Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
The anomaly now known as juxtaposition of the atrial appendages (JAA) was first described, to the best of our knowledge, by Thore in 1843 in an 11-day-old boy with left-sided JAA. Thore wrote that this neonate had “une anomalie du coeur” (an anomaly of the heart), which he described but did not name.
The first paper concerning this malformation in the English literature was that of Birmingham in 1893. His patient was a 20-year-old woman, a “dealer in fish,” who also had left-sided JAA with double-outlet right ventricle (DORV) {S,D,L}, dextrocardia, a large secundum atrial septal defect (ASD) (septum primum was absent), a persistent left superior vena cava to the coronary sinus that drained into the right atrium, a conoventricular type of ventricular septal defect, a bilateral conus (subaortic and subpulmonary), pulmonary outflow tract stenosis (infundibular and valvar), and a right aortic arch. She also had bilateral pulmonary tuberculosis. Understandably, Birmingham called this an extreme anomaly of the heart and great vessels .
In the first half of the 20th century, multiple case reports appeared in the literature concerning left-sided JAA ( Table 10.1 ).
Author(s) | Year | Sex | Age | Diagnosis |
---|---|---|---|---|
Wenner | 1909 | F | 10½ y | Dextrocardia; DORV {S,D,D}; ASD II; LSVC to RA (not via coronary sinus); AV valve with direct insertion into small RV (no chordae tendinae); left AV valve larger, tricuspid, inserting into large LV; bilateral conus (subaortic and subpulmonary); pulmonary outflow tract stenosis, infundibular and valvar (bicuspid pulmonary valve); high conoventricular VSD and low muscular VSDs; left aortic arch; left probe patent ductus arteriosus |
Dünner | 1914 | NK | 10 wk | TGA {S,D,D}; two ASDs: PFO and fenestrations in septum I; RA-to-LV communication; double-inlet left ventricle; large LV and small RV: functionally or anatomically single LV; bilateral conus without PS; low tunnel-like VSD leading to small RV; three coronary ostia, one from each sinus of Valsalva; large left-sided PDA |
Huebschmann | 1921 | M | 5½ mo | Dextrocardia or mesocardia, apex pointing rightward; {S,D,S}; tricuspid atresia; ASD II; Ao from LV with AoV-MV fibrous continuity; small subpulmonary conus with PS, infundibular and valvar (bicuspid PV); small VSD (or BVF) from large LV to subpulmonary conus; MPA arises posteriorly and slightly to the right of the Ao; left aortic arch with right-sided closing PDA |
Kettler | 1933 | M | 15 mo | Left-sided heart with rightward pointing ventricular apex; DORV {S,D,D}; tricuspid atresia; absent RSVC, but with right-sided SA node; LSVC to CoS to RA; bilateral conus (sub-Ao and sub-PA) with PS |
Ngai | 1935 | F | 100 days | Died in a blue spell. Cardiomegaly (3× normal size); dextrocardia; DORV {S,D,A}; ASD II; large LV; small RV; concordant AV alignments (MV into LV, TV into RV); sinistroposition of right atrial appendage; bilateral conus with no outflow tract obstruction; absent left coronary ostium; right aortic arch; small right-sided PDA; lobar pneumonia of azygos lobe |
Bredt | 1936 | F | 5¾ mo | Case 1 (Museum specimen #9): TGA {S,D,D}; small RV and large LV; VSD; bilateral conus; pulmonary atresia. (Case 2 is Dünner’s case 4.) |
Harris and Farber | 1939 | M | 1 y | Case 12: DORV {S,D,D}; ASD II; LVH and E; RV thick-walled, small-chambered, RV sinus small or possibly absent; VSD; bilateral conus; pulmonary outflow tract atresia; absent left coronary ostium, that is, single right coronary artery; small PDA |
Taussig | 1947 | F | 25 y | Case 40, Chapter 33, Fig. 171: Dextrocardia; TGA {I,L,L} (may have had visceral heterotaxy with polysplenia; left-sided JAA with situs inversus of atria, that is, levomalposition of left atrial appendage (not levomalposition of right atrial appendage, as in all previous cases); common atrium; ipsilateral pulmonary veins (right-sided pulmonary veins to right of atrial septal remnant, left-sided pulmonary veins to left of atrial septal remnant); interrupted inferior vena cava; common AV valve and common AV canal; single LV (absence of RV sinus) with common inlet LV; bilateral conus; PS, severe; left aortic arch (abnormal in visceroarterial situs inversus) |
