof32patients,whichisthusfarthelargestpublishedsingle-center study (and our unpublished results add 17 new patients until end of2014).Of the32GCMdiagnoses, 29weremadeover the latter10 years of the study

89

period, with an annual detection rate of approximately 0.1 per 100 000. No priorepidemiologic lifetime data exists for comparison. Earlier autopsy studies reportedincidences of 6.6-23.4 per 100 000 autopsies (25,102), but these figures might beoverestimates since autopsies are usually performed in patients who have sufferedunexpecteddeath.Ontheotherhand,itispossiblethataclinicallymorebenignformofGCMexistsand itsprevalence ishigherthanwhat isdetectedbyEMBsandautopsies.Theheterogeneityofdiseasemanifestationsobserved inthisstudy,as inthepreviousworks(8,37),supportsthistheory.

6.2.2Patientcharacteristicsanddiseasemanifestations In general, our findings considering patient characteristics in CS are consistent withearlierobservations.Wefoundafemalepredominanceof65%andameanageof51±9years, in line with previous studies (5,113). Moreover, this study confirms theassociationofCSandGCMwithautoimmunediseases.Inlinewithpreviousstudies,wefound associated autoimmune disorders in 16% of CS patients and in 19% of GCMpatients,whichishigherthanthe5-10%inthegeneralpopulation(273).

6.2.2.1AV-blockOneofthemostsignificantobservationsfromthisstudyisthehighprevalence(25%)ofCSandGCMascausesofAV-blocksinadultsaged<55years.Thus,CSandGCMcauseafar greater proportion ofAV-blocks in young andmiddle-aged adults thanpreviouslythought.ThisresultwaslaterendorsedbyaCanadiangroupthatfoundCSin34%(11out of 32) patients with initially unexplained AV-block in patients aged 18-60 years(274).Furthermore,likeus,Neryetal.foundthatthepatientswithAV-blockduetoCShad more severe cardiac events on follow-up compared to those in whom AV-blockremained idiopathic (274).Consequently, it is important for theclinicians toconsiderthepossibilityofCSinallpatientsunder55-60yearswithunexplainedAV-block,evenwithoutarrhythmiasorLVdysfunctionatpresentation.Thediagnosticwork-upofthesepatientsshouldincludeLGE-CMRand/or18F-FDG-PET.

6.2.2.2IsolatedCSEven thoughCSwithoutapparentextracardiac involvementhasbeenreportedbefore(5,16,22,87,109),ourstudywasthefirsttodiscussisolatedCSasaclinicalentity.Inournationwide cohort, two-thirds of patients had no apparent extracardiac involvement,which isconsistentwithapreviousmulticenterstudy(5).However,silentsarcoidosisoutsidetheheartwasnotuncommonlydetectedin18F-FDG-PET.Still,aboutone-thirdofpatientsundergoingwhole-body18F-FDG-PETscansdidnothaveanyextracardiacFDGaccumulation. Sinceourstudywaspublished,isolatedCShasattractedattentionandvariablefrequenciesof5.5%-74%inCShavebeenpresented(99,108,137,164,274).The highly variable percentages are likely due to different screening methods with

90

manycenters routinelyusingwhole-body18F-FDG-PET, chest-computed tomography,ophthalmologicexaminationandclose inspectionof theskin inclinicalwork-upofCSand also actively screening patients with extracardiac sarcoidosis (164,274). In thelargest multicenter CS study of 235 patients, chest-computed tomography wasfrequently performed and no less than 83.8% of patients had pulmonary sarcoidosis(129).Itispossiblethatonamicroscopicallevel,allsarcoidosispatientshavelesionsinextracardiacorgans,but fromacardiologist’sperspective, it is important torecognizethe clinical entity of CS, as it makes diagnosis muchmore challenging. There are nosensitive biomarkers thus far to help in the detection of cardiac involvement insarcoidosis,andfurtherinvestigationisneededtoassessthevalueofhs-cTnT/Iinthisrespect.

Inourcohort,moresevereLVinvolvementandthefemalegenderwereassociatedwithisolated CS. Recently, Tezuka et al. also found a correlation between lower LVEF andisolated CS (164). The female predominancemight be explained by different diseasemanifestationsbetweengendersinmyocarditisandcardiomyopathy(275).Wefoundaworse event-free survival in isolated CS patients compared to systemic CS patientswhenabortedsuddendeathwasincludedasanoutcomeevent.Thismightbebecauseof higher frequency of ventricular arrhythmias as the presenting manifestation inisolated CS. Although the reason for the VT proneness is unknown, similarly to ourfindingarecentlargestudyofCSpatientswithICDdemonstrated69%ofpatientswithisolated CS receiving appropriate ICD therapies compared to 34% of patients withsystemicCS(99).Finally, theoverallclinicalconsequencesof isolatedCSremaintobeelucidatedinlargerstudies.

6.2.2.3GCMandcomparisonwithCSInGCM,themeanageof53±13yearsinourcohortwasslightlyhigherthaninpreviousstudies,with amean age of 42-48 years (5,9,11,63). Thismight have implications onprognosis, since young age is possibly linked to worse outcome in myocarditis(11,64,105).TheGCMpatients inour studywere female in69%,whereas the largestpatient series showed no gender difference (5,11). It has been speculated that sexdifferencesmightplayaroleinmyocarditis,withtestosteronepromotingmoreseverediseaseinmen,butthistheoryhasnotbeenconfirmed(275).InGCMpatients,LVEFatdiagnosis was reduced in 74%with a median of 38 ±13%, which is consistent withearlier reports (9,11,63). Moreover, we discovered some support for the geneticpredisposition speculated in earlier studies (11,26,60), as twopatientswithas rareadisorderasGCMweresiblings.

GCMisassociatedwithfulminantonsetandfrequentlyfataloutcome(5,11),althoughalatentandprolongedclinicalcoursehasbeendescribed(5,8,37).Ourstudyconfirmstheheterogenic nature of the disease. The presenting manifestations, including heartfailure, ventricular arrhythmias, sudden death and conduction disturbances, were

91

comparable with previous studies, but their frequency was somewhat different. WefoundAV-blockas the firstmanifestation in31%ofcases,which ishigherthanthe5-15%reportedinpriormulticenterstudies(5,11).Thisdifferencemightbeexplainedbydifferencesindiagnosticprotocolsbetweentheearlierstudiesandours,giventhatweappliedanexceptionallyproactivework-upinpatientswithAV-block.Thus,incontrastto earlier data (5), our finding implies that presentation with AV-block does notdifferentiateGCMfromCS,althoughAV-blockismorecommoninCS.

Previous data indicates that GCM and CS have many resemblances in etiology,pathophysiology and clinical picture, although they are distinctive clinicopathologicentities (5,7,26,27). Our findings on clinical manifestations and outcome support thepreviousconceptthatdespiteoverlappingfeatures,GCMandCSareseparatediseases.Morespecifically,weobservedashortermediantimefromfirstcardiacmanifestationtodiagnosis of threemonths (range 0-16) in GCM, compared to 9.5months (range 0.3-168) in CS, pointing towards a more rapid clinical course in GCM. At diagnosis, thepatientswithGCMhadonlyslightlylowerLVEFof38±13%,incomparisonwith45%±14 inCS,butduring thediseasecourse,GCMpatientsmoreoftendevelopedsystolicheartfailure(90%vs67%).Consistently,severeheartfailureleadingtotransplantationwas more common in GCM (31% vs 10%), and the arrhythmia burden was higher.Finally, the marked differences in survival rates discussed below are in line withpreviousworks.

6.2.3DiseasecourseandtheeffectoftreatmentinCS6.2.3.1LVfunctionInlinewithotherlargestudies(77,97,99,129),themedianLVEFatthetimeofdiagnosiswasmildlyreduced(45%±14%)andpatientserieswithbothlower(115)andhigher(6) initial EF exist. During the disease course, two-thirds of our patients had systolicdysfunction(definedasEF<50%)butdataforcomparisonarelacking.Priortoclinicalawarenessandmoderndiagnosticmethods,CScommonlymanifestedassymptomaticheart failure (19), but in our study as in other recent studies (93,146), overt heartfailurewasmorerare.Similarlytoothertypesofcardiomyopathies,oursandotherCSstudies (6,97,98,129) show that LVEF is an important prognostic factor in CS. In ourstudy,theoutcomewasworst inpatientswithseveresystolicdysfunction(EF<35%),butthedifferencewassmallbetweenthepatientswithmildlyreduced(EF35-50%)andnormal(>50%)EF.Furthermore,RVdysfunctionhasrecentlybeenreportedtobenotonly common but also predisposing to ventricular arrhythmias in CS (98). However,since evaluating RV function by echocardiography is challenging and not routinelyquantified,wedidnotanalyzeit.

6.2.3.2AV-blockOverall,thehighfrequencyofAV-conductiondisturbancesof52%observedduringthediseasecourseofCSwascomparablewithpreviousreports(5,16,77,97).Insmallseries

92

ofupto12patients,thenumberofpatientswhoseAV-conductionrecoveredintimeorwithcorticosteroidswasreportedtobe33-75%(124,246,248) incomparisonto20%inourstudy.Besidessmallsamplesizes, thesedifferencesmightresult fromdifferentdefinitions of AV-recovery, as in our study it was roughly estimated by <10% ofventricularpacingandinthestudiesfocusingonassessingconduction,morespecificallyexaminedbyHoltermonitoringsandECGs.

6.2.3.3.ArrhythmiasCS presented as ventricular arrhythmias in 33% of patients, which is in line withprevious and preceding studies (5,93,97,128,129). ICD was implanted in 54% ofpatientsandtheindicationwassecondarypreventionintwo-thirdsofthem,whereasinthe recent studies from the U.S. over two-thirds of ICDswere implanted for primaryprevention (98,128,129). After the study period, with accumulating information, theimplantation protocol in Finland has changed towards a more active primarypreventionstrategy.Atthemedianfollow-upof4.4yearsfromICDimplantation,41%ofthe 59 Finnish CS patients with ICD received an appropriate ICD therapy, whichcorrespondstothetherapyrateinrecentstudies(98,129).Furthermore,theincreaseinnon-sustained VT frequency on follow-up should be regarded as a sign of potentialdiseaseactivation.

6.2.3.4AssessmentofimmunosuppressivetreatmentresponseAllthepublishedclinicalseriessuggestthatcorticosteroidtherapymodifiestheclinicalcourse in CS (5,6,97,115,124,246,248,249). If treatment is initiated early on, systolicheart failureseemstodeveloprarely,whereasthosewithreducedLVfunctionhaveapoor prognosis despite initiation of treatment (6,97). The corticosteroid treatmentprotocolinourstudywasnotuniformbutsimilartoolderstudies(6,97,244).Ourstudyconfirms that reduced LVEF is associated with worse prognosis despiteimmunosuppressivetreatment.However, incontrasttoearlierfindingsdemonstratingbenefitfromcorticosteroidsinmildtomoderatesystolicheartfailure(6,115)ourdatademonstrated a small yet significant improvement in LVEF in patients with severelyreduced LVEF (<35%), but no change in normal ormoderately depressed LVEF. Thedeficiency in our study as in others is that we do not know how LVEF would haveresultedintheabsenceofcorticosteroidtreatment.Furthermore,theechocardiographicevaluation of LVEF is prone to divergence between operators, notably when LV isdilated. Similarly to earlier studies, no conclusions on survival benefit fromimmunosuppressantscouldbedrawn,sinceallthepatientsdiagnosedbeforedeathortransplantation received immunomodulatory treatment and respectively the oneswithouttreatmentwerediagnosedatdeathortransplantation.

93

Concerning steroid-sparingagents,methotrexate is themostwidely studiedandusedantimetaboliteinsarcoidosistreatment(29,276).Incontrast,inourcohortazathioprinewassignificantlymorecommonlyused(50vssixpatients).Therearedifferencesintheside-effect profile between the two drugs. Azathioprine is associated with moreinfections, leadingtodiscontinuationofthedrug, leucopenia,nausea,andpancreatitis,whereasmethotrexateinvolvesmorefrequenthepatotoxiceffects(276).Animportantaspect in choosingdrug treatment is clinicians’userexperience.Forexample, atHUHazathioprinehasbeenusedforalongtimeintreatingcardiactransplantpatients.

