the morphologic variability in atrioventricular valvar atresia

Cardiology in the Younghttp://journals.cambridge.org/CTY

Additional services for Cardiology in the Young:

Email alerts: Click hereSubscriptions: Click hereCommercial reprints: Click hereTerms of use : Click here

The morphologic variability in atrioventricular valvar atresia

Robert H. Anderson, Siew Yen Ho and Michael L. Rigby

Cardiology in the Young / Volume 10 / Issue 01 / January 2000, pp 32 41DOI: 10.1017/S1047951100006351, Published online: 19 August 2008

Link to this article: http://journals.cambridge.org/abstract_S1047951100006351

How to cite this article:Robert H. Anderson, Siew Yen Ho and Michael L. Rigby (2000). The morphologic variability in atrioventricular valvar atresia. Cardiology in the Young,10, pp 3241 doi:10.1017/S1047951100006351

Request Permissions : Click here

Downloaded from http://journals.cambridge.org/CTY, IP address: 144.82.107.39 on 23 Aug 2012

Cardiol Young 2000; 10: 32-41© Greenwich Medical Media Ltd.

ISSN 1047-9511

Continuing Medical Education

The morphologic variability in atrioventricular valvar atresia

Robert H. Anderson,* Siew Yen Ho and Michael L. Rigby

*Cardiac Unit, Institute of Child Health, University College, London, UK; Paediatrics, National Heart & Lung Institute,Royal Brompton Campus, Imperial College School of Medicine, London, UK

THE ESSENCE OF HEARTS WITH ATRIOVENTRICULARvalvar atresia is that, of necessity, there iscomplete admixture of systemic and

pulmonary venous returns at atrial level, since oneof the atriums has no direct communication withthe ventricular mass. Traditionally, for the purposesof clinical diagnosis, such hearts are divided intothe categories of tricuspid and mitral atresia.Within the context of this division, the essence oftricuspid atresia is complete blockage of the directventricular return of the systemic venous pathways.Mitral atresia, when used in this fashion, invokesthe picture of the complete obstruction of directpulmonary venous return. The potential anatomicsubstrates which can produce these haemodynamicarrangements, however, are complicated. This isbecause, in some situations, blockage of directreturn of the systemic venous pathways can beproduced by an atretic valve which, rather thanbeing a tricuspid valve, has the morphology of themitral valve. Similarly, in other cases, it is atresia ofa morphologically tricuspid valve, rather than amitral valve, which can block completely the directventricular egress from the pulmonary venousatrium. There are then further complexitiesconcerning the underlying anatomic substrates.This is because, although an imperforate valvarmembrane has traditionally been considered theexemplar of valvar atresia, examination of spec-imens in the majority of cases with tricuspid atresiashows that the valve itself, along with itssupporting atrioventricular junction, has failedtotally to form during cardiac development (Fig.1). The morphologic feature underscoring thevalvar atresia is absence of the atrioventricular

Correspondence to: Prof R.H. Anderson, Cardiac Unit, Institute of ChildHealth, 30 Guilford Street, London WC1N 1EH, UK. Tel: 020 7905 2295;Fax: 020 7905 2324; E-mail: [email protected]

Accepted for publication 20 August 1999

connection (Fig. 2). Identification of the valveinvolved as being morphologically tricuspid, ormitral as the case may be, is based upon inferencesconcerning the presumed embryogenesis of theventricular mass. It is not possible to distinguishthe morphology of the valve which is missing onthe basis of the anatomy of the atretic atrium. Sucha situation is less than ideal, and gives appreciablepotential for disagreement.

