Editorial (700KB)	McCarthy KP*, Ho SY*, Anderson RH**. Defining the morphologic phenotypes of atrial septal defects and interatrial communications. Images Paediatr Cardiol 2003;15:1-24
*	Paediatrics, National Heart & Lung Institute, Imperial College and Royal Brompton Hospital, London, UK
**	Cardiac Unit, Institute of Child Health, University College London, Guilford Street, WC1, London, UK

Proper assessment is crucial to determine the strategy for treatment, as some defects are unpredictable, and can present as small insignificant defects, but over time they can develop into haemodynamically important conditions.⁸ In one study conducted over an eight-year period, two thirds of atrial septal defects within the oval fossa were found to enlarge over time, so becoming inappropriate for transcatheter closure.⁹ There have been further technological advances in imaging with the introduction of 3-dimensional echocardiography. This imaging technique allows the entire morphology of the defect to be observed in real time, and therefore permits study of the pathophysiology of atrial shunting.^10,11 This technique provides a more accurate description of the varying dynamic morphology, dimensions, and spatial relationship of the interatrial communications.^12,13 This is important, as a recent clinical study, which measured the size of the defect, found that this did not always correlate to the degree of shunting, or the resulting clinical symptoms.¹⁴ Further advances using magnetic resonance technology are now being trailed in animal studies as an alternative to using ionizing radiation when closing a patent oval fossa.¹⁵ But in order to appreciate the spectrum of phenotypes responsible for interatrial shunting, in relation to their clinical management, it is first essential to appreciate the normal septal anatomy.

Development of the normal atrial septum
The development of the normal atrial septum occurs following the initial looping of the heart. In the embryonic heart, the normal atrial septum, and the surrounding atrial structures, are formed from several embryological tissue components that develop, remodel and fuse in the correct sequence.

As the initial step in septation, a ridge of tissue develops from the superior aspect of the primary atrial component of the heart tube. This ridge is the primary septum (septum primum), and its the leading edge is covered by cushion-like mesenchymal tissue that is continuous over the dorsal mesocardium. This dividing crest of tissue is part of the atrial chamber expressing genes demonstrating morphologically leftness.¹⁶ As it grows into the atrial cavity, it extends down towards the endocardial cushions that are developing concomitantly within the atrioventricular canal. Normal septal development also involves incorporation of another mass of tissue derived from the dorsal mesocardium. This is known as the vestibular spine (spina vestibuli), and it, too, carries on its leading edge a mesenchymal cap. As the primary septum approaches the atrioventricular endocardial cushions, the various mesenchymal structures fuse together.¹⁷ The mass derived from the vestibular spine then muscularises, eventually forming the prominent infero-anterior border of the oval foramen.¹⁸ During the process of development the ventricular septum also ‘moves’ up towards the endocardial cushions, resulting in the septation of the ventricular chambers.¹⁹

Subsequent to the fusion between the primary septum and the endocardial cushions of the atrioventricular canal, the upper part of the primary septum disintegrates to form the ‘ostium secundum’. The remaining part of the primary septum becomes the flap valve of the oval fossa. This flap valve, along with the muscularised antero-inferior rim, forms the true septum that separates the cavities of the atrial chambers. Only after integration of the pulmonary veins into the left atrium do the superior walls of the two atriums ‘infold’, creating the so-called "septum secundum" in the superior portion of the atriums. The flap valve overlaps, but is not completely adherent to, the rims of this superior atrial fold, also known as Waterston’s or Sonderggard’s groove, providing a passage during fetal life for blood to pass from the right to the left atrium (Figure 1). In postnatal life, this deep superior interatrial fold becomes filled with extracardiac fibro-fatty tissue (Figure 2).
 

