Totally Anomalous Pulmonary Venous Connection/Drainage
Definition
Totally anomalous pulmonary venous connection (TAPVC) is an anomaly in which none of the pulmonary veins connect normally with the morphologically left atrium (LA).
TAPVC used to be known as total anomalous pulmonary venous drainage. Then, Dr. Jesse Edwards and colleagues proposed that total anomalous pulmonary venous connection would be a more anatomically accurate diagnosis. This proposal seemed correct and was generally adopted. Hence, TAPV C and TAPV D were initially regarded as synonymous.
As will be seen, later we would realize that it is possible for the pulmonary veins to be normally connected, but to have totally anomalous pulmonary venous drainage (TAPVD). Examples include marked levomalposition of the atrial septum to the left of the left pulmonary veins (LPVs), and mitral atresia with an intact atrial septum and decompression of the LA via a levoatrial cardinal vein into a left vertical vein (LVV) and then into the left innominate vein (LIV) and the right superior vena cava (RSVC) and thence to the right atrium (RA). Such cases are presented subsequently ( Figs. 7.1 and 7.2 ).
Classification
In 1957, Darling, Rothney, and Craig classified TAPVC into four anatomic types: (1) supracardiac, (2) cardiac, (3) infracardiac, and (4) mixed.
Historical Note
Dr. John Craig, the senior author of this landmark paper, was my principal teacher of pediatric pathology at the Children’s Hospital Medical Center (as it was then called) in Boston, from 1956 to 1957. Dr. Bill Rothney was a Senior Resident in Pathology. I did not know Dr. Darling.
In 1962, when I was taking the oral part of the examination for the American Sub-board of Pediatric Cardiology, Dr. Abraham Rudolph asked me, “What is the embryologic basis of anomalous pulmonary venous drainage?” I replied by telling Dr. Rudolph the Darling, Rothney, and Craig classification of TAPVC, which seemed to satisfy him. Years later, I realized that my answer had been not wrong but certainly superficial, as will soon be seen.
Embryologically, TAPVC represents failure of development of the common pulmonary vein, as was appreciated by Lucas et al in 1962. As a consequence of failure of development of the common pulmonary vein, an anastomosis almost always persists and enlarges between the pulmonary venous plexus of the lung buds and the systemic veins.
As the classification of TAPVC indicates, such pulmonary venous–to–systemic venous anastomoses can occur at the supracardiac level between the lungs and the anterior cardinal veins ( Fig. 7.3 ), at the cardiac level between the lungs and the sinus venosus, at the infracardiac level between the lungs and the ductus venosus (see Fig. 7.3 ), or at several of the previously mentioned levels in the mixed form of TAPVC.
These normal anastomoses between the pulmonary venous plexus and the systemic venous plexus usually undergo involution after the development of the common pulmonary vein. But if the common pulmonary vein fails to develop, these pulmonary-systemic venous anastomoses can persist, resulting in TAPVC.
It will be seen that the previously mentioned classification of TAPVC is a very general classification. There are multiple different anatomic subtypes of supracardiac, cardiac, infracardiac, and mixed TAPVC (see Fig. 7.3 ).
Embryology
The normal development of the common pulmonary vein and of the sinus venosus in humans is depicted diagrammatically in Fig. 7.4 . We were able to identify the common pulmonary vein with certainty in human embryos at the 4.6-mm stage (crown-rump length), when the estimated age since ovulation was 27 days ( Fig. 7.5 ). Normally, the common pulmonary vein appears to grow outward from the dorsal atrial wall into the lung buds. Note that the common pulmonary vein is located within a “horseshoe” formed by the right and left horns of the sinus venosus (see Fig. 7.4 ).
When the TAPVC pathway is supracardiac or infracardiac, the anomalous venous pathway is much longer than normal. The common pulmonary vein (see Figs. 7.4 and 7.5 ) provides a much shorter route for the pulmonary venous return from the lungs to the LA. This normal shortcut provided by the common pulmonary vein appears to lead to the involution through disuse atrophy of the various anastomoses between the pulmonary and the systemic veins.
From a developmental standpoint, these so-called anomalous pulmonary venous connections are neither anomalous nor pulmonary. They are normal embryonic pathways. But in the postnatal individual, they are abnormal, that is, not usual. These totally anomalous pulmonary venous pathways are not pulmonary veins. In TAPVC, the basic problem is usually that the common pulmonary vein is absent or atretic.
It is also noteworthy that one can have normal pulmonary venous connections to the LA coexisting with anomalous pulmonary venous connection in the same patient: to the RSVC in one of our patients and to the left superior vena cava (LSVC) and thence to the coronary sinus in another patient. Thus, normal and anomalous pulmonary venous connections rarely can coexist in the same patient, if the normal early embryonic anastomoses between the lungs and the systemic veins fail to involute. Although such rare patients have anomalous pulmonary venous connections that resemble TAPVC, they cannot be said to have TAPVC because the pulmonary veins are also normally connected. Can such rare patients be said to have partially anomalous pulmonary venous connection (PAPVC)? We think that the answer is no—because all parts of both lungs appear to have both normal and anomalous pulmonary venous connection and drainage. There is no part of either lung that drains only anomalously.
Consequently, we think that such rare patients have a newly recognized anomaly: the coexistence of normally connected and anomalously connected pulmonary veins.
What really is TAPVC? The short (and incomplete) answer to this question is absence of the common pulmonary vein and the sequelae thereof (i.e., the persistence of various anastomoses between the pulmonary veins and the systemic veins). This is not the whole story.
But what do we really mean when we say that the common pulmonary vein is “absent”? Failure of the common pulmonary vein to develop in TAPVC may be due to several different processes: (1) agenesis—failure to appear; (2) involution—disappearance; or (3) atresia—appearance, but remaining an uncanalized cordlike strand ( Fig. 7.6 ). In the great majority of cases (97%), no remnant of the common pulmonary vein was found, supporting the hypotheses that most patients with TAPVC have agenesis, or involution of the common pulmonary vein.
In dealing with postnatal anatomy, most of us are used to the notion that normally there are four or five pulmonary veins—one from each lobe of the lungs: two pulmonary veins from the left lung and two or three pulmonary veins from the right lung.
Developmentally, however, there is really only one pulmonary vein, which normally is incorporated into the LA up to just beyond the primary division of each branch. The incorporation of the common pulmonary vein into the LA in normal human embryos is shown in Fig. 7.7 .
The horizontal vein (see Fig. 7.3 ) that runs from pulmonary hilum to hilum in TAPVC—except in some mixed types—is not the common pulmonary vein, which is absent in TAPVC. ,
Thus, this is why we talk about the horizontal pulmonary vein (see Fig. 7.3 ), not the common pulmonary vein, because the common pulmonary vein typically is absent in TAPVC.
Obstruction
In a detailed study published in 1976 of 93 autopsied cases of TAPVC, Delisle et al found that TAPVC is often a rapidly lethal disease and that obstruction of the anomalous venous pathway adversely effects longevity. This was perhaps best indicated by the median ages at death: whole series, 7 weeks; isolated TAPVC with obstruction of the anomalous venous pathway, 3 weeks; and isolated TAPVC without obstruction of the anomalous venous pathway, 3 months.
In the supracardiac type of isolated TAPVC, we were surprised to find that the incidence of obstruction was remarkably high, at 50%.
In the “snowman” type of supracardiac isolated TAPVC, obstruction occurred at two sites: (1) behind the left pulmonary artery (LPA) and in front of the left mainstem bronchus, which Dr. Jesse Edwards and colleagues graphically called the vascular vise, as in Fig. 7.3A ; and (2) at the junction of the LIV with the RSVC, as in Fig. 7.3B .
Why is the type of TAPVC shown in Fig. 7.3A–B known as the snowman type? Because, after about 4 months of age, one can often see a shadow above the cardiac silhouette on the plain posteroanterior chest x-ray film that is formed by the prominent LVV, LIV, and RSVC. The heart shadow forms the “body,” and the venous shadow forms the “head” of the snowman.
In the snowman type of TAPVC, the vascular vise (between the LPA anteriorly and the left bronchus posteriorly, see Fig. 7.3A ) was the more common form (75%). When the LVV passed from the horizontal vein upward and in front of the LPA, as it usually does, the snowman type of TAPVC was not associated with obstruction at this site.
In Fig. 7.3A , the anomalous pulmonary venous pathway is viewed from behind. From the left end of the horizontal vein, the anomalous pathway heads superiorly, just beneath the pleura of the left lung. Then the venous pathway emerges into the mediastinum, turns inferiorly, and then curls superiorly and passes between the LPA anteriorly and the left bronchus posteriorly. In contrast with Fig. 7.3A , Fig. 7.3B is viewed from the front. Noting the labels of the lungs ( LL = left lung, RL = right lung; see Fig. 7.3 ) makes it immediately obvious whether the TAPVC is being viewed from the front (see Fig. 7.3B–D , F ) or from behind ( Fig. 7.3A, E , G–I ).
In one-third of these obstructed snowmen, there was marked poststenotic dilatation of the LVV above the level of the LPA and the left mainstem bronchus ( Figs. 7.8 and 7.9 ), as was also observed by Kauffman et al.
