ContentslistsavailableatSciVerseScienceDirect
Journal
of
Cardiology
j our na l h o me p ag e :w w w . e l s e v i e r . c o m / l o c at e / j j c c
Original
article
Role
of
transthoracic
left
atrial
appendage
wall
motion
velocity
in
patients
with
persistent
atrial
fibrillation
and
a
low
CHADS2
score
Naoyasu
Yoshida
(PhD)
a,∗,
Mitsunori
Okamoto
(MD,
FJCC)
b,
Hidekazu
Hirao
(MD)
b,
Kiyomi
Nanba
(MSc)
a,
Hiroki
Kinoshita
(MD)
b,
Hiroya
Matsumura
(MD)
b,
Yukihiro
Fukuda
(MD)
b, Hironori
Ueda
(MD,
FJCC)
baEchocardiographyDivision,DepartmentofLaboratoryMedicine,HiroshimaPrefecturalHospital,Hiroshima,Japan bDepartmentofCardiology,HiroshimaPrefecturalHospital,Hiroshima,Japan
a
r
t
i
c
l
e
i
n
f
o
Articlehistory: Received1March2012
Receivedinrevisedform17April2012 Accepted8May2012
Availableonline26June2012 Keywords:
Leftatrialfunction
TissueDopplerechocardiography Cerebralembolism
Atrialfibrillation CHADS2score
a
b
s
t
r
a
c
t
Backgroundandpurpose:Thromboembolicriskhasbeenexaminedbysemi-invasivetransesophageal echocardiography.Weassessedtheriskofthrombogenesisinpatientswithpersistentatrialfibrillation (AF)noninvasivelybyusingtransthoracictissueDopplerechocardiography(TDE)inrelationtoalow CHADS2score.
Methods:EightypatientswithpersistentAFunderwentbothtransthoracicandtransesophageal echocar-diography.Peakleftatrialappendage(LAA),wallmotionvelocity(WV)duringLAAcontractionwas measuredbytransthoracicandtransesophagealTDE.LAAflowvelocitywasalsodeterminedby trans-esophagealechocardiography.
Results:TransthoracicLAAWVcouldbemeasuredin78ofthe80patients,andthevalueswereclosely cor-relatedwithtransesophagealTDEvalues(r=0.98)andwithtransesophagealLAAflowvelocity(r=0.82). TransthoracicLAAWVwassignificantlylowerwithincreasingspontaneousechocontrast(SEC)severity (severeSEC,mildSEC,noSEC:5.7±2.4,10.2±3.3,and14.5±5.5cm/s,respectively).SevereSECwas notedin31of61patientswithaCHADS2score≤2,in19of46patientswithaCHADS2score≤1andin6 of21patientswithaCHADS2score=0.FordiagnosingsevereSEC,atransthoracicLAAWV<10cm/shad asensitivityof81%andspecificityof92%inthepatientswithaCHADS2score≤2,asensitivityof74%and specificityof91%inthepatientswithaCHADS2score≤1andasensitivityof44%andspecificityof83% inthepatientswithaCHADS2score=0.
Conclusions:AtransthoracicLAAWV<10cm/sinpersistentAFpatientswithalowCHADS2scoremaybea veryspecificdiagnostictoolforevaluatingsevereSEC,oneofthehighriskfactorsforthromboembolism. ©2012JapaneseCollegeofCardiology.PublishedbyElsevierLtd.Allrightsreserved.
Introduction
Cerebral infarction induced by cardiogenic embolism is observedin 13–34%of strokepatients[1].Cardiogeniccerebral embolismismostfrequentlycausedbyatrialfibrillation(AF).The reducedleftatrialfunctioninpatientswithAFmayresultin throm-busformationintheleftatrialappendage(LAA).Thefunctional parametersoftheLAAsuchasflowvelocity,fractionalchange dur-ingatrialcontraction,andspontaneousechocontrast(SEC)have beenpreviouslyassessed by transesophagealechocardiography. AnLAAflowvelocityof<20cm/sandprominentSECarehigh-risk
∗ Correspondingauthorat: EchocardiographyDivision,Departmentof
Labo-ratoryMedicine,HiroshimaPrefecturalHospital,1-5-54Ujinakanda,Minami-Ku, Hiroshima734-8530,Japan.Tel.:+822541818;fax:+822530659.
E-mailaddress:n yoshida0210@ybb.ne.jp(N.Yoshida).
factors for cerebral embolism [2,3]. LAA wall motion velocity (LAAWV),whichisanindicatorofLAAfunctionand thromboge-nesisrisk,hasbeenmeasuredbytransesophagealtissueDoppler echocardiography (TDE) [4–9]. However, since this is a semi-invasive method, its application as a screening tool is limited. Therefore, analternativenoninvasive methodforassessing LAA functionisrequired.