The normal postnatal heart
In the definitive postnatal heart, when the wall separating the atriums is viewed from its right atrial aspect, the septal aspect, at first sight, seems extensive. The atrial septum, nonetheless, is best defined as the tissues which directly separate the atrial cavities, and which can be removed without exiting the heart. When viewed in the light of this definition, the septum is confined to the thin flap of fibromuscular valvar tissue which forms the floor of the oval fossa, along with the immediate infero-anterior muscular rims of the fossa derived from the vestibular spine (Figure 3).
 
 

Figure 3: Close-up four chamber view of the RA and RV, showing the muscular superior and inferior rims, and the flap valve of the oval fossa dividing the two atriums   RA, right atrium. RV, right ventricle. LA, left atrium

 
Thus, the septal area is only a small part of the wall dividing the atriums. The aortic mound, in contrast, which is to the right of the oval fossa when observed from the right atrial aspect, and which seems to represent an apparently solid muscular structure, is part of the external wall of the heart. Passage of an instrument through this area does not take one into the left atrium, but rather into the transverse sinus of the pericardial cavity, in front of the bulging right coronary sinus at the base of the aortic root (Figure 4).
 

 
Similarly, passing a needle from the right to left atrial chambers superiorly to the oval fossa also passes through extracardiac tissue, as this part of the dividing wall is in reality an infolding of the right atrial wall and hence is not septal (Figure 5). In adult life, this fold is filled with fatty extracardiac tissue. It is within this area that the sinus nodal artery usually takes its course (Figure 6).
 

True defects of the atrial septum, of necessity confined within the bounds of the oval fossa, are often referred to as ‘secundum’ defects. The defect, nonetheless, is a deficiency of the floor of the oval fossa, and this floor is derived from the primary atrial septum. Thus, the defects are "ostium secundum" defects, and not deficiencies of the secondary atrial septum. As we have shown, the secondary "septum" is no more than an infolding of the atrial roof. Such defects within the confines of the fossa account for over three-quarters of all holes between the atriums. In a study of 160,480 live births, atrial septal defects were found to account for one-twentieth of all cases of congenital heart disease diagnosed in infancy.²⁰ The occurrence of congenital heart disease in a total of 43 studies was reviewed recently, and the overall incidence of atrial septal defects was found to be 3.4/10,000 of live births.²¹ In adults with a congenital heart defect, ‘secundum’ atrial septal defects comprise nearly a third of all cases.³

The probability of spontaneous closure of an isolated atrial septal defect in patients below the age of 5 years is high, with closure commonly occuring in up to four-fifths of the small to moderate defects that are diagnosed within the first three months of life.²² Spontaneous obstruction of the hole has also been observed beyond infancy. In one study, spontaneous closure was found in three-fifths of the population studied after the age of 18 months, and in total two-fifths of all the patients studied exhibited a natural closure after the age of five years.²³ The mechanism of spontaneous closure is unknown. Small atrial septal defects can remain completely asymptomatic and hemodynamically insignificant throughout life.

Morphologically, there are notable variations in the structure of these atrial septal defects. The differences impact on transcatheter closure. Intervention may not be possible if the defect is too large, or if there is a lack of suitable rims around the entire circumference to provide anchorage for a device. Other contraindications included the defect being displaced towards the posterior wall, and its proximity to the entrance of the superior or inferior caval veins. Also, if the eustachian valve is thick, this structure can obscure the postero-inferior rim of the oval fossa.²⁴ When examined morphologically, seven-tenths of defects were centrally located defect within the confines of the oval fossa.²⁵ Even within a group having these characteristics, the borders of the defect in two-fifths of the cases were deemed insufficient to provide firm anchorage for an occluding device.²⁶

The extent of deficiency of the flap valve can produce a spectrum of morphology, ranging from the flap valve failing to cover fully the oval fossa completely (Figure 8), to its complete absence (Figure 9). Absence of the flap valve altogether, along with effacement of the rim, leads to a physiologically functional ‘common atrium’. The morphology of the floor itself can also be variable, with some flap valves being very thin and membranous, composed of connective tissues, to others that have integration of myocardium and so become thick and muscular. The Rashkind procedure for atrial septostomy will clearly be more difficult if a thicker and more muscular floor of the oval fossa is encountered.
 

