Mitral Valve Anomalies

Salient Anatomic Features

For clarity and brevity, the salient anatomic features of this type of mitral atresia are summarized in Table 14.4 .

Why Did This Anatomic Type of Mitral Atresia Have Such an Unfavorable History With a Median Age at Death of Only 10 Days?

It is not easy for us to answer this important question because our data reflect not only the natural history but also the iatrogenic history of the foregoing combination of anomalies. Of these 80 patients (see Table 14.4 ), 43 died after a modified Norwood procedure (53.75%) and 3 died as the result of a medically induced abortion (3.75%). Consequently the age at death of these patients does not reflect the natural history of this anatomic type of mitral atresia in 46 of these 80 patients (57.5%).

However, the age at death does reflect the natural history in 31 of these 80 patients (38.75%) who had no surgical procedure of any kind and whose ages at death are known to us: mean age at death, 9.65 ± 22.53 days; range, from 0 (spontaneously occurring stillbirth) to 120 days; and median, 2.5 days. Thus, the median age at death of our unoperated patients with this anatomic type of mitral atresia was even worse (younger) than that found in the total series of all patients (operated and unoperated), that is., 2.5 days versus 10 days of age, respectively.

Some of the anatomic features presented in Table 14.4 appear highly relevant to the early lethality of this anatomic type of mitral atresia (type 1; see Table 14.2 and Fig. 14.1 ). The atrial septum was observed to be well formed and restrictive or obstructive to 100% left-to-right atrial shunting in 26 of these 80 cases (32.5%; see Table 14.4 ). An aneurysm of the septum primum—the flap valve of the foramen ovale—was observed to be bulging into the right atrium (RA), confirming that the atrial septum was obstructive to left-to-right shunting at the atrial level, in 9 of these 80 patients (11.25%; see Table 14.4 ).

In an additional 6 patients, the atrial septum was found to be intact at postmortem examination and hence totally obstructive (7.5%; see Table 14.4 ). Thus, the atrial septum was significantly obstructive to left-to-right atrial shunting in at least 32 of these 80 patients (40%). A restrictive atrial septum results in pulmonary venous congestion and hypertension, typically leading in turn to the clinical picture of congestive heart failure early in the neonatal period.

However, a secundum type of atrial septal defect (ASD) coexisted with mitral atresia in 15 of these 80 patients (18.75%), tending to reduce or eliminate left atrial and pulmonary venous hypertension. The size of the secundum ASD typically determined how much left atrial and pulmonary venous hypertension resulted.

One of these patients with mitral atresia also had cor triatriatum sinistrum, that is, stenosis of the common pulmonary vein (see Table 14.4 ). Consequently, in this patient the presence of pulmonary venous hypertension was unrelated to the anatomic state of the atrial septum.

Other anomalies in addition to a secundum type of ASD can reduce or eliminate left atrial and pulmonary venous hypertension in association with mitral atresia.

A levoatrial cardinal vein ^, was present in 1 patient (1.25%; see Table 14.4 ). As its name suggests, this anomalous vein ran between the LA inferiorly and the left innominate vein superiorly. A levoatrial cardinal vein is different from a persistent left superior vena cava (LSVC). A levoatrial cardinal vein does not connect with the coronary sinus, and the course of a levoatrial cardinal vein is more posterior than that of a persistent LSVC. In this patient, the levoatrial cardinal vein was thought to help to decompress the LA.

This patient, a 20-day-old boy, came from a family with congenital anomalies. Both parents and the patient’s brother had congenital deafness. The patient’s father represented a fourth-generation Waardenburg syndrome; the cause of the mother’s deafness was not known. The Waardenburg syndrome ^, is characterized by lateral displacement of the medial canthi, partial albinism, and deafness because of atrophic changes in the spiral ganglion and nerve of the organ of Corti.

This patient also had familial congenital heart disease: a cousin with congenital heart disease (type not known by us). This patient had multiple congenital anomalies (MCAs) in addition to congenital heart disease: a prominent helix of the right ear; dysplastic toe nails; and a sacral dimple. This patient also had a partially anomalous pulmonary venous connection: the left upper lobe pulmonary veins communicated with the levoatrial cardinal vein, and this pulmonary venous blood returned to the RA via the left innominate vein and the right superior vena cava (RSVC). Both the levoatrial cardinal vein and the partially anomalous pulmonary venous connection tended to decompress the LA and reduce the pulmonary venous hypertension.