6.2.4Seriousoutcomeeventsandsurvival6.2.4.1InCSPriorknowledgeonCSprognosisisbasedonretrospectiveseriesandsmallprospectivestudies (5,16,22,97,109) mostly involving patients with extracardiac sarcoidosisscreened for cardiac involvement (91,93,96). In the largest retrospectivemulticenterstudybyYazakietal.fromJapan,theoverallsurvivalrateinCSwas60%atfiveyearsinall 95 patients and 75% in patients diagnosed before death (97). Similarly, theinternationalcollaborationstudyfromOkuraetal.reportedtransplant-freesurvivalof60%at fiveyears inall42patientsand70%inpatientsdiagnosedatEMB(5). Inourcohort, the five-year transplant-free cardiac survivalwas90% inall110patientsand95%in102patientsdiagnosedpriortodeathortransplantation.Althoughthesefigurescannot be directly compared since our numbers denote cardiac survival and thepreviousstudiesincludednon-cardiacdeaths,theoutcomeseemsmorefavorableinthepresentstudy.Thepossiblereasonsforbettersurvivalcouldbeearlierdiagnosiswithshorterdelayfromsymptomonsettodiagnosisinourstudy(median9.5months,range0.3-168vsmean29.7±53.3inthestudyofOkuraetal.)(5).Alternatively,comparedtothe earlier studies, the immunosuppressive treatment has not changedmarkedly buttherehasbeendevelopmentinrespecttoarrhythmiatreatment.InourstudyICDwasimplantedin54%vs24%inthestudybyOkuraetal.Thehighfrequencyofventriculararrhythmias andparticularly thehighnumberof aborted suddendeaths inour studysupports this theory. Lastly, differences in genetic backgrounds between thepopulationsmighthaveanimpactontheoutcome(49,75,277).

6.2.4.2InGCMConsideringtheoutcomeofGCMinstead,ourresultssuggestimprovedprognosiswithcombined immunosuppressive treatment.Previousdataon thenaturalcourseofGCMdemonstratedagrimprognosis,withamediantransplant-freesurvivalofthreemonthsfrom symptom onset but a significant survival benefit with a combination ofimmunosuppressiontherapyincludingcyclosporine(11).Comparably,aninternationalmulticenterstudydemonstratedafive-yeartransplant-freesurvivalprobabilityof10%in all 73 patients and 22% transplant-free survival in 38 patients diagnosed before

94

deathortransplantationandtreatedwithat leastcorticosteroids(5).Inourstudythecorresponding five-year survivalprobabilitiesof transplant-free survivalwerehigher;52%inall32patientsand63%in26patientsreceiving immunosuppressive therapy,respectively. Besides active and potent drug treatment with multipleimmunosuppressants,thepromptnessindiagnosticsandpossiblydemographicfactorsmightexplaintheamelioratedsurvivalratesinourstudy.Furthermore,similarlytoCS,theincreasedfrequencyofICDsmighthaveimprovedthesurvivalofGCMpatients.

6.3ClinicalimplicationsandfuturedirectionsTheresultsofthisworkcouldhaveanimpactonthediagnosticsandmanagementofCSand GCMpatients in clinical practice.Most importantly, the finding that CS and GCMtogethercause25%ofinitiallyunexplainedAV-blocksinyoungtomiddle-agedpatients(II) has been acknowledged by the authors of the AHA/ACCF/ESC EMB guidelines(234,278). Based on our results, they stated that EMB should probably be a class 2Arecommendation (a reasonableprocedure) inAV-blockpatientswitha similar risk toCS/GCM (depending on age, ethnicity and exclusion of more common causes) (278).Second, in case of high suspicion of CS, imaging-guided biopsies of the heart ormediastinallymphnodes,repeatedifnecessary,canbeusefulastheymarkedlyincreasethe diagnostic yield (I,II). Sampling “hot” 18F-FDG-PET positive lymph nodes inmediastinoscopyhasbecomeclinicalpracticeatHUH,sinceourfindingsdemonstratedthe high (close to 100%) sensitivity of this approach (279). Third, hs-cTnT/I haveemergedassensitiveandeasilyrepeatablemarkersthatareroutinelyusedatHUHasadditionaltoolsinassessingdiseaseactivityandtreatmentmodificationsinthefollow-up of CS. Fourth, referring to our study (V), the 2015 ESC guidelines for themanagement of ventricular arrhythmias stated that the presence of malignantventricular arrhythmias or heart block in GCM or CS might warrant earlierconsiderationofanICDduetotheknownhighriskofarrhythmicdeathortheneedfortransplantation(280).

An important aspect of this study lies in raising awareness and diagnostic activitytowards CS and GCM. At best, increased knowledge enables earlier diagnosis andtreatmentprior toLVdysfunction,which in turn is likely to result inbetteroutcome.Althoughrarediseases,CSandGCMshouldbekeptinmindincaseswithunexplainedDCM, ventricular arrhythmias and AV-block, particularly in young to middle-agedadults.

Further research on CS and GCM is needed to elucidate the underlying mechanismsbehindtheseenigmaticdiseasesandultimatelyto improvetheoutcome. Inenhancingdiagnosticaccuracy,newimagingmethods,suchasT1mappingCMR(281)andnovelPET isotopes (282) offer interesting targets. Moreover, more sensitive EMBs, usingimmunohistochemistryandmolecularbiologymethods(genesequencing,RNAprofilingetc.) are highly expected. Considering the treatment, assessing optimal drug regimen

95

and combinations should be studied. Possibly a similar treatment approach as inrheumatoid arthritis, involving rapid reductions in corticosteroid doses and a lowthreshold for monoclonal antibodies in treatment resistant cases, could be applied.However, the best regimen should be tested in randomized controlled trials. LargeprospectivemulticenterregistriesareneededfortheimplicationofthestudiesinbothCSandGCM.

96

7CONCLUSIONSThepresent studywas initiatedby the clinical observation that thenumberof youngpeoplecontractingCS,adiseasewithindeterminateetiology,diagnosticsandoutcome,wasincreasingatourinstitution.Wefoundthat:

1.ThedetectionrateofCS increasedover50-foldoverthe26-yearstudyperiodfrom1988to2014.Theannualdetectionrateofbiopsy-confirmedCSwas0.6per100000adults (>18 years) in the last two-year study period between 2012 and 2014. TheprevalenceofCSin2012was2.2per100000.

2.CSandGCMtogetherexplain25%of2ndto3rddegreeAV-blocksinadultsaged18-55yearsinwhomtheetiologyofAV-blockisunknownatpresentation.

3. Repeated, imaging-guided EMBs improve the diagnostic accuracy in CS. Thesensitivity of first EMB was 31%, but second and third EMB sessions increased thesensitivityto55%.

4. Sarcoidosis can manifest as a clinically isolated cardiac disorder without signs orsymptoms from extracardiac organs. Two-thirds of CS patients had solely cardiacmanifestations.ClinicallyisolatedCSwascharacterizedbymoresevereLVdysfunctionandfemalepredominance.

5. The outcome of CS with current diagnostics and treatment appears better thanpreviouslyreported. InCSpatientsdiagnosedbeforedeathortransplantation, the10-yearcardiactransplant-freesurvivalwas91%.

6.Hs-cTnT/Iare frequentlyelevated innew-onsetCS.Theconcentrationsrapidlyandconsistentlydecreasewithcorticosteroidtreatment.

7. The prognosis of GCM with current diagnostics and treatment appears bettercomparedtoolderstudies.InGCMpatientsdiagnosedbeforedeathortransplantation,thefive-yearcardiactransplantfreesurvivalwas63%.

97

8ACKNOWLEDGEMENTSThisstudywascarriedoutatHelsinkiUniversityHospital`sHeartandLungCenterandattheUniversityofHelsinkiin2008-2015.

Thisscientific journeyhastaughtmeagreatdeal; firstlyaboutscientificmethodsandwaysof thinking,andsecondlyaboutmyselfandthewaysof theworld.Here Iwouldlike to express my acknowledgements to all the people with whom I have had thepleasuretoconnectwithonthisjourney.

IowemydeepestgratitudetoProfessorMarkkuKupari,whoseideasandvisionshavebeentheinspirationforthesestudies.Hisuniqueunderstandingofcardiologyandhisprofound skills in statistics and scientific writing have been the cornerstones of thisstudy.

I am indebted to Docent Jukka Lehtonen for his mentoring. His intelligence andmultitasking skills never cease to amazeme.His unique sense of humorhas been anirreplaceablesourceofjoyduringthisproject.

A growing group of researchers related to inflammatory heart diseases research hasformed during the past few years. I wish to thank Piia Simonen for her energeticcompanionshipandencourageDianaVelikanova,KajEkström,KaisaNordenswanandJuliaRaaktocontinuethegoodworktogether.Hopefullythisthesisisjustthebeginningof a long and prosperous project. I would also like to thank all the co-authors andassociate investigators in the hospitals across Finland who make up the MIDFINresearchnetwork.Iowegratitudetoallthepatientswhoparticipatedinthisstudy.

IamsincerelygratefultoDocentTuomasRissanenandDocentVilleKytöfortheirfastyet precise comments which helped me finalize this thesis. I deeply admire yourscientificenthusiasm.IalsowishtoextendmygratitudetoNickMoonforrevisingthelanguage.

I also wish to express my gratitude to my clinical colleagues in the cardiologydepartment for their friendship, teaching and collaboration during my cardiologyspecialization.

I want to thank the cordial Dean Kimmo Kontula for letting me to use his researchfacilitiesatBiomedicum.Furhermore,IwishtothankthewholeofKontula’smolecularmedicine laboratory group, especially Annukka Lahtinen and Kati Donner for theirsupportandlettingmefeellikeIwasaparttheirstudygroup.

IwouldliketothankDocentMaritGranerandDocentTimoHiltunenforpracticalhelpinstatisticsandscientificwritingintheearlyimportantstepsofmyresearchcareer.

98

Aspeersupportispricelessinlife,IsendmywarmestthoughtstoLauraMäkinen,SanjaSädevirta, Virpi Koskela-Niska, Milla Rosengård-Bärlund and Sanna Kaye, who areworkingonsimilarprojects.

Heartfeltthanksgotomydearfriends,especiallyMarika,Hanna,Taina,andMaaritforsharing the joys and sorrows of life. I extend my deepest gratitude to my sisterAnnamari;inknowingthatsheisaround,Ineverfeelaloneinthisworld.

Iwould not be herewithout the strong foundations of independency, hardwork andperseveranceprovidedbymybelovedparentsLeenaandTimo.StilltodayIcanalwayscount on their support, both mentally and in everyday matters which has recentlyinvolvedalotoftakingcareofAdrian.IalsowanttothankSusanneandIsmo,KristiinaandChristos,HeikkiandEilaforbeingthere.

Most importantly, my husband and best friend Johannes. Words cannot express thelove, gratitude, and appreciation I feel towards you. You have supportedme throughgoodtimesandbad.Letthecardiacimageyoudrewformeonthepage25ofthisthesissymbolize our love, fragile yet strong as it is. My son Adrian, you were born in themiddle of this study project and took a lot of attention away from it forwhich I amforeverthankful.Iloveyou.

This study was supported financially by the Finnish Foundation for CardiovascularResearch, Helsinki University EVO grants, the Finnish Cultural Foundation, and theFinnish-Norwegian Medical Foundation, and the congress presentations weresupportedbytheFinnishCardiacSociety.

