JACCVol.28,No.4October1996:1017-23

1017

PEDIATRIC CARDIOLOGY

Morphometric Analysis of Unbalanced Common AtrioventricularCanal Using Two-Dimensional Echocardiography

MERYL S. COHEN, MD, MARSHALL L. JACOBS, MD, PAUL M. WEINBERG, MD, FACC,

JACK RYCHIK, MD, FACC

Philadelphia,Pennsylvania

objectives.This study was designedto define morphometricechocardiographicvariablesof unbalancedcommonatrioventric-ular canal (CAVC)that could aid in appropriate referral forsurgicalrepair.

Background.UnbalancedCAVChas a high surgicalmortalityrate. This may be secondaryto inappropriatereferral of somepatientsfortwo-ventriclerepair (closureofseptaldefects)insteadofsingle-ventriclerepair (Norwoodpalliationand Fontanoperation),

Methods.Theechocardiogramsof103patientswithCAVCwereretrospectivelyreviewed.In the subcostal left anterior obliqueview,the area of the atrioventricular (AV)valveaportionedovereach ventriclewas measured,and an AVvalveindex (AVVI)wascalculated as Ieftkight valve area. The ventricular cavity ratiobetweenthe twoventricleswasestimatedas leftventricularlengthtimeswidthdividedbyright ventricularlengthtimeswidth.Thesevariableswerecorrelatedwith surgicalreferral and outcome.

Results.Patients previouslycategorizedas having balancedCAVCall had AVVI >0.67 (n = 77). Of the patients with

In commonatrioventricularcanal(CAVC),thecommonatrio-ventricular(AV) valvemaybe positionedequallyoverbothventricles(balanced)or unequallyoverthe rightor leftventri-cle (unbalanced),withvariabledegreesof associatedventricu-lar hypoplasia.Unbalanceof the commonAVvalveoccursin-10% of all patientswith CAVC (l). Althoughsuccessfulrepair of balancedCAVCthroughclosureof the atrial andventricularseptaldefectcomponents(two-ventriclerepair) iswell established(4-6), surgicaloutcome for unbalancedCAVC is poor (2,3). Reparativestrategiesfor unbalancedCAVChaveincludedclosureof the septaldefectsor, in casesofmarkedunbalance,palliativeprocedures,suchaspulmonaryarterial banding(2) or the Norwoodoperation (1,7).Pooroutcomefor this lesionmaybe due in part to the inabilitytocorrectlychoosethe appropriateoperativeprocedure.Diffi-cultyexistsin reliablypredictingwhichpatientswilltolerate

Fromthe Divisionsof Cardiologyand CardiothoracicSurgery,The Chil-dren’sHospitalof Philadelphiaand Departmentsof Pediatricsand Surgery,UniversityofPennsylvaniaSchoolof Medicine,Philadelphia,Pennsylvania.

ManuscriptreeeivedDecember14,1995;revisedmanuscriptreceivedMay9,1996,acceptedMay14,1996.

Addressfor correspondence:Dr.JackRychik,NoninvasiveCardiovascularLaboratory,Divisionof Cardiology,The Children’sHospitalof Philadelphia,Philadelphia,Pennsylvania19104.E-mail:jrycbik@mail.med.upenn.edu.

01996bythe AmericanCollegeofCardiologyPublishedbyElsevierScienceInc.

—.

unbalancedCAVC(n = 26), 11had ductal-dependentcirculationand underwentNorwoodpalliation (AVVI0,21 & 0.13,mean &SD), and 15 had two-ventriclerepair (AVVI0.51 * 0.12, p c0.0001).Ofthese15patients,9havesurvived,withnodifferenceinmean AVWbetween survivorsand nonsurvivors(0.52 & 0,11versus 0.49 & 0.13, p = 0.72). For all 103 patients, AVWcorrelated with ventricular cavityratio. However,of the unbal.anced CAVCgroup who underwent two-ventriclerepair, threenonsurvivorshad a discrepancybetweenAVVIand ventricularcavity ratio (low AVVIbut normal ventricular size). A largeventricularseptaldefectwaspresent in all sixnonsurvivorsbut inonlyfour of nine survivors(p < 0.05).