Partially anomalous pulmonary venous connection was found in 3 of these 80 patients with this type of mitral atresia (3.75%; see Table 14.4 ). One patient (1.25%; see Table 14.4 ) had a sinus venosus defect with unroofing of the right pulmonary veins. Both the posterosuperior interatrial communication and the unroofing of the right pulmonary veins would tend to reduce left atrial hypertension. The interatrial communication would permit left-to-right shunting at the atrial level, and the unroofing of the right pulmonary veins functions as partially anomalous drainage of the right pulmonary veins into the RA.

A coronary sinus septal defect is another way of decompressing the LA and the pulmonary veins when mitral atresia is present. Four patients (5.0%; see Table 14.4 ) had a so-called coronary sinus septal defect between the coronary sinus posteriorly and the LA anteriorly. Such a defect permits pulmonary venous blood in the LA to flow posteriorly into the coronary sinus and then to the RA, or retrogradely via the LSVC and the left innominate vein to the RSVC and RA.

Three of our patients with the latter type of anomalous pulmonary venous drainage (i.e., decompressing coronary sinus septal defect, permitting pulmonary venous blood to return to the RA via the coronary sinus, or via the LSVC, left innominate vein, and RSVC) also had stenosis or atresia of the right atrial ostium of the coronary sinus (3.75%; see Table 14.4 ).

An incomplete form of common atrioventricular (AV) canal was found in 1 patient (1.25%; see Table 14.4 ). The mitral component of the common AV valve was atretic, and a large ostium primum type of defect (i.e., the most common form of incomplete AV septal defect) permitted left-to-right shunting at the atrial level, decompressing the LA. The ventricular septum was intact.

Leftward displacement of the superior portion of the septum primum (see Fig. 14.2 ) was documented in 8 of these patients (10%; see Table 14.4 ). We speculate that this curious leftward malposition of the superior portion of the septum primum may have reduced the embryonic and fetal via sinistra (blood flow into the left heart), thereby perhaps contributing to the morphogenesis of mitral atresia. I think that Dr. Paul Weinberg (circa 1976, when Dr. Weinberg was our Cardiac Registry fellow in cardiac pathology) was the first to point out the leftward malposition of the septum primum in patients with mitral atresia, and to draw attention to the potential morphogenetic importance of this atrial septal malposition.

To summarize thus far, the status of the atrial septum, which can be significantly or totally obstructive, and the presence or absence of other anomalies that may influence left atrial and pulmonary venous hypertension, appear important concerning the very unfavorable natural history of this anatomic type of mitral atresia.

The functional status of the ductus arteriosus is another important variable. At autopsy, the ductus arteriosus usually appears to be widely patent, but not always. In the present study, the ductus arteriosus was observed to be restrictive or functionally closing in 16 patients (20%; see Table 14.4 ). We suspect that this may well be a significant underestimate.

Functional narrowing or near closure of the patent ductus arteriosus (PDA) is thought to be an important factor contributing to the massive right ventricular and right atrial hypertrophy and enlargement that was found in all cases, plus tricuspid regurgitation that was found in 9 patients (11.25%; see Table 14.4 ) (probably another important underestimate). Functional narrowing of the PDA leads to excessive pulmonary blood flow, increasing pulmonary venous and left atrial hypertension, increasing tricuspid regurgitation, and leading to a clinical picture of increasing left-sided and right-sided congestive heart failure.

Functional narrowing of the large PDA also can contribute to discrete coarctation of the aorta (preductal or juxtaductal), that was observed in 10 of these patients (12.5%; Table 14.4 ). The ductal medial musculature can encircle the small aortic isthmus, or the aorta opposite the ductal opening, creating a juxtaductal shelf. When the ductus constricts, the coarctation (juxtaductal shelf) becomes more marked; and when the ductal constriction relaxes, the coarctation lessens or disappears.

Functional narrowing of the large PDA acts like an acute “coarctation” of the arterial blood flow in an anterograde direction to the lower body, and in a retrograde direction to the brachiocephalic arteries, very small ascending aorta, and the coronary arteries.

One may wonder why the ascending aorta is so small in the anatomic complex of mitral atresia, intact ventricular septum, and aortic valvar atresia {S,D,S}, as is shown in Fig. 14.2E . Our understanding is that to the degree to which form is determined by function, in this anatomic complex, the ascending aorta typically functions as a common coronary artery . This may explain why the ascending aorta in this type of mitral atresia is usually only about two coronary arteries wide ( Fig. 14.2E ).