The overall understanding of the nature of atri-oventricular valvar atresia is then clouded stillfurther by ongoing morphological controversies.Some stili seem reticent to accept the concept ofabsence of one atrioventricular connection as amorphological entity.1 Yet, as already demon-strated, it is the recognition of the nature of anabsent connection (Fig. 3), and its distinction fromthe imperforate valvar membrane (Fig. 4), which isthe key to appreciation of much of the anatomicheterogeneity to be found in patients with atri-oventricular valvar atresia.2 Another confoundingproblem, nonetheless, is very nature of theventricular mass in such patients. In the past, wehad argued that the heart itself in patients withclassical tricuspid atresia was "univentricular".3'4 Inmaking this assertion, we were unequivocallywrong. Although the heart is functionally univen-tricular in the commonest form of this lesion, theanatomic arrangement is that of a big left ventriclealong with a small rudimentary and incompleteright ventricle (Fig. 5). Our emphasis on theapparent morphological "univentricular" nature ofsuch hearts, therefore, was misplaced. Despite this,there can be no question but that the concept of afunctionally univentricular entity remains totallyvalid. Thus, the comparability between theventricular mass in hearts with atrioventricularvalvar atresia (Fig. 6), and that seen in patientswith double inlet left ventricle (Fig. 7), is obviousand well established.5 In this review, we discuss all

Vol. 10, No. 1 Anderson et al: Atrioventricular valvar atresia 33

Figure 1.This specimen with "classical" tricuspid atresia is seen from behind,having removed the parietal wall of the right atrium. The essence of themalformation is complete absence of the right atrioventricular junction.Note the presence of the artery at the acute margin delimiting the posteriorextent of the septum between the dominant left and the rudimentary rightventricle. The septum does not extend to the crux.

these morphological problems. We show how use ofa descriptive, rather than a nominative, approach tonomenclature is helpful in resolving the questionsconcerning the anatomic variability of patientswith atrioventricular valvar atresia.

Basic definitions of atrioventricular valvaratresia

"Atresia" is a greek word. When translated literally,it means blockage of a natural channel of the body.In the normal heart, there are two such channelsproviding access from the atriums to the ventricles.These are guarded by the tricuspid and mitralvalves, respectively. Thus, the normal heartpossesses paired atrioventricular junctions (Fig. 8).Atresia of either of these channels can then beproduced by one of two discrete morphologicalmechanisms. The commonest abnormality, irre-spective of whether it involves the morphologicallytricuspid or mitral valves, is to find completeabsence not only of the valve itself, but also the areanormally occupied by the muscular atrioventricularjunction. The floor of the afflicted atrium is thenexclusively muscular (Fig. 1), and is separated by thefibrofatty atrioventricular groove from the under-lying ventricular chambers (Fig. 3). Thisarrangement is well demonstrated by cross-

Figure 2.The same heart as shown in Fig. 1. The muscular floor of the right

atrium has been stripped back from the ventricular mass, revealing the

deep fibro-fatty atrioventricular groove which interposes between right

atrium and the ventricular mass.

Figure 3.This cut in four-chamber projection demonstrates absence of the rightatrioventricular connection, highlighted by the dots, in tricuspidatresia. Note that the "dimple"points to the subaortic outflow tract ofthe left ventricle, and is the atrioventricular component of themembranous septum.

34 Cardiology in the Young January 2000

Figure 4.27>ir specimen has pulmonary atresia with intact ventricular septum,and concordant atnoventricular connections. The tricuspid valve isimperforate, producing tricuspid atresia. Note the hypoplastic rightatrioventricular junction blocked by the imperforate valvar membrane.

Figure 6.This angiogram is from a patient with absent left atrioventricularconnection, and with the left atrium connected to a dominant leftventricle. The rudimentary right ventricle, supporting the aorta, isleft-sided. This is indicative of left hand ventricular topology. Madthe valve developed between left atrium and right ventricle, it wouldhave been of tricuspid morphology.

Figure 5.In this specimen with tricuspid atresia, a long axis section shows thedominant left ventricle and the rudimentary right ventricle. Only theleft atrium is connected to a ventricle (univentricular atrioventricularconnection).

Figure 7.In this patient, there is double inlet to the dominant left ventricle withrudimentary right ventricle in left-sided position and discordantventriculo-arterial connections. The picture is directly comparable tothat shown in Fig. 6.