Conclusions.Echocardiographicmorphometryis useful in de-finingunbalancein CAVC.If AVVIis <0,67in the presenceof alarge ventricular septal defect, a single-ventricleapproach torepair shouldbe considered.

(JArnCoilCardiol1996;28:1017-23)

two-ventriclerepair and which shouldhave single-ventriclepalliationwithsubsequentFontanoperation.

In the past, cardiaccatheterizationand angiographicesti-mates of relativeventricularsize and volumewere used todeterminethe degree of unbalancein CAVC (8,9).Thesemethods,however,are subjectto error becauseseverityofunbalanceof the commonAV valve may not necessarilycorrelatewiththe degreeofventricularhypoplasiapresent.Inaddition,rightventricularvolumehasbeenfoundto be largerthan left ventricularvolumein most patientswith balancedCAVC(9-11);hence,it maybe difficultto distinguishbalancefrom unbalancesolelyon the basisof ventricularsize.Fur-thermore, angiographicvisualizationof the common AVvalve,the definingstructureof balanceor unbalance,can bedifficult.

Becauseultrasoundprovidesexcellentresolutionof AVvalvetissueas wellas adjacentseptalstructures,it shouldbesuperiorto angiographyin delineatingAVvalveposition.Wetherefore used echocardiographyto developan index thatquantifiesthe degree of commonAV valve unbalanceinrelationto the ventricularseptum.The objectiveof thisstudywasto morphometricallydefineby two-dimensionalechocar-diographythevariablesofAVvalvepositionthat couldaid intheappropriatediagnosticclassificationofunbalancedCAVC,

0735-1097/96/$15.00PII SO735-1097(96)OO262-8

1018 COHENETAL. JACCVol.28,No.4UNBALANCEDATRIOVENTRICULARCANAL October1996:1017-23

AbbreviationsandAcronyms

AV = atrioventricularAW1 = atrioventricularvalveindexCAVC= commonatrioventriculucanal

thusresuhingin correctrecommendationsfor operationandimprovedoutcome.

MethodsPatients. The echocardiogramsand medical records of

patients who underwent surgicalrepair of CAVC at theChildren’sHospitalofPhiladelphiafromJanuary1990to June1994wereretrospectivelyreviewed.Patientswereexcludedifthey had heterotaxysyndrome,dextrocardiaor conotruncalanomalies(i.e., tetralogyof Fallot),aloneor in combination.Of 195patients identified,103 met these criteria and hadpreoperativeechocardiogramsof goodqualityfor review.

Patientswere classifiedinto two groups.Of the 103pa-tients,77(75%)werejudgedat the timeofinitialpresentationby angiographyor operativeinspection,or both, concurrentwiththe impressionofthepreoperativetwo-dimensionalecho-cardiogram(conventionalmeans)to havea balancedCAVC(groupI). The remaining26 patients(25%)were judgedtohavesomedegreeof commonAVvalveunbalancebyangiog-raphy,echocardiographyor surgicalinspection(groupII, 23unbalancedto the right,3 unbalancedto the left).Echocar-diographicmorphometricanalysisof the AVvalve(see later)wasnotperformeda prioriandthuswasnotconsideredin thediagnosticcategorizationof thesepatientsbeforeoperation.

Duringreviewof the medicalrecords,the followingwasnotedfor eachpatient:1)the presenceof Downsyndrome;2)thedegreeofAVvalveinsufficiencybeforeoperation(patientswere categorizedas havinghemodynamicallysignificantAVvalveinsufficiencyifmoderateor severeinsufficiencywasseenbyDopplercolor[12]);and3) thesizeof theventricularseptaldefect(patientswerecategorizedas havinga largeunrestric-tive defect or a small restrictiveor no defect by two-dimensionaland Dopplercolorflowechocardiography).