To return to the question concerning why the natural history of this anatomic type of mitral atresia is so lethal (median age at death in unoperated patients, 2.5 days of age), we have seen thus far that the inflow into the heart can be significantly obstructed (typically by a restrictive atrial septum) and that the outflow from the heart also can be importantly obstructed (often by a functionally closing PDA). But what happens between the inflow tract and the outflow tract is also hemodynamically relevant.

The morphologically LV was specifically described as rudimentary or tiny in 39 of these 80 patients (48.75%; see Table 14.4 ) and as absent in 16 of these 80 cases (20%; see Table 14.4 ). Absence of the LV means that after making careful serial sections in the expected site of the LV, no vestige of the LV could be found and identified by careful gross examination using loupes.

This means that this anatomic type of mitral atresia (see Table 14.4 ) can be associated with anatomically single right RV (absence of the LV), and with functionally single RV (marked hypoplasia of the LV). In turn, this also means that the classic definition of single ventricle is inadequate; that is, that single ventricle is present when both AV valves or a common AV valve open into one ventricular chamber (univentricular AV connection). This old definition of single ventricle arbitrarily excludes tricuspid atresia and mitral atresia, which we now understand is unsatisfactory.

Thus, in 55 of these 80 cases, the LV was specifically described as markedly hypoplastic or absent (68.75%; see Table 14.4 ). Again, this statistic is regarded as conservative (i.e., as an underestimate).

In only 1 patient (1.25%; see Table 14.4 ) was the LV described as thick-walled and small-chambered, with endocardial fibroelastosis (EFE) of the LV. A careful search for evidence of a ventricular septal defect (VSD) was negative. This patient, a 22-day-old boy (Case 12), had a modified Norwood procedure at 4 days of age. Autopsy revealed precoronary stenosis: the opening into the coronary compartment at the aortic root was slit-like, nonpatulous, but probe patent (1- × 2-mm internal diameter). Multiple antemortem thrombi were also found in the modified Blalock-Taussig anastomosis. Right-sided juxtaposition of the atrial appendages, with dextromalposition of the left atrial appendage (LAA) that is known to be associated with HLHS, was also found ( Fig. 14.3 ).

However, when the LV is thick-walled and small-chambered, this raises the question: Is the LV really hypoplastic? The answer to this question has proved elusive because it is difficult to know precisely how much of the ventricular septum should be included with the left ventricular free wall—if one wishes to weigh such an LV and compare it with normal control LVs. However, this was the only patient with a thick-walled and small-chambered LV that we encountered in these 80 cases of this anatomic type of mitral atresia (see Table 14.4 ). The overwhelming majority of these patients had LVs that were definitely hypoplastic or absent, as is shown diagrammatically in Fig. 14.1 .

In Case 12, the mitral atresia was described as “membranous.” A membranous form of mitral atresia was found in 14 of these 80 patients (17.5%; see Table 14.4 ). More often, the mitral atresia was described as “muscular,” that is, looking at the floor of the LA, in the expected site of the mitral valve, one sees a muscular floor, with perhaps a small dimple, where the mitral valve “should” be.

Our speculation concerning the thick-walled and small-chambered LV with EFE of Case 12 is that perhaps the mitral atresia occurred later in intrauterine development than it usually does. If the membranous mitral valve were patent for some time in utero, this might perhaps explain why the free wall of the LV was thick and why there was EFE of the small-chambered LV’s endocardium. Or perhaps a VSD had been present but closed premortem. These are hypotheses that we can neither prove nor disprove.

But to return to the main point—a consideration of the hemodynamics in this anatomic type of mitral atresia (see Table 14.4 ), these patients all had a functionally single RV. The tricuspid valve is not optimally designed to occlude an approximately circular systemic AV orifice, whereas the mitral valve is. For further consideration of this important point, please see Chapter 11 .

Tricuspid regurgitation was documented in 9 of these patients (11.25%; see Table 14.4 ); again we suspect that this statistic is probably a significant underestimate. (Autopsy is a good way of establishing structural anomalies, but not of all functional abnormalities.)