sectional echocardiography (Fig. 9). The pattern,accurately described as absence of one atrioven-tricular connection, means that, of necessity, one ofthe ventricles lacks totally its inlet component. Itresults, therefore, in one variant6 of univentricularatrioventricular connection (Figs 3,5). Thisarrangement is to be contrasted with the second,much rarer, anatomic lesion producing valvaratresia, namely presence of an imperforate valvarmembrane (Fig. 4). An imperforate valve, perforce,must block a discrete atrioventricular junction.Such imperforate valves, therefore, can be found inthe setting of biventricular atrioventricular connec-tions (Fig. 4). The imperforate valve can also befound in the setting of double inlet ventricle, givinganother example of univentricular atrioventricularconnection (Fig. 10). The potential anatomic vari-ability for imperforate valves, therefore, is consid-erably greater than for absence of one or otheratrioventricular connection. It is the potentialcombinations of segmental connections which isresponsible for the anatomic heterogeneity.7

Sequential segmental analysis

The variability to be found in patients with atri-oventricular valvar atresia involves all segments ofthe heart. The different patterns at each level arethen pertinent to the most appropriate way ofdescribing the malformations. Since the basic effectof the valvar atresia is to block the return from oneor other atrium to the ventricular mass, thearrangement of the atrial chambers themselves isthe first important variant. For example, takingtricuspid atresia as a paradigm, the "classical"variant of this lesion blocks the direct return ofsystemic venous blood to the right ventricle.Typically, this produces atresia of the right-sidedatrioventricular junction. In patients with mirror-imaged atrial arrangement, however, it is the left-sided atrium which is abnormal when there istricuspid atresia. As we have already discussed, ifthe atrial chambers are arranged in usual fashion,but in a heart in which the ventricular mass ismirror-imaged, either absence or imperforatenessof the left-sided atrioventricular junction can justi-fiably be described as tricuspid atresia, but nowproducing blockage of the pulmonary venousreturn (Fig. 11). And, in this setting, theventricular mass (Fig. 6) is indistinguishable fromclassical double inlet left ventricle with left-sidedrudimentary right ventricle (Fig. 7). Properly todiagnose and describe these malformations,therefore, it is best to account for variabilityinitially at atrial level, then to dissect the structureof the atrioventricular junctions, and finally to

combine this analysis with similar review of thepotential variability at the ventriculo-arterial junc-tions, since the way the arterial trunks take originfrom the ventricular mass will also markedly affectthe clinical presentation.

Variability at atrial level

Most patients seen with atrioventricular valvaratresia will have usual atrial arrangement("solitus"). In this setting, atresia of the right-sidedatrioventricular junction, be it due to absence of theconnection (Fig. 3) or an imperforate valvarmembrane (Fig. 4), will block the direct return tothe ventricles of the blood from the systemicvenous tributaries. Left atrioventricular valvaratresia, in contrast, will block the direct egress ofthe pulmonary venous return. In most cases, butnot all, the right-sided atrioventricular valve isactually or potentially of tricuspid morphology,while the abnormal left atrioventricular valve is animperforate mitral valve, or would have been ofmitral morphology had it been formed. Thesetypical cases are traditionally described as tricuspidand mitral atresia, and are immediately recog-nisable. There is then considerable potential forcomplexity in the ventricular mass which, as wewill see, also requires full segmental analysis.Complexity at the atrial level, in contrast, is rare. Inthe infrequent cases with mirror-imaged atrialarrangement, the patterns, as anticipated, aresimply the reverse of the usual situation.Atrioventricular valvar atresia in the setting ofisomerism, however, can produce problems indescription. Thus, when the atrial appendages areisomeric, absence of an atrioventricular connectioncan produce comparable atrial anatomy on eitherthe right or left side of the heart. The haemody-namic problem will then depend primarily on theprecise venoatrial connections, which alwaysrequire full description in the setting of isomerism.

Perhaps the major problem in providingaccurate description at atrial level comes whenthe atrioventricular septum is deficient. This isbecause it can be very difficult to separate the rareentity of tricuspid atresia with straddling left atri-oventricular valve (Fig. 12) from atrioventricularseptal defect with common atrioventricularjunction and so-called "double outlet left atrium".Similarly, it can be difficult to determine whetherthere is mitral atresia with straddling right valve(Fig. 13, right), or atrioventricular septal defectwith "double outlet right atrium" (Fig. 13, left).As far as is possible, we use anatomic criterions tomake this distinction. Thus, we have argued thatthe paradigm of the commonest form of atrioven-


Figure 8.This simulated four-chamber sea ion of the normal heart reveals thepresence of discrete and separate atrioventricular (AV) junctions forthe right and left sides of the heart.