Once the defect is assessed, and it is determined that closure by device is recommended, then in ideal cases, this is a relatively straightforward procedure (Figure 10).
 

Contraindications to closure include a large defect, even when a suitably large device is available. If the defect is unduly large, the device is likely to be adjacent to important structures with the atriums (Figure 11).
 

 
The final position of the device is also of crucial importance, irrespective of the size of the defect, as its edges could impinge inferiorly in the right atrium on the coronary sinus, the atrioventricular nodal area, and the inferior caval vein (Figure 12), or the device could encroach on the leaflets of the mitral valve in the left atrium. In one study, in one-third of cases the narrowest boarder of the defect was found to be between the oval fossa and the aortic mound, at the antero-superior rim.²⁷ The same study showed the superior rim to be furthest from the defect. A short rim in this position could result in the right upper and lower pulmonary veins being occluded on the left atrial aspect, or the superior caval vein being obscured on the right atrial aspect (Figure 13).
 

Holes can occur at various locations within the flap valve itself (Figures 14,15). In a recent review of post-mortem specimens, over half were found to have a fenestrated flap valve.²⁷ The degree of fenestration can vary greatly. If widespread, the floor of the fossa can have a fishnet-like appearance, with the tissue becoming quite extensive.
 

In these circumstances, the flap valve can take the appearance of a windsock blowing into the left atrium (Figure 16). Such extensive flap valves also have the potential to become aneurysmal in the opposite direction, producing a left-to-right shunt with right ventricular overload.²⁵ Conversely, a few widely separated holes may complicate the strategies for treatment (Figure 17).

Defects outside the confines of the atrial septum
Superior and Inferior sinus venosus defects
These defects are both examples of interatrial communications in which the atriums communicate through a channel outside the boundaries of the true atrial septum (Figure 18). In cases of superior sinus venosus defect, the hole is located superiorly to the oval fossa, which can itself either be intact or deficient. In the typical morphology seen in hearts with a superior sinus venosus defect, the superior caval vein usually overrides the crest of the defect. The caval channel then has biatrial connections, opening into the right and left atriums.²⁸ The phenotypic feature of the lesion is the presence of the defect outside the confines of the normally formed oval fossa (Figure 19). Indeed, it is possible to pass a probe through the extracardiac tissue found within the intact superior muscular rim of the oval fossa, which forms the inferior rim of the defect itself.
 

Figure 19: The defect overrides the superior rim of the oval fossa, resulting in the SCV having direct communication to both the right and left atriums. Note the anomalous connection of the pulmonary vein to the SCV   S-ASD, superior sinus venosus defect. OF, oval fossa. RAA, right atrial appendage. Pulm.vein, pulmonary vein. ICV, inferior caval vein. SCV, superior caval vein

The cases with abnormal pulmonary venous drainage are unsuitable for device closure, and can be complicated to close at surgery. The surgeon must exercise caution and avoid the junction between the superior caval vein and the right atrial appendage, as this is the location of the sinus node (Figure 20). Additionally, the area around the junction of the superior caval vein and the terminal groove must be avoided, as this is the location of the sinus nodal artery.      Figure 20: Anterior view of the right atrium, showing the location of the sinus node and artery   RAA, right atrial appendage. SCV, superior caval vein