Recentdevelopments in transthoracicechocardiography and newtechnologyhavefacilitated theevaluationofthesize,flow velocities,andemptyingfractionoftheLAA[10–12].Moreover, LAAfunctioninpatientswithsinusrhythmandAFhasbeen deter-minedbytransthoracicTDE.TransthoracicTDEcanpredicttherisk ofthromboembolisminpatientswithahistoryofstrokeor tran-sientischemicattack[13–15].Ontheotherhand,theCHADS2score isawidelyusedclinicalindexforevaluatingtheriskof thromboem-bolismand initiating anticoagulanttherapy in patientswithAF
[16].Recently,CHA(2)DS(2)-VAScscoreshavebeenalsoproposed
0914-5087/$–seefrontmatter©2012JapaneseCollegeofCardiology.PublishedbyElsevierLtd.Allrightsreserved. http://dx.doi.org/10.1016/j.jjcc.2012.05.007
Table1
Patientcharacteristics.
Variables Atrialfibrillationpatients
Number 78
Men 56
Age(years) 69±9
HT 24
DM 14
Congestiveheartfailure 26
Warfarin 25 CHADS2score≥3 17 CHADS2score=2 15 CHADS2score=1 25 CHADS2score=0 21 LAAV(cm/s) 31±21 LAAA(mm2) 507±201 LAAFC(%) 26±18 LAAWVbyTEE(cm/s) 8.0±4.7 LAAWVbyTTE(cm/s) 8.3±4.8 LAD(mm) 45±8
HT,hypertension; DM,diabetesmellitus; Warfarin,administration beforethis study;LAAV,leftatrialappendage(LAA)flowvelocityduringatrialcontraction; LAAA,maximalLAAarea;TEE,transesophagealechocardiography;TTE, transtho-racicechocardiography;LAAFC,LAAfractionalchangeduringatrialcontraction; LAAWV,LAAwallmotionvelocityduringLAAcontraction;LAD,leftatrialdimension.
forevaluatingtheriskofischemicstroke/cardiovasculareventsin JapanesepatientswithparoxysmalAForforpredictingclinical out-comesinpatientswithatrialfibrillationaftercatheterablationin Asianpatients[17,18].AhighCHADS2score,especiallyincluding strokewithavalue of2,indicatestheneedfor anticoagulation therapyinpatientswithpersistent AF.Thescoremaybeuseful forsecondarypreventionofstrokeinpatientswithpersistentAF. However,thesescoresarenotalwaysconsistentwiththeleftatrial (LA)function.ReducedLAfunctionmayresultinbloodstasisinthe LAandmaycausecardiogeniccerebralembolism.ThereducedLA functionmaycauseSECintheLA.TheseverityofSECisstrongly associatedwithLAAthrombusandembolicevents[19].Therefore, theabilitytopredicttheseverityofSECintheLAwouldbe clin-icallyusefulbecausewecurrentlyanticoagulateallpatientswith severeSEC.Itisveryimportanttononinvasivelyevaluatesevere SECinpatientswithoutastrokehistoryandwithalowCHADS2 score.Inthisstudy,weexaminedthediagnosticroleof transtho-racicLAAWVinrelationtotheCHADS2scoreforevaluatingtherisk ofthrombosisorbloodstasisinpatientswithpersistentAF.
Materialsandmethods
Participants
Thestudypopulationcomprised80consecutivepatientswith persistentnonvalvularAFwhounderwentbothtransthoracicand transesophagealechocardiographyinourlaboratoryfromOctober 2008 to January 2010. Transesophageal echocardiography was orderedbycardiologistsorneurologistsforexaminingtheriskof thromboembolisminthepatientswithAF.Thedatawere retro-spectivelyanalyzed.PersistentAFwasdefinedasAFlastingmore than1weekaccordingtothepastelectrocardiographyorclinical symptoms.Theirmeanagewas69±9years.Patientspredisposed toaorticarchandcarotidsourceswerenotincludedinthestudy. Thecharacteristicsofthe78patientsinwhomLAAWVcouldbe measuredusingtransthoracicTDEareshowninTable1.Among the78patients,61hadaCHADS2score≤2,and17hadaCHADS2 score≥3.Therewere46patientswithaCHADS2score≤1and21 patientswithCHADS2score=0.Twenty-fivepatientshadreceived anticoagulationtherapybeforeparticipationinthisstudy.There were9casesofnewcerebralembolismthatwerediagnosedonthe basisoftheclinicalsignsandfindingsfrommagneticresonance
imaging,computedX-raytomography,andcerebralangiography, oracombinationofthesemethods.Ifstronganticoagulation ther-apyfortheacutephaseofcerebralinfarctionisstartedtooearly, itmaycausebleedingintheinfarctionarea. Therefore,the neu-rologistin our hospitaldid notadminister warfarin during the super-acutephaseofcerebralinfarction.Furthermore,in persis-tent AF,decisions regardinganticoagulation therapy dependon thetransesophagealechocardiographicfindings;therefore,asmall numberofpatientsunderwentanticoagulationtherapybeforethis study.