Figure 10.In this echocardiographic section, it can be seen that the right atri-oventricular connection is formed, but is blocked by an imperforatevalvar membrane (IMP). The left atrioventricular valve is open, andboth valves are exclusively connected to the dominant left ventricle.The section shows double inlet left ventricle with tricuspid atresiaproduced by an imperforate valvar membrane. The atrioventricularconnections are univentricular despite the presence of the imperforatevalve. Compare with Fig. 9.

Figure 9.The echocardiographic appearance of classical tricuspid atresia.Note the extensive fibro-fatty atrioventricular groove interposedbetween the floor of the right atrium and the rudimentary andincomplete right ventricle (arrowed). This is indicative of absenceof the right atrioventricular connection. Abbreviations: RA — rightatrium; LA — left atrium; RV — right ventricle; LV — leftventricle; LT — open left atrioventricular valve. The abbreviationsare also used for subsequent echocardiograms.

Figure 11.This specimen shows double inlet left ventricle with an imperforate leftatrioventricular (AV) valve. The rudimentary right ventricle is left-sided, indicating left-hand ventricular topology, and suggesting thatthe imperforate valve is of tricuspid morphology.

Vol. 10, No. 1 Anderson et ah Atrioventricular valvar atresia 37

tricular septal defect is the common atrioven-tricular junction.8 This feature, namely thecommon junction, is lacking when one atrioven-tricular connection is absent (Fig. 12). On thisbasis, most cases can readily be assigned to thecategories of absent connection or commonjunction. When one connection is absent, thestraddling and overriding valve will perforce be ofright or left morphology, and not common. Theatrioventricular connection will accurately bedescribed as uniatrial but biventricular. Incontrast, when there is a common junction, theatrioventricular connections will be biatrial,albeit draining predominantly from one or otheratrium. But it is also well recognised thatcongenital heart malformations represent spec-trums of abnormality. Thus, when an "ostiumprimum" defect is recognised as providing theexit from one atrium (Fig. 14), then the otheratrium continues to drain to the ventricular massthrough a common atrioventricular junction. Ifthe "ostium primum" defect becomes increasinglysmall, at some point it could block off completelythe atrium draining through it. Would the strad-dling common valve extending into bothventricles be any different morphologically from astraddling right valve (Fig. 15) if the "ostiumprimum" in this spectrum closed spontaneously?Furthermore, it is easy for the morphologist toargue that, properly to define tricuspid atresia, itis necessary to find the atrial septum walling offthe right atrium from the left atrioventricularjunction (Fig. 12). It may be possible, however, tofind hearts in which the right atrioventricularconnection has failed to form when the left atri-oventricular junction is guarded by a trifoliate leftatrioventricular valve, and the ventricular massshows the typical morphology of an atrioven-tricular septal defect. In the presence of an"ostium primum" defect, this lesion would not fitthe anatomic definition proposed above for"tricuspid atresia". Yet the anatomy observedwould be exactly as expected for formation of halfof a common atrioventricular junction. Indeed,this arrangement has now been consistentlyrecognised9 in mice known to have deficient atri-oventricular septation (Fig. 16). It seems that,however hard we try to provide exclusiveanatomic definitions, the heterogeneitydiscovered in malformed human hearts defies ourbest efforts to achieve consistency. We mustrecognise the validity of various interpretations ofthe abnormal morphology. Careful and completedescription remains the best solution to these rarecombinations of malformations (see below).

Variability at the atrioventricular junctions

The basic variation found at the atrioventricularjunctions has already been described, namely thedifferentiation of absence of one atrioventricularconnection (Fig. 3) from an imperforate valvarmembrane (Fig. 4). This variability is itself inex-tricably linked with further variability in theventricular mass, but once more the heterogeneityis considerable. Completely to account for all thevariations in atrioventricular valvar atresia, it isnecessary to permutate the potentials for vari-ability at atrial level, within the atrioventricularjunction, within the ventricular mass, and at theventriculo-arterial junctions. In our opinion, it isno longer possible to account for all this variationsimply in terms of "tricuspid" and "mitral" atresia.Thus, at first sight it seems possible to categorisepatients with tricuspid atresia into those withuniventricular atrioventricular connection, whenone connection is absent, as opposed to those withbiventricular connections when the tricuspidvalve is formed but imperforate. This simpledivision fails when it is recognised that somepatients with double inlet left ventricle can have

Figure 12.In this specimen, sectioned in four-chamber fashion, there is absence of

the right atrioventricular connection, but the left atrioventricular

valve straddles so as to be connected in both ventricles. This produces

an atrioventricular connection which is uniatrial but biventricular.