Helsinki,November2015

RiinaKandolin

99

REFERENCES

1. AyoubC,PenaE,OhiraH,DickA,LeungE,NeryPB,etal.AdvancedImagingofCardiacSarcoidosis.CurrCardiolRep.2015;17:17–12.

2. UemuraA,MorimotoS,HiramitsuS,KatoY,ItoT,HishidaH.Histologicdiagnosticrateofcardiacsarcoidosis:evaluationofendomyocardialbiopsies.AmHeartJ.1999;138:299–302.

3. LaganaSM,ParwaniAV,NicholsLC.Cardiacsarcoidosis:apathology-focusedreview.ArchPatholLabMed.2010;134:1039–46.

4. BirnieDH,SauerWH,BogunF,CooperJM,CulverDA,DuvernoyCS,etal.HRSexpertconsensusstatementonthediagnosisandmanagementofarrhythmiasassociatedwithcardiacsarcoidosis.HeartRhythm.2014;11:1305–23.

5. OkuraY,DecGW,HareJM,KodamaM,BerryGJ,TazelaarHD,etal.Aclinicalandhistopathologiccomparisonofcardiacsarcoidosisandidiopathicgiantcellmyocarditis.JAmCollCardiol.2003;41:322–9.

6. ChiuC-Z,NakataniS,ZhangG,TachibanaT,OhmoriF,YamagishiM,etal.Preventionofleftventricularremodelingbylong-termcorticosteroidtherapyinpatientswithcardiacsarcoidosis.AmJCardiol.2005;95:143–6.

7. RosensteinED,ZuckerMJ,KramerN.Giantcellmyocarditis:mostfatalofautoimmunediseases.SeminArthritisRheum.2000;30:1–16.

8. RenH,PostonRS,HrubanRH,BaumgartnerWA,BaughmanKL,HutchinsGM.Longsurvivalwithgiantcellmyocarditis.ModPathol.1993;6:402–7.

9. DavidoffR,PalaciosI,SouthernJ,FallonJT,NewellJ,DecGW.Giantcellversuslymphocyticmyocarditis.Acomparisonoftheirclinicalfeaturesandlong-termoutcomes.Circulation.1991;83:953–61.

10. LitovskySH,BurkeAP,VirmaniR.Giantcellmyocarditis:anentitydistinctfromsarcoidosischaracterizedbymultiphasicmyocytedestructionbycytotoxicTcellsandhistiocyticgiantcells.ModPathol.1996;9:1126–34.

11. CooperLT,BerryGJ,ShabetaiR.Idiopathicgiant-cellmyocarditis-naturalhistoryandtreatment.MulticenterGiantCellMyocarditisStudyGroupInvestigators.NEnglJMed.1997;336:1860–6.

12. SharmaOP.Sarcoidosis:ahistoricalperspective.ClinDermatol.2007;25:232-241.

13. DoughanAR,WilliamsBR.Cardiacsarcoidosis.Heart.2006;92:282–8.

14. SchaumannJ.Sarcoidosiswithprurigo.AnnDermatolSyphiligr.1953;80:457–76.

100

15. LongcopeWT,FreimanDG.Astudyofsarcoidosis;basedonacombinedinvestigationof160casesincluding30autopsiesfromTheJohnsHopkinsHospitalandMassachusettsGeneralHospital.Medicine.1952;31:1–132.

16. MatsuiY,IwaiK,TachibanaT,FruieT,ShigematsuN,IzumiT,etal.Clinicopathologicalstudyoffatalmyocardialsarcoidosis.AnnNYAcadSci.1976;278:455–69.

17. SilvermanKJ,HutchinsGM,BulkleyBH.Cardiacsarcoid:aclinicopathologicstudyof84unselectedpatientswithsystemicsarcoidosis.Circulation.1978;58:1204–11.

18. FlemingHA,McMahonJN,McCarthyCF,KelehanP.Sarcoidheartdisease.BrHeartJ.1983;50:498.

19. FlemingHA.Sarcoidheartdisease.BrMedJ.1986;292:1095–6.

20. FlemingHA,BaileySM.Sarcoidheartdisease.JRCollPhysiciansLond.1981;15:245–53.

21. RatnerSJ,FenoglioJJ,UrsellPC.Utilityofendomyocardialbiopsyinthediagnosisofcardiacsarcoidosis.Chest.1986;90:528–33.

22. FlemingHA,BaileySM.TheprognosisofsarcoidheartdiseaseintheUnitedKingdom.AnnNYAcadSci.1986;465:543–50.

23. CooperLT.Giantcellandgranulomatousmyocarditis.HeartFailClin.2005;1:431–7.

24. SaltykowS.UberdiffuseMyokarditis.VirchowsArchivfurPathologischeAnatomie.1905;182:1–39.

25. WhiteheadR.Isolatedmyocarditis.BrHeartJ.1965;27:220–30.

26. BlauwetLA,CooperLT.Idiopathicgiantcellmyocarditisandcardiacsarcoidosis.HeartFailRev.2013;18:733–46.

27. DaviesMJ,PomeranceA,TeareRD.Idiopathicgiantcellmyocarditis-adistinctiveclinico-pathologicalentity.BrHeartJ.1975;37:192–5.

28. MantiniN,WilliamsB,StewartJ,RubinsztainL,KacharavaA.Cardiacsarcoid:aclinician'sreviewonhowtoapproachthepatientwithcardiacsarcoid.ClinCardiol.2012;35:410–5.

29. ValeyreD,PrasseA,NunesH,UzunhanY,BrilletP-Y,Müller-QuernheimJ.Sarcoidosis.Lancet.2014;383:1155–67.

30. BaughmanRP,TeirsteinAS,JudsonMA,RossmanMD,YeagerH,BresnitzEA,etal.Clinicalcharacteristicsofpatientsinacasecontrolstudyofsarcoidosis.AmJRespirCritCareMed.2001;164:1885–9.

101

31. JohnsCJ,MicheleTM.Theclinicalmanagementofsarcoidosis.A50-yearexperienceattheJohnsHopkinsHospital.Medicine.1999;78:65–111.

32. SaidhaS,SotirchosES,EcksteinC.Etiologyofsarcoidosis:doesinfectionplayarole?YaleJBiolMed.2012;85:133–41.

33. IannuzziMC,RybickiBA,TeirsteinAS.Sarcoidosis.NEnglJMed.2007;357:2153–65.

34. DubreyS,ShahS,HardmanT,SharmaR.Sarcoidosis:thelinksbetweenepidemiologyandaetiology.PostgradMedJ.2014;90:582–9.

35. BargoutR,KellyRF.Sarcoidheartdisease:clinicalcourseandtreatment.IntJCardiol.2004;97:173–82.

36. TchernevG,TanaC,SchiavoneC,CardosoJ-C,AnanievJ,WollinaU.Sarcoidosisvs.Sarcoid-likereactions:TheTwoSidesofthesameCoin?WienMedWochenschr.2014;164:247–59.

37. DaviesRA,VeinotJP,SmithS,StruthersC,HendryP,MastersR.Giantcellmyocarditis:clinicalpresentation,bridgetotransplantationwithmechanicalcirculatorysupport,andlong-termoutcome.JHeartLungTransplant.2002;21:674–9.

38. PapadopoulosKI,HörnbladY,LiljebladhH,HallengrenB.Highfrequencyofendocrineautoimmunityinpatientswithsarcoidosis.EurJEndocrinol.1996;134:331–6.

39. FuseK,KodamaM,ItoM,OkuraY,KatoK,HanawaH,etal.PolarityofhelperTcellsubsetsrepresentsdiseasenatureandclinicalcourseofexperimentalautoimmunemyocarditisinrats.ClinExpImmunol.2003;134:403–8.

40. HunninghakeGW,CrystalRG.Pulmonarysarcoidosis:adisordermediatedbyexcesshelperT-lymphocyteactivityatsitesofdiseaseactivity.NEnglJMed.1981;305:429–34.

41. MollerDR.Cellsandcytokinesinvolvedinthepathogenesisofsarcoidosis.SarcoidosisVascDiffuseLungDis.1999;16:24–31.

42. GrunewaldJ,EklundA.RoleofCD4+Tcellsinsarcoidosis.ProcAmThoracSoc.2007;4:461–4.

43. GruttersJC,vandenBoschJMM.Corticosteroidtreatmentinsarcoidosis.EurRespirJ.2006;28:627–36.

44. ChenES,MollerDR.Sarcoidosis-scientificprogressandclinicalchallenges.NatRevRheumatol.2011;7:457–67.

45. GrunewaldJ,EklundA,HedJ,TakahashiH,WigzellH.AspectsonthealveolaraccumulationofTcellsinsarcoidosis.Sarcoidosis.1992;9:142–4.

102

46. ZhouT,ZhangW,SweissNJ,ChenES,MollerDR,KnoxKS,etal.Peripheralbloodgeneexpressionasanovelgenomicbiomarkerincomplicatedsarcoidosis.PLoSONE.2012;7:1-13.

47. SverrildA,BackerV,KyvikKO,KaprioJ,MilmanN,SvendsenCB,etal.Heredityinsarcoidosis:aregistry-basedtwinstudy.Thorax.2008;63:894–6.

48. RybickiBA,IannuzziMC,FrederickMM,ThompsonBW,RossmanMD,BresnitzEA,etal.Familialaggregationofsarcoidosis.Acase-controletiologicstudyofsarcoidosis(ACCESS).AmJRespirCritCareMed.2001;164:2085–91.

49. IwaiK,SekigutiM,HosodaY,DeRemeeRA,TazelaarHD,SharmaOP,etal.Racialdifferenceincardiacsarcoidosisincidenceobservedatautopsy.Sarcoidosis.1994;11:26–31.

50. SatoH,WoodheadFA,AhmadT,GruttersJC,SpagnoloP,vandenBoschJMM,etal.SarcoidosisHLAclassIIgenotypingdistinguishesdifferencesofclinicalphenotypeacrossethnicgroups.HumMolGenet.2010;19:4100–11.

51. SpagnoloP,GrunewaldJ.Recentadvancesinthegeneticsofsarcoidosis.JMedGenet.2013;50:290–7.

52. IkezoeK,HandaT,TanizawaK,KuboT,ItoI,SokaiA,etal.Atoll-likereceptor3singlenucleotidepolymorphisminJapanesepatientswithsarcoidosis.TissueAntigens.2015;85:204–8.

53. LiuY,LiH,XiaoT,LuQ.Epigeneticsinimmune-mediatedpulmonarydiseases.ClinRevAllergyImmunol.2013;45:314–30.

54. ChristophiGP,CazaT,CurtissC,GumberD,MassaPT,LandasSK.Geneexpressionprofilesingranulomatissuerevealnoveldiagnosticmarkersinsarcoidosis.ExpMolPathol.2014;96:393–9.

55. GuptaD,AgarwalR,AggarwalAN,JindalSK.Molecularevidencefortheroleofmycobacteriainsarcoidosis:ameta-analysis.EurRespirJ.2007;30:508–16.

56. EishiY.EtiologiclinkbetweensarcoidosisandPropionibacteriumacnes.RespirInvestig.2013;51:56–68.

57. NewmanLS,RoseCS,BresnitzEA,RossmanMD,BarnardJ,FrederickM,etal.Acasecontroletiologicstudyofsarcoidosis:environmentalandoccupationalriskfactors.AmJRespirCritCareMed.2004;170:1324–30.

58. NewmanKL,NewmanLS.Occupationalcausesofsarcoidosis.CurrOpinAllergyClinImmunol.2012;12:145–50.

59. CrowleyLE,HerbertR,MolineJM,WallensteinS,ShuklaG,SchechterC,etal.“Sarcoidlike”granulomatouspulmonarydiseaseinWorldTradeCenterdisasterresponders.AmJIndMed.2011;54:175–84.

60. KittlesonMM,MinhasKM,IrizarryRA,YeSQ,EdnessG,BretonE,etal.Gene

103

expressioningiantcellmyocarditis:Alteredexpressionofimmuneresponsegenes.IntJCardiol.2005;102:333–40.

61. KodamaM,MatsumotoY,FujiwaraM,MasaniF,IzumiT,ShibataA.Anovelexperimentalmodelofgiantcellmyocarditisinducedinratsbyimmunizationwithcardiacmyosinfraction.ClinImmunolImmunopathol.1990;57:250–62.