Writteninformedconsentwasobtainedfromallparticipants.
Echocardiographicapparatus
Both transthoracic and transesophageal echocardiographic examinationswereconductedwithanultrasoundsystem(Vivid 7;General ElectricHealthcare, Milwaukee,WI,USA). The mea-surement specifications for transthoracic TDE were as follows: framerate,141frames/s;frequency,2.6MHz;andsampling vol-umewidth,3.3mm.Asdescribed previously,thecorresponding measurements fortransesophageal TDE wereas follows: frame rate,78frames/s;frequency,3.9MHz;andsamplingvolumewidth, 3.2mm[20].
Transesophagealechocardiography
The peak LAA flow velocity (LAAV) during LAA contraction andthechangeintheLAAfractionalareaduringatrial contrac-tion(LAAFC)wererecordedusingconventionaltransesophageal echocardiographyinthelong-axisview.Weexaminedthe pres-enceofSECintheLAbodyandLAA.Inthepresentstudy,ifSEC wasvisibleonlywhentheechogainwasincreasedby4dBfrom thenormalgainsetting,wedefineditasmildlypositive.IfSECwas visibleatthenormalgainsetting,wedefineditassevere.The nor-malgainsettingwasdefinedasnoexcessivenoiseechointheother cardiaccavity[20].WhentheseverityofSECdifferedbetweenthe LAbodyandLAA,thestrongergradewasrecordedasthe sever-ity.ThepresenceofthrombiandtheseverityofSECwereassessed by3independentinvestigatorsusingtransesophageal echocardio-graphy.Furthermore,theLAAWVatthetipoftheLAAduringLAA contractionwasdeterminedusingpulsedtransesophagealTDE,and themaximumpositivepeakwavevelocitywithineachRRinterval wasaveragedover10cardiaccycles.Atthattime,themitralring motionartifactswerecarefullyneglected.
AnglecorrectionwasnotperformedbecausetheDopplerbeam wasalmostparalleltothemotionofthelongitudinallycontracting appendage(Fig.1).
Transthoracicechocardiography
LAAWVwasmeasuredusing transthoracicTDEwithin1hof thetransesophagealechocardiographyexamination.The triangle-shapedLAAontheleftsideoftheaorticrootintheparasternal short-axisviewwasidentified.Thesamplevolumewasplacedvery closetothetipoftheLAA,andtheLAAWVwasmeasuredusing pulsedtransthoracicTDE(Fig.1).
LAAWVwascalculatedastheaverageofthemaximumnegative peakwavevelocitywithineachRRintervalover10cardiaccycles excludingthemitralringmotionsignals(Fig.2).TheDopplerbeam anglewascorrectedbyobservingthedirectionofthelongitudinally contractingappendage(Fig.1).TheLAdimensionwas convention-allymeasuredusingM-modetransthoracicechocardiography.
Fig.1. MeasurementofLAAWVduringtheLAAcontractionphaseusingpulsedtissueDopplerechocardiographywithtransesophageal(leftpanel)andtransthoracic(right panel)echocardiographyinpatientswithatrialfibrillation.Thesamplevolume(dottedarrows)forLAAWVmeasurementwasplacedatthetipoftheLAA.Theaverage maximumpositivepeakwavewithineachRRintervalduringLAAcontractionwasdesignatedasLAAWV.AO,aorta;LA,leftatrium;LAA,leftatrialappendage;LAAWV,left atrialappendagewallvelocity;LV,leftventricle;ROI,regionofinterest.
Fig.2.LAAWVduringtheLAAcontractionphasewasmeasuredusingpulsedtissueDopplerechocardiography(leftpanel),andthemitralannularmotionsignals(right panel)weremeasuredusingtransthoracicechocardiographyinpatientswithatrialfibrillation.WhentheLAAWV(solidlines)islow,mitralringmotionsignals(dottedlines) aresometimesmistakenforLAAWVreadings.ElectrocardiographicmonitoringisnecessarytodifferentiateLAAWVfrommitralringsignals(dottedlines).AO,aorta;LA,left atrium;LAA,leftatrialappendage;LAAWV,leftatrialappendagewallvelocity;LV,leftventricle;RA,rightatrium;ROI,regionofinterest;RV,rightventricle.
Statisticalanalysis
Simplelinearregressionanalysiswasusedtodetermine cor-relations between different parameters. The differences in the
valuesbetweenthetwogroupswereexaminedusingunpairedt -tests,andthedifferencesinthevaluesamong thethree groups wereexaminedusinganalysisofvariance(ANOVA)witha Bonfer-ronicorrection.TheLAAWVcutoffvaluefordiagnosing cerebral
Table2
CharacteristicsofatrialfibrillationpatientswithorwithoutSEC.