Common AV valve Right AV valve (straddling)

Figure 13.This cartoon illustrates the potential spectrum which exists between,

on the one hand (left panel) a common atrioventricular junction and,

on the other hand (right panel), absence of the left atrioventricular

connection with straddling right atrioventricular valve. The

distinction between the two entities depends on whether the atrial

septum walls off the left atrium from the atrioventricular junction.

Figure 14.In this heart, shown from the right atrium, there is a common atrioven-

tricular (AV) valve equally shared between the right and left ventricles

(see Fig. 13, left panel). The only exit for blood from the left atrium is

through an "ostium primum" defect. Note that a persistent left superior

caval vein (LSCV) drains through an enlarged coronary sinus.

an imperforate atrioventricular valve (Fig. 10).Hence, the imperforate valve co-exists with theuniventricular connection. Furthermore, thisarrangement with double inlet and an imperforateright valve can be found when the incompleteright ventricle is either right-sided or left-sided.Would it help, when the ventricle is left-sided, todescribe the patient as having mitral atresia, whenthe imperforate right atrioventricular valveproduces the clinical picture of tricuspid atresia?Surely a full account of the segmentalarrangement is needed?

Figure 15.In this specimen, the atrial septum walls off the left atrium from theventricular mass (see Fig. 13, right panel). The atrioventricular(AV) valve, therefore, is a right valve, but it connects the right atriumto both the left and the right ventricles. It is difficult to distinguish themorphology from the common valve shown in Fig. 14.

Figure 16.This scanning electron microscope shows a heart from a mouse with

trisomy 16. The floor of the right atrium is muscular, and a trifoliate

left atrioventricular (AV) valve connects the left atrium to the left

ventricle. This heart had an "ostium primum" defect. Reproduced by

kind permission ofDr Sandra Webb and Professor Nigel Brown ofSt

George's Hospital Medical School.

The distinction of univentricular versus biven-tricular connections is also found wanting in thoseunusual patients considered above who have


unequivocal absence of one atrioventricularconnection, but with straddling of the solitary atri-oventricular valve (Fig. 12). Here again, it isessential to give a full segmental description, sincethe anatomic heterogeneity to be encountereddefies the potential to describe these lesions simplyin terms of "mitral" or "tricuspid" atresia.

usually sufficiently severe to preclude any attemptat biventricular repair.

It follows from the above discussion that,although most cases with atrioventricular valvaratresia have a ventricular septal defect, somepatients can have intact ventricular septums. Inthose with deficient ventricular septation, the

Variability in the ventricular mass

Almost all patients with atrioventricular valvaratresia will possess one big and one small ventricle.Very rarely, patients may be encountered when oneatrioventricular connection is absent and there is asolitary ventricle present of indeterminate apicalmorphology (Fig. 17). These are the only patientswith atrioventricular valvar atresia who have trulyuniventricular hearts. In the remaining greatermajority of patients, the ventricular mass is func-tionally univentricular. Anatomically, nonetheless,it is made up of one big and one small ventricle.Most usually, the big ventricle is of leftmorphology, while the small right ventricle is rudi-mentary and incomplete. This is the arrangementseen in so-called "classical" tricuspid atresia, butalso when the left-sided atrioventricular connectionis absent and the rudimentary right ventricle is left-sided. The atrial anatomy so typical of "classical"tricuspid atresia, however, can be replicated whenthe left atrium is connected to a dominant rightventricle, and the rudimentary and incomplete leftventricle is found posteriorly and to the right (lefthand ventricular topology — Fig. 18). Further vari-ations in ventricular morphology are found whenthe atresia is the consequence of an imperforatevalve. If the valve is imperforate in the setting ofdouble inlet ventricle, then the ventricularmorphology is exactly as expected for absence of anatrioventricular connection. When the atrioven-tricular connections are biventricular, theventricular mass is still typically unbalanced, butboth ventricles are normally constituted, albeit thatthe one with an imperforate valve is hypoplastic(Fig. 19). This pattern of imperforate atrioven-tricular valve is typically seen when the ventricularseptum is intact (Fig. 4), and the outflow tract isadditionally atretic. The example shown in Figure19, nonetheless, has a co-existing ventricular septaldefect. It is also possible to envisage the situation inwhich an imperforate atrioventricular valve couldbe found in a heart with ventricles of equal size,potentially permitting surgical excision of theabnormal valve and replacement with a valvarprosthesis. We have seen the potential for this onlyin patients with coexisting Ebstein's malformation(Fig. 20), and then associated malformations are