62. ZhangS,KodamaM,HanawaH,IzumiT,ShibataA,MasaniF.Effectsofcyclosporine,prednisoloneandaspirinonratautoimmunegiantcellmyocarditis.JAmCollCardiol.1993;21:1254–60.

63. ElezkurtajS,LassnerD,SchultheissHP,EscherF.Vascularinvolvementincardiacgiantcellmyocarditis:anewpathophysiologicalaspect.ClinResCardiol.2013;103:161–3.

64. CaforioALP,CalabreseF,AngeliniA,TonaF,VinciA,BottaroS,etal.Aprospectivestudyofbiopsy-provenmyocarditis:prognosticrelevanceofclinicalandaetiopathogeneticfeaturesatdiagnosis.EurHeartJ.2007;28:1326–33.

65. AsimakiA,TandriH,DuffyER,WinterfieldJR,Mackey-BojackS,PickenMM,etal.Altereddesmosomalproteinsingranulomatousmyocarditisandpotentialpathogeniclinkstoarrhythmogenicrightventricularcardiomyopathy.CircArrhythmElectrophysiol.2011;4:743–52.

66. SabatineMS,PohK-K,MegaJL,ShepardJ-AO,StoneJR,FroschMP.CaserecordsoftheMassachusettsGeneralHospital.Case36-2007.A31-year-oldwomanwithrash,fever,andhypotension.NEnglJMed.2007;357:2167–78.

67. MitoffPR,MesanaTG,MielniczukLM,GrenonJ,VeinotJP,CooperLT,etal.Giantcellmyocarditisinapatientwithaspondyloarthropathyafteradrughypersensitivityreaction.CanJCardiol.2013;29:7–8.

68. DanielsPR,BerryGJ,TazelaarHD,CooperLT.Giantcellmyocarditisasamanifestationofdrughypersensitivity.CardiovascPathol.2000;9:287–91.

69. BaigM,HatrickR.Giantcellmyocarditiswithincessantventriculararrhythmiastreatedsuccessfullywithmethylprednisoloneandratantithymocyteglobulin.CardiolResPract.2011;2011:1-4.

70. ReichJM.Acriticalanalysisofsarcoidosisincidenceassessment.MultidiscipRespirMed.2013;8:57.

71. ZipseMM,SauerWH.Cardiacsarcoidosis.CurrCardiolRep.2014;16:514–10.

72. GribbinJ,HubbardRB,LeJeuneI,SmithCJP,WestJ,TataLJ.IncidenceandmortalityofidiopathicpulmonaryfibrosisandsarcoidosisintheUK.Thorax.2006;61:980–5.

73. ReichJM,JohnsonRE.IncidenceofclinicallyidentifiedsarcoidosisinanorthwestUnitedStatespopulation.SarcoidosisVascDiffuseLungDis.1996;13:173–7.

104

74. HillerdalG,NöuE,OstermanK,SchmekelB.Sarcoidosis:epidemiologyandprognosis.A15-yearEuropeanstudy.AmRevRespirDis.1984;130:29–32.

75. RybickiBA,MajorM,PopovichJ,MaliarikMJ,IannuzziMC.Racialdifferencesinsarcoidosisincidence:a5-yearstudyinahealthmaintenanceorganization.AmJEpidemiol.1997;145:234–41.

76. PietinalhoA,HiragaY,HosodaY,LöfroosAB,YamaguchiM,SelroosO.ThefrequencyofsarcoidosisinFinlandandHokkaido,Japan.Acomparativeepidemiologicalstudy.Sarcoidosis.1995;12:61–7.

77. SmedemaJ-P,SnoepG,vanKroonenburghMPG,vanGeunsR-J,DassenWRM,GorgelsAP,etal.CardiacinvolvementinpatientswithpulmonarysarcoidosisassessedattwouniversitymedicalcentersintheNetherlands.Chest.2005;128:30–5.

78. SharmaOP,MaheshwariA,ThakerK.Myocardialsarcoidosis.Chest.1993;103:253–8.

79. HunninghakeGW,CostabelU,AndoM,BaughmanR,CordierJF,BoisduR,etal.ATS/ERS/WASOGstatementonsarcoidosis.AmericanThoracicSociety/EuropeanRespiratorySociety/WorldAssociationofSarcoidosisandotherGranulomatousDisorders.SarcoidosisVascDiffuseLungDis.1999;16:149–73.

80. JoyceE,NinaberMK,KatsanosS,DebonnaireP,KamperidisV,BaxJJ,etal.Subclinicalleftventriculardysfunctionbyechocardiographicspeckle-trackingstrainanalysisrelatestooutcomeinsarcoidosis.EurJHeartFail.2015;17:51–62.

81. DarlingtonP,GabrielsenA,SörenssonP,CederlundK,EklundA,GrunewaldJ.CardiacinvolvementinCaucasianpatientswithpulmonarysarcoidosis.RespirRes.2014;15:15.

82. AkashiH,KatoTS,TakayamaH,NakaY,FarrM,ManciniD,etal.Outcomeofpatientswithcardiacsarcoidosisundergoingcardiactransplantation-single-centerretrospectiveanalysis.JCardiol.2012;60:407–10.

83. ZhangM,TavoraF,HuebnerT,HeathJ,BurkeA.Allograftpathologyinpatientstransplantedforidiopathicdilatedcardiomyopathy.AmJSurgPathol.2012;36:389–95.

84. RobertsWC,RobertsCC,KoJM,FilardoG,CapehartJE,HallSA.Morphologicfeaturesoftherecipientheartinpatientshavingcardiactransplantationandanalysisofthecongruenceorincongruencebetweentheclinicalandmorphologicdiagnoses.Medicine.2014;93:211–35.

85. LukA,MetaweeM,AhnE,GustafssonF,RossH,ButanyJ.Doclinicaldiagnosescorrelatewithpathologicaldiagnosesincardiactransplantpatients?Theimportanceofendomyocardialbiopsy.CanJCardiol.2009;25:48–54.

105

86. ArdehaliH,HowardDL,HaririA,QasimA,HareJM,BaughmanKL,etal.Apositiveendomyocardialbiopsyresultforsarcoidisassociatedwithpoorprognosisinpatientswithinitiallyunexplainedcardiomyopathy.AmHeartJ.2005;150:459–63.

87. TavoraF,CresswellN,LiL,RippleM,SolomonC,BurkeA.Comparisonofnecropsyfindingsinpatientswithsarcoidosisdyingsuddenlyfromcardiacsarcoidosisversusdyingsuddenlyfromothercauses.AmJCardiol.2009;104:571–7.

88. HamzehN,SteckmanDA,SauerWH,JudsonMA.Pathophysiologyandclinicalmanagementofcardiacsarcoidosis.NatRevCardiol.2015;24:1–11.

89. AbelerV.Sarcoidosisofthecardiacconductingsystem.AmHeartJ.1979;97:701–7.

90. NagaiT,KohsakaS,OkudaS,AnzaiT,AsanoK,FukudaK.Incidenceandprognosticsignificanceofmyocardiallategadoliniumenhancementinpatientswithsarcoidosiswithoutcardiacmanifestation.Chest.2014;146:1064–72.

91. PatelMR,CawleyPJ,HeitnerJF,KlemI,ParkerMA,JaroudiWA,etal.Detectionofmyocardialdamageinpatientswithsarcoidosis.Circulation.2009;120:1969–77.

92. PatelAR,KleinMR,ChandraS,SpencerKT,DecaraJM,LangRM,etal.Myocardialdamageinpatientswithsarcoidosisandpreservedleftventricularsystolicfunction:anobservationalstudy.EurJHeartFail.2011;13:1231–7.

93. MehtaD,LubitzSA,FrankelZ,WisniveskyJP,EinsteinAJ,GoldmanM,etal.Cardiacinvolvementinpatientswithsarcoidosis:diagnosticandprognosticvalueofoutpatienttesting.Chest.2008;133:1426–35.

94. VignauxO,DhoteR,DubocD,BlancheP,DevauxJ-Y,WeberS,etal.DetectionofmyocardialinvolvementinpatientswithsarcoidosisapplyingT2-weighted,contrast-enhanced,andcinemagneticresonanceimaging:initialresultsofaprospectivestudy.JComputAssistTomogr.2002;26:762–7.

95. DhôteR,VignauxO,BlancheP,DubocD,DusserD,BrezinA,etal.ValueofMRIforthediagnosisofcardiacinvolvementinsarcoidosis.RevMedInterne.2003;24:151–7.

96. GreulichS,DeluigiCC,GloeklerS,WahlA,ZürnC,KramerU,etal.CMRimagingpredictsdeathandotheradverseeventsinsuspectedcardiacsarcoidosis.JACCCardiovascImaging.2013;6:501–11.

97. YazakiY,IsobeM,HiroeM,MorimotoS-I,HiramitsuS,NakanoT,etal.Prognosticdeterminantsoflong-termsurvivalinJapanesepatientswithcardiacsarcoidosistreatedwithprednisone.AmJCardiol.2001;88:1006–10.

98. SchullerJL,ZipseM,CrawfordT,BogunF,BeshaiJ,PatelAR,etal.Implantable

106

cardioverterdefibrillatortherapyinpatientswithcardiacsarcoidosis.JCardiovascElectrophysiol.2012;23:925–9.

99. KronJ,SauerW,MuellerG,SchullerJ,BogunF,SarsamS,etal.Outcomesofpatientswithdefiniteandsuspectedisolatedcardiacsarcoidosistreatedwithanimplantablecardiacdefibrillator.JIntervCardElectrophysiol.2015;13:1–10.

100. OtsukaK,TerasakiF,EishiY,ShimomuraH,OguraY,HoriiT,etal.Cardiacsarcoidosisunderliesidiopathicdilatedcardiomyopathy:importanceofmediastinallymphadenopathyindifferentialdiagnosis.CircJ.2007;71:1937–41.

101. Chapelon-AbricC,deZuttereD,DuhautP,VeyssierP,WechslerB,HuongDLT,etal.CardiacSarcoidosis.Medicine.2004;83:315–34.

102. OkadaR,WakafujiS.Myocarditisinautopsy.HeartVesselsSuppl.1985;1:23–9.

103. KytöV,SaukkoP,LignitzE,SchwesingerG,HennV,SarasteA,etal.Diagnosisandpresentationoffatalmyocarditis.HumPathol.2005;36:1003–7.

104. LassnerD,KühlU,SiegismundCS,RohdeM,ElezkurtajS,EscherF,etal.Improveddiagnosisofidiopathicgiantcellmyocarditisandcardiacsarcoidosisbymyocardialgeneexpressionprofiling.EurHeartJ.2014;35:2186–95.

105. CooperLT.GiantCellMyocarditisinChildren.ProgPediatrCardiol.2007;24:47–9.

106. DasBB,RectoM,JohnsrudeC,KleinL,OrmanK,ShoemakerL,etal.Cardiactransplantationforpediatricgiantcellmyocarditis.JHeartLungTransplant.2006;25:474–8.

107. CooperLT,HareJM,TazelaarHD,EdwardsWD,StarlingRC,DengMC,etal.UsefulnessofImmunosuppressionforGiantCellMyocarditis.AmJCardiol.2008;102:1535–9.

108. BlanksteinR,OsborneM,NayaM,WallerA,KimCK,MurthyVL,etal.CardiacPositronEmissionTomographyEnhancesPrognosticAssessmentsofPatientswithSuspectedCardiacSarcoidosis.JAmCollCardiol.2013;63:329-36.

109. RobertsWC,McAllisterHAJr,FerransVJ.Sarcoidosisoftheheart:aclinicopathologicstudyof35necropsypatients(groupI)andreviewof78previouslydescribednecropsypatients(groupII).AmJMed.1977;63:86-108.