Coronary sinus defects
The coronary sinus is a venous channel that is located within the left atrioventricular groove, above the annulus of the mitral valve. Distally located tributaries join to convey the deoxygenated blood back to the right atrium. Defects within the wall of the coronary sinus are a rarity. They are usually found associated with anomalous connection of the left superior caval vein to the roof of the left atrium, which enters between the right pulmonary veins and the left atrial appendage. In the normal heart, the muscular wall of the coronary sinus is separated by extracardiac tissue from the inferior wall of the left atrium. Thus, the coronary sinus is an independent venous structure with a distinct wall from the atrium. Developmentally, therefore, the existence of this type of interatrial defect is remarkable. In essence, the defect necessitates breakdown not only of the wall of the discrete venous channel, but also partial dissolution of the wall of the adjacent left atrium. The degree of disintegration of the two walls can vary widely, from small distinct fenestrations (Figures 21,22), to complete ‘unroofing’ of both walls, producing complete mixing of the venous deoxygenated blood and the oxygenated blood within the left atrium (Figure 23). The key phenotypic feature is again the presence of the defect outside the confines of the oval fossa.
 

 
The Ostium Primum Defect
Another lesion producing an interatrial communication is the "ostium primum" variant of atrioventricular septal defect. But here the phenotypic feature in these hearts is the presence of a common atrioventricular junction, albeit with separate valvar orifices into the right and left ventricles (Figures 24).
 

These anomalies, therefore, should be categorised as atrioventricular septal defects, even though shunting across the defect is exclusively at atrial level. The hearts do not have separate atrioventricular junctions as in the normal heart, and as in all other hearts described thus far with interatrial communications. In reality, the "ostium primum" defect, possesses all the phenotypic features of atrioventricular septal defect with common atrioventricular junction. The most important characteristic is that the valve guarding the common junction has five leaflets. Only two leaflets are exclusively within the right ventricle; and one is solely within the left ventricle. The remaining two leaflets, the superior and inferior bridging leaflets, are shared between both ventricles, straddling the ventricular septum (Figure 25).
 

The distinguishing feature of the "primum" defect is that the two bridging leaflets are joined to each other by a tongue of fibrous tissue positioned directly on top of the crest of the ventricular septum, thus dividing the junction into discrete and separate left and right valvar inlets to the ventricles.³⁰ Almost always, the bridging leaflets and the tongue are also fused to the ventricular septal crest. It is this feature which confines shunting across the septal defect at atrial level, albeit with much of the shunting being below the level of the atrioventricular junction (Figure 26,27). The outflow tract of the left ventricle, nonetheless, retains the phenotypic anterior displacement, resulting in a longer and narrower left ventricular outflow tract as the aorta is ‘unfastened’ from its central position at the base of the heart. Indeed, the tethering of the superior bridging leaflet to the ventricular septal crest elongates the length of the outflow tract when compared to hearts with common atrioventricular valvar orifice.
 

 

 Associated considerations
Atrial septal defects and interatrial communications can be accompanied by a variety of other congenital cardiac defects, such as coarctation of the aorta,³¹ mitral valvar prolapse,³² and partially anomalous drainage of the pulmonary veins into the right atrium.³³ Regardless of their anatomical location, the physiologic consequence of atrial septal defects and atrial communications result in the shunting of blood from one atrium to the other. Ultimately, the direction and magnitude of shunting is determined by the size of the defect, and the relative compliance of the ventricles. These vary according to age, with children having a normal compliance into infancy and a faster heart rate,³⁴ compared to adults who have a comparatively slower rate and a ventricular compliance that decreases with age.³⁵ Typically a sizable defect permits a large shunt, with significant hemodynamic consequences.

Conclusion
The spectrum of interatrial communications is very broad, covering deficiencies of the flap valve of the oval fossa, to formation at various locations within the atrium of an extra-septal channel linking the right and left sides of the heart. Although generally grouped together as atrial septal defects, their phenotypic features are quite different and require careful distinction between defects that are located within the true atrial septum from those that are outside the confines of the oval fossa. Taken together, interatrial communications comprise one of the most common of all congenital cardiac defects, and often with an associated cardiac defect. The use today of advanced imaging techniques to assess the morphology and physiology of these frequent cardiac anomalies has led to efficient planning of treatment through surgical or interventional closure. These procedures now carry very low rates of morbidity and mortality, providing a positive outlook for the patient.
 