Why are we talking about the LVV (see Fig. 7.3 )? Isn’t this vein derived embryologically from the left anterior cardinal vein? We think that the answer is yes. Why then do we not call this vein the LSVC? The answer is because it does not return to the heart as a persistent LSVC does, either to the coronary sinus in visceroatrial situs solitus or to the left-sided morphologically RA in visceroatrial situs inversus. We think there must be developmental differences between the LVV (see Fig. 7.3 ) and LSVC. Inferiorly, at the level of the left bronchus and the LPA, the LVV may not be derived from the left anterior cardinal vein, accounting for the anatomic differences between the LVV in TAPVC (see Fig. 7.3 ) and a typical persistent LSVC.
In the less common site of obstruction of the snowman type of supracardiac TAPVC, the opening of the LIV into the RSVC can be remarkably small: only 2 mm in diameter, with marked poststenotic dilatation of the RSVC opposite the stenotic orifice of the LIV. The cause of this stenosis remained unclear.
In the RSVC type of supracardiac isolated TAPVC, the majority were obstructed (67%). Two patients had severe stenosis at the point of entry of the oblique connecting vein into the RSVC (see Fig. 7.3C ). Another had stenosis (hypoplasia) of the anomalous venous pathway, both within the right lung and within the mediastinum—between the right lung and the RSVC (see Fig. 7.3D ).
An additional patient had a right-sided vascular vise because the oblique connecting vein ran from the horizontal vein inferiorly and passed superiorly and rightward between the right pulmonary artery (RPA) anteriorly and the right bronchus posteriorly to reach the RSVC (see Fig. 7.3E ).
In the azygos vein type of supracardiac isolated TAPVC, both patients had severe obstruction. In one, there was severe hypoplasia of the oblique connecting vein, the diameter being less than 1 mm (see Fig. 7.3 ). The other patient had marked stenosis of the entry of the azygos vein into the RSVC.
In the coronary sinus type of cardiac isolated TAPVC ( Fig. 7.10 ), none had obstruction. An example of familial TAPVC was found in this type: a 1½-month-old white male infant had a sibling who had died 2 years previously from subdiaphragmatic TAPVC.
In the ductus venosus type of infracardiac isolated TAPVC (see Fig. 7.3G–I ), obstruction was thought to be 100% (n = 14). The anomalous pulmonary venous pathway always led to the ductus venosus and then continued as follows: (1) to the left portal vein in 8 patients (see Fig. 7.3G ); (2) to the inferior vena cava (IVC) in 3 cases (see Fig. 7.3H ); (3) to the left portal vein and the IVC in 1 patient, because both portions of the ductus venosus (leading to the left portal vein and to the IVC) were patent; (4) to the left portal vein, to the IVC, and to the left gastric vein in 1 patient (see Fig. 7.3I ); and (5) with no continuation whatever in 1 case because of closure , of the ductus venosus leading both to the left portal vein and to the IVC ( Fig. 7.11 ). This patient was a male identical twin; the other twin was normal.
How was it possible for this patient with atresia of both parts of the ductus venosus (see Fig. 7.11 ) to survive for 6 days postnatally? ( Atresia literally means a “absence of” tresis “a hole,” Greek.) We really do not know the answer to this question. We speculated that perhaps pulmonary-to-bronchial venous anastomoses with retrograde blood flow may have facilitated this infant’s survival. It is also true that we do not know when the ductus venosus became atretic; the ductus venosus may have been patent for some portion of this patient’s postnatal life. Alternatively, the ductus venosus may have closed prenatally, the circulation depending on a patent ductus arteriosus (PDA). If the ductus venosus closed prenatally, the patient had a ductus arteriosus–dependent circulation. Hence, postnatal patency of the ductus arteriosus may well have contributed to this patient’s postnatal survival.
This patient (see Fig. 7.11 ) makes it clear that TAPVC below the diaphragm is a ductus venosus–dependent anomaly. The natural history is if the ductus venosus closes completely ( Fig. 7.12 ), and the ductus arteriosus also closes, the patient dies.
The infracardiac type of TAPVC can be obstructed not only by the narrowing or closure of the ductus venosus, but also by extrinsic compression of the paraesophageal vertical vein at the diaphragm. Stenosis at the diaphragm (see Fig. 7.3G ) was found in 3 of these 14 patients (21%).
When the obstruction of the anomalous pulmonary venous connection is very severe or complete (i.e., when atresia is present), diagnostic studies such as cardiac catheterization and angiocardiography can be diagnostically misleading, as occurred in the patient shown in Figs. 7.11 and 7.12 . The posteroanterior chest x-ray film ( Fig. 7.13 ) showed a normal-sized heart with “ground-glass” lung fields, accurately suggesting the correct diagnosis of TAPVC with atresia of the anomalous connection. However, cardiac catheterization revealed no localized oxygen step-up and angiocardiography showed no anomalous pulmonary venous pathway—both because this connection was atretic (see Figs. 7.11 and 7.12 ). Information from the cardiac catheters led to the erroneous conclusion that this patient probably had lung disease, not congenital heart disease.
Dr. Ed Neuhauser, our Chief of Radiology at that time, took one look at the chest x-ray film (see Fig. 7.13 ) and said, “Totally anomalous pulmonary venous connection below the diaphragm with obstruction, right?” It took an autopsy to prove that this diagnosis was indeed right (see Fig. 7.11 and 7.12 ). In retrospect, we understood that one should expect negative cardiac catheterization and angiocardiography findings in TAPVC with severe obstruction.
Twins occurred twice in these 14 patients with TAPVC to the ductus venosus. Both were identical twins, and in both pairs the co-twin was normal.
The mixed type of TAPVC was always isolated and was occasionally obstructed (20%). In the study of Delisle et al, 5 of 93 cases (5%) had anomalous pulmonary venous connections at more than one level:
- 1.
a snowman pathway from both lungs to the RSVC that also had a stenotic connection from both lungs to the coronary sinus;
- 2.
a snowman connection that also had anomalous connections from the right upper lobe to the RSVC;
- 3.
the left lung draining via a snowman pathway and the right lung draining into the coronary sinus;
- 4.
an infracardiac connection to the portal vein and a small snowman connection; and
- 5.
a subdiaphragmatic connection to the portal vein, with connections from the right upper lobe to the RSVC.
The latter 2 patients with anomalous connections to the ductus venosus were both thought to have an obstruction. Isolated TAPVC means that no other congenital heart disease is present. Nonisolated TAPVC means that another congenital heart disease coexists.
Delisle et al appreciated this distinction between isolated and nonisolated TAPVC. They also understood that there are two different groups within nonisolated TAPVC: without heterotaxy and with heterotaxy.
Nonisolated Totally Anomalous Pulmonary Venous Connection Without Heterotaxy
In this series of 93 autopsied cases, there were 12 (13%) that had nonisolated TAPVC with congenital heart disease but without heterotaxy: double-outlet right ventricle (DORV) with pulmonary stenosis in 3; D-transposition of the great arteries (TGA) in 3; mitral atresia in 2; and tricuspid atresia, double-inlet left ventricle (LV), and pulmonary artery sling in 1 case each.
Extracardiac anomalies included cat eye syndrome, conjoined twins, and agenesis of the right lung with anomalous pulmonary venous drainage. The latter situation may be called total partially anomalous pulmonary venous connection.
TAPVC with intact atrial septum is very rare. Such a case was found in this group. A persistent LSVC drained into the coronary sinus, as did all of the pulmonary veins. The cephalic end of the LSVC was markedly dilated ( Fig. 7.14 ). Although the atrial septum was intact, there were multiple ventricular septal defects (VSDs), making it possible for oxygenated blood to reach the left heart and the systemic circulation. Hastreiter et al published a similar case: TAPVC with an intact atrial septum and a PDA.
Nonisolated Totally Anomalous Pulmonary Venous Connection With Heterotaxy
There were 23 such cases (25% of the series) : asplenia, 14 cases; polysplenia and rudimentary spleen, 8 cases; and heterotaxy with a normally formed spleen, 1 patient.