Thus, in patients with mitral atresia, intact ventricular septum, aortic valvar atresia, and normal {S,D,S} segmental anatomy, there are multiple hemodynamic handicaps: inflow obstruction, outflow obstruction, and a functionally single pulmonary ventricle (morphologically RV), with a suboptimal AV (tricuspid) valve that is prone to regurgitation when called upon to function as a systemic AV valve.

From a therapeutic standpoint, several other points merit mention:

• Because of the importance of narrowing or functional closure of the PDA in the disastrous natural history of this form of HLHS, it is important to know that prostaglandin E1 can prevent the ductus arteriosus from closing. It is also important to know not to give oxygen to a neonate who may have HLHS, because “everything good” (e.g., oxygen) tends to close a PDA, whereas “everything bad” (e.g., hypoxemia) tends to open the neonatal ductus arteriosus. And therapeutically in this situation, one wants to keep the ductus arteriosus open, to avoid functional “coarctation” of the aorta.

• Regarding the coronary arteries in this anatomic type of mitral atresia, the surgeon should know that the coronary arteries are usually normal, but not always. In the present series, an aberrant third coronary artery originated from the inferior surface of the proximal left pulmonary artery in 1 patient (1.25%; see Table 14.4 ), and an aberrant left circumflex coronary artery arose from the inferior surface of the proximal right pulmonary artery in 1 patient (1.25%; see Table 14.4 ). This aberrant left circumflex coronary artery was oversewn surgically, an important error.

Thus, the surgeon should resist the urge to “tidy things up” by interrupting such mysterious, small, unidentified arteries that course down to the heart from the proximal left or right pulmonary artery. It may be helpful to know that in sharks the coronary arteries arise from the branchial or gill arches that correspond to the aortic arches in humans: the sixth aortic arches contribute to the pulmonary artery branches and the ductus arteriosi bilaterally. Thus, comparative anatomy helps make coronary arteries arising from the proximal pulmonary artery branches comprehensible.

Aortic atresia can be valvar and subvalvar in this complex of anomalies, as was found in 9 patients (11.25%; see Table 14.4 ). What does this mean developmentally? We know that mitral valve tissue can become adherent to the left ventricular septal surface, resulting in severe fibrous subaortic stenosis.

Indeed, we now think that discrete fibrous subaortic stenosis is really a mitral valve malformation, in the sense that this fibrous subaortic tissue appears to be “ un tidied up” AV endocardial cushion tissue, intimately related to the mitral valve. This subaortic fibrous tissue may also help explain why there was no high (subaortic, conoventricular) type of VSD in this complex of anomalies. In our experience, typical fibrous subaortic stenosis always extends over onto the anterior mitral leaflet. This obstructive tissue always remains fibrous and never becomes muscular. If this subaortic tissue were conal (infundibular), it would become muscular. But it never does. Instead, fibrous subaortic stenosis appears to be fibroelastic AV endocardial cushion tissue, closely related to the mitral valve.

In this type of mitral atresia (see Table 14.4 ), the fibrous subaortic tissue can extend 2 or 3 mm below the atretic aortic valve. This important detail merits further study. We speculate that perhaps this totally obstructive subaortic fibrous tissue may be responsible for the atresia of the overlying aortic valve.

Does the combination of mitral atresia, intact ventricular septum, aortic valvar atresia, with normal {S,D,S} segmental anatomy usually occur in isolation, or is it often found with other multisystem congenital anomalies?

MCAs —meaning malformations of other organ systems in addition to the cardiovascular system—were present in 9 of those 80 patients (11.25%; see Table 14.4 ). The salient data may be summarized as follows:

• Sex: Males, 4; females, 5; males to females = 4/5 (0.8:1.0).

• Age at Death: Mean, 5.08 ± 6.57 days; range, 0 (spontaneous stillbirth) to 20 days of age; and median, 2.67 days.

• Findings: The salient findings in these 9 patients with mitral atresia, intact ventricular septum, aortic valvar atresia, normal segmental anatomy {S,D,S}, and MCAs (polysystem malformations) are summarized in Table 14.5 .