Variables SevereSEC MildSEC NoSEC pvalue
Number 48 17 13 – Men 33 13 10 NS Age(years) 69±9 67±9 60±19 NS HT 17 8 2 NS DM 9 3 2 NS Warfarin 21 6 3 NS ThrombusinLAA 13 0 0 0.004* Cerebralembolism 9 0 0 0.028* CHADS2score 2.1±1.4 0.8±0.7 0.5±0.7 <0.001* LAAV(cm/s) 19±13 38±16 60±19 <0.001* LAAA(mm2) 533±224 489±164 441±150 NS LAAFC(%) 16±12 36±19 43±12 <0.001* LAAWVbyTEE(cm/s) 5.3±1.8 10.6±3.5 14.1±5.3 <0.001* LAAWVbyTTE(cm/s) 5.7±2.4 10.2±3.3 14.5±5.5 <0.001* LVEF(%) 61±12 66±6 61±14 NS LAD(mm) 48±7 42±8 40±7 <0.001*
SEC,spontaneousechocontrast;NS,notstatisticallysignificant;p-valuesindicatedifferencesamong3groupsaccordingtoanalysisofvariancewithaBonferroniadjustment; HT,hypertension;DM,diabetesmellitus;Warfarin,administrationbeforethisstudy;LAAV,leftatrialappendage(LAA)flowvelocityduringatrialcontraction;LAAA,maximal LAAarea;TEE,transesophagealechocardiography;TTE,transthoracicechocardiography;LAAFC,LAAfractionalchangeduringatrialcontraction;LAAWV,LAAwallmotion velocityduringLAAcontraction;LAD,leftatrialdimension.
Fig.3.RegressionanalysisofLAAWVmeasuredbytransesophagealand transtho-racicechocardiographyinpatientswithatrialfibrillationwhenanglecorrection wasappliedfortransthoracicechocardiography.LAAWV,leftatrialappendagewall velocity;TEE,transesophagealechocardiography;TTE,transthoracic echocardiog-raphy.
embolismand/or LAAthrombus and for diagnosing severe SEC wasapproximatelyhalf(10cm/s)thenormalvalue(22±3cm/s) mentionedinastudyinvolvinghealthyindividuals[21].The sen-sitivity,specificity,andpredictivevaluesfordiagnosing cerebral embolism,LAAthrombus,andsevereSECweredeterminedusing LAAWVandCHADS2scores(Table3).Wegroupedthepatientswith embolismand/orLAthrombosistogetherashavingLAthrombus. TheStatisticalPackagefortheSocialSciences(SPSSversion11.0J; SPSSJapanInc., Tokyo,Japan) wasusedforallstatistical analy-sesexceptdiagnosticaccuracy.Statisticalsignificancewassetat
p<0.05.
Interobserver variability in the LAAWV measurements was calculated asthedifferencebetween twomeasurements ofthe samesubjectperformedbytwodifferentobserversdividedbythe meanvalue.Intraobservervariabilitywascalculatedasthe differ-encebetweentwomeasurementsofthesamesubjectbyasingle observer,dividedbythemeanvalue.
Thestudyprotocolwasapprovedbyourinstitutionalreview board.
Results
ValidationoftransthoracicLAAWVmeasurement
LAAWV was successfully measured using transthoracic TDE in 78 of the 80 patients (98%). The transthoracic LAAWV
patternswerereproducibleandsimilartothoseoftransesophageal LAAWVin theseAFpatients(Fig.1).The LAAWVvalues deter-minedbytransthoracicTDErangedfrom2to26cm/sinpatients withpersistent nonvalvular AF. The LAAWVvalues determined bytransthoracicand transesophagealTDEdidnotdiffer signifi-cantly(9.0±5.9vs.8.7±5.9cm/s,respectively)andwerestrongly correlated (r=0.98)(Fig.3).LAAWVwassignificantlycorrelated withtransesophagealLAAV(r=0.82,p<0.001)andLAAFC(r=0.56,
p<0.001)(Fig.4).
Theintra-andinterobservervariabilitiesinLAAWV measure-mentbytransthoracicTDEwere5.6%and6.8%,respectively.
LAAWVmeasuredbytransthoracicTDEinrelationtothrombosis
andbloodstasis
Transthoracic LAAWV values were significantly lower in patients with cerebral embolism or LAA thrombus than in patients without these conditions (5.0±2.2 vs. 9.3±4.9cm/s,
p<0.001).TransthoracicLAAWVvaluesweresignificantlylower with increasing SEC severity (severe SEC, mild SEC, with-out SEC: 5.7±2.4, 10.2±3.3, and 14.5±5.5cm/s, respectively) (Table2).
ThrombosisriskevaluationusingtransthoracicLAAWVand
CHADS2score
ACHADS2score≥3had highsensitivityand specificity(88% and 97%,respectively)for diagnosing cerebralembolism and/or LAA thrombus, but low specificity (49%) for diagnosing severe SEC.