Operation. All patientsin groupI (balanced)underwenttwo-ventriclerepairconsistingofpatchor sutureclosureoftheatrialandventricularseptaldefectcomponents(n = 77,ageatoperation17~ 28months[meant SD]).Patientsin groupIIwith a diagnosisof unbalancedCAVCwere noted to havevaryingdegreesof ventricularhypoplasiaand either 1) pre-sentedin earlyinfancywithmarkedleftventricularhypoplasiaand a ductaldependentsystemiccirculationandhenceunder-wentNorwoodstageI palliation(group11A,n = 11,age atoperation9 f 11 days);or 2) presented in later infancy,withoutductaldependentsystemiccirculation,wereperceivedto have two adequatelysizedventriclesand underwenttwo-ventriclerepair(groupIIB,n = 15,ageat operation4.8f 3.6

months).Allpatientsunbalancedto theleft(n = 3)underwenttwo-ventriclerepair.

Echocardiography: morphometric analysis of the commonAVvalve and estimation of ventricular size. The echocardio-gramsofall103patients,acquiredbeforesurgicalinterventionwere analyzedin blinded manner with regard to surgicalreferral.Patientsbetween3 weeksand 3 monthsof ageweresedatedwith 60 to 75 mgkg bodyweightof chloralhydratebeforethestudy.StudieswereinitiallyrecordedonVHS0.5-in.formatvideotapeusingeithera Hewlett-PackardSonos1000or AcusonXP128ultrasoundsystemcoupledwitheithera 5.0-or 3.5-MHztransducer.

The utilityof the subcostalleft anteriorobliqueplane invisualizingthe ventricularseptum on end, as well as therelationoftheAVvalveto theseptumhasbeenwelldescribed(13,14).The degreeof balanceor unbalanceof the AVvalvewasthereforeestablishedbyperformingthefollowing,utilizingan off-lineanalysissystem(Digisonics)interfacedwitha per-sonalcomputer:1)At end-diastole,thecommonAVvalvewasvisualizedin the left anteriorobliqueplane (Fig. 1). 2) Thecrestof the muscularseptumand the tip of the infundibularseptumwereidentified,and a linewasdrawnfromone to theother,bisectingthe commonAVvalve,therebydesignatingaportionofthevalveto eachventricle.3)WiththecommonAVvalvedividedbetweenthe twoventricles,the respectivedesig-natedvalveareasoverthe left andrightventricleweretraced,and a ratio of the two areas (leftvalvearea dividedby rightvalvearea)wasgenerated(AVvalveindexIAWI]). In threecaseswherethecommonvalvewasunbalancedto the left,thisratiowasinverted(rightvalveareafleftvalvearea).An AWIof 1.0thereforerepresentsequalbalanceof the commonAVvalveoverbothventricles.

To estimatethe cavitysizeof the rightand left ventriclesrelativeto eachother,measurementsweremadein the apicalfour-chamberviewof ventricularlength(fromthe AV valveannulusto apex)andwidth(fromthe crestof the ventricularseptumto the free wall) for each ventricle.A ratio of theventricularcavitydimensionswasthen calculatedas leftven-tricularlengthtimeswidthdividedby rightventricularlengthtimeswidth.

Allmeasurementsweremadeinat leastthreecardiaccyclesat end-diastoleby single-frameanalysis,and the averagewasusedforcalculations.Theframebeforetheonsetofclosureofthe AVvalvewasdesignatedas end-diastole.

Data analysis and statistics. Resultsare reportedasmeanvaluet SD.Linearregressionanalysiswasusedto test for acorrelationbetweenthe degree of unbalanceof the CAVC(AWI) and the ventricularcavityratio.The Mann-Whitneyrank sum test was used to assessdifferencesbetween thegroups for the echocardiographicindexesmeasured. Chi-squareanalysis(or theFisherexacttest,ifn < 10)wasusedtodeterminedifferencesin the presenceof Down syndrome,degree of preoperativeAV valve insufficiency,size of theventricularseptaldefectcomponentand survivalbetweenthegroups.A p value<0.05wasconsideredsignificant.

JACCVol.28,No.4October1996:1017–23

COHENETAL. 1019UNBALANCEDATRIOVENTRICULARCANAL

Table1. Comparisonof PatientsWith Balanced Versus UnbalancedCommonAtrioventricular Canal According to Presence of DownSyndrome,Unrestrictive Ventricular Septal Defect andHemodynamicallySignificantPreoperative AtrioventricularValve Insufficiency

BalarmdCAVC UnbalCAVC(groupI) (groupII)(n= 77) (n= 26) p Value

Down Synd 39(51%) 9(35%) 0.23UnrestrictiveVSD 44(57%) 19(73%) 0.23AW insuffic 10(13%) 9(35%) 0.03

Datapresentedarenumber(%)ofpatients.AVVinsuffic= atrioventricrrlarvalveinsufficiency;CAVC= commonatrioventricularcanal;Synd= syndrome;Unbal= unbalanced;VSD= ventricularseptaldefect.

