cross echocardiographicall patients recognition of the criss cross had been assisted either byright...

Br Heart J 1985; 54: 61-7

Cross sectional echocardiographic andangiocardiographic correlation in criss cross heartsPETER J ROBINSON, VACHERA KUMPENG, FERGUS J MACARTNEY

From the Hospital for Sick Children, London

SUMMARY Cross sectional echocardiography can provide accurate anatomical diagnosis in congenitalheart diseases and therefore should be able reliably to identify criss cross hearts and enable theanalysis of their sequential arrangement non-invasively. The cross sectional echocardiographic diag-noses in eight consecutive patients with this condition were compared with those made at cardiaccatheterisation and cineangiocardiography (five retrospectively, three prospectively). The mean

number of invasive studies required to reach the diagnosis was 1-9 (range 1-4). Complete anatomicaldiagnosis was achieved with cross sectional echocardiography in all patients, but identification ofventricular morphology was much more straightforward using cineangiocardiography. If the trans-ducer was held steady in either a precordial or subcostal positiou and rocked anteriorly and pos-

teriorly the characteristic crossing over of the ventricular inflows tould easily be seen. In no planewas there normal parallel arrangement of ventricular inflows.A complete diagnosis should be possible in these patients using cross sectional echocardiography

in experienced hands and at a single session in the cardiac catheterisation laboratory.

Criss cross heart is a rare rotational abormality of theheart in which systemic and pulmonary venousstreams cross without mixing at atrioventricular valvelevel. I As this definition suggests, it is an abnormalityof relation rather than connection2 but is often associ-ated with abnormalities of connection, and usuallyaccompanies other complex intracardiac anoma-lies.3-8 The ventricles are most often arranged in arelatively superoinferior fashion with the right sidedatrium connecting with the left sided ventricle, andthe left sided atrium with the right sided ventricle.Establishment of the precise type and mode ofatrioventricular connection may not be achievedunless multiple angiocardiograms, with varying pro-jections and sites of injection, are used. Thus repeatedcardiac catheterisations may be required before theprecise diagnosis is established.4A single case report of the cross sectional echocar-

diographic diagnosis of criss cross heart has recentlybeen published.9 To assess the relative roles of crosssectional echocardiography and angiocardiography inthis condition we reviewed the cases seen at this hos-pital since 1980 and compared our cross sectional

Requests for reprint to Professor F J Macartney, The Hospital forSick Children, Great Ormond Street, London WC1N 3JH.

Accepted for publication 5 March 1985

echocardiographic diagnoses with the findings at car-diac catheterisation and angiocardiography.

Patients and methods

Eight patients with this condition have been seensince January 1980 (four male, four female) aged 6weeks to 12 years (median 3-5 years) at initial presen-tation. Cross sectional echocardiography, cardiaccatheterisation, and cineangiocardiographic recordswere retrospectively reviewed in five patients. Threefurther patients were examined prospectively duringthe time of the study using our standard echocardio-graphic protocol.10

Cardiac catheterisation was performed using stan-dard techniques with 1-4 (mean 1.9) studies requiredper patient before a definitive diagnosis was made. Inall patients recognition of the criss cross had beenassisted either by right atrial or caval injections or byfollow through opacification of the left atrium andventricle. Cross sectional echocardiographic studieswere performed with an ATL (Advanced TechnologyLaboratory) Mark V mechanical sector scanner and a3 or 5 MHz transducer in four patients, a ToshibaSSH 20 system with a 2-5 MHz transducer in onepatient, and an ADR 4000S mechanical sector scannerwith a 3 MHz transducer in the last three patients.The resulting videotapes were reviewed. Echocar-

61

on August 27, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.54.1.61 on 1 July 1985. Dow

nloaded from

http://heart.bmj.com/

Table Invasive and non-invasivefindings

Case Age Sex Cross sectional echocardiographyNo

Situs AV RAVV LAVV VSD RVOTO VA Gt vesselconnection connecuion posion

1 5 mnth M Solitus C N N Perimembranous Valvar, subvalvar D Ao anterior aninlet outlet