Classification of Totally Anomalous Pulmonary Venous Connection
The salient anatomic findings in these 93 postmortem cases of TAPVC are summarized in Tables 7.1 to 7.4 .
| Frequency | Site of Anomalous Connection | No. of Cases | % of Series |
|---|---|---|---|
| 1 | To left innominate vein (“snowman”) | 24 | 26 |
| 2 | To ductus venosus (infracardiac) | 22 | 24 |
| 3 | To coronary sinus | 17 | 18 |
| 4 | To right superior vena cava | 14 | 15 |
| 5 | To right atrium a | 7 | 8 |
| 6 | To more than one level (mixed) | 5 | 5 |
| 7 | To azygos vein | 2 | 2 |
| 8 | To left superior vena cava | 2 | 2 |
a Totally anomalous pulmonary venous drainage (TAPVD), not connection, to the right atrium.
| % of Series | No. of Cases | |
|---|---|---|
| Isolated TAPVC | 58 | 62 |
| Nonisolated TAPVC | 35 | 38 |
| Without heterotaxy | 12 | 13 |
| With heterotaxy a | 23 | 25 |
a Asplenia, 14 cases; polysplenia and rudimentary spleen, 8 cases; asplenia-like heterotaxy with normally formed spleen, 1 case.
| Site of Connection | Isolated | Nonisolated | No. of Cases | % |
|---|---|---|---|---|
| Supracardiac | ||||
| To left innominate vein (snowman) | 20 | 4 | 24 | 26 |
| To right superior vena cava | 6 | 8 | 14 | 15 |
| To azygos vein | 2 | 0 | 2 | 2 |
| To left superior vena cava | 0 | 2 | 2 | 2 |
| Cardiac | ||||
| To coronary sinus | 11 | 6 | 17 | 18 |
| To right atrium a | 0 | 7 | 7 | 8 |
| Infracardiac | ||||
| To ductus venosus | 14 | 8 | 22 | 24 |
| Mixed | ||||
| To more than one level | 5 | 0 | 5 | 5 |
a Totally anomalous pulmonary venous drainage (not connection) to the right atrium.
| Site of Connection | Obstructed | Not Obstructed | Median Age at Death | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| No. of Cases | % of Group | No. of Cases | % of Group | Obstructed | Not Obstructed | |||||
| Left innominate vein (snowman) | 8 | 40 | 12 | 60 | 4 wk | 14 wk | ||||
| Right superior vena cava | 4 | 67 | 2 | 33 | 2 wk | 2 1/2 y a | ||||
| 9 1/3 y a | ||||||||||
| Azygos vein | 2 | 100 | 0 | 0 | 36 h | |||||
| 23 d | ||||||||||
| Coronary sinus | 0 | 0 | 11 | 100 | 0 | 7 wk | ||||
| Ductus venosus | 14 | 100 | 0 | 0 | 3 wk | 0 | ||||
| Mixed | 2 | 40 | 3 | 60 | 17 d | 5 mo | ||||
| 8 wk | ||||||||||
a Surgical deaths. When there are only 2 cases in a group, both ages at death are given, rather than the median.
Surgical Considerations
From the surgical standpoint, how true is the generalization that it does not really matter what anatomic type of TAPVC is present or whether it is obstructed, because there always is a horizontal vein running from pulmonary hilum to hilum (see Fig. 7.3 ) that the surgeon can anastomose to the dorsal wall of the LA, as is seen in Fig. 7.9 ? Specifically, is the “surgeon’s friend”—the horizontal vein (see Fig. 7.3 )—always present, just “waiting” to be anastomosed to the LA? The answer is that typically there is a horizontal vein, but with the following exceptions:
- 1.
TAPVC to the RA,
- 2.
TAPVC to the coronary sinus, and
- 3.
occasionally in the mixed type of TAPVC.
TAPVC to RA
In TAPVC to the RA, there is no discrete horizontal vein that can be anastomosed to the LA in the usual way. The pulmonary veins connect directly with what has been interpreted in the past as the dorsal wall of the RA, without forming a horizontal vein. From the anatomic standpoint, one approach to surgical repair is atrial septectomy, followed by surgical construction of a new atrial septum that will direct the pulmonary venous return into the LA and/or to the physiologically appropriate atrioventricular (AV) valve.
In 1995, S. Van Praagh et al published a study of 36 patients, 21 with postmortem confirmation and 15 living patients, in which there was partially anomalous (44%) or totally anomalous (56%) pulmonary venous drainage into the RA because of displacement of the septum primum into the LA. Displacement of the septum primum—leftward in atrial situs solitus or rightward in atrial situs inversus—was present in all patients and appeared to be responsible for the anomalous pulmonary venous drainage.
For example, in visceroatrial situs solitus, when the septum primum was displaced to the left so that the atrial septum lay between the right and LPVs, this created the appearance known as ipsilateral pulmonary veins, which is characteristic of the polysplenia syndrome and has led to the (we think erroneous) interpretation of bilaterally left atria or left atrial isomerism ( Fig. 7.15 ) in visceral heterotaxy with polysplenia.
When the atrial septum is displaced even further to the left so that the septum primum lies to the left of all of the pulmonary veins, this creates the appearance generally known as TAPVC to the RA (see Fig. 7.1 ) (see Tables 7.1 to 7.3 ). The atrial septum can be displaced so far to the left that it resembles a supramitral membrane and can result physiologically in significant supramitral stenosis and LV hypoplasia. Surgically, such a markedly malpositioned atrial septum should be excised and replaced with a normally located atrial septum—currently fashioned from glutaraldehyde-fixed pericardium.
Does this mean that the conventional diagnosis of “TAPVC to the RA” is wrong? We now think that the answer is yes. This diagnosis should be modified to TAPVD to the RA. Accurately speaking, ipsilateral pulmonary veins have partially anomalous pulmonary venous drainage (not connection).
Whenever the pulmonary veins connect with the atria, within the “horseshoe” formed by the right and left horns of the sinus venosus (see Fig. 7.4 ), we think that the pulmonary veins are connected normally . However, the pulmonary venous drainage can be partially anomalous (see Fig. 7.15 ) or totally anomalous ( Fig. 7.16 ), depending on how malpositioned the septum primum is. In Figs. 7.1 and 7.15 , note how malpositioned the atrial septum is relative to the ventricular septum (which is normally located). Normally, the atrial septum is approximately parallel with the ventricular septum, not markedly angulated as it is in Figs. 7.1 and 7.15 . Displacement of septum primum into the LA often can be more obvious echocardiographically than it is anatomically.
From the right atrial view, the septum primum is much too easy to see, that is, much better seen from the right atrial perspective than normally is the case (see Fig. 7.16 ). (This intriguing fact was first pointed out to us by Dr. Luis Alday of Cordoba, Argentina.) Why is this so? We think that the answer is because the superior limbic band of the septum secundum, which normally covers the upper part of the septum primum, is poorly developed or absent. This is particularly true in patients with polysplenia and visceral heterotaxy. Rarely, this can also occur in patients with asplenia or with a normally formed spleen and visceral hetertoaxy.
Why can septum primum be displaced into the morphologically LA, resulting in partially or TAPVD? Our hypothesis is as follows: When the superior limbic band of the septum secundum is absent or very poorly formed, the upwardly growing septum primum has no septum secundum to attach to superiorly, and consequently systemic venous blood flow from the IVC and SVC can displace the unattached septum primum into the LA.
This new understanding of the role of malposition of the septum primum has several significant sequelae:
- 1.
TAPVC and TAPVD are two different things. What was formerly thought to be TAPV C to the RA is now thought to be TAPV D to the RA, with normally pulmonary venous connections and malposition of the septum primum.
- 2.
Septum primum malposition of the atrial septal defect (ASD) (see Fig. 7.1 ) is a newly recognized anatomic type of ASD that lies between the posterior margin of the malpositioned septum primum and the posterior wall of the LA.
- 3.
This is a new understanding of TAPVC/D. The supracardiac and infracardiac forms of TAPVC do indeed appear to result from failure of development of the common pulmonary vein, almost always leading to the persistence of early embryonic anastomoses between the pulmonary and the systemic veins, as described earlier.
But what is TAPVC to the coronary sinus? The common pulmonary veins may have developed and connected abnormally with the left sinus horn, because of anomalous development of the common pulmonary vein, abnormal development of the left sinus horn (coronary sinus), or both. Does the common pulmonary vein grow out dorsally from the atrial wall to tap into the pulmonary venous plexus, as Fig. 7.4 suggests? Or does the common pulmonary vein grow in both directions (dorsally from the heart and ventrally from the developing lungs)? Are there species differences? All of these questions have been answered positively. Consequently, we think that definitive resolution of these questions embryologically may well help clarify the morphogenesis of TAPVC to the coronary sinus.
The pulmonary veins in TAPVC to the coronary sinus look like pulmonary veins, suggesting that the common pulmonary vein did not fail to develop. By contrast, the common pulmonary vein does appear to be absent in the supracardiac and infracardiac forms of TAPVC.
Hypothesis
If the foregoing hypothesis is correct, this means that the supracardiac and infracardiac forms of TAPVC are indeed characterized by failure of development of the common pulmonary vein, whereas the cardiac forms of TAPVC and TAPVD are not. TAPVD to the RA appears to have normally connected pulmonary veins, with malposition of the septum primum into the LA. In TAPVC to the coronary sinus, the common pulmonary vein appears to have developed and connected abnormally to the immediately subjacent left sinus horn (see Fig. 7.4 ).
For us, the foregoing is a new developmental insight, namely, that TAPVC and TAPVD appear to result from three different morphogenetic processes:
- 1.
failure of development of the common pulmonary vein, resulting in the supracardiac and infracardiac forms of TAPVC;
- 2.
atrial septal malposition, resulting in partially or TAPVD to the RA; and
- 3.
abnormal connection of the common pulmonary vein, resulting in TAPVC to the coronary sinus.
It is understood that there may ultimately prove to be more than three developmental mechanisms that can result in TAPVC and TAPVD. However, the foregoing is our best present understanding.