Inthe78patients,atransthoracicLAAWV<10cm/shada sen-sitivity,specificity,andpositiveandnegativepredictivevaluesof 86%,88%,94%,and77%fordiagnosingsevereSEC.Amongthe61 persistentAFpatientswithaCHADS2score≤2,embolismand/or LAAthrombuswereobservedin2andsevereSECwasnotedin 31.SevereSECwasnotedin13ofthe46persistentAFpatients withaCHADS2score≤1,andin6ofthe21persistentAFpatients withaCHADS2score0.Inthe61patientswithaCHADS2score≤2, atransthoracicLAAWV<10cm/shadasensitivity,specificity,and positiveandnegativepredictivevaluesof81%,92%,94%,and77% fordiagnosingsevereSEC,respectively.Inthe46patientswitha CHADS2score≤1,atransthoracicLAAWV<10cm/shada sensitiv-ity,specificity,andpositiveandnegativepredictivevaluesof74%, 91%,89%,and77%fordiagnosingsevereSEC,respectively(Table3). Inthe21persistentAFpatientswithaCHADS2score0,a transtho-racicLAAWVwas6.8±4.1cm/sin6patientswithsevereSECand
Fig.4.LAAWVplottedagainstLAAflowvelocityduringLAAcontraction(LAAV)inpatientswithatrialfibrillation(leftpanel).LAAWVplottedagainstLAAfractionalarea duringatrialcontraction(LAAFC)(rightpanel).LAA,leftatrialappendage;LAAWV,leftatrialappendagewallvelocity.
Table3
AccuracyofdiagnosisofsevereSECpatientsusingtransthoracicLAAWV<10cm/s.
Variables Sensitivity(%) Specificity(%) Positivepredictive
value(%)
DiagnosingsevereSEC 86 88 94
DiagnosingsevereSECinpatientswithaCHADS2score≤2 81 92 94
DiagnosingsevereSECinpatientswithaCHADS2score≤1 74 91 89
DiagnosingsevereSECinpatientswithaCHADS2score=0 44 83 66
SEC,spontaneousechocontrast;LAAWV,leftatrialappendagewallmotionvelocityduringleftatrialappendagecontraction.
was13.0±6.0cm/sin15patientswithoutsevereSEC(p=0.015). Inthe21persistentAFpatientswithaCHADS2score0,a transtho-racicLAAWV<10cm/shadasensitivity,specificity,andpositiveand negativepredictivevaluesof44%,83%,66%,and67%fordiagnosing severeSEC.
Discussion
ValidationoftransthoracicLAAWVforassessingLAAfunction PreviousstudiesusingtransthoracicpulsedDoppler echocar-diographysuccessfullymeasuredDopplerflowvelocitiesin62–88% of patients [10,11]. Recently, transthoracic TDE, a noninvasive method,wasdemonstratedtobeusefulforanalyzingLAAfunction in patients with AF [13–15,20]. In the present study, we suc-cessfullydeterminedLAAWVin 98%of thepatientswithAFby usingtransthoracicTDE.Inaddition,thestrongcorrelationbetween transthoracicandtransesophagealLAAWVmeasurementsinthe presentstudymayvalidatetheuseoftransthoracicTDE.Both intra-andinter-observervariabilitiesintransthoracicLAAWV measure-mentswerelow. These findingssuggest that transthoracicTDE ismoreusefulandreliablethanflowvelocitymeasurementwith transthoracicpulsedDopplerechocardiography.Thesuperiorityof TDEoverpulsedDopplerflowmetrymaybeattributedtothe differ-enceintheamplitudeofultrasoundreflectionpowerbetweenthe myocardiumandbloodcellsbecausethemyocardiumisamuch strongerultrasoundreflector.However,caremustbeexercisedto eliminatethemitralannularmotionsignalsconcomitantwiththe TDEsignalofLAAmotion.
LAAWVmeasuredbytransthoracicTDEinrelationtoLA
thrombosisandbloodstasis
SECdeterminedbytransesophagealechocardiographyimplies bloodstasisoraggregationandmaybeanimportantriskfactorfor cerebralembolism[22,23].Uretskyetal.detectedSECinpatients withatransthoracicLAAWV≤13cm/swithasensitivityand speci-ficityof97%and60%,respectively[14].Tamuraetal.showedthat LAAWV<8.7cm/shadasensitivityof77%andaspecificityof76%for diagnosingLAAthrombusformationandwasanindependent pre-dictorofcerebro-vascularevents(hazardratio,3.460)[13,24].We previouslystudiedtheLAAWVduringcontractionandrelaxation inhealthyJapanesesubjectsanditsmeanvalueduring contrac-tionwas22±3cm/s[20].Wetook10cm/sastheLAAWVcutoff forabnormallowvalue,becauseitwasasimpleindexandwas approximatelyhalftheJapanesenormalvalueanditsvalue posi-tionedbetween13cm/stakenbyUretskyetal.and8.7cm/staken byTamuraetal.[13,14].AtransthoracicLAAWV<10cm/salone hadasensitivityandspecificityof86%and88%,respectively,for diagnosingsevereSECinJapanesepatients.Inthepresentstudy, transthoracicLAAWVvaluesweresignificantlylowerwith increas-ingSECseverity.Thus,alowLAAWVvaluemaybeanimportant indicatorofbloodstasisoraggregationintheLAA.