2 4 yr F Solitus C N S Perimembranous - C Ao anterior aninlet outlet

3 5 yr M Solitus C N N Trabecular inlet Subvalvar D Side by side, IL

4 6 wk F Solitus C N S Perimembranous - D Ao anterior aninlet

5 1 yr M Solitus C N N Perimembranous - D Ao anterior aninlet outlet

6 3 yr F Solitus D N N Perimembranous - DORV Ao anterior adinlet outlet

7 12 yr F Solitus C N N Perimembranous - C Ao anterior anl8 10 yr M Solitus C N S Perimembranous - C Ao posterior al

inlet outlet

-, absent; Ao, aorta; AV, atrioventricular; C, concordance; D, discordance; DORV, double outlet right ventricle; Gt vessel position, greatposition in relation to each other; L, left; LAVV, left atrioventricular valve; N, norual; 0, override; R, right; RAW, right atrioventricularRVOTO, right ventricular outflow tract obstruction; S, stenotic; VA, ventriculoarterial; VSD, ventricular septal defect.

diographic and angiocardiographic opinions werereached by consensus between us. Two patients died;in one the heart was available for detailed necropsy.

Results (Table)

CROSS SECTIONAL ECHOCARDIOGRAPHYEach patient had situs solitus of the atria as shown bya left sided aorta and right sided inferior vena cavadraining to the right atrium in subcostal cuts.11 Con-

cordant atrioventricular connections were present inseven patients and discordant atrioventricular connec-tions in one. In each case it was impossible to find aplane showing simultaneous and parallel arrangementof the atrioventricular valves and ventricular inflowregions, as described by van Mill et al.9 A moredefinitive criterion, in our opinion, was the directdemonstration of a criss cross atrioventricular rela-tion. This could be obtained by positioning thetransducer in such a way as to demonstrate from

Fig. 1 COoss sseionaledwavdiogramfirn the apcx (a) show cmeczion ofdt nghtatni (RA) to dthightvmetick (RV) uithdte aortic root (Ao) cut in its sht axis. The right (RV) and left vetides (LV) re in superoifenorTrangent and the trisidvale has a bightechio on dte septal lealt (see text). (b) shows,fromn dte same rasduferfp tion at the apex but wnth more postwrianuidaion, aputnoy vein (pv) raini to dte left atrvin (LA) wuich is connctd to dt left vi1ice (LV). The miral vale (my)opens aoos theplae of ight ventiadar inflow (see (a)). Note dte woderator band (mb) in de rightvemic. I, infenor; L, left; R,ight; S, sqptrior; di, chod e.

62 .Robinson, Kumpeng, Macartney



j.com/

Br H


nloaded from


Echocardiography in criss cross hearts

diac cadtetensation .

s AV RAVV LAVV VSD RVOTO VA Gt vesselconnecton concordance position

tus C N N Perimembranous Valvar, subvalvar D Ao anterior and L

tus C N S Perimembranous - C Ao anterior and L

tus C N N Perimembranous Valvar, subvalvar D Side by side, Ao to L

tus C N S Perimembranous - D Ao anterior and L

tus C N N Perimembranous - D Ao anterior and L

tus D N N Perimembranous outlet - ?DORV Ao anterior and L

tus C 0 N Perimembranous - C Ao anterior and L[us C N S Perimembranous outlet - C Ao posterior and R

below each atrium crossing to the contralateral ven-tricle, with one atrioventricular connection crossingbehind the other (Figs. 1 and 2). The precise trans-ducer position required varied from case to case inthese highly complex hearts but was usually from theapex or subcostal region. One patient had a straddlingtricuspid valve, but no patient had mitral straddling.Three patients, all with concordant atrioventricularconnections, had cross sectional echocardiographicevidence of stenosis of the mitral valve (Fig. 3), andone showed a mobile bright echo on the septal leaflet

_Q a1

Fig. 2 Cros secnl eccardiogrmfwn dse sametransduerposiion h more anerior in (a) andmorepostro anuauon in (b) showingdc crossam angenet ofveniclr inflow region. hv, hepatc vein; I, inferior; ivc,infeior vena cava; L, left; LA, left amrn; my, nntral vave;R,nght; RA, ightamwn; RV, ight wvide; S, penior; tv,tiaspid valve; vsd,vviar sepal defect.