We would like to pay tribute to Dr. Jesse Edwards, who in 1953 proposed the concept that TAPVD may be due to malposition of the atrial septum:
Anomalous connection [of the pulmonary veins] with the superior portion of the RA may be explained on the basis of abnormality of the atrial septum. If the septum develops farther to the left than is normal, that outpouching of the sinoatrial region which joins the pulmonary vessels may lie to the right of the atrial septum and the entire venous system of the lungs will then connect with the right atrium. Lesser degrees of abnormal positioning of the atrial septum may account for cases in which the left pulmonary veins enter the LA while those of the right lung enter the right atrium.
This concept was reaffirmed by Moller et al in 1967. After our study had been completed in 1994, we were delighted to rediscover Edwards’ forgotten hypothesis , and to be able to demonstrate that the pathologic anatomic findings do indeed show septum primum malposition, just as Edwards had foreseen. Now we must continue with those anomalies in which there is no horizontal vein for surgical anastomosis with the LA.
TAPVC to the Coronary Sinus
In TAPVC to the coronary sinus, there is no discrete horizontal vein. However, the coronary sinus can function as a horizontal vein from the surgical standpoint, and a large “window” can readily be made between the coronary sinus and the LA (see Figs. 7.10 and 7.17 ). The operative steps are as follows (see Fig. 7.17 ).
The RA is opened horizontally, from the right atrial appendage anteriorly and extending the incision posteriorly, stopping anterior to the sulcus terminalis to avoid injury to the tail of the sinoatrial (SA) node and to the posterior internodal tract or preferential internodal pathway. From the surgeon’s perspective in the operating room, the SVC is to the left, the IVC is to the right, and the right atriotomy incision appears to be in a longitudinal or vertical direction, as opposed to a latitudinal or horizontal direction. This opening incision avoids both the SA node lateral to the entry of the SVC and the sulcus terminalis and crista terminalis where the posterior internodal tract or preferential pathway runs (see Figs. 7.10 and 7.17A ). The aim of this careful placement of the right atriotomy is to avoid sick sinus syndrome or other atrial arrhythmias postoperatively.
The coronary sinus is relatively huge and has a cornucopia-like shape. A right-angle clamp is inserted into the markedly enlarged coronary sinus (see Figs. 7.17A and 7.17B ). The apposed (conjoined) anterior wall of the coronary sinus and the posterior wall of the LA are then pushed up with the tip of the clamp so that the conjoined wall is displaced and can be seen through the patent foramen ovale (see Fig. 7.17B ).
This tented-up wall is then grasped with toothed forceps and a piece is cut out, leaving an opening of at least 15 mm in diameter (see Fig. 7.17C ). In this manner, a wide and sutureless opening is created between the coronary sinus posteriorly and the LA anteriorly (see Fig. 7.17D ). This opening is made well within the coronary sinus, to the left of the orifice of the coronary sinus, so as not to injure the posterior internodal pathway or preferential tract that courses near the ostium of the coronary sinus (see Fig. 7.10 ).
The ostium of the coronary sinus is then closed with a running horizontal mattress stitch, placing the sutures at least 4 to 5 mm inside the ostium (see Fig. 7.17E ), to avoid injury to the AV node and the internodal preferential conduction pathways (see Fig. 7.10 ). Alternatively, the coronary sinus ostium may be closed using a patch, placed well inside the large coronary sinus, to avoid any distortion of the surgically created coronary sinus septal defect—the “window” between the coronary sinus posteriorly and the LA anteriorly.
The patent foramen ovale is then sutured closed (see Fig 7.17F ), placing the sutures through the left side of the superior limbic band of septum secundum to spare the middle internodal preferential conduction pathway that runs along the right side of the superior limbic band of septum secundum.
Parenthetically, it should be added that whether the internodal tracts shown in Fig. 7.10 actually exist as anatomically discrete internodal tracts or alternatively are preferential internodal conduction pathways (but not anatomically discrete tracts) remains controversial. This is why we have referred to them both as internodal tracts and as internodal preferential conduction pathways . Within either interpretation, these internodal regions are thought to be electrophysiologically important and hence should not be inadvertently damaged surgically.
The foregoing operation for the correction of TAPVC to the coronary sinus ( Fig. 7.17 ) is now known as the Van Praagh procedure. This operation was “born” in the Cardiac Registry of Children’s Hospital Boston. The surgeon had attempted to redirect the coronary sinus blood flow through a surgically enlarged patent foramen ovale using a large U-shaped conduit between the coronary sinus ostium below and enlarged patent foramen ovale above. Autopsy revealed that the large U -shaped conduit had obstructed the IVC blood stream almost totally, leading to the death of the patient. The operation described previously was designed (1) to avoid vena caval obstruction, (2) to avoid pulmonary venous obstruction, and (3) to avoid interruption of the preferential internodal electrophysiologyic pathways between the SA and AV nodes.
On March 17, 1971, Dr. Robert E. Gross, the legendary pioneer of congenital heart surgery who first ligated a PDA in Lorraine Sweeney on August 26, 1938, sent me the following note ( Fig. 7.18 ):
Dr. Van Praagh. Richard! Hail! Your operation is simply superb. It went just like fine clockwork. There was no difficulty making the window between the coronary sinus and the left auricle. There is another aspect to this—it greatly increases the volume of the left auricle. And the coronary sinus was so huge there was no trouble at all moving around inside of it and closing it off. There is no A-V block. Many thanks. REG
Mixed Type of TAPVC
In the mixed type of TAPVC, occasionally there is no horizontal vein for the surgeon to suture to the LA. In the study by Delisle et al, this situation was found in only 1 of 5 cases of mixed TAPVC (20%): the left lung drained via a snowman connection, and the right lung drained into the coronary sinus.
Agenesis of the right lung with anomalous pulmonary venous connection of the left lung merits mention. There were two such patients in the study by Delisle et al : a snowman connection in one, and a coronary sinus connection in the other. However, the reason that this situation is noteworthy is that it can mimic a vascular ring . Agenesis of the right lung resulted in extrinsic dextrocardia. Because the heart was abnormally right sided, the normal left aortic arch compressed the tracheobronchial tree anteriorly and superiorly. To make matters worse, the large LPA was posterior to the left bronchus. (Normally, the LPA is anterior to the left bronchus.) Hence, the large LPA compressed the left bronchus posteriorly and inferiorly. The ligamentum arteriosum, and in one case an aberrant right subclavian artery, facilitated the external tracheobronchial compression.
From the surgical standpoint, in addition to correction of the anomalous pulmonary venous connection, steps to relieve the vascular tracheobronchial compression also may well be necessary, such as:
- 1.
division of the ligamentum arteriosum, and
- 2.
division of an aberrant right subclavian artery, if present, and aortopexy—attaching the aorta anteriorly in a subcostal or substernal location to reduce or eliminate the anteroposterior tracheobronchial compression.
Thus, with agenesis of the right lung, tracheobronchial compression appeared to be as important as the TAPVC, suggesting that both problems should be managed surgically.
With the exception of the aforementioned new morphogenetic understanding (three different mechanisms), the foregoing is essentially what we knew before undertaking the present study for this book. That which follows is what we have learned very recently by reviewing all of our data.
Findings
This is a study of 204 postmortem cases of TAPVC and TAPVD that includes the 93 cases of Delisle et al referred to previously.
Prevalence: TAPVC/D constitutes 6.34% of the cases of congenital heart disease in this study (204 of 3216).
Sex: The sex was known in 199 cases. Males = 121/199 (60.8%). Females = 78/199 (39.2%). Males-to-females = 121/78 = 1.55:1.0. Thus, there was a male predominance (61% versus 39%).
Age at Death: The age at death was known in 196 patients: mean = 448 · 959 ±
1239 · 519 days, that is, 1.23 years ± [1 standard deviation] 3 · 40 years, ranging from 0 (fetal demise) to 22.63 years. The median age at death was 39 days (1.3 months).
Heart position: The heart position was known in 202 patients: levocardia, 161 cases (79.70%); dextrocardia, 38 cases (18.81%); and mesocardia, 3 cases (1.49%).
Types of Patient: What types of patient had TAPVC/D? In other words, how many had isolated TAPVC/D, that is, with no other congenital heart disease and with no other associated malformations? How many had nonisolated TAPVC/D, that is, with other congenital heart disease and/or with associated malformations? And in the nonisolated TAPVC/D group, what were the other forms of congenital heart disease and/or associated malformations? These questions are answered briefly in Table 7.5 .
TABLE 7.5
Types of Patients With TAPVC/D (n = 11)
Type of Patient
No. of Cases
% of Series
Isolated
102
51
Nonisolated
98
49
Heterotaxy with asplenia
58
29
Multiple congenital anomalies
23
11.5
Heterotaxy with polysplenia
8
4
Conjoined twin
5
2.5
Heterotaxy with normal spleen
2
1
Congenital heart block
1
0.5
Ellis-van Creveld syndrome
1
0.5
TAPVC/D, Totally anomalous pulmonary venous connection/drainage.