Figure 17.This echocardiographic section shows absence of the right atrioven-tricular connection (arrowed — tricuspid atresia) in a patient with asolitary ventricle of indeterminate morphology (IND V).

Figure 18.These panels show (upper) absence of the right atrioventricularconnection, but with the left atrium connected to a dominant rightventricle. The rudimentary left ventricle is posterior and to the right(left hand topology). The lower panel shows that the left atrioven-tricular valve straddles and overrides the ventricular septum. Theright valve in this patient would have been of tricuspid morphology,yet the clinical picture is that of classical tricuspid atresia.


Figure 19.This echocardiographic four-chamber section shows an imperforate(IMP) right atrioventricular valve in a patient with concordant atri-oventricular connections. The mitral valve is open. Compare with Fig. 4-

Right atrium

ventricular septum and the outlet septum. Theycan be of variable size, but show a tendencytowards spontaneous closure with growth. Thisproduces subpulmonary obstruction when theventriculo-arterial connections are concordant, butsubaortic obstruction with discordant ventriculo-arterial connections (Fig. 21). The structure of thedefect is basically comparable in these situations,although the subpulmonary infundibulum isusually much longer than the subaortic one.5 Thebasic arrangement, and the disposition of theconduction tissues, is also comparable irrespectiveof the atrioventricular connection, and irrespectiveof the location of the rudimentary and incompleteright ventricle (Fig. 22). It is possible to find apicalmuscular defects in this setting, and also for thedefect to be doubly committed and juxta-arterialwhen the muscular outlet septum is lacking.Perimembranous defects, in contrast, do not existin hearts with dominant left and rudimentary andincomplete right ventricles. Perimembranousdefects are found only when there are biventricularatrioventricular connections, or else when the rightventricle is dominant.

Variability at the ventriculo-arterial junctions

Traditionally, tricuspid atresia was divided intopatients with concordant as opposed to discordantventriculo-arterial connections, with further sub-division according to the size of the ventricular

I m perforatetricuspid > al\i"

Figure 20.In this specimen, the imperforate right atrioventricular valve in a patient

with concordant atrioventricular connections shows Ebstein's malfor-

mation. It is only in this setting that we have encountered imperforate

valves blocking an atrioventricular junction potentially of suitable

magnitude to permit the insertion of a right atrioventricular valve.

morphology of the defect reflects the ventricularmorphology. Thus, almost always those withdominant left and rudimentary right ventricleshave muscular ventricular septal defects. Typically,these defects are positioned between the apical

Figure 21.This heart shows tricuspid atresia with discordant ventriculo-arterial

connections. Restriction at the level of the ventricular septal defect

(VSD) produces subaortic obstruction.


Figure 22.Whenever there is univentricular connection to a dominant leftventricle, irrespective of the precise atrioventricular connection or theventriculo-arterial connections, the conduction axis is disposed asshown in the cartoon. The inset shows the view of the rudimentaryright ventricle as would be obtained by the surgeon in the operatingroom. It is the septal component adjacent to the obtuse margin whichcan safely be resected to enlarge the defect.

septal defect.10 Mitral atresia most usually is seen aspart of the hypoplastic left heart syndrome, or withdouble outlet right ventricle. With increasing expe-rience, it has become plain that any type ofventriculo-arterial connection can accompany thevarious forms of atrioventricular valvar atresia.Some have attempted to legislate for this increasedvariability in tricuspid atresia by expanding thetraditional alpha-numeric classification,11 but suchcategorisations become increasingly unwieldy asmore and more variations are included.12 Becauseof this, it remains our preference to use a descriptiveapproach. Thus, we can describe in simple fashion,for example, those who have tricuspid atresia asso-ciated with common arterial trunk, or with doubleoutlet from the dominant left ventricle, or anyother variation.