Fig. 3 Cross sectioal echocardiogram from the apex showingdoming ofthe stenosed mitral valve (mv) in a patient with a crisscross heart. I, inferior; L, kft, LA, left atrium, laa, kft atrialappendage; my, mitral valve; R, right, RV, right ventricle; S,supenor; vsd, ventricular septal defect.

63



j.com/

Br H


nloaded from


64

Fig. 4 Cross sectioal echocardiogram showing the relativepositions of the right (RV) and left ventricles (LV) with leftventricle inferior and slighdy rightward of the right venticle.Note thefishmouth appearance of the mitral valve (mv) andseptal attachment of the ricuid valve (tv). I, inferior; L, left;R, right, S, superior; vsd, venticular septal defect.

Robinson, Kumpeng, Macartney

of the tricuspid valve (Fig. 4). This echo was verysimilar to one seen in another patient that proved atoperation to be ruptured chordae tendineae that hadcoiled back on themselves..Because we were uncertain of the precise trans-

ducer position used in retrospective studies the crosssectional echocardiographic determination of relativeventricular position was usually not possible retro-spectively. In the three patients studied prospectivelybefore angiocardiography (Fig. 5), however, the ven-tricular positions were correctly predicted as rightventricle slightly anterior, superior, and leftward andleft ventride slightly posterior, inferior, and right-ward.As all our patients had ventricular septal defects

that extended to the inlet portion of the septum wecould not use offsetting of the atrioventricular valvesto identify ventricular morphology. We thereforeused either the moderator band as a marker of rightventricle or the fishmouth appearance of the mitralvalve in short axis cuts to identify a left ventricle (Figs.lb and 4). The ventricular septal defect extendedthrough the perimembranous region to the muscularoutlet region in five patients and was subaortic in two.In one patient tricuspid tissue protruded through the

Figs. 5(a) and (b) Long axis cross sectional echocdiogram showing spid valme (tc) tisse prolapsmg into dte righ eicLe(RV) m systole and ito the left vmriaar (LV) ouflow tract in diastole. A, anterior; Ao, aorticroot; I, infeior; LA, left atvenwmv, mitral vahve; P, postenor, S, supaeor.



j.com/

Br H


nloaded from


Echocardiography in criss cross hearts

PT AO

vsd

LV

65

RA

LV

LA

Figs. 6(a) and (b) Antrpostri cineangwogramfimn an

mijectom mto the left ventcl (LV) with follow through leftatial (LA) opactfication m a patnthwi cnss hea,

avno dar con ce, and venvicul discordance.

Ao, aorta; I, inferior; L, left; PT, pubnonary mnk; R, ight;RV, right wenicle; S, supeior; vsd, ventriular septal defect.

ventricular septal defect to the left ventricular outflowtract in diastole and into the right ventricle in systole(Fig. 5). No patient had subaortic obstruction. Twohad subpulmonary obstruction due to fibromusculartunnel formation, one with associated valvar stenosis.The origin of the great arteries from the ventricles

Fig. 7 Cineogranfromn thesamepant as in Figs. 6(a)and (b), wh injeton into the right arwrn. The projectio isanteroposterior and cmnpasiso with Figure 6(b) shows the crisscross arrangenn of ventricuar inflow regions. Ao, aorta; I,inferior; L, left; LV, left venticle; R, right; RA, right atiun;RV, night ventncle; S, supnor.

was easily recognised by echocardiography. Therewere concordant ventriculoarterial connections inthree patients, two with the aorta anterior and slightlyleftward of the pulmonary trunk and one with theusual great vessel relation. One patient had doubleoutlet right ventricle with the aorta anterior andslightly to the left of the pulmonary trunk, and fourpatients had discordant ventriculoarterial connec-tions, three with aorta anterior and to the left of thepulmonary trunk and one with side by side great ves-sels with aorta to the left. In the three patients studiedprospectively the aortic arch was left sided and a per-sistent ductus arteriosus was excluded.