110. KimJS,JudsonMA,DonninoR,GoldM,CooperLT,PrystowskiENetal.Cardiacsarcoidosis.AmHeartJ.2009;157:9–21.

111. DubreySW,FalkRH.Diagnosisandmanagementofcardiacsarcoidosis.ProgCardiovascDis.2010;52:336–46.

112. MoralesAR,LevyS,DavisJ,FineG.Sarcoidosisandtheheart.PatholAnnu.1974;9:139–55.

107

113. YazakiY,IsobeM,HiramitsuS,MorimotoS,HiroeM,OmichiC,etal.Comparisonofclinicalfeaturesandprognosisofcardiacsarcoidosisandidiopathicdilatedcardiomyopathy.AmJCardiol.1998;82:537–40.

114. SchullerJL,OlsonMD,ZipseMM,SchneiderPM,AleongRG,WienbergerHD,etal.Electrocardiographiccharacteristicsinpatientswithpulmonarysarcoidosisindicatingcardiacinvolvement.JCardiovascElectrophysiol.2011;22:1243–8.

115. NagaiT,NaganoN,SuganoY,AsaumiY,AibaT,KanzakiH,etal.EffectofCorticosteroidTherapyonLong-TermClinicalOutcomeandLeftVentricularFunctioninPatientsWithCardiacSarcoidosis.CircJ.2015;79:1593-600.

116. SekhriV,SanalS,DelorenzoLJ,AronowWS,MaguireGP.Cardiacsarcoidosis:acomprehensivereview.ArchMedSci.2011;7:546–54.

117. YoshidaY,MorimotoS,HiramitsuS,TsuboiN,HirayamaH,ItohT.IncidenceofcardiacsarcoidosisinJapanesepatientswithhigh-degreeatrioventricularblock.AmHeartJ.1997;134:382–6.

118. NaruseY,SekiguchiY,NogamiA,OkadaH,YamauchiY,MachinoT,etal.Systematictreatmentapproachtoventriculartachycardiaincardiacsarcoidosis.CircArrhythmElectrophysiol.2014;7:407–13.

119. UusimaaP,YlitaloK,AnttonenO,KerolaT,VirtanenV,PääkköE,etal.Ventriculartachyarrhythmiaasaprimarypresentationofsarcoidosis.Europace.2008;10:760–6.

120. FurushimaH,ChinushiM,SugiuraH,KasaiH,WashizukaT,AizawaY.Ventriculartachyarrhythmiaassociatedwithcardiacsarcoidosis:itsmechanismsandoutcome.ClinCardiol.2004;27:217–22.

121. KoplanBA,SoejimaK,BaughmanK,EpsteinLM,StevensonWG.Refractoryventriculartachycardiasecondarytocardiacsarcoid:electrophysiologiccharacteristics,mapping,andablation.HeartRhythm.2006;3:924–9.

122. JeficD,JoelB,GoodE,MoradyF,RosmanH,KnightB,etal.Roleofradiofrequencycatheterablationofventriculartachycardiaincardiacsarcoidosis:reportfromamulticenterregistry.HeartRhythm.2009;6:189–95.

123. ZipseMM,SauerWH.Electrophysiologicmanifestationsofcardiacsarcoidosis.CurrOpinPulmMed.2013;19:485–92.

124. BanbaK,KusanoKF,NakamuraK,MoritaH,OgawaA,OhtsukaF,etal.Relationshipbetweenarrhythmogenesisanddiseaseactivityincardiacsarcoidosis.HeartRhythm.2007;4:1292–9.

125. FreemanAM,Curran-EverettD,WeinbergerHD,FensterBE,BucknerJK,GottschallEB,etal.Predictorsofcardiacsarcoidosisusingcommonlyavailablecardiacstudies.AmJCardiol.2013;112:280–5.

126. KumarS,BarbhaiyaC,NagashimaK,ChoiE-K,EpsteinLM,JohnRM,etal.

108

VentricularTachycardiainCardiacSarcoidosis:CharacterizationofVentricularSubstrateandOutcomesofCatheterAblation.CircArrhythmElectrophysiol.2014;8:87–93.

127. McArdleBA,BirnieDH,KleinR,deKempRA,LeungE,RenaudJ,etal.Isthereanassociationbetweenclinicalpresentationandthelocationandextentofmyocardialinvolvementofcardiacsarcoidosisasassessedby¹⁸F-fluorodoexyglucosepositronemissiontomography?CircCardiovascImaging.2013;6:617–26.

128. BetenskyBP,TschabrunnCM,ZadoES,GoldbergLR,MarchlinskiFE,GarciaFC,etal.Long-termfollow-upofpatientswithcardiacsarcoidosisandimplantablecardioverter-defibrillators.HeartRhythm.2012;9:884–91.

129. KronJ,SauerW,SchullerJ,BogunF,CrawfordT,SarsamS,etal.Efficacyandsafetyofimplantablecardiacdefibrillatorsfortreatmentofventriculararrhythmiasinpatientswithcardiacsarcoidosis.Europace.2013;15:347–54.

130. ForbesG,UsherA.FatalMyocardialSarcoidosis.BrMedJ.1962;2:771–1.

131. NeryPB,McArdleBA,RedpathCJ,LeungE,LemeryR,DekempR,etal.Prevalenceofcardiacsarcoidosisinpatientspresentingwithmonomorphicventriculartachycardia.PacingClinElectrophysiol.2014;37:364–74.

132. WintersSL,CohenM,GreenbergS,SteinB,CurwinJ,PeE,etal.Sustainedventriculartachycardiaassociatedwithsarcoidosis:assessmentoftheunderlyingcardiacanatomyandtheprospectiveutilityofprogrammedventricularstimulation,drugtherapyandanimplantableantitachycardiadevice.JAmCollCardiol.1991;18:937–43.

133. MehtaD,MoriN,GoldbargSH,LubitzS,WisniveskyJP,TeirsteinA.Primarypreventionofsuddencardiacdeathinsilentcardiacsarcoidosis:roleofprogrammedventricularstimulation.CircArrhythmElectrophysiol.2011;4:43–8.

134. AizerA,SternEH,GomesJA,TeirsteinAS,EckartRE,MehtaD.Usefulnessofprogrammedventricularstimulationinpredictingfuturearrhythmiceventsinpatientswithcardiacsarcoidosis.AmJCardiol.2005;96:276–82.

135. VasaiwalaSC,FinnC,DelprioreJ,LeyaF,GagermeierJ,AkarJG,etal.Prospectivestudyofcardiacsarcoidmimickingarrhythmogenicrightventriculardysplasia.JCardiovascElectrophysiol.2009;20:473–6.

136. DecheringDG,KochhäuserS,WasmerK,ZellerhoffS,PottC,KöbeJ,etal.Electrophysiologicalcharacteristicsofventriculartachyarrhythmiasincardiacsarcoidosisversusarrhythmogenicrightventricularcardiomyopathy.HeartRhythm.2013;10:158–64.

137. NeryPB,KerenA,HealeyJ,LeugE,BeanlandsRS,BirnieDH.Isolatedcardiacsarcoidosis:establishingthediagnosiswithelectroanatomicmapping-guided

109

endomyocardialbiopsy.CanJCardiol.2013;29:1015.e1–3.

138. PhilipsB,MadhavanS,JamesCA,RieleteASJM,MurrayB,TichnellC,etal.Arrhythmogenicrightventriculardysplasia/cardiomyopathyandcardiacsarcoidosis:distinguishingfeatureswhenthediagnosisisunclear.CircArrhythmElectrophysiol.2014;7:230–6.

139. CainMA,MetzlMD,PatelAR,AddetiaK,SpencerKT,SweissNJ,etal.Cardiacsarcoidosisdetectedbylategadoliniumenhancementandprevalenceofatrialarrhythmias.AmJCardiol.2014;113:1556–60.

140. WillnerJM,Viles-GonzalezJF,CoffeyJO,MorgenthauAS,MehtaD.Catheterablationofatrialarrhythmiasincardiacsarcoidosis.JCardiovascElectrophysiol.2014;25:958–63.

141. Viles-GonzalezJF,PastoriL,FischerA,WisniveskyJP,GoldmanMG,MehtaD.Supraventriculararrhythmiasinpatientswithcardiacsarcoidosisprevalence,predictors,andclinicalimplications.Chest.2013;143:1085–90.

142. CordovaFC,D'AlonzoG.Sarcoidosis-associatedpulmonaryhypertension.CurrOpinPulmMed.2013;19:531–7.

143. DesaiMY,FallertMA.Rapidlyprogressingcongestiveheartfailureduetocardiacsarcoidosisinvolvingpapillarymuscles:acasereportandbriefreviewoftheliterature.CardiolRev.2003;11:163–8.

144. MiyazawaK,YoshikawaT,TakamisawaI,MaharaK,InoueK,IguchiN,etal.Presenceofventricularaneurysmpredictspoorclinicaloutcomesinpatientswithcardiacsarcoidosis.IntJCardiol.2014;177:720–2.

145. SchatkaI,BengelFM.Advancedimagingofcardiacsarcoidosis.JNuclMed.2014;55:99–106.

146. SchullerJL,LoweryCM,ZipseM,AleongRG,VarosyPD,WeinbergerHD,etal.Diagnosticutilityofsignal-averagedelectrocardiographyfordetectionofcardiacsarcoidosis.AnnNoninvasiveElectrocardiol.2011;16:70–6.

147. HirotaM,YoshidaM,HoshinoJ,KondoT,IsomuraT.SublocalizationofCardiacInvolvementinSarcoidosisandSurgicalExclusioninPatientsWithCongestiveHeartFailure.AnnThoracSurg.2015;100:81–7.

148. NakajimaS,HanazawaK,KaitaniK,IzumiC,NakagawaY,AndoK,etal.Cardiacresynchronizationtherapyislesseffectiveinthepatientswithcardiacsarcoidosis:insightsfromCubicstudy.JAmCollCardiol.2013;61:E661.

149. GasparovicH,BuenoR,RybickiFJ,BaughmannKL,CapodilupoRC,LeaccheM,etal.Sarcoidosisoftheaorticroot.JThoracCardiovascSurg.2004;128:761–2.

150. CrossB,NicolarsenJ,BullockJ,SugengL,BardoD,LangR.Cardiacsarcoidosispresentingasmitralregurgitation.JAmSocEchocardiogr.2007;20:906.e9–13.

110

151. ZoneraichS,GuptaMP,MehtaJ,ZoneraichO,WesselyZ.Myocardialsarcoidosispresentingasacutemitralinsufficiency.Chest.1974;66:452–4.

152. KinneyE,MurthyR,AscunceG,DonohoeR,ZelisR.Pericardialeffusionsinsarcoidosis.Chest.1979;76:476–8.

153. PesolaG,TeirsteinAS,GoldmanM.Sarcoidoispresentingwithpericardialeffusion.Sarcoidosis.1987;4:42–4.

154. ZelcerAA,LeJemtelTH,JonesJ,StahlJ.Pericardialtamponadeinsarcoidosis.CanJCardiol.1987;3:12–3.

155. DardaS,ZughaibME,AlexanderPB,MachadoCE,DavidSW,SabaS.Cardiacsarcoidosispresentingasconstrictivepericarditis.TexHeartInstJ.2014;41:319–23.

156. ButanyJ,BahlNE,MoralesK,ThangaroopanM,RossH,RaoV,etal.Theintricaciesofcardiacsarcoidosis:acasereportinvolvingthecoronaryarteriesandareviewoftheliterature.CardiovascPathol.2006;15:222–7.

157. WaitJL,MovahedA.Anginalchestpaininsarcoidosis.Thorax.1989;44:391–5.

158. KinneyEL,JacksonGL,ReevesWC,ZelisR.Thallium-scanmyocardialdefectsandechocardiographicabnormalitiesinpatientswithsarcoidosiswithoutclinicalcardiacdysfunction.Ananalysisof44patients.AmJMed.1980;68:497–503.

159. TellierP,PaychaF,AntonyI,NitenbergA,ValeyreD,FoultJM,etal.Reversibilitybydipyridamoleofthallium-201myocardialscandefectsinpatientswithsarcoidosis.AmJMed.1988;85:189–93.