ClinicalvalueoftransthoracicLAAWVinrelationtoCHADS2score
TheCHADS2scoreisawidelyusedclinicalindexfor evaluat-ingtheriskofthromboembolismanddecidingwhethertoinitiate anticoagulanttherapyinpatientswithnonvalvularAF[16].Wealso foundthataCHADS2score≥3hadarelativelyhighsensitivityand specificityfordiagnosingcerebralembolismand/orLAAthrombus. TheseresultssupporttheuseofahighCHADS2scorenotonlyas adiagnosticaidforevaluatingtheriskofthrombosisbutalsoasa clinicalindicatorforstartinganticoagulanttherapyinpatientswith persistentAF.However,sincetheCHADS2scoreincludesstroke withavalueof2,usingittodeterminethesensitivityofpredicting priorstrokeistautological.Virtuallyallclinicianswould antico-agulateapatientwithchronicAFwhohasalreadyhadastroke. ACHADS2scoreof≥3maybeusefulforthesecondary preven-tionofstroke,butitmaynotbeenoughforprimarypreventionof strokeinAFpatients,because,aCHADS2scoremaynotbealways consistentwithLAAfunction.Therefore,wetestedtheclinical sig-nificanceofthetransthoracicLAAWVasanadditionalnoninvasive aidforevaluatingLAAfunctionandthrombogenesis.SevereSECin theleftatriumisawell-knownmarkerofahighriskforthrombosis. ThepredictionofSECseveritywouldbeclinicallyusefulbecausewe currentlyanticoagulateallpatientswithsevereSECregardlessof thepresenceorabsenceofpriorstroke[25–27].Wefoundthata LAAWV<10cm/shadhighsensitivityandspecificityfor diagnos-ingsevereSEC;inaddition,LAAWVwassignificantlylowerwith increasingSECseverity.Itisimportanttononinvasivelyevaluate SECfor primarypreventionofstrokein AFpatientswitha low CHADS2score.Inthepresentstudy,severeSECwasnotedin13 ofthe46persistentAFpatientswithaCHADS2score≤1andin6 patientswithsevereSECeveninthe21patientswithaCHADS2 score0.Inthe46patientswithaCHADS2score≤1,atransthoracic LAAWV<10cm/shadasensitivityandspecificityof74%and91% fordiagnosingsevereSEC.Inthe21persistentAFpatientswith aCHADS2score0,atransthoracicLAAWV<10cm/shada sensi-tivityand specificityof44%and 83%for diagnosingsevereSEC. Thus,atransthoracicLAAWV<10cm/smaybeaspecific diagnos-tictoolforevaluatingsevereSECinthepatientsjudgedashaving alow riskforthrombosisaccordingtotheCHADS2score.From thisperspective,atransthoracicLAAWV<10cm/smaybeasimple andconvenientindicatorforstartinganticoagulationtherapyand maypreventprimarystrokeinthepatientsjudgedaslowriskfor thromboembolismaccordingtotheirCHADS2score.Moreover,we couldsparepatientswithAFwithaCHADS2score≤2andnormal LAAWVfromanticoagulationtherapytoavoidrelated complica-tions,whileoptimalanticoagulationstrategyforstrokeprevention hasbeennowinvestigatedinJapanesepatientswithAF,J-RHYTHM Registrystudydesign[28].
Limitations
Oneofthemajorlimitationsofthisstudyisitsretrospective designasopposedtoarandomizedprospectiveone.Thenumber ofpatientsenrolledwasrelativelysmall,whichmayhavecauseda selectionbias.
Itis necessarytocorrecttheangleoftheultrasonicDoppler beamwhenmeasuringtransthoracicLAAWV;theDopplerbeam anglerangedfrom 10◦ to50◦ inthe presentstudy, whichmay have caused some errors. However, in the present study, the LAAWVvalueswerealmostequalbetweentransthoracicand trans-esophagealmeasurements.Wepreviouslyreportedthecorrection of the Doppler beam angle by observing the direction of the longitudinally contracting appendage in normal subjects [20]. Anotherlimitationofthisstudymightbetheultrasonicviewused toobtainthetransthoracicLAAWV;wechosetheparasternal short-axisviewfordeterminingLAAWV.Uretskyetal.andSallachetal. foundnodifferencesintheassessmentofLAAwallmotionbetween theparasternalandapicalviewsintransthoracicechocardiography
[14,15].Inthepresentstudy,thesamplevolumewasplacedasclose aspossibletothetipoftheLAAsinceitisthemostmobilepointin theLAA.However,theregionalLAAwallmotionmayinfacthave differentmyocardialmotionvelocities.Furtherstudiesarerequired toelucidatethesignificanceofregionalwallmotionintheLAA.