(Table1). However,a significantlyhigherpercentageof pa-tientswithunbalancedCAVChadhemodynamicallyimportantAV valveinsufficiencythan thosewithbalancedCAVC(p =0.03)(Table1).PatientswithDownsyndromeweremorelikelyto have an unrestrictiveventricularseptal defect (completeCAVC)than thosewithoutDownsyndrome(42 [86%]of 49versus21[64%]of33,p < 0.0001),but thedegreeofAVvalveinsufficiencywasnot signitlcantlydifferentbetweenthesetwogroups(p = 0.344).

Morphometric analysis: balanced (group I) versus unbal-anced (group II) CAVC. The AWI for all 103 patientscorrelatedwellwiththeventricularcavityratio (r = 0,75,p <0.001)(Fig.2).MeanAWI andventricularcavityratioswerenot statisticallydifferentin relationto the sizeor presenceof aventricularseptaldefect(Table2).

Patientscategorizedby conventionalmeansas havingbal-anced CAVChad morphometricvariablesthat were signifi-cantlyhigherthan patientscategorizedas havingunbalancedCAVC.TheAVVIforthebalancedgrouprangedfrom0.67to1.19,and ventricularcavityratios rangedfrom 0.46to 1.19.Patientsdeterminedto have unbalanceddefectspresentedwith a widespectrumrangingfrom mild commonAV valveunbalanceto severeunbalancewithmarkedventricularhypo-plasia.TheAVVIfor the unbalancedgrouprangedfrom0.07to 0.65;ventricularcavityratiosrangedfrom0.01to 0.87.Nooverlapwas present in AWI between the balanced andunbalancedgroups(no balancedgrouppatienthad an AWIlower,andno unbalancedgrouppatienthad an AWI higher,than0.66).However,significantoverlapwasnotedforventric-ular cavityratio,withthree patientsin the unbalancedgrouphavingventricularcavityratiosgreaterthanthemeanvalueforpatients in the balancedgroup (Fig. 2). One patient hadmarked discrepancybetween AWI and ventricularcavityratio manifestedby significantcommonAV valveunbalance(AWI 0.27)but relativelyequal right and left ventricularcavitydimensions(ventricularcavityratio 0.77)(Fig.2).

Unbalanced CAVC. GroupIZ:single-versw two-ventriclerepair.For the 11patientswith unbalancedCAVCwho un-derwentNorwoodpalliation,both mean AVVI and meanventricularcavityratioweresignificantlysmallerthan for the15who underwenttwo-ventriclerepair (Table2). However,

1020 COHENETAL.UNBALANCEDATRIOVENTRICULARCANAL

1.3 x

JACCVol.28,No.4October1996:1017-23

1,2

1.1

4 1

● ● ☛

●

●

●✎☛☛

0.9

0.8

j 0.70 ●

~ 0.6 Figure 2. Correlation betweenu AVVI and ventricular cavity ratio

0.5

0.4❑ n

0.3A

.x

o0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4

Ventricular Cavity Ratio

overlapwaspresentbecausepatientswithan AVVIbetween0.27and 0.41and a ventricularcavityratiobetween0.26and0.34underwenteither a Norwoodoperationor two-ventriclerepair(Fig.2).Nopatientwithan AWI <0.3or a ventricularcavityratio<0.26hada two-ventriclerepair;allpresentedwithductal-dependentcirculationin infancy.