Isolated TAPVC/D accounted for only slightly more than half of this series (51%; see Table 7.5 ). Nonisolated TAPVC/D was prominent (49%; see Table 7.5 ). It is interesting how different these numbers are from our earlier study of 93 postmortem cases of TAPVC/D (see Table 7.2 ), in which isolated TAPVC accounted for 62% and nonisolated TAPVC for 38% ( p = .07, i.e., not significant, but close).
In the present study, done in 2003, no case was omitted for any reason. The size of the nonisolated group was somewhat increased by the number of patients with visceral heterotaxy (n = 68; see Table 7.5 ). Heterotaxy accounted for 69% of the nonisolated group of TAPVC (see Table 7.5 ), similar to the earlier study (66%) (see Table 7.2 ).
Table 7.6 conveys the anatomic complexity of these cases of TAPVC:
- 1.
7 different kinds of heart with normally related great arteries (solitus normally related, or inversus normally related);
- 2.
6 different kinds of heart with TGA;
- 3.
12 different kinds of heart with DORV;
- 4.
2 different types of double-outlet infundibular outlet chamber (with absence or marked hypoplasia of the right ventricular sinus, body, or inflow tract); and
- 5.
2 different types of anatomically corrected malposition of the great arteries.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Anatomic Types of TAPVC/D
Perhaps the most important realization of the present study is that the anatomic types of TAPVC/D are linked to the type of visceroatrial situs that is present. The anatomic types of some forms of TAPVC/D are significantly different, depending on the anatomic type of visceroatrial situs that coexists ( Tables 7.7 to 7.11 ).
| Anatomic Type of TAPVC/D | No. of Cases | % of 126 Cases |
|---|---|---|
| 38 | 30.16 |
| 32 | 25.40 |
| 24 | 19.05 |
| 13 | 10.32 |
| 11 | 8.73 |
| 5 | 3.97 |
| 2 | 1.59 |
| 1 | 0.79 |
| Anatomic Type of TAPVC/D | No. of Cases | % of 58 Cases |
|---|---|---|
| 31 | 53.45 |
| 16 | 27.59 |
| 6 | 10.34 |
| 3 | 5.17 |
| 1.72 | |
| 1 | 1.72 |
| Anatomic Type of TAPVC/D | No. of Cases | % of 8 Cases |
|---|---|---|
| 5 | 62.5 |
| 2 | 25.0 |
| 1 | 12.5 |
| Anatomic Type of TAPVC/D | No. of Cases | % of 2 Cases |
|---|---|---|
| 1 | 50 |
| 1 | 50 |
| Anatomic Type of TAPVC/D | No. of Cases | % of 4 Cases |
|---|---|---|
| 2 | 50 |
| 1 | 25 |
| 1 | 25 |
For example, in visceroatrial situs solitus, the most common form of TAPVC/D was the snowman type to the LIV (38 cases, 30.16% of 126 cases; Table 7.7 ). By contrast, the snowman type of TAPVC was the rarest form in heterotaxy with asplenia (1 case, 1.72% of 58 cases; Table 7.8 ). TAPVC to the LIV (the snowman type) did not occur at all in the other nonsolitus types of visceroatrial situs: heterotaxy with polysplenia (n = 8; Table 7.9 ); heterotaxy with a normally formed spleen (n = 2; Table 7.10 ); and situs inversus of the viscera and atrial (n = 4; Table 7.11 ). This difference in the prevalence of the snowman type of TAPVC in visceroatrial situs solitus compared with nonsolitus types of visceroatrial situs is statistically highly significant ( p < .0001, x 2 = 23.97).
Conclusions
TAPVC to the LIV (snowman type) was common in visceroatrial situs solitus (38; see Table 7.7 ), but rare in heterotaxy with asplenia (1.7%; see Table 7.8 ), and was not observed in small series of heterotaxy with polysplenia (see Table 7.9 ), heterotaxy with normally formed spleen (see Table 7.10 ) or in visceroatrial situs inversus (see Table 7.11 ).
By contrast, the prevalence of TAPVC to the ductus venosus was approximately the same in visceroatrial situs solitus (25.4%; see Table 7.7 ) and visceral heterotaxy with asplenia (27.6%; see Table 7.8 ). TAPVC to the ductus venosus was not observed in heterotaxy with polysplenia (see Table 7.9 ) or in heterotaxy with a normally formed spleen (see Table 7.10 ), but did occur in visceroatrial situs inversus (25%; see Table 7.11 ).
The data suggest that TAPVC to the ductus venosus is approximately equally as frequent in solitus and nonsolitus visceroatrial situs.
TAPVC to the coronary sinus occurred only in visceroatrial situs solitus (19%; see Table 7.7 ) but not in visceral heterotaxy (see Tables 7.8 to 7.10 ) and not in situs inversus (see Table 7.11 ) ( p < .0001).
TAPVC to the coronary sinus occurred only in visceroatrial situs solitus.
The mixed type of TAPVC occurred both in visceroatrial situs solitus (10.3%; see Table 7.7 ) and in heterotaxy with asplenia (5.2%; see Table 7.8 ) and polysplenia (12.5%; see Table 7.9 ), but not in visceroatrial situs inversus (see Table 7.11 ) ( p = NS, i.e., 0.25).
The prevalence of the mixed type of TAPVC in visceroatrial situs solitus and in visceroatrial situs nonsolitus (heterotaxy and situs inversus) was not significantly different.
TAPVC to SVC occurred in visceroatrial situs solitus (9.5%; see Table 7.7 ) but was much more common in heterotaxy with asplenia (53.5%; see Table 7.8 ), heterotaxy with polysplenia (25%; see Table 7.9 ), heterotaxy with a normally formed spleen (50%; see Table 7.10 ), and visceroatrial situs inversus (50%; see Table 7.11 ). Indeed, TAPVC to SVC (right or left) was by far the most common type of TAPVC in visceral heterotaxy (34/68, i.e., 50%; see Tables 7.8 through 7.10 ). These differences in the prevalence of TAPVC to SVC in visceroatrial situs solitus and in visceroatrial situs nonsolitus are very highly statistically significant ( p < .0001, x 2 = 38.62).
TAPVC to the SVC is common in visceral heterotaxy and in situs inversus (see Tables 7.8 to 7.11 ) but is much less frequent in visceroatrial situs solitus (see Table 7.7 ).
TAPVC/D to the RA is infrequent in visceroatrial situs solitus (4%; see Table 7.7 ), but is more common in visceral heterotaxy with asplenia (10%; see Table 7.8 ), in heterotaxy with polysplenia (62.5%; see Table 7.9 ), in heterotaxy with a normally formed spleen (50%; see Table 7.10 ), and in visceroatrial situs inversus (25%; see Table 7.11 ). In our earlier study of 93 postmortem cases with Dr. Georges Delisle et al in 1976, we never found TAPVC/D to the RA in the isolated form with visceroatrial situs solitus (see Table 7.3 ). Consequently, we concluded that TAPVC/D to the RA is characteristic of the nonisolated form, typically with visceral heterotaxy and asplenia or polysplenia.
The present larger study of 198 (of a total of 204) autopsied cases of TAPVC/D shows that it is indeed possible to have TAPVC to the RA in visceroatrial situs solitus, that is, without visceral heterotaxy or situs inversus (3.97%; see Table 7.7 ). Nonetheless, the prevalence of TAPVC/D to the RA in visceroatrial situs nonsolitus (13/72, i.e., 18.05%) (see Tables 7.8 to 7.11 ) remains a statistically highly significant difference ( p < .001, x 2 = 10.98) compared with the findings in visceroatrial situs solitus (see Table 7.7 ).
TAPVC/D to the RA is significantly more frequent in visceral heterotaxy (with asplenia, or polysplenia, or with a normal spleen) and in visceroatrial situs inversus than in visceroatrial situs solitus.
To summarize, TAPVC/D more frequently found in visceroatrial situs solitus includes:
- 1.
TAPVC to the LIV (snowman type) ( p < .0001) and
- 2.
TAPVC to the coronary sinus ( p < .0001).
TAPVC/D more frequently found in visceral heterotaxy and in situs inversus includes:
- 1.
TAPVC to a SVC ( p < .0001) and
- 2.
TAPVC/D to the RA ( p < .001).
TAPVC/D with prevalences that were not significantly different in visceroatrial situs solitus and nonsolitus (heterotaxy with asplenia, heterotaxy with polysplenia, heterotaxy with normal spleen, and visceroatrial situs inversus) include:
- 1.
mixed TAPVC/D ( p = .25, i.e., NS) and
- 2.
TAPVC to the ductus venosus ( p = .78, i.e., NS).
Associated Malformations
In Visceroatrial Situs Solitus (n = 126)
- 1.
TAPVC to the LIV (snowman type). Obstruction was present in 9 of 38 cases (23.7%). Agenesis of the right lung was observed in 1 patient, and agenesis of the left lung in another case.
- 2.
TAPVC to the ductus venosus. Atresia of the anomalous pulmonary venous pathway was present in 3 of 32 patients (9.4%). A conjoined twin was present in 3 other cases (9.4%), and multiple congenital anomalies (MCAs) were found in 2 patients (6.25%).
- 3.