Miskall and Fraser | 1948 | F | 11 mo | TGA {S,D,D} with intact ventricular septum; ASD II (1.0 × 0.5 cm); no outflow tract obstruction; meningocele at bifid T8 vertebra, operated on successfully at 5th day of life. Death from pneumonia of left lower lobe |
Rogers, Cordes, and Edwards | 1950 | M | 12 y | TGA {S,D,D}; tricuspid atresia; ASD II, valve incompetent patent foramen ovale with fenestrated septum primum; bilateral conus; PS, severe, subvalvar and valvar with bicuspid pulmonary valve |
The complexity of the congenital heart disease associated with left-sided JAA is very impressive (see Table 10.1 ), leading to at least two questions: Is there any pattern here? Developmentally, what does JAA mean? These are the questions that we and others would subsequently explore. As pointed out by Harris and Farber in 1939, Kettler in 1933 proposed that what would subsequently be called JAA was caused by “insufficient torsion of the cardiac loop so that the bulbous portion fails to migrate to the left and join correctly with the remainder of the heart.” Many other works have also contributed to our present understanding of JAA.
There is much that could be said about Table 10.1 . First, this is a meta-analysis, an analysis based on the literature, not on first-hand examination of these heart specimens. Although I did my best to be accurate, I am not entirely confident about some of these cases. Taussig’s case (see Table 10.1 ) is diagrammed, but the description is scanty.
Were the great arteries in Huebschmann’s case (see Table 10.1 ) really normally related? The main pulmonary artery arose posteriorly and slightly to the right of the aorta, which certainly does not sound like normally related great arteries. However, there was a small and stenotic subpulmonary conus from which the pulmonary artery arose, the aorta originated above the left ventricle (LV), and there was aortic-mitral fibrous continuity (see Table 10.1 ). Hence, the internal relationships between the great arteries and the conus, between the great arteries and the LV, and between the great arteries and the atrioventricular (AV) canal all appear to have been normal; hence, our conclusion that the segmental anatomy was {S,D,S}, despite the abnormal relationships between the great arteries and external spatial coordinates (anterior-posterior and right-left).
Huebschmann’s case (see Table 10.1 ) leads to the question: Is it possible for normally related great arteries to have an anterior aortic valve and a posterior pulmonary valve? Surprising to relate, the answer is yes, as we learned in our first study of JAA that was published in 1968 ( Fig. 10.1 ). Although the internal relations of the great arteries are normal, as in Huebschmann’s case (see Table 10.1 ), the external relationships of the great arteries relative to fixed external spatial coordinates (such as anterior-posterior and right-left) can be very abnormal when the ventricles—from which the great arteries arise—are significantly malpositioned. In our patient (Case 21), there was isolated ventricular inversion, that is, {S,L,S} (see Fig. 10.1 ).
Thus, this is why I diagnosed Huebschmann’s case as I did (see Table 10.1 ); but because this is a meta-analysis, I also want to acknowledge that my diagnosis of this case could be erroneous. Nonetheless, it is helpful to understand that normally related great arteries rarely can have very abnormal spatial relationships relative to fixed external spatial coordinates when the ventricles are very malpositioned. Concerning whether the great arteries are normally or abnormally related, the relationships that matter are the internal intracardiac ones, not the external spatial relationship ones—particularly when the ventricles are malpositioned.
In 1954, Dixon introduced the term juxtaposition of the atrial appendages, which subsequently has been widely used. In 1963, Ellis and Jameson proposed the designation congenital levoposition of the right atrial appendage . In 1976, Park et al introduced the term congenital levojuxtaposition of the right atrial appendage.