HAEMODYNAMICS AND ANGIOCARDIOGRAPHY(Figs. 6 and 7)Atrial situs, atrioventricular and ventriculoarterialconnections were confirmed by angiocardiography inall patients. (Cineangiocardiography showed the crisscross arrangement when appropriate sites of injectionwere used.) Pressure measurements confirmed thepresence of mitral stenosis in the correct threepatients and excluded atrioventricular valve stenosisin the others. Subpulmonary stenosis was confirmedby pressure measurements in one patient andangiocardiography in the other. Though there wereno mistakes with echocardiography, ventricular mor-phology was more obvious on angiocardiograms thanon cross sectional echocardiograms. Precise localisa-

_:s

... Ao.z

4:

..:I,v

:,.:M



j.com/

Br H


nloaded from


66

tion of the ventricular septal defect was not alwayspossible with certainty angiocardiographicallybecause of problems in profiling the plane of theinterventricular septum.'012 Likewise atrioventricu-lar valve straddling was better appreciated echocar-diographically.13 The aortic arch was left sided and apersistent ductus arteriosus was excluded in allpatients.

NECROPSY FINDINGSNecropsy confirmed the premortem diagnoses andfound no new abnormalities.

Discussion

This study shows how the systematic cross sectionalechocardiographic examination of hearts with crisscross arrangement of their inflow portions can provideaccurate anatomical information and highlights howwell suited this technique is to -the segmentalapproach in the diagnosis of complex congenital heartdisease. In criss cross heart both ventricles and greatvessels may be in unexpected positions, but a carefulsystematic approach will demonstrate the connec-tions, particularly if continuous sweeps betweenroutinely used cross sectional echocardiographic cutsare used. This was exemplified in our study particu-larly in one patient in whom the aorta was almostdirectly anterior to the pulmonary trunk and yet thegreat vessels connected in concordant fashion.

There has previously been only one case reportshowing the ability of cross sectional echocardiogra-phy to make a nearly complete anatomical diagnosis inthis condition. Those authors used subcostal cutsalone, parasternal cuts having been unsuccessful.9While we agree that much of the anatomical informa-tion may be obtained from these cuts, we found thatthe precordial views were useful in confirming theimpressions gained from the subcostal cuts. Further-more, precordial cuts were superior in identifyingsubpulmonary obstruction, great vessel position, andin some cases the actual recognition of the criss crossarrangement of the atrioventricular junction. Afurther disadvantage of relying on subcostal cuts isthat they may not provide a plane showing opening ofboth atrioventricular valves. If the transducer is notangled sufficiently either anteriorly or posteriorly afalse impression of either double inlet ventricle orabsent atrioventricular connection may be given, as inthe initial echocardiographic diagnosis of the patientstudied by van Mill and colleagues.9 Although weagree that the normal parallel arrangement of the rightventricular and left ventricular inflow tract is not seenin this condition, we have shown how the actualdemonstration of the criss cross arrangement is poss-ible echocardiographically.