TAPVC to the coronary sinus. Obstruction of the anomalous pulmonary venous pathway was observed in 1 patient (A76-008), who had agenesis of the LPVs. Where the right pulmonary veins (RPVs) joined the coronary sinus, the junction was obstructive. This patient also had DORV {S,D,D}, with the DiGeorge syndrome and MCAs. This case proves that it is indeed possible to have TAPVC to the coronary sinus with congenital obstruction (stenosis or atresia), unrelated to prior surgery. Prior to this patient, we had thought (erroneously) that congenital obstruction did not occur with TAPVC to the coronary sinus.
- 4.
Mixed TAPVC. MCAs were present in 2 of these 13 patients (15.4%).
- 5.
TAPVC to the RSVC. Obstruction was present in 3 of these 11 patients (27.3%). MCAs were found in 2 other patients (18.2%).
- 6.
TAPVD to the RA. Of these 5 patients, 2 (40%) had MCAs, 1 of which had the Ellis-van Creveld syndrome.
- 7.
TAPVC to the azygos vein. Both of these patients had severe stenosis of the anomalous pulmonary venous pathway (100%).
In Visceral Heterotaxy With Asplenia (n = 58)
- 1.
TAPVC to the RSVC. Obstruction was present in 6 of these 20 patients (30%).
- 2.
TAPVC to the LSVC. Obstruction was present in 4 of these 11 patients (36.4%).
In these 31 cases with TAPVC to a right or left SVC, obstruction was present in a total of 10 patients (32.3%).
- 3.
TAPVC to the ductus venosus. Although some degree of obstruction may well have been present in all 16 cases (100%), it was particularly marked in 2 (12.5%).
Systemic Veins
Just as the types of pulmonary venous anomalies were linked to the types of visceroatrial situs that were present (as earlier), so too the types of systemic venous malformations were also linked to the types of visceroatrial situs that coexisted, as follows.
Situs Solitus of the Viscera and Atria (Present in 122 of 198 Cases of TAPVC/D [61.6%])
- 1.
Normal systemic veins were present in 97 of 122 patients (79.5%). Abnormalities of the systemic veins were found in 25 cases (20.5%).
- 2.
Persistent LSVC to the coronary sinus to the RA was present in 18 of 122 patients (14.75%). Although an abnormality, this anomaly led to no physiologic derangement because all of the systemic venous blood did indeed return to the morphologically RA.
- 3.
Interruption of the IVC was found in 2 of these 122 patients with visceroatrial situs solitus (1.6%). The interrupted IVC was right-sided in 1 patient and left-sided in the other.
- 4.
Bilateral SVC with unroofing of the coronary sinus was present in 1 patient (0.8%). Because of the large coronary sinus septal defect (unroofing of the coronary sinus), the blood of the left SVC drained into the LA.
- 5.
Bilateral SVC with hypoplasia of the right SVC was found in 1 patient (0.8%). Bilateral SVC was observed in 2 of 122 patients with situs solitus of the viscera and atria (1.6%).
- 6.
Left SVC to LA, absence of an identifiable coronary sinus, and atresia of the right SVC were present in 1 patient (0.8%). Unroofing of the coronary sinus explains why the left SVC drained into the LA and why a discrete coronary sinus was not found.
- 7.
Absence of the coronary sinus was observed in 1 patient (0.8%) with visceroatrial situs solitus and TAPVC/D.
- 8.
Absence of the ductus venosus was noted in 1 patient (0.8%).
- 9.
A small arteriovenous fistula between the descending aorta and the IVC was present in 1 patient (0.8%).
Systemic Veins in Visceroatrial Heterotaxy With Asplenia
Of the 198 patients with TAPVC in which the data were suitable for analysis, visceral heterotaxy with asplenia was present in 59 cases (29.8%). The complexity of the systemic venous anatomy is so great that it almost defies brief summary. Of these 59 cases with heterotaxy and asplenia, 1 patient was excluded because the systemic veins were not well described. Hence, the following is an analysis of 58 patients with TAPVC, heterotaxy, and asplenia.
In how many of these 58 cases of the asplenia syndrome did we think we could identify the anatomic type of atrial situs? The answer is in 46 of 58 patients (79.3%). We were not able to diagnose the atrial situs with confidence in 12 of 58 cases (20.7%).
The key to the understanding of the systemic veins in these patients is to diagnose, when possible, the basic anatomic type of visceroatrial situs that is present, despite the coexistence of visceral heterotaxy (anomalies of lateralization or asymmetry). The anatomic pattern of the systemic veins is linked to the visceroatrial situs and indeed is an expression of the visceroatrial situs.
What is (are) the basic type (s) of visceroatrial situs in the heterotaxy syndrome with asplenia? This is one of the mysteries of contemporary pediatric cardiology. Let us see what these cases suggest:
The atrial situs was diagnosed in 46 patients:
- 1.
basically situs solitus of the atria, that is, {A(S,-,-}, in 26 of 46 cases (56.52%); and
- 2.
basically situs inversus of the atria, that is, {A(I),-,-}, in 20 of 46 patients (43.48%).
This ratio of the proportions of atrial situs solitus to atrial situs inversus suggests an almost random distribution:
Now let us look at the 12 cases in which the atrial situs was not diagnosed but was recorded as situs ambiguus, not otherwise qualified, that is, {A,-,-}. The sidedness of the IVC was recorded in 9 patients:
- 1.
left-sided IVC6
- 2.
right-sided IVC3
If one accepts that the sidedness of the IVC strongly suggests the basic type of visceroatrial situs that is present (right-sided IVC = probable situs solitus, and left-sided IVC = probable situs inversus), the findings become:
Hence, the ratio of the proportions of situs solitus/situs inversus in this sample of the asplenia syndrome becomes 1 · 115, that is., quite close to 1:1, a randomized distribution of atrial situs. Thus, these findings suggest that in visceroatrial heterotaxy with asplenia, the basic types of situs are approximately 1 to 1, or 50/50 (in percentages), that is, essentially randomized.
If the aforementioned data are representative and the inferences are valid, this means that visceroatrial situs ambiguus with asplenia is much less ambiguous than it used to be just a few years ago. However, it is also noteworthy that there are some cases of the asplenia syndrome in which we were not able to diagnose the basic type of atrial situs with confidence, that is, {A,-,-} = 12/58 (20.7%).
The systemic venous anomalies found within each group, that is, {A(S),-,-}, {A(I),-,-}, and {A,-,-} are summarized in Tables 7.12, 7.13, and 7.14 , respectively.
| With Solitus Atria: {A(S),-,-}, n = 26/58 (44.8%) |
a “Absent” and “unroofed” coronary sinus are essentially the same anomaly. The status of coronary sinus was not specified in the other 19 cases.
| With Inversus Atria: {A(I),-,-}, n = 20/58 (34.5%) |
| Anomaly | No. of Cases | % of Series |
|---|---|---|
| Bilateral SVC | 7 | 12.1 |
| Atresia of RSVC | 5 | 8.6 |
| Absent CoS | 4 | 6.9 |
| Unroofed CoS | 4 | 6.9 |
| Right-sided hepatic veins | 5 | 8.6 |
| RSVC to LA (R) | 2 | 3.4 |
| R→L switch of IVC to RA (L) | 1 | 1.7 |
| Interrupted IVC | 1 | 1.7 |
| With Undiagnosed Atrial Situs: {A,-,-}, n = 12/58 (20.7%) |
| Anomaly | No. of Cases | % of Series |
|---|---|---|
| Bilateral SVC | 8 | 13.8 |
| Left-sided IVC | 6 | 10.3 |
| Right-sided IVC | 3 | 5.2 |
| Absent CoS | 3 | 5.2 |
| Unroofed CoS | 1 | 1.7 |
| Right-sided hepatic vein(s) | 2 | 3.4 |
| Left-sided hepatic vein(s) | 1 | 1.7 |
| Left-sided hepatic vein to LIVC | 1 | 1.7 |
Systemic Veins in Visceroatrial Heterotaxy With Polysplenia
In 198 well-documented cases of TAPVC, 9 had visceral heterotaxy with polysplenia (4.5%). The atrial situs was solitus, that is, {A(S),-,-}, in 6 patients (66.7%) and was inversus, or {A(I),-,-}, in 3 cases (33.3%). Although atrial situs solitus was more predominant in polysplenia (66.7%) than in asplenia (56.5%), this difference was not statistically significant ( p = .25, NS, Fisher’s exact test).
Thus, in this sample of visceral heterotaxy, patients with asplenia and patients with polysplenia both had atrial situs solitus and atrial situs inversus. The status of the spleen (asplenia versus polysplenia) cannot be used to infer the type of atrial situs present.
It is noteworthy that in this small series of patients with polysplenia, the atrial situs was always diagnosed with confidence. This is a characteristic difference between polysplenia (atrial situs can be diagnosed with confidence always, or almost always) and asplenia (atrial situs may not be diagnosed with confidence in ≈ 20% of cases). It also should be understood that because the concept of atrial isomerism is erroneous, it is therefore possible to diagnose the atrial situs of polysplenia patients almost always and to diagnose the atrial situs of asplenic patients usually (≈ 80%).
The salient associated systemic venous anomalies in the 6 patients with TAPVC, heterotaxy with polysplenia, and atrial situs solitus were as follows:
- 1.