Conclusionsandimplications
Despitesomemethodologicallimitations,transthoracicLAAWV maybeausefulnoninvasivetoolforestimatingLAAfunctionand thrombogenesisespeciallyinpatientswithalow CHADS2score withpersistentAF.
References
[1]PalazzuoliA,RicciD,LenziC,LenziJ,PalazzuoliV.Transesophageal echocardio-graphyforidentifyingpotentialcardiacsourcesofembolisminpatientswith stroke.NeurolSci2000;21:195–202.
[2]PollickC,TaylorD.Assessmentofleftatrialappendagefunctionby trans-esophagealechocardiography:implicationsforthedevelopmentofthrombus. Circulation1991;84:223–31.
[3]KamaleshM,CopelandTB,SawadaS.Severelyreducedleftatrialappendage function:acauseofembolicstrokeinpatientsinsinusrhythm.JAmSoc Echocardiogr1998;11:902–4.
[4] ParvathaneniL,MahenthiranJ,JacobS,FoltzJ,GillWJ,GhummanW, Gradus-PizloI,FeigenbaumH,SawadaSG.ComparisonoftissueDopplerdynamicsto Dopplerflowinevaluatingleftatrialappendagefunctionbytransesophageal echocardiography.AmJCardiol2005;95:1011–4.
[5] GurlertopY,YilmazM,AcikelM,BozkurtE,ErolMK,SenocakH,AlpN.Tissue Dopplerpropertiesoftheleftatrialappendageinpatientswithmitralvalve disease.Echocardiography2004;21:319–24.
[6]TopsakalR,EryolNK,Ozdo˘gruI,SeyfeliE,AbaciA,O˘guzhanA,Bas¸arE,Ergin A,CetinS.ColorDopplertissueimagingtoevaluateleftatrialappendage functioninpatientswithmitralstenosisinsinusrhythm.Echocardiography 2004;21:235–40.
[7]SahinT,UralD,KilicT,BildiriciU,KozdagG,AgacdikenA,UralE.Evaluation ofleftatrialappendagefunctionsaccordingtodifferentetiologiesofatrial fib-rillationwithatissueDopplerimagingtechniquebyusingtransesophageal echocardiography.Echocardiography2009;26:171–81.
[8]SevimliS,GundogduF,ArslanS,AksakalE,GurlertopHY,IslamogluY,TasH, AcikelM,ErolMK,SenocakH,KarakelleogluS,AtesalS,AlpN.Strainandstrain rateimaginginevaluatingleftatrialappendagefunctionbytransesophageal echocardiography.Echocardiography2007;24:823–9.
[9]Cayli M,Acartürk E,Demir M,Kanadas¸i M. Systolictissue velocity is a usefulechocardiographicparameterinassessmentofleftatrialappendage function in patients with mitral stenosis. Echocardiography 2007;24: 816–22.
[10]FukudaN,ShinoharaH,SakabeK,OnoseY,NadaT,TamuraY.Transthoracic Dopplerechocardiographicmeasurementofleftatrialappendagebloodflow velocity:comparisonwithtransesophagealmeasurement.EurJEchocardiogr 2003;4:191–5.
[11]ColettaC,InfusinoT,SciarrettaS,SestiliA,TrambaioloP,CianfroccaC,De MarchisE,AuritiA,SalustriA.TransthoracicDopplerechocardiographyforthe assessmentofleftatrialappendagesizeandbloodflowvelocity,amulticentre study.JCardiovascMed2008;9:147–52.
[12]IwamaM,KawasakiM,TanakaR,OnoK,WatanabeT,HiroseT,NagayaM,Noda T,WatanabeS,MinatoguchiS.Leftatrialappendageemptyingfractionassessed byafeature-trackingechocardiographicmethodisadeterminantofthrombus inpatientswithnonvalvularatrialfibrillation.JCardiol2012;59:329–36. [13]TamuraH,WatanabeT,HironoO,NishiyamaS,SasakiS,ShishidoT,Miyashita
T,MiyamotoT,NitobeJ,KayamaT,KubotaI.Lowwallvelocityofleftatrial appendagemeasuredbytrans-thoracicechocardiographypredictsthrombus formationcausedbyatrialappendagedysfunction.JAmSocEchocardiogr 2010;23:545–52.
[14] UretskyS,ShahA,BangaloreS,RosenbergL,SarjiR,CantalesDR, Macmillan-MarottiD,ChaudhryFA,SherridMV.Assessmentofleft atrialappendage functionwithtransthoracictissueDopplerechocardiography.EurJ Echocar-diogr2009;10:363–71.