160. LamCSP,TolepKA,MetkeMP,GlocknerJ,CooperLT.Coronarysarcoidosispresentingasacutecoronarysyndrome.ClinCardiol.2009;32:e68–71.

161. TreabaDO,BensonMD,AssadLW,DainauskasJR.SarcoidosisandimmunoglobulinlambdaIIlight-chainamyloidosisdiagnosedafterorthotopichearttransplantation:acasereportandreviewoftheliterature.ModPathol.2005;18:451–5.

162. IijimaK,ChinushiM,FurushimaH,AizawaY.IntramuralinflammationasacauseoftransientST-segmentelevationinapatientofcardiacsarcoidosis.Europace.2012;14:300–2.

163. GibbonsWJ,LevyRD,NavaS,MalcolmI,MarinJM,TardifC,etal.Subclinicalcardiacdysfunctioninsarcoidosis.Chest.1991;100:44–50.

164. TezukaD,TerashimaM,KatoY,ToriiharaA,HirasawaK,SasaokaT,etal.ClinicalCharacteristicsofDefiniteorSuspectedIsolatedCardiacSarcoidosis:ApplicationofCardiacMagneticResonanceImagingand(18)F-Fluoro-2-deoxyglucosePositron-EmissionTomography/ComputerizedTomography.JCardFail.2015;21:313-22.

111

165. IsobeM,TezukaD.Isolatedcardiacsarcoidosis:Clinicalcharacteristics,diagnosisandtreatment.IntJCardiol.2015;182:132–40.

166. MaleszewskiJJ,OrellanaVM,HodgeDO,KuhlU,SchultheissH-P,CooperLT.Long-TermRiskofRecurrence,MorbidityandMortalityinGiantCellMyocarditis.AmJCardiol.2015;115:1733-8.

167. GranerM,LommiJ,KupariM,Raisanen-SokolowskiA,ToivonenL.Multipleformsofsustainedmonomorphicventriculartachycardiaascommonpresentationingiant-cellmyocarditis.Heart.2007;93:119–21.

168. LarsenBT,MaleszewskiJJ,EdwardsWD,CooperLT,SobonyaRE,ThompsonVE,etal.Atrialgiantcellmyocarditis:adistinctiveclinicopathologicentity.Circulation.2013;127:39–47.

169. BurkeJS,MedlineNM,KatzA.Giantcellmyocarditisandmyositis.Associatedwiththymomaandmyastheniagravis.ArchPathol.1969;88:359–66.

170. PalmerHP,MichaelIE.Giant-cellmyocarditiswithmultipleorganinvolvement.ArchInternMed.1965;116:444–7.

171. NambaT,BrunnerNG,GrobD.Idiopathicgiantcellpolymyositis.Reportofacaseandreviewofthesyndrome.ArchNeurol.1974;31:27–30.

172. CooperLT,OkuraY.IdiopathicGiantCellMyocarditis.CurrTreatOptionsCardiovascMed.2001;3:463–7.

173. PerkelD,CzerLSC,MorrisseyRP,RuzzaA,RafieiM,AwadM,etal.Hearttransplantationforend-stageheartfailureduetocardiacsarcoidosis.TransplantProc.2013;45:2384–6.

174. Statementonsarcoidosis.JointStatementoftheAmericanThoracicSociety(ATS),theEuropeanRespiratorySociety(ERS)andtheWorldAssociationofSarcoidosisandOtherGranulomatousDisorders(WASOG)adoptedbytheATSBoardofDirectorsandbytheERSExecutiveCommittee.AmJRespirCritCareMed.1999;160:736–55.

175. HiragaH,YuwaiK,HiroeM,etal.GuidelineforDiagnosisofCardiacSarcoidosis:StudyReportonDiffusePulmonaryDiseasesFromtheJapaneseMinistryofHealthandWelfare.Tokyo:JapaneseMinistryofHealthandWelfare,1993;23–24.

176. SoejimaK,YadaH.TheWork-UpandManagementofPatientswithApparentorSubclinicalCardiacSarcoidosis:WithEmphasisontheAssociatedHeartRhythmAbnormalities.JCardiovascElectrophysiol.2009;20:578–83.

177. JudsonMA,CostabelU,DrentM,WellsA,MaierL,KothL,etal.TheWASOGSarcoidosisOrganAssessmentInstrument:Anupdateofapreviousclinicaltool.SarcoidosisVascDiffuseLungDis.2014;31:19–27.

178. Diagnosticstandardandguidelinesforsarcoidosis.JpnJSarcoidosis

112

GranulomatousDisord.2007;27:89–102.

179. KulS,OzcelikHK,UyarelH,KarakusG,GuvencTS,YalcınsoyM,etal.Diagnosticvalueofstrainechocardiography,galectin-3,andtenascin-Clevelsfortheidentificationofpatientswithpulmonaryandcardiacsarcoidosis.Lung.2014;192:533–42.

180. MatsumotoS,HirayamaY,SaitohH,InoT,MiyauchiY,IwasakiY-K,etal.NoninvasivediagnosisofcardiacsarcoidosisusingmicrovoltT-wavealternans.IntHeartJ.2009;50:731–9.

181. SuzukiT,KandaT,KubotaS,ImaiS,MurataK.Holtermonitoringasanoninvasiveindicatorofcardiacinvolvementinsarcoidosis.Chest.1994;106:1021–4.

182. BurstowDJ,TajikAJ,BaileyKR,DeRemeeRA,TaliercioCP.Two-dimensionalechocardiographicfindingsinsystemicsarcoidosis.AmJCardiol.1989;63:478–82.

183. UemuraA,MorimotoS-I,KatoY,HiramitsuS,OhtsukiM,KatoS,etal.Relationshipbetweenbasalthinningoftheinterventricularseptumandatrioventricularblockinpatientswithcardiacsarcoidosis.SarcoidosisVascDiffuseLungDis.2005;22:63–5.

184. NaganoN,NagaiT,SuganoY,MoritaY,AsaumiY,AibaT,etal.AssociationBetweenBasalThinningofInterventricularSeptumandAdverseLong-TermClinicalOutcomesinPatientsWithCardiacSarcoidosis.CircJ.2015;79:1601–8.

185. AydinKaderliA,GulluluS,CoskunF,YilmazD,UzaslanE.Impairedleftventricularsystolicanddiastolicfunctionsinpatientswithearlygradepulmonarysarcoidosis.EurJEchocardiogr.2010;11:809–13.

186. FahyGJ,MarwickT,McCreeryCJ,QuigleyPJ,MaurerBJ.Dopplerechocardiographicdetectionofleftventriculardiastolicdysfunctioninpatientswithpulmonarysarcoidosis.Chest.1996;109:62–6.

187. SköldCM,LarsenFF,RasmussenE,PehrssonSK,EklundAG.DeterminationofcardiacinvolvementinsarcoidosisbymagneticresonanceimagingandDopplerechocardiography.JInternMed.2002;252:465–71.

188. AngomachalelisN,HourzamanisA,VamvalisC,GavrielidesA.Dopplerechocardiographicevaluationofleftventriculardiastolicfunctioninpatientswithsystemicsarcoidosis.PostgradMedJ.1992;68:S52–6.

189. FocardiM,PicchiA,NikiforakisN,BargagliE,FossiA,MaggiorelliC,etal.Assessmentofcardiacinvolvementinsarcoidosisbyechocardiography.RheumatolInt.2009;29:1051–5.

190. YoussefG,BeanlandsRSB,BirnieDH,NeryPB.Cardiacsarcoidosis:applicationsofimagingindiagnosisanddirectingtreatment.Heart.2011;97:2078–87.

113

191. TsujiT,TanakaH,MatsumotoK,MiyoshiT,HiraishiM,KanekoA,etal.Capabilityofthree-dimensionalspeckletrackingradialstrainforidentificationofpatientswithcardiacsarcoidosis.IntJCardiovascImaging.2013;29:317–24.

192. AggeliC,FelekosI,TousoulisD,GialafosE,RaptiA,StefanadisC.Myocardialmechanicsfortheearlydetectionofcardiacsarcoidosis.IntJCardiol.2013;168:4820–1.

193. OhiraH,TsujinoI,IshimaruS,OyamaN,TakeiT,TsukamotoE,etal.Myocardialimagingwith18F-fluoro-2-deoxyglucosepositronemissiontomographyandmagneticresonanceimaginginsarcoidosis.EurJNuclMedMolImaging.2008;35:933–41.

194. SmedemaJ-P,TruterR,deKlerkPA,ZaaimanL,WhiteL,DoubellAF.Cardiacsarcoidosisevaluatedwithgadolinium-enhancedmagneticresonanceandcontrast-enhanced64-slicecomputedtomography.IntJCardiol.2006;112:261–3.

195. YoshidaA,Ishibashi-UedaH,YamadaN,KanzakiH,HasegawaT,TakahamaH,etal.Directcomparisonofthediagnosticcapabilityofcardiacmagneticresonanceandendomyocardialbiopsyinpatientswithheartfailure.EurJHeartFail.2013;15:166–75.

196. MatohF,SatohH,ShirakiK,OdagiriK,SaitohT,UrushidaT,etal.Theusefulnessofdelayedenhancementmagneticresonanceimagingfordiagnosisandevaluationofcardiacfunctioninpatientswithcardiacsarcoidosis.JCardiol.2008;51:179–88.

197. HamzehNY,WamboldtFS,WeinbergerHD.ManagementofcardiacsarcoidosisintheUnitedStates:aDelphistudy.Chest.2012;141:154–62.

198. YilmazA,FerreiraV,KlingelK,KandolfR,NeubauerS,SechtemU.Roleofcardiovascularmagneticresonanceimaging(CMR)inthediagnosisofacuteandchronicmyocarditis.HeartFailRev.2012;18:747–60.

199. TadamuraE,YamamuroM,KuboS,KanaoS,SagaT,HaradaM,etal.EffectivenessofdelayedenhancedMRIforidentificationofcardiacsarcoidosis:comparisonwithradionuclideimaging.AmJRoentgenol.2005;185:110–5.

200. SerraJJ,MonteGU,MelloES,CoralGP,AvilaLFR,PargaJR,etal.Imagesincardiovascularmedicine.Cardiacsarcoidosisevaluatedbydelayed-enhancedmagneticresonanceimaging.Circulation.2003;107:e188–9.

201. NadelJ,LancefieldT,VoskoboinikA,TaylorAJ.Lategadoliniumenhancementidentifiedwithcardiacmagneticresonanceimaginginsarcoidosispatientsisassociatedwithlong-termventriculararrhythmiaandsuddencardiacdeath.EurHeartJCardiovascImaging.2015;16:634-41.

202. CrawfordT,MuellerG,SarsamS,PrasitdumrongH,ChaiyenN,GuX,etal.Magneticresonanceimagingforidentifyingpatientswithcardiacsarcoidosis

114

andpreservedormildlyreducedleftventricularfunctionatriskofventriculararrhythmias.CircArrhythmElectrophysiol.2014;7:1109–15.

203. IseT,HasegawaT,MoritaY,YamadaN,FunadaA,TakahamaH,etal.ExtensivelategadoliniumenhancementoncardiovascularmagneticresonancepredictsadverseoutcomesandlackofimprovementinLVfunctionaftersteroidtherapyincardiacsarcoidosis.Heart.2014;100:1165–72.

204. ShafeeMA,FukudaK,WakayamaY,NakanoM,KondoM,HasebeY,etal.Delayedenhancementoncardiacmagneticresonanceimagingisapoorprognosticfactorinpatientswithcardiacsarcoidosis.JCardiol.2012;60:448–53.

205. WatanabeE,KimuraF,NakajimaT,HiroeM,KasaiY,NagataM,etal.Lategadoliniumenhancementincardiacsarcoidosis:characteristicmagneticresonancefindingsandrelationshipwithleftventricularfunction.JThoracImaging.2013;28:60–6.