GroupIZB:survivorsversusnonsurvivorsof two-ventriclerepair.Nine(60%)of the 15patientswithunbalancedCAVCwhounderwenttwo-ventriclerepair(groupIIB)havesurvived(Table3).NosignificantdifferencewaspresentinmeanAVVIor meanventricularcavityratiosbetweensurvivorsand non-survivors(Table 2). AlthoughDown syndromeas well ashemodynamicallyimportant(moderateto severe)preopera-tiveAVvalveinsufficiencywasmoreprevalentinnonsurvivors,thesewerenot statisticallysignificantriskfactors.Five(83%)ofsixnonsurvivorsandfour(44%)ofninesurvivorshadDownsyndrome(p = 0.29);four nonsurvivors(67%)and onlyonesurvivor(11%) had moderateAV valveinsufficiency(p =0.08).However,the presenceof a large,unrestrictiveventric-

Table2. AtrioventricularValveIndexandVentricularCavityRatio{mean(SD)]fortheVariousStudyGroups

No.ofPatients AVVI VCR

BalancedCAVC(groupI)UnbalCAVC(groupII)p valueSingle-ventriclerepair(groupHA)Two-ventriclerepair(groupIIB)p valueSurvivors(groupIIB)Nonsurvivors(groupIIB)p value

77 0.90(0.11)26 0.38(0.19)

<0.000111 0.21(0.13)15 0.51(0.12)

<0.00019 0.52(0.11)6 0.49(0.13)

0.72

0.75(0.16)0.38(0.25)<0.0001

0.16(0.11)0.54(0.20)<0.0001

0.50(0.15)0.64(0.25)

0.32

VCR= ventricularcavityratio;otherabbreviationsas in Table1.

(r= 0.75,p < 0.001). -AUnbeleneadShrgleVentrkle

Rapdr

❑ lhrbekneed Tw@hntrlcieRepair- Survivors

xUnbalancedTwo-VentricleRepair- Non-Survlors

ularseptaldefectwasa significantriskfactorforpoorsurgicaloutcomein patientswithunbalancedCAVCundergoingtwo-ventriclerepair. All six nonsurvivors(1OO%)and only four(44%)ofninesurvivorshad an unrestrictiveventricularseptaldefect(p< 0.05).

Of the sixnonsurvivors,three had a markeddiscrepancybetweenAWI andventricularcavityratio (lowAWI withaventricularcavity ratio equal to the mean value for thebalancedgroup)(Table3, Patients11,12and 13).

One survivorof two-ventriclerepairhas developedcoarc-tationpostoperativelyrequiringreoperationand has acquiredleft-sidedpulmona~veno-occlusivedisease(Table3, Patient4). Anotherpatient underwentsuccessfulvalvereplacementfor severepostoperativeAV valveinsufficiency(Patient6).The other sevenare doingwellat follow-up.

DiscussionSurgicaloutcomefor patientswithunbalancedCAVChas

been extremelypoor,withmortalityrates in somereportsashighas57%to 100%(l-3). Two-ventriclerepairbyclosureofthe atrial and ventricularseptaldefectsmaybe successfullyachievedin somecases;however,in thosewithmarkedunbal-ance of the AV valve,alternate strategiesare needed. Im-provedsurvivalfor patientswith hypoplasticIeft heart syn-drome and other forms of aortic outflowobstructionbyNorwoodpalliation(7) followedby hemi-Fontanand Fontanoperationshas allowedfor the successfulapplicationof thisapproachin patientswith markedlyunbalancedCAVCandassociatedleft ventricularhypoplasia.However,in cases ofmoderateunbalanceof the AVvalve,decidingwhichsurgicalstrategyto usemaybedifficult.Performanceof a two-ventriclerepairinpatientswithextensiveunbalanceofthecommonAVvalvemayresultin significantmortality.Predictivecriteriafor

JACCVol.28,No.4October1996:1017-23

COHENETAL. 1021UNBALANCEDATRIOVENTRICULARCANAL

Table3. Demographics,SurgicalProcedure and Outcome in Group IIB (patients with unbalanced common atnoventricular canal andtwo-ventriclerepair)