Robinson, Kumpeng, Macartney

In understanding the underlying anatomy of heartswith normal and criss cross atrioventricular relationsit is helpful to think of atrioventricular.channels run-ning from each atrium to the inlet and apex of theventricle to which it is connected. In normal heartsthese two channels are roughly parallel, with theresult that it is possible to obtain a four chamber sec-tion of the heart in which each atrium is connected tothe ipsilateral ventricle and the atrioventricular valvesopen in the same direction. By contrast, in criss crosshearts the atrioventricular channels cross one anotherwith the result that the atrioventricular valves open indifferent directions, and views can be obtained inwhich one or other atrium connects to the contralat-eral ventricle as seen in the echocardiographic section.This may well not correspond to the conventionalfrontal, sagittal, or horizontal plane of the body.When the atrioventricular connection is biventricular,one atrioventricular channel must cross in front of theother. This means that it is impossible to demonstratethe entire criss cross in a single tomographic section.Thus cross sectional echocardiography performed

systematically and following a preset routine byexperienced operators should allow complete sequen-tial anatomical diagnosis of these complicated heartsnon-invasively. Cardiac catheterisation can be plan-ned much more precisely and should be needed onlyfor gaining pressure and oximetric data and to rule outadditional ventricular septal defects. A completeanatomical and physiological diagnosis should now bepossible in these patients using cross sectionalechocardiography and a single session in the cardiaccatheterisation laboratory.

PJR was supported by the Wellcome Foundation andFJM is supported by the Vandervell and British HeartFoundations.

References

1 Anderson RH. Criss cross hearts revisited. Pediatr Car-diol 1982; 3: 305-13.

2 Tynan MJ, Becker AE, Macartney FJ, Quero-JimenezM, Shinebourne EA, Anderson RH. Nomenclature andclassification of congenital heart disease. BrHeartJ 1979;41: 544-53.

3 Freedom RM. Superoinferior ventricle and criss crossatrioventricular connections: an analysis of the myth andmystery. Modern Problems in Paediatrics 1983; 22: 48-62.

4 Freedom RM, Culham G, Rowe RD. The criss-cross andsuperoinferior ventricular heart: an angiocardiographicstudy. Am J7 Cardiol 1978; 42: 620-8.

5 Sato K, Ohara S, Tsukaguchi I, et al. A criss-cross heartwith concordant atrioventricular-arterial connections.Report of a case. Circulation 1978; 57: 396-400.

6 Schneeweiss A, Shem-Tov A, Blieden LC, Deutsch V,Neufeld HN. Criss-cross heart- A case with horizontalseptum, complete transposition, pulmonary atresia and



j.com/

Br H


nloaded from


Echocardiography in criss cross heartsventricular septal defect. Pediair Cardiol 1982; 3: 325-8.

7 Marino B, Chiariello L, Bosman C, et al. Criss-crossheart with discordant atrioventricular connections.Pediat Cardiol 1982; 3: 315-8.

8 Attie F, Muftoz-Castellanos L, Ovseyevitz J, et al. Cros-sed atrioventricular connections. Am Heart J 1980; 99:163-72.

9 van Mill G, Moulaert A, Harinck E, Wenink A,Oppenheimer-Dekker A. Subcostal two-dimensionalechocardiographic recognition of a criss-cross heart withdiscordant ventriculoarterial connection. Pediatr Cardiol1982; 3: 319-23.

10 Macartney FJ, Smallhorn JF. Cross-sectional echocar-diography in neonates and children. In: Rowlands D, ed.

67

Recent advances in cardiology-9. Edinburgh: ChurchillLivingstone, 1984: 27-75.

11 Huhta JC, Smalihorn JF, Macartney FJ. Two dimen-sional echocardiographic diagnosis of situs. Br Heart J1982; 48: 97-108.

12 Bargeron LM Jr, Elliott LP, Soto B, Bream PR, CurryGC. Axial cineangiography in congenital heart disease.Section I. Concept, technical and anatomic considera-tions. Circulation 1977; 56: 1075-83.

13 Smallhorn JF, Tommasini G, Macartney FJ. Detectionand assessment of straddling and overriding atrioven-tricular valves by two dimensional echocardiography. BrHeartJ 1981; 46: 254-62.



j.com/

Br H


nloaded from


cross echocardiographicall patients recognition of the criss cross had been assisted either byright...

Documents