In 1996, Stella Van Praagh et al introduced the following designations:
juxtaposition of the morphologically right atrial appendage in solitus and inversus atria ; and
juxtaposition of the morphologically left atrial appendage in solitus and inversus atria
The reasons for the introduction of these modified designations were twofold:
JAA should be described not only positionally, but also morphologically. In congenital heart disease, this is true of all parts of the heart; that is, this is a basic principle.
Juxtaposition of the atrial appendages with malposition of the right atrial appendage (JRAA) and juxtaposition of the atrial appendages with malposition of the left atrial appendage (JLAA) are two very different and largely opposite syndromes, as will be seen.
Juxtaposition of the atrial appendages means that both atrial appendages lie side by side, rather than being separated by the great arteries as they normally are ( juxta, near and positio, place, Latin). In addition to being apposed, the malpositioned atrial appendage is often superior to or even on top of the normally located atrial appendage ( Fig. 10.2 ).
This chapter is based on 74 postmortem cases of JAA: 15 cases from the Congenital Heart Disease Research and Training Center, Hektoen Institute for Medical Research, Chicago, Illinois ; 6 cases from the Hospital for Sick Children, Toronto, Ontario, Canada ; and 53 cases from the Cardiac Registry of Children’s Hospital Boston. ,
Until 1994, JAA was classified positionally:
left JAA , meaning that both atrial appendages lie to the left of the vascular pedicle (aorta and pulmonary artery) (see Fig. 10.2 ) ; and
right JAA, indicating that both atrial appendages lie to the right of the vascular pedicle (see Fig. 10.1 ).
Morphologic and positional classification , asks three questions of the diagnostician:
Morphologically, which appendage is malpositioned relative to the great arteries? Is it the morphologically right atrial appendage that is on the wrong side of the vascular pedicle, as in Fig. 10.2 ? Or is it the morphologically left atrial appendage that is on the wrong side of the vascular pedicle, as in Fig. 10.1 ?
Positionally what is the sidedness of the abnormally located atrial appendage that results in JAA? Is it malpositioned to the left of the great arteries? Or is it malpositioned to the right of the great arteries?
What is the atrial situs of the patient? Is it situs solitus (usual or normal pattern of atrial organization)? Or is it situs inversus (the mirror-image pattern of atrial organization)? As will be seen, the type of atrial situs helps to make the JAA understandable.
There are four morphologic and positional anatomic types of JAA ( Fig. 10.3 ):
Malposition of the morphologically right atrial appendage (RAA) to the left of the vascular pedicle in atrial situs solitus (see Fig. 10.2 ). This is the classic type of malposition of the RAA to the left of the vascular pedicle, resulting in typical left-sided JAA, that occurred in 34 of our 35 cases with malposition of the RAA (97%).
Malposition of the RAA to the right of the vascular pedicle in atrial situs inversus ( Fig. 10.4 ). This is the rare form of malposition of the RAA, resulting in right-sided JAA, that we observed in only 1 of these 35 patients (3%).
Malposition of the left atrial appendage (LAA) to the right of the vascular pedicle in atrial situs solitus ( Fig. 10.5 ). This was the more frequent form of malposition of the LAA, which resulted in right-sided JAA, and was found in 16 of these 18 cases (89%).
Malposition of the LAA to the left of the vascular pedicle in atrial situs inversus ( Fig. 10.6 ). This was the rare form of malposition of the LAA, found in only 2 of these 18 cases (11%).
Thus, each atrial appendage was malpositioned in only one way in each type of atrial situs (see Fig. 10.3 ):
In visceroatrial situs solitus, the RAA can be malpositioned in only one way: to the left of the vascular pedicle (see Fig. 10.2 ).
In visceroatrial situs inversus, the left-sided RAA can be malpositioned in only one way: to the right of the vascular pedicle (see Fig. 10.4 ).