206. VignauxO,DhoteR,DubocD,BlancheP,DusserD,WeberS,etal.Clinicalsignificanceofmyocardialmagneticresonanceabnormalitiesinpatientswithsarcoidosis:a1-yearfollow-upstudy.Chest.2002;122:1895–901.

207. YangY,SafkaK,GrahamJJ,RoifmanI,ZiaMI,WrightGA,etal.CorrelationoflategadoliniumenhancementMRIandquantitativeT2measurementincardiacsarcoidosis.JMagnResonImaging.2014;39:609–16.

208. YoussefG,LeungE,MylonasI,NeryP,WilliamsK,WisenbergG,etal.Theuseof18F-FDGPETinthediagnosisofcardiacsarcoidosis:asystematicreviewandmetaanalysisincludingtheOntarioexperience.JNuclMed.2012;53:241–8.

209. OkumuraW,IwasakiT,ToyamaT,IsoT,AraiM,OriuchiN,etal.Usefulnessoffasting18F-FDGPETinidentificationofcardiacsarcoidosis.JNuclMed.2004;45:1989–98.

210. McArdleBA,LeungE,OhiraH,CockerMS,deKempRA,DaSilvaJ,etal.TheroleofF18-fluorodeoxyglucosepositronemissiontomographyinguidingdiagnosisandmanagementinpatientswithknownorsuspectedcardiacsarcoidosis.JNuclCardiol.2013;20:297–306.

211. YamagishiH,ShiraiN,TakagiM,YoshiyamaM,AkiokaK,TakeuchiK,etal.Identificationofcardiacsarcoidosiswith(13)N-NH(3)/(18)F-FDGPET.JNuclMed.2003;44:1030–6.

212. IshimaruS,TsujinoI,TakeiT,TsukamotoE,SakaueS,KamigakiM,etal.Focaluptakeon18F-fluoro-2-deoxyglucosepositronemissiontomographyimagesindicatescardiacinvolvementofsarcoidosis.EurHeartJ.2005;26:1538–43.

213. AhmadianA,BroganA,BermanJ,SverdlovAL,MercierG,MazziniM,etal.QuantitativeinterpretationofFDGPET/CTwithmyocardialperfusionimagingincreasesdiagnosticinformationintheevaluationofcardiacsarcoidosis.JNucl

115

Cardiol.2014;21:925–39.

214. OsborneMT,HultenEA,SinghA,WallerAH,BittencourtMS,StewartGC,etal.Reductionin¹⁸F-fluorodeoxyglucoseuptakeonserialcardiacpositronemissiontomographyisassociatedwithimprovedleftventricularejectionfractioninpatientswithcardiacsarcoidosis.JNuclCardiol.2014;21:166–74.

215. TakayaY,KusanoKF,NakamuraK,KajiM,ShinyaT,KanazawaS,etal.Reductionofmyocardialinflammationwithsteroidisnotnecessarilyassociatedwithimprovementinleftventricularfunctioninpatientswithcardiacsarcoidosis:predictorsoffunctionalimprovement.IntJCardiol.2014;176:522–5.

216. Sobic-SaranovicD,ArtikoV,ObradovicV.FDGPETimaginginsarcoidosis.SeminNuclMed.2013;43:404–11.

217. LangahR,SpicerK,GebregziabherM,GordonL.Effectivenessofprolongedfasting18f-FDGPET-CTinthedetectionofcardiacsarcoidosis.JNuclCardiol.2009;16:801–10.

218. BulkleyBH,RouleauJR,WhitakerJQ,StraussHW,PittB.Theuseof201thalliumformyocardialperfusionimaginginsarcoidheartdisease.Chest.1977;72:27–32.

219. HiroseY,IshidaY,HayashidaK,MaenoM,TakamiyaM,OhmoriF,etal.Myocardialinvolvementinpatientswithsarcoidosis.Ananalysisof75patients.ClinNuclMed.1994;19:522–6.

220. LeGuludecD,MenadF,FaraggiM,WeinmannP,BattestiJP,ValeyreD.Myocardialsarcoidosis.Clinicalvalueoftechnetium-99msestamibitomoscintigraphy.Chest.1994;106:1675–82.

221. KiuchiS,TeraokaK,KoizumiK,TakazawaK,YamashinaA.UsefulnessoflategadoliniumenhancementcombinedwithMRIand67-Gascintigraphyinthediagnosisofcardiacsarcoidosisanddiseaseactivityevaluation.IntJCardiovascImaging.2007;23:237–41.

222. TawaraharaK,KurataC,OkayamaK,KobayashiA,YamazakiN.Thallium-201andgallium67singlephotonemissioncomputedtomographicimagingincardiacsarcoidosis.AmHeartJ.1992;124:1383–4.

223. OkayamaK,KurataC,TawaraharaK,WakabayashiY,ChidaK,SatoA.Diagnosticandprognosticvalueofmyocardialscintigraphywiththallium-201andgallium-67incardiacsarcoidosis.Chest.1995;107:330–4.

224. BlauwetLA,CooperLT.Idiopathicgiantcellmyocarditisandcardiacsarcoidosis.HeartFailRev.2012;18:733–46.

225. Berthelot-RicherM,O'ConnorK,BernierM,TrahanS,CoutureC,DuboisM,etal.Whenshouldweconsiderthediagnosisofgiantcellmyocarditis?Revisiting

116

“classic”echocardiographicandclinicalfeaturesofthisrarepathology.ExpClinTransplant.2014;12:565–8.

226. HoriT,FujiwaraH,TanakaM,MatsudaM,TsuyuguchiN,SuohM.Idiopathicgiantcellmyocarditisaccompaniedbyasymmetricseptalhypertrophy.JpnCircJ.1987;51:153–6.

227. VaideeswarP,CooperLT.Giantcellmyocarditis:clinicalandpathologicalfeaturesinanIndianpopulation.CardiovascPathol.2013;22:70–4.

228. AzarineA,GuillemainR,BrunevalP.Differentfocaldelayedgadolinium-enhancementpatternsusingcardiacmagneticresonanceinacaseofdiffusegiantcellmyocarditis.EurHeartJ.2009;30:1485.

229. SujinoY,KimuraF,TannoJ,NakanoS,YamaguchiE,ShimizuM,etal.Cardiacmagneticresonanceimagingingiantcellmyocarditis:intriguingassociationswithclinicalandpathologicalfeatures.Circulation.2014;129:e467–9.

230. PöyhönenP,HolmströmM,KivistöS,HänninenH.LategadoliniumenhancementonCMRandsustainedventriculartachycardiapredictseverecardiacinflammation.ActaCardiol.2014;69:637–47.

231. CaforioALP,PankuweitS,ArbustiniE,BassoC,Gimeno-BlanesJ,FelixSB,etal.Currentstateofknowledgeonaetiology,diagnosis,management,andtherapyofmyocarditis:apositionstatementoftheEuropeanSocietyofCardiologyWorkingGrouponMyocardialandPericardialDiseases.EurHeartJ.2013;34:2636–48.

232. ChildsH,FriedrichMG.Cardiovascularmagneticresonanceimaginginmyocarditis.ProgCardiovascDis.2011;54:266–75.

233. MahrholdtH,GoedeckeC,WagnerA,MeinhardtG,AthanasiadisA,VogelsbergH,etal.Cardiovascularmagneticresonanceassessmentofhumanmyocarditis:acomparisontohistologyandmolecularpathology.Circulation.2004;109:1250–8.

234. CooperLT,BaughmanKL,FeldmanAM,FrustaciA,JessupM,KuhlU,etal.Theroleofendomyocardialbiopsyinthemanagementofcardiovasculardisease:ascientificstatementfromtheAmericanHeartAssociation,theAmericanCollegeofCardiology,andtheEuropeanSocietyofCardiology.EndorsedbytheHeartFailureSocietyofAmericaandtheHeartFailureAssociationoftheEuropeanSocietyofCardiology.JAmCollCardiol.2007;50:1914–31.

235. LiangJJ,HeblVB,DeSimoneCV,MadhavanM,NandaS,KapaS,etal.ElectrogramGuidance:AMethodtoIncreasethePrecisionandDiagnosticYieldofEndomyocardialBiopsyforSuspectedCardiacSarcoidosisandMyocarditis.JACCHeartFail.2014;2:466–73.

236. ShieldsRC,TazelaarHD,BerryGJ,CooperLT.Theroleofrightventricularendomyocardialbiopsyforidiopathicgiantcellmyocarditis.JCardFail.

117

2002;8:74–8.

237. YilmazA,KindermannI,KindermannM,MahfoudF,UkenaC,AthanasiadisA,etal.Comparativeevaluationofleftandrightventricularendomyocardialbiopsy:differencesincomplicationrateanddiagnosticperformance.Circulation.2010;122:900–9.

238. ChimentiC,FrustaciA.Contributionandrisksofleftventricularendomyocardialbiopsyinpatientswithcardiomyopathies:aretrospectivestudyovera28-yearperiod.Circulation.2013;128:1531–41.

239. KonecnyT,NoseworthyPA,KapaS,CooperLT,MulpuruSK,SandhuGS,etal.Endomyocardialbiopsy-integratingelectrodeatthebioptometip.TherAdvCardiovascDis.2015;9:66-9.

240. YasutakeH,SeinoY,KashiwagiM,HonmaH,MatsuzakiT,TakanoT.Detectionofcardiacsarcoidosisusingcardiacmarkersandmyocardialintegratedbackscatter.IntJCardiol.2005;102:259–68.

241. BabaY,KuboT,KitaokaH,OkawaM,YamanakaS,KawadaY,etal.Usefulnessofhigh-sensitivecardiactroponinTforevaluatingtheactivityofcardiacsarcoidosis.IntHeartJ.2012;53:287–92.

242. TanadaY,SatoY,SawaT,FujiwaraH,TakatsuY.Serialmeasurementofhigh-sensitivitycardiactroponinIandN-terminalproB-typenatriureticpeptideinapatientpresentingwithcardiacsarcoidosis.InternMed.2012;51:3379–81.

243. DeBoerRA,vanderVeldeAR,MuellerC,vanVeldhuisenDJ,AnkerSD,PeacockWF,etal.Galectin-3:amodifiableriskfactorinheartfailure.CardiovascDrugsTher.2014;28:237–46.

244. SadekMM,YungD,BirnieDH,BeanlandsRS,NeryPB.Corticosteroidtherapyforcardiacsarcoidosis:asystematicreview.CanJCardiol.2013;29:1034–41.

245. AdcockIM.Molecularmechanismsofglucocorticosteroidactions.PulmPharmacolTher.2000;13:115–26.

246. KatoY,MorimotoS-I,UemuraA,HiramitsuS,ItoT,HishidaH.Efficacyofcorticosteroidsinsarcoidosispresentingwithatrioventricularblock.SarcoidosisVascDiffuseLungDis.2003;20:133–7.

247. KudohH,FujiwaraS,ShiotaniH,KawaiH,HirataK.Myocardialwashoutof99mTc-tetrofosminandresponsetosteroidtherapyinpatientswithcardiacsarcoidosis.AnnNuclMed.2010;24:379–85.

248. YodogawaK,SeinoY,ShiomuraR,TakahashiK,TsuboiI,UetakeS,etal.Recoveryofatrioventricularblockfollowingsteroidtherapyinpatientswithcardiacsarcoidosis.JCardiol.2013;62:320–5.

249. YodogawaK,SeinoY,OharaT,TakayamaH,KatohT,MizunoK.Effectofcorticosteroidtherapyonventriculararrhythmiasinpatientswithcardiac

118

sarcoidosis.AnnNoninvasiveElectrocardiol.2011;16:140–7.

250. NagaiS,YokomatsuT,TanizawaK,IkezoeK,HandaT,ItoY,etal.Treatmentwithmethotrexateandlow-dosecorticosteroidsinsarcoidosispatientswithcardiaclesions.InternMed.2014;53:427–33.