[15] SallachJA,PuwanantS,DrinkoJK,JafferS,DonalE,ThambidoraiSK,Asher CR,JaberWA,StoddardMF,ZoghbiWA,WeissmannNJ,MulvaghSL,MaloufJF, JasperSE,BorowskiAG,etal.Comprehensiveleftatrialappendageoptimization ofthrombususingsurfaceechocardiography:TheCLOTSMulticenterPilotTrial. JAmSocEchocardiogr2009;22:1165–72.
[16]GageBF,WatermanAD,ShannonW,BoechlerM,RichMW,RadfordMJ. Vali-dationofclinicalclassificationschemesforpredictingstroke:resultsfromthe NationalRegistryofAtrialFibrillation.JAMA2001;285:2864–70.
[17]KomatsuT,TachibanaH,SatohY, OzawaM,KunugitaF,Ueda H, Naka-mura M. Relationship between CHA(2)DS(2)-VASc scores and ischemic stroke/cardiovasculareventsinJapanesepatientswithparoxysmalatrial fib-rillationwithoutreceivinganticoagulanttherapy.JCardiol2012;59:321–8. [18]ChaoTF,LinYJ,TsaoHM,TsaiCF,LinWS,ChangSL,LoLW,HuYF,TuanTC,
SuenariK,LiCH,HartonoB,ChangHY,AmbroseK,WuTJ,etal.CHADS(2)and CHA(2)DS(2)-VAScscoresinthepredictionofclinicaloutcomesinpatientswith atrialfibrillationaftercatheterablation.JAmCollCardiol2011;29:2380–5. [19]FatkinD,KellyRP,FeneleyMP.Relationsbetweenleftatrialappendageblood
flowvelocity,spontaneousechocardiographiccontrastandthromboembolic riskinvivo.JAmCollCardiol1994;23:961–9.
[20]YoshidaN,OkamotoM,NanbaK,YoshizumiM.TransthoracictissueDoppler assessmentofleftatrialappendagecontractionandrelaxation:theirchanges withaging.Echocardiography2010;27:839–46.
[21] OkamotoM,FujiiY,MakitaY,KajiharaK,YamasakiS,IwamotoA,Hashimoto M,SuedaT,YoshidaN.Leftatrialappendagefunctioninpatientswithsystemic embolisminspiteofinsinusrhythm.JAmSocEchocardiogr2006;19:211–4. [22]vonder ReckeG, Schmidt H,Illien S,OmranH.Useof transesophageal
contrastechocardiography forexcludingleftatrial appendagethrombiin patientswithatrialfibrillationbeforecardioversion.JAmSocEchocardiogr 2002;15:1256–61.
[23]DawnB,VarmaJ,SinghP,LongakerRA,StoddardMF.Cardiovasculardeathin patientswithatrialfibrillationisbetterpredictedbyleftatrialthrombusand spontaneousechocardiographiccontrastascomparedwithclinicalparameters. JAmSocEchocardiography2005;18:199–205.
[24]TamuraH,WatanabeT,NishiyamaS,SasakiS,WanezakiM,ArimotoT, Taka-hashiH,ShishidoT,MiyashitaT,MiyamotoT,KubotaI.Prognosticvalueoflow leftatrialappendagewallvelocityinpatientswithischemicstrokeandatrial fibrillation.JAmSocEchocardiogr2012;25:576–83.
[25]MaltagliatiA,GalliCA,TamboriniG,CelesteF,MuratoriM,PepiM.Incidence ofspontaneousechocontrast,‘sludge’andthrombibeforecardioversionin patientswithatrialfibrillation:newinsightsintotheroleoftransesophageal echocardiography.JCardiovascMed2009;10:523–8.
[26]KleemannT,BeckerT,StraussM,SchneiderS,SeidlK.Prevalenceofleft atrial thrombus and dense spontaneous echo contrast in patients with short-termatrialfibrillation<48hoursundergoingcardioversion:valueof transesophagealechocardiographytoguidecardioversion.JAmSoc Echocar-diogr2009;22:1403–8.
[27] BernhardtP,SchmidtH,HammerstinglC,LüderitzB,OmranH.Patientswith atrialfibrillationanddensespontaneousechocontrastathighriskaprospective andserialfollow-upover12monthswithtransesophageal echocardiogra-phyandcerebralmagneticresonanceimaging.JAmCollCardiol2005;45: 1807–12.
[28]AtarashiH,InoueH,OkumuraK,YamashitaT,OrigasaH,J-RHYTHMRegistry Investigators.Investigationofoptimalanticoagulationstrategyforstroke pre-ventioninJapanesepatientswithatrialfibrillation—theJ-RHYTHMRegistry studydesign.JCardiol2011;57:95–9.