LSVC to coronary sinus, to RA, 4 of 6 (66.7%);
- 2.
interruption of the IVC, 2 of 6 (33.3%); and
- 3.
bilateral superior venae cava, 1 of 6 (16.7%).
One of these patients with polysplenia and visceroatrial situs solitus did not have heterotaxy. Polysplenia without visceroatrial heterotaxy is noteworthy and rare.
The salient associated systemic venous anomalies in the 3 patients with TAPVC, heterotaxy with polysplenia, and atrial situs inversus were as follows:
- 1.
interruption of the IVC in all 3 (100%),
- 2.
RSVC to coronary sinus to left-sided RA in 2 (66.7%),
- 3.
bilateral SVC in 1 (33.3%), and
- 4.
hepatic veins draining into the morphologically LA in 1 patient (33.3%).
Interruption of the IVC with an enlarged azygos vein to a SVC was observed in 5 of these 9 polysplenic patients (56%).
Systemic Veins in Visceral Situs Inversus
In 198 well-documented cases of TAPVC, situs inversus of the viscera was present in 6 patients (3.03%). The salient visceroatrial findings were the following:
- 1.
The atrial situs was inversus in 5 of 6 patients (83%), just as one would expect.
- 2.
In 1 patient, a 3-day-old boy, the IVC was left-sided, as expected, but it switched from left to right at the level of the liver to connect with a right-sided morphologically RA. In other words, there was visceroatrial situs discordance of the {I(S),-,-} type. The segmental anatomy was DORV {I(S),D,D}. The patient had a RSVC, with nonobstructive TAPVC to the RSVC, bilaterally trilobed lungs, a normally formed right-sided spleen, common gastrointestinal mesentery, common atrium, common AV valve opening into a single morphologically RV, absent morphologically LV, bilateral conus, pulmonary outflow tract atresia, absent left anterior descending coronary artery, left aortic arch (abnormal for visceral situs inversus), and a PDA (3 mm). This case is recorded in detail because of its rarity.
- 3.
A persistent RSVC, resulting in bilateral SVCs, was found in 2 patients (33.3%).
- 4.
Congenital complete heart block was observed in 1 patient (16.7%).
Systemic Veins in Visceral Heterotaxy With a Normally Formed Spleen
Of 198 well-documented cases of TAPVC, 2 (1.01%) had visceral heterotaxy with a normally formed spleen. The salient features of these 2 rare cases were as follows.
One was a 6-month-old girl with dextrocardia. The segmental anatomy was {A(I),D,I}, i.e., isolated ventricular noninversion. The IVC was right-sided, but it switched from right to left at the level of the liver and connected with the left-sided morphologically RA. The stomach and normally formed spleen were left sided. The SVC was left-sided and the coronary sinus was absent. There was TAPVC to the left-sided morphologically RA. The lobation of the lungs was solitus. A common atrium was present in association with completely common AV canal, type A of Rastelli. A single LV was present due to absence of the right ventricular sinus. Common-inlet LV was present. From the infundibular outlet chamber an atretic pulmonary artery originated. The aortic arch was right-sided, but the descending thoracic aorta was left-sided. A left-sided PDA connected the LPA and the left innominate artery (left PDA being inappropriate for situs inversus).
The other patient was a 7 4/12–year-old boy with dextrocardia and DORV {A(I),D,D}. The right-sided IVC switched from right to left at the liver and connected with the left-sided morphologically RA. The stomach and normally formed spleen were left-sided (echocardiographic observations, the autopsy being limited to heart and lungs). The coronary sinus was absent and there were bilateral SVCs. There was TAPVC to the junction of the LSVC with the left-sided morphologically RA, without obstruction. The right lung had 4 lobes and the left lung had 3. There was a partially common AV canal with an ostium primum type of ASD. The tricuspid component of the common AV valve was large and regurgitant. The mitral component of the common AV valve was adherent to the crest of the ventricular septum and was atretic. There was no VSD. The patient had a functionally single morphologically RV. The morphologically LV was minute, with neither inlet nor outlet. There was a subaortic conus, pulmonary atresia, and a left aortic arch.
Thus, visceral heterotaxy rarely does occur with a normally formed spleen, as these two cases of complex congenital heart disease illustrate.
Partially Anomalous Pulmonary Venous Connection/Drainage
The main questions for consideration are:
- 1.
What really is PAPVC/D?
- 2.
How many anatomic types of PAPVC/D are there?
- 3.
What are their relative prevalences?
- 4.
Is PAPVC/D usually isolated? Or is it usually accompanied by other associated malformations (nonisolated PAPVC/D)?
- 5.
Among the patients with nonisolated TAPVC/D, are there any patterns among the associated cardiac or noncardiac malformations?
- 6.
Are some types of PAPVC/D clinically and therefore surgically important, whereas other types are far less important?
This is a study of 45 postmortem cases of PAPVC/D.
Prevalence: PAPVC/D constituted 1.40% of the cases of congenital heart disease in this study (45 of 3216). PAPVC/D was much less frequent than TAPVC/D: 1.4% versus 6.34%.
Sex: The sex was known in all 45 cases: males, 21 (47%) and females 24 (53%). The male-to-female ratio was 21:24 (0.875). Thus, PAPVC/D was characterized by a female preponderance.
These findings are the opposite of those encountered with TAPVC/D, which had a male predominance: males, 121/199, 60.8%; females, 78/199, 39.2%; and male-to-female ratio = 1.55:1.0. However, these sex differences between PAPVC/D and TAPVC/D did not quite reach statistical significance ( p = .08).
Age at Death: mean = 966 ∙ 15 days, that is, 2.65 years ± [1 standard deviation] 2332 ∙ 59 days, that is, ± 6.39 years, ranging from 0 (fetal death) to 25 years. The median age at death of patients with PAPVC/D was 75 days (2.5 months).
For comparison, the median age at death of patients with TAPVC/D was 39 days (1.3 months).
Scimitar Syndrome: Because the scimitar syndrome seems to be a real entity, whereas many of the other anatomic types of PAPVC/D may be random events, that is, failures of separation of the pulmonary venous and systemic venous plexuses (see later), we thought it might be useful to look at the data for patients with the scimitar syndrome in isolation, not admixed with data from the other anatomic types of PAPVC/D:
Sex: males, 3; females, 10; and male-to-female ratio = 3/10 (0.3:1.0).
Age at Death: mean = 151 days, that is, 5 months ± [1 standard deviation] 294 days, that is, ± 9.8 months, ranging from 0 postnatal days (fetal abortion) to 820 days, that is, 2.25 years. The median age at death of these 13 patients with scimitar syndrome was 10 days.
Comment:
- 1.
The median age at death in patients with the scimitar syndrome (10 days) was much less than the median age at death in patients with PAPVC/D who did not have the scimitar syndrome (116 days, or 3.87 months).
- 2.
The male-to-female ratio in patients with the scimitar syndrome (0.3) was very different from the ratio found in patients with PAPVC/D who did not have scimitar syndrome (18/14 = 1.3) ( p < .05).
The anatomic types of partially anomalous pulmonary venous connection or drainage (PAPVC/D) that were found in 46 postmortem cases are summarized, in order of decreasing frequency, in Table 7.15 . As Table 7.15 shows, we found 14 anatomically different types of PAPC/D, with the scimitar syndrome being the most common type in this series:
- 1.
typical right-sided scimitar syndrome in 12 patients (26%), and
- 2.
1 rare case of left-sided scimitar syndrome (2%).
| Anatomic Type | No. of Cases | % of Series a |
|---|---|---|
| 12 | 27 |
| 12 | 27 |
| 6 | 13 |
| 4 | 9 |
| 1 | 2 |
| 2 | 4 |
| 1 | 2 |
| 1 | 2 |
| 1 | 2 |
| 1 | 2 |
| 1 | 2 |
| 1 | 2 |
| 1 | 2 |
| 1 | 2 |
Hence, the total prevalence of the scimitar syndrome in this series was 28%.
Scimitar Syndrome
The scimitar syndrome was christened in 1960 by Drs. Catherine Neill, Charlotte Ferencz, David Sabiston, and H. Sheldon because of the curvilinear density that may be seen in the lower right lung field in the plain chest x-ray film of such patients. This curvilinear shadow—that is somewhat reminiscent of a curved oriental sword (scimitar)—is cast by the anomalous RPV that drains most or all of the right lung into the IVC just above or just below the diaphragm ( Fig. 7.19 ).