References

1. Samanek M, Voriskova M. Congenital heart disease among 815,569 children born between 1980 and 1990 and their 15-year survival: a prospective Bohemia survival study. Ped Cardiol 1999;20:411-417
2. Feldt RH, Avasthey P, Yoshimasu F, et al. Incidence of congenital heart disease in children born to residents of Olmsted County, Minnesota, 1950-1969. Mayo Clin Proc 1971; 46: 794-799
3. Brickner EM, Hillis DL, Lange RA. Congenital Heart Disease in Adults: First of Two Parts. The New England Journal of Medicine 2000; 342: 256-263
4. Boutin C, Musewe NN, Smallhorn JF, Dyck JD, Kobayashi T, Benson LN. Echocardiographic follow-up of atrial septal defect after catheter closure by double-umbrella device. Circulation 1993;88:621-627
5. Chan KC, Godman MJ. Morphological variations of fossa ovalis atrial septal defects (secundum): feasibility for transcutaneous closure with the clam-shell device. Br Heart J 1993;69:52-55
6. Magni G, Hijazi ZM, Marx G, Das G, Delabays A, Levine J, et al. Utility of 3-D echocardiography in patient selection and guidance for atrial septal defect (ASD) closure by the new Das-angel wings occluder device. J Am Coll Cardiol 1996;27(suppl):769-765
7. Mehta RH, Helmcke F, Nanda NC, Pinheiro L, Samdarshi TE, Vinod KS. Uses and limitations of transthoracic echocardiography in the assessment of atrial septal defect in the adult. Am J Cardiol 1991;67:288-294
8. Tortoriello TA, McMahon C, Kovalchin JP, Bricker JT, Grifka RG. Growth of an atrial septal defect: Missing the window for transcatheter closure. Pediatr Cardiol 2002;23:542-544
9. McMahon CJ, Feltes TF, Fraley JK, Bricker JT, Grifka RG, Tortoriello TA, Blake R, Bezold LI. Natural history of growth of secundum atrial septal defects and implications for transcatheter closure. Heart 2002;87:256-259
10. Maeno YV, Benson LN, McLaughlin PR, Boutin C. Dynamic morphology of the secundum atrial septal defect evaluated by three dimensional transoesophageal echocardiography. Heart 2000; 83:673-677
11. King DL, Harrison MR, King DL, Gopal AS, Martin RP, DeMaria AN. Improved reproducibility of left atrial and left ventricular measurements by guided three-dimensional echocardiography. J Am Coll Cardiol 1992;20:1238-1245
12. Lange A, Walayat M, Turnbull CM, Palka P, Mankad P, Sutherland GR, Godman MJ. Assessment of atrial septal defect morphology by transthoracic three dimensional echocardiography using standard grey scale and Doppler myocardial imaging techniques: comparison with magnetic resonance imaging and intraoperative findings. Heart 1997;78:382-389
13. Acar P, Saliba Z, Bonhoeffer P, Aggoun Y, Bonnet D. Influence of atrial septal defect anatomy in patient selection and assessment of closure with the Cardioseal device. A three-dimensional transoesophageal echocardiographic reconstruction. European Heart Journal 2000;21:573-581
14. Tabery S, Daniëls O. How classical are the clinical features of the "ostium secundum" atrial septal defect? Cardiol Young 1997;7:294-301
15. Buecker A, Spuentrup E, Grabitz R, Freundenthal F, Muehler EG, Schaeffter T, van Vaals JJ, Gunther RW. Magnetic resonance-guided placement of atrial septal closure device in animal model of patent foramen ovale. Circulation 2002;106:511-515
16. Wessels A, Anderson RH, Markwald RR, Webb S, Brown NA, Viragh SZ, Moorman AFM, Lamers WH. Atrial development in the human heart: An immunohistochemical study with emphasis on the role of mesenchymal tissues. Anat Rec 2000;259:288-300.