251. BarnabeC,McMeekinJ,HowarthA,MartinL.Successfultreatmentofcardiacsarcoidosiswithinfliximab.JRheumatol.2008;35:1686–7.

252. UthmanI,ToumaZ,KhouryM.Cardiacsarcoidosisrespondingtomonotherapywithinfliximab.ClinRheumatol.2007;26:2001–3.

253. ChungES,PackerM,LoKH,FasanmadeAA,WillersonJT,Anti-TNFTherapyAgainstCongestiveHeartFailureInvestigators.Randomized,double-blind,placebo-controlled,pilottrialofinfliximab,achimericmonoclonalantibodytotumornecrosisfactor-alpha,inpatientswithmoderate-to-severeheartfailure:resultsoftheanti-TNFTherapyAgainstCongestiveHeartFailure(ATTACH)trial.Circulation.2003;107:3133–40.

254. YancyCW,JessupM,BozkurtB,ButlerJ,CaseyDE,DraznerMH,etal.2013ACCF/AHAguidelineforthemanagementofheartfailure:areportoftheAmericanCollegeofCardiologyFoundation/AmericanHeartAssociationTaskForceonPracticeGuidelines.JAmCollCardiol.2013;62:e147–239.

255. EchtDS,LiebsonPR,MitchellLB,PetersRW,Obias-MannoD,BarkerAH,etal.Mortalityandmorbidityinpatientsreceivingencainide,flecainide,orplacebo.TheCardiacArrhythmiaSuppressionTrial.NEnglJMed.1991;324:781–8.

256. EpsteinAE,DimarcoJP,EllenbogenKA,EstesNAM,FreedmanRA,GettesLS,etal.2012ACCF/AHA/HRSfocusedupdateincorporatedintotheACCF/AHA/HRS2008guidelinesfordevice-basedtherapyofcardiacrhythmabnormalities:areportoftheAmericanCollegeofCardiologyFoundation/AmericanHeartAssociationTaskForceonPracticeGuidelinesandtheHeartRhythmSociety.JAmCollCardiol.2013;61:e6–75.

257. BrignoleM,AuricchioA,Baron-EsquiviasG,BordacharP,BorianiG,BreithardtO-A,etal.2013ESCGuidelinesoncardiacpacingandcardiacresynchronizationtherapy:theTaskForceoncardiacpacingandresynchronizationtherapyoftheEuropeanSocietyofCardiology(ESC).DevelopedincollaborationwiththeEuropeanHeartRhythmAssociation(EHRA).EurHeartJ.2013;34:2281–329.

258. EpsteinAE,DimarcoJP,EllenbogenKA,EstesNAM,FreedmanRA,GettesLS,etal.ACC/AHA/HRS2008guidelinesforDevice-BasedTherapyofCardiacRhythmAbnormalities:executivesummary.HeartRhythm.2008;6:934–55.

259. ZaidiAR,ZaidiA,VaitkusPT.Outcomeofhearttransplantationinpatientswithsarcoidcardiomyopathy.JHeartLungTransplant.2007;26:714–7.

260. KhanR,TweedieEJ,PflugfelderPW,WhiteJA.Cardiacsarcoidinahearttransplantrecipient:detectionwithcardiacmagneticresonanceimaging.

119

TransplantProc.2010;42:1976–8.

261. YagerJEE,HernandezAF,SteenbergenC,PersingB,RussellSD,MilanoC,etal.Recurrenceofcardiacsarcoidosisinahearttransplantrecipient.JHeartLungTransplant.2005;24:1988–90.

262. BurkeWM,KeoghA,MaloneyPJ,DelpradoW,BryantDH,SprattP.Transmissionofsarcoidosisviacardiactransplantation.TheLancet.1990;336:1579.

263. FrustaciA,ChimentiC,PieroniM,GentiloniN.Giantcellmyocarditisrespondingtoimmunosuppressivetherapy.Chest.2000;117:905–7.

264. DesjardinsV,PelletierG,LeungTK,WatersD.Successfultreatmentofsevereheartfailurecausedbyidiopathicgiantcellmyocarditis.CanJCardiol.1992;8:788–92.

265. MenghiniVV,SavcenkoV,OlsonLJ,TazelaarHD,DecGW,KaoA,etal.Combinedimmunosuppressionforthetreatmentofidiopathicgiantcellmyocarditis.MayoClinProc.1999;74:1221–6.

266. ChaudhryMA,CorreaA,LeeC,YoonA,GrazetteL,SaremiF,etal.Moderndaymanagementofgiantcellmyocarditis.IntJCardiol.2015;178:82–4.

267. NieminenMS,SalminenUS,TaskinenE,HeikkiläP,PartanenJ.Treatmentofseriousheartfailurebytransplantationingiantcellmyocarditisdiagnosedbyendomyocardialbiopsy.JHeartLungTransplant.1994;13:543–5.

268. BrilakisES,OlsonLJ,BerryGJ,DalyRC,LoisanceD,ZuckerM,etal.Survivaloutcomesofpatientswithgiantcellmyocarditisbridgedbyventricularassistdevices.ASAIOJ.2000;46:569–72.

269. IwaiK,TachibanaT,TakemuraT,MatsuiY,KitaichiM,KawabataY.Pathologicalstudiesonsarcoidosisautopsy.I.Epidemiologicalfeaturesof320casesinJapan.ActaPatholJpn.1993;43:372–6.

270. TakayaY,KusanoKF,NakamuraK,ItoH.ComparisonofOutcomesinPatientsWithProbableVersusDefiniteCardiacSarcoidosis.AmJCardiol.2015;115:505-9.

271. AntonelliA,FerrariSM,CorradoA,DiDomenicantonioA,FallahiP.Autoimmunethyroiddisorders.AutoimmunRev.2015;14:174–80.

272. BlanksteinR,OsborneMT,WallerAH,MurthyVL,DorbalaS,StevensonWG,etal.Reply:(18)F-FDGimaginginpatientswith"suspected,"butnot“proven,”sarcoidosis.JAmCollCardiol.2014;64:631.

273. Ramos-CasalsM,Brito-ZerónP,KostovB,Sisó-AlmirallA,BoschX,BussD,etal.Google-drivensearchforbigdatainautoimmunegeoepidemiology:analysisof394,827patientswithsystemicautoimmunediseases.AutoimmunRev.2015;14:670–9.

120

274. NeryPB,BeanlandsRS,NairGM,GreenM,YangJ,McArdleBA,etal.Atrioventricularblockastheinitialmanifestationofcardiacsarcoidosisinmiddle-agedadults.JCardiovascElectrophysiol.2014;25:875–81.

275. FairweatherD,CooperLT,BlauwetLA.Sexandgenderdifferencesinmyocarditisanddilatedcardiomyopathy.CurrProblCardiol.2013;38:7–46.

276. BaughmanRP,GruttersJC.Newtreatmentstrategiesforpulmonarysarcoidosis:antimetabolites,biologicaldrugs,andothertreatmentapproaches.LancetRespirMed.2015;3:813–22.

277. RybickiBA,HirstK,IyengarSK,BarnardJG,JudsonMA,RoseCS,etal.Asarcoidosisgeneticlinkageconsortium:thesarcoidosisgeneticanalysis(SAGA)study.SarcoidosisVascDiffuseLungDis.2005;22:115–22.

278. CooperLT,BlauwetLA.Whenshouldhigh-gradeheartblocktriggerasearchforatreatablecardiomyopathy?CircArrhythmElectrophysiol.2011;4:260–1.

279. SimonenP,LehtonenJ,KandolinR,SchildtJ,MarjasuoS,MiettinenH,etal.F-18-FluorodeoxyglucosePositronEmissionTomography-GuidedSamplingofMediastinalLymphNodesintheDiagnosisofCardiacSarcoidosis.AmJCardiol.2015;116:1581-5.

280. Authors/TaskForceMembers,PrioriSG,Blomstrom-LundqvistC,MazzantiA,BlomN,BorggrefeM,etal.2015ESCGuidelinesforthemanagementofpatientswithventriculararrhythmiasandthepreventionofsuddencardiacdeath:TheTaskForcefortheManagementofPatientswithVentricularArrhythmiasandthePreventionofSuddenCardiacDeathoftheEuropeanSocietyofCardiology(ESC)Endorsedby:AssociationforEuropeanPaediatricandCongenitalCardiology(AEPC).EurHeartJ.2015Aug29:ehv316–87.

281. GonzalezJA,KramerCM.RoleofImagingTechniquesforDiagnosis,PrognosisandManagementofHeartFailurePatients:CardiacMagneticResonance.CurrHeartFailRep.2015;12:276–83.

282. SarikayaI.CardiacapplicationsofPET.NuclMedCommun.2015;36:971-85.

Cardiac Sarcoidosis and Giant Cell Myocarditis in Finland

HELSINKI UNIVERSITY HOSPITALDIVISION OF CARDIOLOGYHEART AND LUNG CENTERFACULTY OF MEDICINEDOCTORAL PROGRAMME IN CLINICAL RESEARCHUNIVERSITY OF HELSINKI

RIINA KANDOLIN

DISSERTATIONES SCHOLAE DOCTORALIS AD SANITATEM INVESTIGANDAM UNIVERSITATIS HELSINKIENSIS 104/2015

104/2015

Helsinki 2015 ISSN 2342-3161 ISBN 978-951-51-1768-7

RIIN

A K

AN

DO

LIN

Card

iac Sarcoid

osis

Gian

t C

ell

Myocard

itis in

Fin

land

Recent Publications in this Series

85/2015 And DemirClinical Use of Urinary Gonadotropin Determinations in Children and Adolescents86/2015 Raisa HaavikkoSynthesis of Betulin Derivatives with New Bioactivities87/2015 Hongxin ChenHeterobasidion annosum sensu stricto Pathogenesis: Bioinformatic and Functional Study of Cerato-Platanin Family Proteins88/2015 Kristiina RajamäkiRole of the Infl ammasome Pathway in Atherosclerosis89/2015 Pirkko KriikkuToxicological Abuse Profi le of New Recreational Drugs in Driving-Under-The-Infl uence and Post-Mortem Cases in Finland90/2015 Eeva MustonenRed Clover Isofl avonoids in Feed, Plasma and Milk of Ruminants91/2015 Flavia RestituttiCentral, Cardiovascular and Metabolic Effects of Dexmedetomidine Associated with the Selective Peripheral Alpha2-adrenoceptor Antagonist MK-467 in Dogs92/2015 Katariina NurmiPrevention of Infl ammatory Cellular Responses by Ethanol and Hemin – Interplay Between Infl ammasomes and Processes Inhibiting Infl ammation93/2015 Liang HeRare Genetic Variants and Complex Diseases - a Bayesian Approach94/2015 Jani SaksiStroke-Associated Carotid Atherosclerosis - a Role for FABP4 in Atheroma Vulnerability and Cardiometabolic Risk95/2015 Otto HemminkiCancer Immunotherapy with a Gene Modifi ed Serotype 3 Oncolytic Adenovirus96/2015 Jukka KoskelaIndividual Trajectories in Asthma and COPD: A Longitudinal Perspective to Obstructive Lung Disease97/2015 Miira KlemettiTrends In Obstetric and Perinatal Outcomes of Women with Type 1 Diabetes During 1988-2011 – A Finnish Population-Based Observational Study98/2015 Ileana B. QuinteroPhysiological Functions of Prostatic Acid Phosphatase99/2015 Minna MatikainenLaunching New Products in the Finnish Pharmaceutical Industry: A Relationship Approach100/2015 Päivi Ylä-AnttilaPhagophore Membrane Connections and RAB24 in Autophagy101/2015 Kaisa KyöstiläIdentifi cation of Novel Mutations and Molecular Pathways for Canine Neurodegeneration and Chondrodysplasia102/2015 Emmi JoensuuEpigenetic Alterations in Sporadic and Familial Cancers103/2015 Elina ReponenPreoperative Risk-Assessment Methods, Short-Term Outcome, and Patient Satisfaction in Elective Cranial Neurosurgery

and