Ageat AVVOperation Down Large Unbal Irrsuffic

Pt No. (me) Synd VSD CAVC AVV1 VCR (preop) Operation PostopCourse

Survivors1234

31.5

112.5

No NoYes YesNo YesNo No

RightRightRightRight

0.520.430.440.30

0.590.390.440.30

MildMildMildMild

MildModerateto

severeTrivialMildMild

Moderate

Moderate

Moderate

MildModerate

Mild

Single-patchrepairSingle-patchrepairSingle-patchrepairSingle-patchrepair

Single-patchrepairSingle-patchrepair

Single-patchrepairTwo-patchrepairSingle-patchrepair

Two-patchrepair

Two-patchrepair,COArepair

PABthensingle-patchrepair

Single-patchrepairSingle-patchrepair

Single-patchrepair

MildMR,doingwellDoingwellMildTR,doingwellLeftpulmonaryvein

stenosis,COArepairModerateTR,doingwellMitralvalvereplacement,

pacemakerDoingwellModerateMR,doingwellDoingwell

Completeheartblock,diedin 1st24h

Postopdeathwithin24h

Diedat home,1wkpostoperatively

Postopdeathwithin24hSevereMR,died2wk

postopPostopdeathat 5 mo

56

5.58

No NoNo No

RightRight

0.550.65

0.260.59

789

Nonsurvivors10

32

11

Yes YesYes YesYes No

RightLeftRight

0.580.570.60

0.510.570.61

6 No Yes Right 0.57 0.49

Right11 0.7 Yes Yes 0.27 0.77

12 Yes Yes Left 0.56 0.951

1314

311

Yes YesYes Yes

LeftRight

0.620.39

0.870.43

15 5 Yes Yes Right 0.53 0.34

COA= coarctationoftheaorta;MR= mitralregurgitation;PAB= pulmonaryarteryband;Postop= postoperative;preop= preoperative;Pt = patient;TR =tricmpidregurgitation;otherabbreviationsask Tabjesl and2. “

determiningtheviabilityoffunctionoftheleft-sidedstructuresafter two-ventriclerepairof thislesionare lacking.

Echocardiographic morphometry. In thepresentstudy,weinvestigatedan echocardiographicallydefined, quantifiablevariableof unbalanceof the commonAV valve(AVVI)andanalyzedthe relationof this variableto relativeventricularcavitysize,surgicalreferraland outcome.In all 77 patientswithanAWI >0.67,diagnosticcategorizationbymorphomet-ricanalysisconcurredwiththepreviousdiagnosisofa balancedCAVCbyangiographyor subjectiveechocardiographicassess-ment,whereasall26patientswithAWI <0.67concurredwiththe previousdiagnosisof an unbalancedCAVC.In addition,morphometricmeasurementswerehelpfulin furtherclassi~-

ingpatientsbysurgicalreferral.Therewasa lowerlimitof theAWI (0.27)at whichallpatientshad ductal-dependentcircu-lation and henceunderwentNorwoodoperation(sevenpa-tients).FourpatientswithanAVVIbetween0.27and0.40hadductal-dependentcirculationandalsounderwenttheNorwoodoperation.All remainingpatientsunderwenta two-ventriclerepair.

TheAWI predictedrelativeventricularcavitysizein mostbutnotallpatients(Fig.2).SomepatientswithawellbalancedCAVC(AWI 0.9to 1.19)had a largerightventricularcavityrelativeto the left (ventricularcavityratio 0.7 to 0.8).Thisfindinghas been previouslyreported(9) and maybe due topredominanceof atrial septal defect physiologywith right

8, :’1’1‘: 1AVVI >0.67 AVVI <0.67

Balanced Unbalanced

Two-Ventricle Repair Non-Ductal Dependent Circulation Ductal Dependent Circulation(AVVI > 0,27) (AVVI < 0.27)

Figure3. Algorithm for stratitjingpatients with unbalanced CAVCforsurgicalrepair,usingtheAWI.VSD= ventricularseptaldefect.

==-

.-..—-——.——.——

1022 COHENETAL. JACCVol.28,No.4UNBALANCEDATRIOVENTRICULARCANAL October1996:1017-23

ventricularvolumeoverload.Conversely,three patientswerenotedto havemarkedunbalanceoftheAVvalve(AVVI0.3to0.5)witha relativelynormalventricularcavityratio(0.7to 0.9).Ventricularsizeassessmentalonewouldhavemisjudgedthedegree of commonAV valve unbalancepresent in thesepatients.Thisfindingsuggeststhat AVvalveinflowis not thesoledeterminantof ventricularcavitysize.