Summation of variability

With increasing experience, it is becoming evidentthat all potential variations in atrioventricularvalvar atresia will, eventually, be encountered inthe clinical setting. Thus, imperforate atrioven-tricular valves, or absence of one atrioventricularconnection, can be found with any arrangement ofthe atrial appendages. An imperforate valve can befound in the setting of double inlet ventricle, orwith any of the possible biventricular atrioven-tricular connections. Absence of one atrioven-tricular connection can, very rarely, be found in thepattern producing the uniatrial but biventricularatrioventricular connection. Absence of either theright-sided or left-sided atrioventricularconnection, or an imperforate right or left atrioven-tricular valve, can be found in all variants of righthand or left hand ventricular topology, or even

when there is a solitary ventricle of indeterminatemorphology. All possible ventriculo-arterialconnections can then be associated with thesealready huge combinations of anomalies. The onlyway realistically of catering for this heterogeneity isto use a descriptive approach to diagnosis and cate-gorisation, and to assess the potential for variabilityat each segmental level.

Acknowledgements

This work was supported by the British HeartFoundation, together with the Joseph LevyFoundation.

References

1. Rao PS. Terminology: Is Tricuspid atresia the correct term to use?In: Rao PS, ed. Tricuspid Atresia. 2nd Ed. Futura PublishingCompany Inc, Mount Kisco, New York. 1999:3—16.

2. OtieJD, Anderson C, Ettedgui JA, Zuberbuhler JR, AndersonRH. Echocardiographic-anatomical correlations in tricuspidatresia. J Am Coll Cardiol 1995;26:75O-758.

3. Anderson RH, Wilkinson JL, Gerlis LM, Smith A, Becker AE.Atresia of the right atrioventricular orifice. Br Heart J1977;39:4l4^i28.

4. Anderson RH, Becker AE, Macartney FJ, Shinebourne EA,Wilkinson JL, Tynan MJ. Is "tricuspid atresia" a univentricularheart? Pediatr Cardiol 1979;l:51-56.

5. Deanfield JE, Tommasini G, Anderson RH, Macartney FJ.Tricuspid atresia: an analysis of the coronary artery distributionand ventricular morphology. Br Heart J 1982;48:485^492.

6. Anderson RH, Macartney FJ, Tynan M, Becker AE, FreedomRM, Godman MJ, Hunter S, Quero Jim.nez M, Rigby ML,Shinebourne EA, Sutherland GR, Smallhorn JF, Soto B, ThieneG, Wilkinson JL, Wilcox BR, Zuberbuhler JR. Univentricularatrioventricular connection: the single ventricle trap unsprung.Pediatr Cardiol 1983;4:273-280.

7. Anderson RH, Rigby ML. The morphologic heterogeneity of"tricuspid atresia". Int J Cardiol 1987;l6:67-73.

8. Anderson RH, Ho SY, Falcao S, Daliento L, Rigby ML. The diag-nostic features of atrioventricular septal defect with common atri-oventricular junction. Cardiol \bung 1998;8:33—49.

9. Webb S, Brown NA, Anderson RH. Cardiac morphology at latefetal stages in the mouse with trisomy 16: consequences fordifferent formation of the atrioventricular junction whencompared to humans with trisomy 21. Cardiovasc Res1997;34:515-524.

10. Edwards JE, Burchell HB. Congenital tricuspid atresia: a classi-fication. Med Clin North Am 194933:1177-1196.

11. Tandon R, Edwards JE. Tricuspid atresia. A re-evaluation andclassification. J Thorac Cardiovasc Surg 1974;67:530— 542.

12. Rao PS. A unified classification for tricuspid atresia. Am Heart J1980;99:799-804.

the morphologic variability in atrioventricular valvar atresia

Documents