Other cardinal features of the scimitar syndrome are an anomalous systemic arterial blood supply arising from the region of the celiac axis of the abdominal aorta and passing through the diaphragm to supply the lower right lung, a reciprocally small right pulmonary artery (RPA), hypoplasia of the right lung, and dextrocardia secondary to the smallness of the right lung. The detailed findings in these 12 cases of right-sided scimitar syndrome are summarized in Table 7.16 .
| Finding | No. of Cases | % of Series |
|---|---|---|
| Hypoplastic right lung | 11 | 92 |
| Bilaterally unilobed lungs | 4 | 33 |
| All right pulmonary veins to IVC | 7 | 58 |
| All right pulmonary veins to ductus venosus, to hepatic sinusoids, to RA | 1 | 8 |
| Inferior scimitar vein to IVC and superior vein to RSVC | 1 | 8 |
| Inferior scimitar vein to IVC and right upper pulmonary veins to LA | 2 | 17 |
| Right lower pulmonary vein to RA, all other pulmonary veins to LA | 1 | 8 |
| Anomalous arteries from abdominal aorta | 7 | 58 |
| Dextrocardia | 6 | 50 |
| Mesocardia | 1 | 8 |
| Ostium secundum type of ASD | 5 | 42 |
| LSVC to coronary sinus to RA | 4 | 33 |
| Preductal coarctation of the aorta | 3 | 25 |
| Mitral atresia | 2 | 17 |
| Conoventricular type of VSD | 2 | 17 |
| Muscular type of ventricular septal defect | 3 | 25 |
| Polysplenia {S,D,S} | 2 | 17 |
| Polysplenia without heterotaxy | 1 | 8 |
| Bicuspid aortic valve | 2 | 17 |
| Truncus arteriosus type A1 | 1 | 8 |
| Truncal valvar regurgitation | 1 | 8 |
| Eccentric coronary ostia within sinuses of Valsalva | 2 | 17 |
| Aberrant left coronary from pulmonary artery bifurcation | 1 | 8 |
| Stenosis of left upper pulmonary vein | 1 | 8 |
| Right aortic arch | 1 | 8 |
| Wolff-Parkinson-White syndrome | 1 | 8 |
| Absent right pulmonary artery | 1 | 8 |
| Severe pulmonary hypertension, bilateral | 1 | 8 |
| Familial scimitar syndrome | 1 | 8 |
| Multiple congenital anomalies (i.e., more than the cardiovascular system involved) | 7 | 58 |
| Horseshoe kidneys, forme fruste (i.e., fibrous union at lower poles) | 1 | 8 |
| Microcephaly with lissencephaly | 1 | 8 |
| Bilateral cataracts | 1 | 8 |
| Septated vagina with uterus bilocularis | 2 | 17 |
| Multicystic right kidney | 1 | 8 |
| Left diaphragmatic hernia, foramen of Bochdalek type | 2 | 17 |
| Eventration of right leaf of diaphragm | 2 | 17 |
| Left umbilical vein passing through hepatic substance (not fissure) to reach porta hepatis | 1 | 8 |
| Right adrenal separated from right kidney | 1 | 8 |
| Left ovary extending up to left-sided spleen | 1 | 8 |
| Coloboma, left iris | 1 | 8 |
| Absent gallbladder, biliary tree otherwise normal | 1 | 8 |
Left-Sided Scimitar Syndrome
Our findings in what may be called the left-sided scimitar syndrome involved a 3-day-old female infant and may be summarized as follows. A scimitar vein draining all of the left pulmonary venous blood passed through the diaphragm and into a hepatic venous confluence on the superior surface of the liver. From there, the anomalous left pulmonary venous pathway traveled rightward and joined the suprahepatic segment of the IVC via a stenotic orifice that was 3 mm in diameter. A single RPV connected with the LA. Numerous small pulmonary venous collaterals bilaterally connected with the intercostals veins.
Two relatively large systemic arteries from the abdominal aorta above the celiac axis penetrated the diaphragm and supplied the lower portions of the left lung and the right lung. Other smaller systemic collateral arteries from the descending thoracic aorta supplied the lungs bilaterally; that is, major aortopulmonary collateral arteries (MAPCAs) were present both from above and below the diaphragm.
The RPA and LPA branches were strikingly hypoplastic and serpentine, we thought because of the major systemic arterial collateral blood supply to both lungs.
The IVC was interrupted, with absence of its renal-to-hepatic venous segment. An enlarged right-sided azygos vein connected with the RSVC. The spleen was normally located and normally formed.
We reported this rare case in 1979 because it represented a most unusual basis for persistence of the fetal pattern of the circulation. At autopsy, the PDA was small and appeared to be closing.
What Really Is the Scimitar Syndrome?
As Table 7.16 shows, the scimitar syndrome typically is an anomaly of the right lung that almost always has a scimitar vein draining into the IVC (in 11 of 12 patients, 92%; see Table 7.16 ). The one exception had an anomalous RPV draining into the RA. In this case, the hypoplastic right lung received four moderately large systemic arteries from the celiac axis below the diaphragm that supplied 60% to 70% of the arterial blood supply of the hypoplastic right lung. Dextrocardia was present, secondary to the right pulmonary hypoplasia. The normally related great arteries produced severe tracheal compression anteriorly (treated with aortopexy), again secondary to the right pulmonary hypoplasia and the secondary dextrocardia. In view of all of the previously mentioned findings, we thought that this case should be regarded as a closely related variant of the scimitar syndrome, even though the signature finding—the scimitar vein—was not present in this rare case.
Thus, the main point is that the scimitar syndrome really is an anomaly typically of the right lung, with characteristic malformations of the systemic veins, and often with anomalous systemic arterial blood supply (in 58% of these cases; see Table 7.16 ).
The next point is that both lungs can be abnormal, as in 33% of our cases of right-sided scimitar syndrome (see Table 7.16 ) and as in our patient with left-sided scimitar syndrome.
Rarely is it possible to have a scimitar vein with a normal ipsilateral lung. Our case had a left-sided scimitar vein that ran behind the LA and then passed through the diaphragm and connected with the suprahepatic segment of the right-sided IVC. This patient (whom we are currently in the process of reporting) had elective surgical closure of a secundum ASD at 10 years of age. At 22 years of age, she had side-to-side anastomosis of the left pulmonary venous confluence with the LA, with ligation and division of the left-sided scimitar vein to the right of the anastomosis. Six years postoperatively, cardiac magnetic resonance imaging (MRI) showed that the anastomosis was widely patent. Ten years postoperatively, the patient remains asymptomatic. (This is a living and currently unpublished patient.)
Pulmonary pathologists regard typical scimitar syndrome as a special form of intralobar sequestration . We made the diagnosis of sequestration of the right lung—meaning that the right lung did not communicate normally with the right bronchus (hence the right lung was sequestered or separated from its bronchus)—in only 1 case (8%).
Familial congenital anomalies were noted in 1 of these patients (8%) and consanguinity (first cousin marriage) was recorded in 1 case (8%).
Although one usually thinks of the scimitar syndrome as occurring in isolation, it is noteworthy that other forms of congenital heart disease frequently coexisted (see Table 7.16 ): secundum type of ASD in 5 patients (42%); persistent LSVC to the coronary sinus in 4 (33%); preductal coarctation of the aorta in 3 (25%); muscular VSD in 3 (25%); conoventricular type of VSD in 2 (17%); mitral atresia in 2 (17%); polysplenia in 2 (17%), without visceral heterotaxy in 1 (8%); and truncus arteriosus, aberrant left coronary artery, right aortic arch Wolff-Parkiknson-White syndrome, and absent RPA in 1 case each (8%).
Perhaps even more impressive, MCAs, that is, malformations involving not only the cardiovascular system but also one or more additional systems, were present in more than half of these patients with the scimitar syndrome (58%; see Table 7.16 ).
Diagnostic and Surgical Implications
The scimitar syndrome is clearly a very serious clinical form of PAPVC because it is so much more than “just” a PAPVC.
Diagnostically, one should not only document the scimitar vein but also search for additional anomalous right pulmonary venous connections. Also high on one’s mental must-exclude list should be anomalous systemic arteries from below the diaphragm, or from above it, or from both sites of origin; additional types of congenital heart disease; and a wide variety of possible multisystem anomalies (see Table 7.16 ).
Surgically, although one would like to treat such patients as conservatively as possible—by baffling the scimitar venous return into the LA and by interrupting the anomalous systemic arterial blood supply typically from the region of the celiac axis (by surgical ligation or interventional coils)—one may be forced to do a right pneumonectomy, because typical scimitar syndrome is really a sick right lung, not just PAPVC, with or without anomalous systemic arterial blood supply from below the diaphragm.
Thus, in our experience, the scimitar syndrome is not only the most frequent form of PAPVC (13/45 patients, 29%), but also diagnostically and surgically the most serious form (see Table 7.16 ).
RUPV to RSVC
The right upper pulmonary vein (RUPV) draining anomalously into the RSVC was the second most common form of PAPVC that was encountered in the present series: 12 of 45 patients (27%; see Table 7.15 ).
In RUPV to RSVC (see Fig. 7.19 ), only one pulmonary vein drains anomalously. Consequently, RUPV to RSVC is a less severe form of PAPVC than is the scimitar syndrome, in which typically all RPVs drain anomalously.
Other differences from the scimitar syndrome are as follows: RUPV to RSVC typically is not associated with an abnormal right lung, right pulmonary hypoplasia, secondary dextrocardia, or anomalous systemic arterial blood supply.
What, then, is typical of RUPV to RSVC? It often is not the main diagnosis and consequently may be overlooked diagnostically and surgically. Prominent “main” diagnoses associated with RUPV to RSVC are presented in Table 7.17 .