Unbalanced CAVC:survival after two-ventriclerepair. Fif-teen patientswith unbalancedCAVC(AWI <0.67)under-wenttwo-ventriclerepair(Table3), Of the ninesurvivors,thespectrumofunbalancewasrelativelywide(AVVI0.30to 0.65).Neither degreeof unbalanceof the commonAV valvenorrelativesize of the ventricleinfluencedsurvivalwithinthisgroup. Of note, the patient with the smallestAVVI andventricularcavityratio who underwenttwo-ventriclerepairsurvivedthe operationand is presentlyalive.He has subse-quentlyundergonerepairof coarctationof the aorta and hasdevelopedleft pulmonaryveno-occlusivedisease(Patient4,Table3).

HemodynamicallysignificantpreoperativeAVvalveinsuf-ficiencywasmoreprevalentin nonsurvivors;however,statisti-calsignificancewasnot reached(p = 0.08).Of interest,allsixnonsurvivorshad a largeunrestrictiveventricularseptaldefect(p< 0.05).BothsignificantAVvalveinsufficiencyand a largeventricularseptaldefectresult in a volumeload on the leftventricle.It is plausiblethat thisvolumeload may alter thedimensionsof the left ventriclebefore repair and result inspuriouslygreatermorphometricvaluesthanwouldexistwerethese factorsnot present.This phenomenonwas evidentinthree patients (Patients 11, 12 and 13, Table 3) who hadmarked unbalanceof the commonAV valve and normalventricularcavitysize.All three had unrestrictiveventricularseptaldefects,and twoof the three had significantAV valveinsufficiencybefore operation.All three expiredsoon aftertwo-ventriclerepair.

Althoughdimensionalanalysiswasthe focusof our inves-tigation,functionalabnormalitiesinfluencingoutcomesshouldbe consideredas well in patientswith significantAV valveinsufficiencyor an unrestrictiveventricularseptaldefect,orboth. Reducedafterload,such as existswith significantAVvalveinsufficiency,may make the left ventricleof an unbal-ancedCAVCwithpoorventricularfunctionappearto contractwell.In addition,a largeventricularseptaldefectmayactas aconduitfor systemiccirculatorysupportby allowingright toleft shuntingfrom the rightventricledirectlyinto the aorta.Onewouldexpectthesepatientsto havemildaorticdesatura-tion. We reviewedthe cardiac catheterizationdata for allpatientswho underwenttwo-ventriclerepair (12 of 15 hadpreoperativecardiaccatheterizationdata availablefor review)andfoundthat 5 of 12(2 survivors,3 nonsurvivors)had aorticdesaturation.It isdifficultto determinethe siteof the righttoleftshuntin patientswitha largeatrialseptaldefect.Withthepotentialforeitherpulmona~venousdesaturation,atriallevelshuntingor ventricularlevelshuntingpresent,in additiontostreamingeffects,interpretationof aorticdesaturationdata isdifficultandwasnothelpfulin distinguishingbetweenpatients

in our study.Nevertheless,it appearsquitelikelythat givenasmallbut adequateatrioventricularinflow,a poorlyfunction-ingventriclemaysupplygoodsystemicperfusionwiththe aidof rightventricularsupportacrossa ventricularseptaldefect.Onlyafterclosureof the defectwouldleftventriculardysfunc-tion and the potential inabilityto support systemicoutputbecomeunmasked.Morphometricanalysismustthereforebeinterpretedin conjunctionwithan assessmentofvalveinsuffi-ciencyand the presence of a ventricularseptal defect toappropriatelyreferpatientsfor operation.

Limitations of the study. Volumetric estimates of ventric-ular cavity size were not performed. Volume measurementssuch as the Simpson’smethod have not been proved to beaccurate for the right ventricle and are probably not accuratefor the left ventricle in patients with CAVC because the shapeof the left ventricle in this disease is different from that of thenormal ventricle. Linear dimensions in the single-plane viewwere readily available in all cases and were used to estimaterelative cavitydimensionrather than volume between the rightand left ventricle in each patient. These measurements are alsoreadily reproducible.