The same is true of the LAA:
In visceroatrial situs solitus, the LAA can be malpositioned in only one way: to the right of the vascular pedicle (see Fig. 10.1 and 10.5 ).
In visceroatrial situs inversus, the LAA can be malpositioned in only one way: to the left of the vascular pedicle (see Fig. 10.6 ).
The foregoing is why there are four anatomic types of JAA, based on their morphologic and positional anatomy (see Fig. 10.3 ). The frequencies of these four anatomic types are also included in this diagram (see Fig. 10.3 ), based on our two most recent studies. ,
One may ask why did we not include atrial situs ambiguus with visceroatrial heterotaxy and asplenia in the morphologic and positional classification of JAA (see Fig. 10.3 ). The answer is that we did not find any such cases in our data. , In addition, from a logical standpoint, atrial situs ambiguus is either undiagnosed or undiagnosable atrial situs solitus or atrial situs inversus. Consequently, no type of atrial situs has been omitted from Fig. 10.3 because atrial situs ambiguus is not a specific, third anatomic type of atrial situs, to the best of our present understanding.
How can one understand the frequencies of each anatomic type of JAA (see Fig. 10.3 )?
Why is JAA, involving malposition of the RAA, typically left-sided JAA (97%) rather than right-sided JAA (3%)?
Because atrial situs solitus is so much more common than atrial situs inversus. Atrial situs solitus is estimated to be 5000 to 15,000 times more frequent than atrial situs inversus. (See Chapter 3 for discussion of dextrocardia and other cardiac malpositions.)
Why is JAA involving malposition of the LAA so much more frequently right-sided (89%) than left-sided (11%)?
Again, because atrial situs solitus is so much more frequent than atrial situs inversus.
Is it possible to diagnose the atrial situs when visceral heterotaxy with asplenia coexists?
Yes. In some cases this can be done with confidence (see Fig. 10.3 ) (see Chapter 29 ).
Why is it clinically and surgically important to diagnose the morphologic anatomic identity of the malpositioned atrial appendage in JAA?
Because JAA involving malposition of the RAA (left-sided malposition in situs solitus and right-sided malposition in situs inversus) and JAA involving malposition of the LAA (right-sided in situs solitus and left-sided in situs inversus) constitute distinctive and largely opposite syndromes ( Table 10.2 ). ,
MRAA (n = 35) | MLAA (n = 18) | ||||
---|---|---|---|---|---|
Associated Malformations | No. | % | No. | % | p Value |
|
21 | 60 | 0 | 0 | <.001 |
|
26 | 74 | 5 ∗ | 28 | <.01 |
|
0 | 0 | 12 | 69 | <.001 |
|
0 | 0 | 5 | 28 | <.01 |
|
35 | 100 | 4 | 22 | <.01 |
|
1 | 3 | 7 | 39 | <.01 |
|
1 | 3 | 5 | 28 | <.05 |
∗ Double-inlet LV was present in 3 of these 5 cases, and in all 5 the hypoplastic or absent RV sinus was associated with stenosis of the left, not the right, atrial outlet. These cases were hemodynamically and anatomically very different from cases with RV hypoplasia or absence and JAA with malposition of the RAA.
The syndrome of JAAs with malposition of the RAA (JRAA) to the left or right of the vascular pedicle is characterized by multilevel right heart obstruction (see Table 10.2 ). Specifically there was right atrial outflow obstruction, that is, tricuspid atresia or stenosis, or hypoplasia in 21 of 35 cases (60%); right ventricular hypoplasia or absence in 26 of 35 cases (74%); and an abnormal conus with transposition of the great arteries (TGA), DORV, or anatomically corrected malposition of the great arteries in 35 of 35 cases (100%). All of these features of JAA with malposition of the RAA were statistically significantly different from those associated with JAA and malposition of the LAA (see Table 10.2 ).
In somewhat greater detail, the findings in JAA with malposition of the RAA, for convenient brevity called JRAA, may be summarized as follows:
Sex: Males-to-females = 19 of 15 (1.27/1). The sex was not recorded in 1 case.