  TABLE 14.5

  Multiple Congenital Anomalies in Patients With Mitral Atresia, Intact Ventricular Septum, Aortic Valvar Atresia, and Normal Segmental Anatomy {S,D,S} (n = 9 of 80 [11.25%])

  Case No.	Sex	Age at Death	Multiple Congenital Anomalies
  1. Case 49	Male	2 days	Incompletely bilobed lungs bilaterally
  			Normal karyotype, 46XY
  			Abnormally formed left ear with absent antihelix and hypoplasia of tragus
  			Webbed neck
  			Clinodactyly
  			Hyperextensible thumbs
  			Sacral pit with hair
  			Generalized hypotonia
  			Small penis Microcephaly
  2. Case 55	Female	6 days	Hydronephrosis, left-sided
  3. Case 56	Male	Stillborn fetus	Asymmetrical face with frontal (not induced bossing abortion)
  			Abnormal ears and toes
  			Camptodactyly of hands, bilateral
  			Hydronephrosis, bilateral
  			Very small thymus (1/7 ± 3 g)
  4. Case 69	Female	11 days	Pectus excavatum
  5. Case 84	Female	1½ hours (0.0625 day)	Holoprosencephaly with agenesis of optic nerves
  			Agenesis of olfactory tracts
  			Hypotelorism with single orbit
  			Microphthalmia
  			Single naris
  			Choanal atresia
  			Microstomia
  			Narrow palate
  			Low-set ears
  			Short neck
  			Polydactyly of hands and feet
  			Camptodactyly
  			Abnormal vertebral ossification
  			Duodenal stenosis
  			Double vagina, cervix, and uterus
  			Presacral teratoma
  			Right simian crease
  6. Case 138	Female	1 day	Right cleft lip + palate
  			Normal karyotype, 46XX
  			(Paternal cleft lip and palate)
  7. Case 150	Female	2.67 days	Supernumerary fissure, right upper
  			lobe of lung
  			Double left renal artery
  			Bifid uvula
  8. Case 162	Male	3 days	Male pseudohermaphrodite with relatively large phallus
  			Undescended testes
  			Vagina-like structure
  			Uterus-like structure
  9. Case 178	Male	20 days	Prominent helix, right ear
  			Dysplastic toe nails
  			Sacral dimple
  			History of both parents with congenital deafness, patient’s father with fourth-generation Waardenburg syndrome

Anatomic Findings

The salient anatomic findings in these 2 rare cases of mitral atresia with normal {S,D,S} segmental anatomy, intact ventricular septum, and aortic valvar patency and hypoplasia may be summarized as follows:

• Sex: Male, 1; female, 1; and male to female = 1/1 = 1.

• Age at Death: Mean, 302.5 ± 427.80 days, ranging from 0 days (a 22-week abortus) to 605 days (1 8/12 years), and median, 302.5 days (10.08 months, or 10 months and 2 days). These statistics are included only as data. They are thought to have no statistical significance because of the very small size of this group (n = 2).

Because this rare anatomic subset of mitral atresia consists of only 2 cases, each will be summarized individually.

Case 45 was a 1 8/12-year-old boy who died in 1986. Mitral atresia occurred with normal {S,D,S} segmental anatomy. At autopsy, the ventricular septum was intact; however, a jet lesion involving the left ventricular apex suggested that an old apical muscular VSD had been present at some time before death but had undergone spontaneous closure. Hence, at autopsy, the ventricular septum was found to be intact. The left ventricular papillary muscles were absent. The aortic valve was bicuspid (more accurately, bicommissural), because the intercoronary commissure was absent. Although bicommissural, valvar aortic stenosis was thought not to be present. Left atrial hypertrophy was present, without enlargement. A patent foramen ovale was thought to be restrictive. The RA was both hypertrophied and enlarged. A rete Chiari was present in the RA because of the presence of a prominent right venous valve remnant. Although morphologically unremarkable, tricuspid regurgitation had been documented both angiocardiographically and echocardiographically (two-dimensional echo). Right ventricular hypertrophy and enlargement were massive. A discrete juxtaductal coarctation of the aorta was found, associated with hypoplasia of the aortic arch between the innominate and the left common carotid arteries, the internal diameter of this portion of the aortic arch measuring 3 mm. A restrictive PDA was also present. Hydronephrosis of the right kidney was an associated anomaly.

From a therapeutic standpoint, this boy underwent a modified Norwood procedure (i.e., a stage 1 palliation of HLHS) at 21 days of age. A 10-mm Gore-Tex tube was placed between the proximal main pulmonary artery (MPA) and the distal aortic arch where it merges into the descending thoracic aorta. Pulmonary arterial blood flow was supplied by a modified Blalock-Taussig anastomosis using a 4-mm Gore-Tex conduit.