17. Anderson RH, Brown NA, Webb S. Development and structure of the atrial septum. Heart 2002;88:104-110
18. Drose JA. Fetal Echocardiography. 1998. Publisher WB Saunders Co. Pennsylvania. Chapter: Atrial septal defects. ER Stamm.
19. Anderson RH, Webb S, Brown NA. Clinical anatomy of the atrial septum with reference to its developmental components. Clin Anat 1999;12:362-374
20. Dickinson DF, Arnold R, Wilkinson JL. Congenital heart disease among 160 480 liveborn children in Liverpool 1960 to 1969. Implications for surgical treatment. Br Heart J 1981;46:55-62
21. Hoffman JIE, Kaplan S. The incidence of congenital heart disease. Journal of the American College of Cardiology. 2002;39:890-1900
22. Radzik D, Davignon A, van Doesburg N, Fournier A, Marchand T, Ducharme G. Predictive factors for spontaneous closure of atrial septal defects diagnosed in the first 3 months of life. J Am Coll Cardiol 1993;22:851-853
23. Brassard, Myriam MD; Fouron, Jean-Claude MD; van Doesburg, Nicolaas H. MD; Mercier, Lise-Andree MD; De Guise, Pierre MD. Outcome of children with atrial septal defect considered too small for surgical closure. The American Journal of Cardiology 1999; 83:1552-1555
24. Ferreira SMAG, Ho SY, Anderson RH. Morphological study of defects of the atrial septum within the oval fossa: implications for transcatheter closure of left to right shunt. Br Heart J 1992;67:316-320
25. Chau AK, Leung MP, Yung T. Surgical validation and implications for transcatheter closure of quantitative echocardiographic evaluation of atrial septal defect. The American Journal of Cardiology 2000;85:1124-1130
26. Ewert P, Berger F, Vogel M, Dahnert I, Alexi-Meshkishvili V, Lange PE. Morphology of perforated atrial aneurysm suitable for closure by transcatheter device placement. Heart 2000;84:327-331
27. Ussia GP, Momenah TS, Ursell P, Brook MM, Moore P, De Luca F. Evaluation of the morphology of the oval fossa for placement of devices. Cardiol Young 2000;10:502-509
28. Li J, al Zaghal A.M, Anderson RH. The nature of the superior sinus venosus defect. Clinical Anat 1998:11;349-352
29. al Zaghal AM, Li J, Anderson RH, Lincoln C, Shore D, Rigby ML. Anatomical criteria for the diagnosis of sinus venosus defects. Heart 1997;78:298-304
30. Anderson RH, Ho SY, Falcao S, Daliento L, Rigby ML. The diagnostic features of atrioventricular septal defect with common atrioventricular junction. Cardiol Young 199;8:33-49
31. Yeager SB, Keane JF. Fate of moderate and large secundum type atrial septal defect associated with isolated coarctation in infants. The American Journal of Cardiology 1999;84:362-363
32. Leachman RD, Cokkinos DV, Cooley DA. Association of ostium secundum atrial septal defects with mitral valve prolapse. Am J Cardiol 1976;38:167-169
33. Allan LD, Sharland GK. The echocardiographic diagnosis of totally anomalous pulmonary venous connection in the fetus. Heart 2001;85:433-437
34. Harada K, Suzuki T, Tamura M, et al. Role of age on transmitral flow velocity patterns in assessing left ventricular diastolic function in normal infants and children. Am J Cardiol 1995;76:530-532
35. Iwase M, Nagata K, Izawa H, et al. Age-related changes in left and right ventricular filling velocity profiles and their relationship in normal subjects. Am Heart J 1993;126:419-426

Contact information

Ms. Karen P McCarthy National Heart & Lung Institute, Imperial College  and Royal Brompton Hospital  Dovehouse Street London SW3 6LY - UK  k.mccarthy@imperial.ac.uk