Age at Death: Median, 5.5 months; mean, 41.7 months; range 22-week fetus to 19.5 years.
Cardiac Malposition: Dextrocardia, 5 of 35 (14.3%); and mesocardia, 7 of 35 (20%). Total cardiac malpositions, 12 of 35 (34%).
Persistent Left Superior Vena Cava: 11 of 34 with atrial situs solitus (32.35%) and 1 of 1 with atrial situs inversus (100%). Total prevalence, 12 of 35 (34.29%), which was significantly higher than the frequency of persistent LSVC in a control series (10 of 127, 7.9% p < .001).
Eustachian Valve of the Inferior Vena Cava and Thebesian Valve of the Coronary Sinus: Absent in 24 of 33 cases (73%).
Atrial Septum: Intact in only 1 of 35 cases (3%). In 34 of 35 patients (97%), a secundum ASD, a widely patent foramen ovale, or a resected atrial septum was found.
RV Sinus: Hypoplastic or absent in 26 of 35 (74%).
Tricuspid Valve: Tricuspid atresia, 11 of 35 (31.4%); tricuspid stenosis, 7 of 35 (20%); tricuspid hypoplasia, 3 of 35 (8.6%); myoxomatous tricuspid valve, 1 of 35 (2.9%); abnormal attachments to ventricular septum or abnormal chordae tendinae, 3 of 35 (8.6%). The tricuspid valve was normal in 10 of 35 (28.6%). Hence, the tricuspid valve was abnormally formed in 71%.
Mitral Valve: By contrast, with JAA and malposition of the RAA, the mitral valve was usually normally formed: 29 of 35 cases (82.9%). However, abnormalities of the mitral valve were found in 6 of 35 patients (17.1%): myxomatous mitral valve, 3 of 35 (8.6%); abnormal chordal attachments, 1 of 35 (2.9%); thick nodular leaflets, 1 of 35 (2.9%); and mitral regurgitation, 1 of 35 (2.9%).
Ventricular Situs: D-loop ventricles in 35 of 35 cases (100%).
Atrioventricular Alignments and Situs: AV alignment concordance was present in all (35/35 cases, 100%). The right atrium (RA) was aligned with and opened into the RV, and the left atrium (LA) was aligned with and opened into the LV in all patients. However, there was AV situs concordance in only 34 of 35 cases (97%): { S,D, D}.
In one rare patient, there was AV situs discordance with AV alignment concordance. This patient had DORV {I,D,D} with visceral heterotaxy and asplenia (see Fig. 10.4 ). The AV situs was discordant: {I,D,-}. But the AV alignments were concordant: the left-sided RA was aligned with and opened into the right-sided RV, and the right-sided LA was aligned with and opened into the left-sided LV.
As this rare case illustrates, there are two different types of AV concordance and discordance: (1) alignment concordance/discordance (which atrium opens into which ventricle?) and (2) situs concordance/discordance (i.e., is the situs of the ventricles the same as or different from the situs of the atria?). Almost always, alignment concordance or discordance and situs concordance or discordance are the same; that is, alignment and situs concordance or discordance are congruent. But JAA is the exception. , , AV alignment and situs concordance/discordance can be opposites (incongruent), as in this rare case (see Fig. 10.4 ) and others ( Fig. 10.7 ).
Why is incongruence of AV alignment concordance or discordance and AV situs concordance or discordance a feature of JAA (and of no other form of congenital heart disease, to our knowledge)? Our hypothesis is as follows: In complex congenital heart disease, the atria usually are not malpositioned; they are the “straight men.” Usually only the ventricles are malpositioned; they are the “professional contortionists.” Typically therefore, there is only one variable—the position of the ventricles. But with JAA, there are two variables. The positions of the ventricles and of the atria both are abnormal—making possible rare and surprising AV alignments that one would not expect in view of the atrial and the ventricular situs.