The number of patients with unbalanced CAVC was rela-tivelysmall but was proportionately greater within our cohort(25%) than previouslyreported (10%) (l). This findingmayberelated to selectionbias because our center is a referral site forsingle-ventriclepalliation by Norwood operation. Of note, nopatient without a patent ductus arteriosus and right to left flowsupplyingsystemiccirculation underwent Norwood operation.There were only three patients in the studywith unbalance tothe left; thus not many conclusions can be drawn about thisentity. The physiologymay ve~ well be different in thesepatients because the smallventricular chamber is nonsystemic;however,the problem of relative AV valvestenosis after repairas a cause of morbidity or mortality, or both, would exist inright and left dominant ventricles.

Summary. We suggestthat in patients with suspectedunbalancedCAVC, the AWI should be calculatedusingtwo-dimensionalechocardiographyin the subcostalleft ante-rior obliqueview.As a supplementto othermeansof evalua-tion,we believethat the algorithmshownin Figure3 shouldform the basis for appropriatestratificationof patientsforoperation.Becausethisdiseasehas a highmortalityrate, theFontanoperation,whichhasa lowmortalityratein mostexperi-encedinstitutions,maybe abetter optionforsomepatientswithunbalancedCAVC.A prospectiveapplicationof morphometricanalysisin unbalancedCAVCis warrantedand shouldaid inimprovingsurgicaloutcomeforthiscomplexlesion.

References1. ComoA,MarinoB,CatenaG,MarcellettiC.Atrioventricularseptaldefects

withsevereleftventricularhypoplasia.J ThoracCardiovascSurg1988;96:249-52.

2. MehtaS,HirschfeldS,RiggsT,LiebmanJ. Echocardiographicestimationofventricularhypoplasiain completeatrioventricularcanal.Circulation1979;59:888-93.

3. BharatiS,LevM.Thespectmmofcommonatrioventricrdarorifice(canal).ArnHeartJ 1973;86:553-61.

JACCVol.28,No.4October 1996:1017-23

4. MairDD,McGoonDC.Surgicalcorrectionofatrioventricularcanalduringthe firstyearoflife.AmJCardiol1977;40:66-9.

5. ChinAJ, KeaneJF, NorwoodW~ CastenadaAR. Repairof completecommonatrioventricularcanalin infarrcy.JThorac CardiovascSurg1982;84:437-45.

6. StuderM,BlackstoneEH,KirklinJW,et al.Determinantsofearlyandlateresultsof repairof atrioventricularseptal(canal)defects.J TfroracCardio-vascSurg1982;84:523-42.

7. NorwoodWI.Hypoplasticleftheart syndrome.AnnThoracSurg1991;52:688-95.

8. ThiesWR,BargeronLM,BiniRM,ColvinEV,SotoB.Spectrumofheartswith one underdevelopedand one dominantventricle.PediatrCardiol1986;7:129-39.

9. llranopaulosBD, FisherEA, DuBrowIW, HastreiterAR. Ventricularvolumesin commonatrioventricularcanal.Circulation1978;57:991-5.

COHENETAL. 1023UNBALANCEDATRIOVENTRICULARCANAL

10.Esuirrosa-CalianiJS,Alvarez-GuisadoL,Munoz-CastellanosL,et al.Atrio-ventricularseptaldefect:quantitativeanatomyoftherightventricle.PediatrCardiol1991;12:206-13.

11.JarmakarriJM,GeorgeB,WhellerJ. Ventricularvolumecharacteristicsininfantsandchifdrenwithendoeardialcushiondefects.Circulation1978;58:153-7.

12.Wu YT, ChangAC, Chin AJ. Semiquantitativeassessmentof mitralregurgitationbyDopplercolorflowimagingin patientsaged<20 years.AmJ Cardiol1993;71:727-32.

13.ChinAJ, YeagerSB, SandersSP, et al. Accuracyof prospectivetwo-dimensionalechocardiographicevaluationofleftventricularoutflowtractincompletetranspositionof thegreatarteries.AmJ Cardiol1985;55:759-64.

14.SilvermanN.Atrioventricularseptaldefects(atrioventricularcanaldefects).In:PediatricEchocardiography.Baltimore,MD:Williams&Wilkins,1993:143-66.