Dr. John Kirklin once asked me whether it is possible for the AV alignments to predict the ventricular situs wrongly. We all now know that the answer is yes. As in this rare case (see Fig. 10.4 ) of DORV {I,D,D} with AV alignment concordance, the fact that the atria were in situs inversus and that there were concordant AV alignments might well suggest that the ventricles were of the L-loop or inverted type. However, D-loop (noninverted) ventricles were found. Incongruity of alignment and situs concordance/discordance has been known and documented since 1970. , , This is one reason why it is necessary to diagnose both the type of AV alignment (concordant, discordant, double-inlet LV, double-inlet RV, etc.) and the type of ventricular situs (D-loop/L-loop ventricles) because the AV alignments can predict the wrong type and ventricular loop or the AV alignments may be nonpredictive, as in double-inlet LV/RV and as in atrial situs ambiguus.
Why can the AV alignments be wrongly predictive or nonpredictive concerning the ventricular situs (D-/L-loop)? Because the AV alignments and the ventricular situs are two different variables. The only diagnostic method that always works accurately is to diagnose each variable specifically: What are the AV alignments (concordant, discordant, other)? What is the ventricular situs (D-loop/L-loop)?
Anderson, Smith, and Wilkinson have recognized the incongruity phenomenon, which they have called disharmony between AV connections and segmental combinations. They also pointed out that such cases are unusual variants of “crisscross” hearts. In our case (see Fig. 10.4 ), for example, the left-sided RA opens into the right-sided RV, and the right-sided LA opens into the left-sided LV. The AV inflow tracts seem to cross each other, instead of normally being approximately parallel and uncrossed. The appearance of crisscross AV relations is produced by a major malposition of the ventricles. The AV inflow tracts really do not intersect and physically cross each other. Nonetheless the appearance (or illusion) of crisscross AV relations is useful as a sign of ventricular malposition. (Crisscross AV relations are discussed and illustrated in Chapter 18 .)
This rare patient (see Fig. 10.4 ) was the first reported case of JAA with right-sided malposition of the RAA in atrial situs inversus.
Incongruence of AV alignment concordance or discordance relative to AV situs concordance or LA appendage (LLA) discordance almost always has occurred with JAA associated with malposition of the LAA , , , and seldom with JAA and malposition of the RA appendage (RAA). Why this difference? We do not know. It should be added that JAA with incongruence of AV alignments and situs is rare, the numbers are small, and judgment should therefore be deferred, awaiting more data.
In JAA with malposition of the RAA, a ventral septal defect (VSD) was present in 32 of 35 patients (91%). The anatomic types of VSD that were present are summarized in Table 10.3 .
Types of VSD | No. | % |
---|---|---|
|
4 | 12.5 |
|
3 | 9 |
|
11 | 34 |
|
1 | 3 |
|
1 | 3 |
|
4 | 12.5 |
|
4 | 12.5 |
|
4 | 12.5 |
∗ A membranous VSD is a small conoventricular defect (between the conal septum above and the ventricular septum and septal band below), in the region of the interventricular portion of the membranous septum, without significant malalignment of the conal septum.
Conoventricular VSD was by far the most common type. Including membranous VSDs and bulboventricular foramina (with a single LV, infundibular outlet chamber, and absence of the right ventricular sinus), conoventricular defects were found in 24 of 32 patients (75%).
However, a bulboventricular foramen customarily is not regarded as a VSD because “ventricular septal defect” is short for interventricular septal defect. Logically, one cannot have an interventricular septal defect if there is only one ventricular sinus, as typically is the case with a single LV, infundibular outlet chamber, and absence of the RV sinus.
Although I agree with the logic of this view, it should also be understood that the morphologic anatomy of a bulboventricular foramen per se (independent of what lies on either side of the defect) typically is the same as that of a conoventricular VSD: Above, the foramen is limited by the bulbar or conal septum. Below, the foramen is limited by the septal band and ventricular septum, or their remnants.
If one would prefer to remove the cases with a bulboventricular foramen from Table 1.3 , the total number of VSDs becomes 28 and the number with conoventricular VSDs is 20 (71%). An intact ventricular septum was found in only 3 of 35 cases (9%).
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here