The unreconstructed transverse aortic arch, with an internal diameter of only 3 mm, became important after spontaneous closure of the apical muscular VSD. After VSD closure, all of the blood flow to the innominate artery, the pulmonary arteries, and the coronary arteries had to pass through the hypoplastic (3-mm internal diameter) transverse aortic arch. The patient died suddenly at home at 1 8/12 years of age.

Case 127 was a female fetus, aborted at 22 weeks gestational age. Again, the salient anatomic findings were mitral atresia, normal segmental anatomy {S,D,S}, and an intact ventricular septum. There was premature closure of the foramen ovale. Left ventricular hypoplasia was moderate, and the left ventricular papillary muscles were absent. The aortic valve was hypoplastic but patent, with severe congenital aortic hypoplasia, the orifice being only 1 mm in diameter. Right atrial hypertrophy and enlargement, with very marked right ventricular hypertrophy and enlargement, and a moderate-sized PDA were also found.

This fetus had MCAs. The right lung displayed abnormal lobation with partially anomalous pulmonary venous connection of the scimitar syndrome type. The right pulmonary veins did not connect with the LA. Instead, the right pulmonary veins left an abnormal and large lateral lobe of the right lung and coursed downward, passing through the right leaf of the diaphragm and into the abdomen. The precise connection of this anomalous right pulmonary vein within the abdomen was not described (because the autopsy was limited to the thorax). Nonetheless, this fetus was thought to have the scimitar syndrome with the characteristically anomalous right pulmonary venous connection and drainage.

The right lung had a very small upper lobe, a large lower lateral lobe, and a smaller medial lobe. No anomalous pulmonary arterial supply to the right lung was described. So this fetus had MCAs with normal cardiac segmental anatomy, mitral atresia, intact ventricular septum, severe congenital aortic valvar “stenosis” (hypoplasia), the scimitar syndrome, and a dysmorphic spleen with a downward dangling “tail.”

Familial HLHS was also found in this patient’s family. Two previous female infants had autopsy-proved diagnoses of HLHS. Our present patient, this 22-week-old aborted fetus, was the third female infant with a definite diagnosis of HLHS. The first sibling died at 4 days of age in 1978. The second died at 3½ weeks of age in 1981. Our patient died in 1984. It is noteworthy that a normal male infant was born to this family in 1979. Some of the father’s siblings have or had congenital heart disease; no other details are known to us. The history suggests that familial HLHS may have been inherited from the father. Be that as it may, this case history appears definitely to document that familial HLHS can and does occur.

Salient Anatomic Features

What were these 27 cases of mitral atresia with normal segmental anatomy, one or more VSDs, and a patent aortic valve really like? First of all, it will be appreciated that this group is hemodynamically the opposite of the previous much larger group (n = 80 patients, 45% of all of our cases of mitral atresia; see Table 14.2 , item 1, and Fig. 14.1 , item 1) in the sense that when the aortic valve is patent, blood can flow in an anterograde (normal) direction in the ascending aorta, whereas when the aortic valve is atretic, the blood must flow in a retrograde (“backward”) direction to perfuse the coronary arteries and the brachiocephalic arteries. In this hemodynamic sense, the present group is very different from the previous one.

Atrial Septum.

The typical finding was a well-formed atrial septum that was obstructive to left-to-right atrial shunting necessitated by mitral atresia. The worst case was premature closure of the foramen ovale, resulting in an intact or impervious atrial septum, that we found in 3 of these 27 patients (11.1%; Cases 71, 92, and 118). However, in a few patients, the atrial septum was not at all obstructive to left-to-right shunting at the atrial level.

Common atrium was found in 1 patient (1/27 = 3.7%; Case 66) who also had MCAs that included Meckel diverticulum, tracheo-esophageal fistula with esophageal atresia, duodenal atresia, a web at the ampulla of Vater, a two-vessel umbilical cord, low-set ears, and a vertebral anomaly (T8 was bifid because of an incompletely fused vertebral body).

The incomplete form of common AV canal with mitral atresia occurred in 1 patient (3.7%; Case 66), who also had MCAs that included Meckel diverticulum, tracheo-esophageal fistula with esophageal atresia, duodenal atresia, a web at the ampulla of Vater, a two-vessel umbilical cord, low-set ears, and a vertebral anomaly: T8 was bifid because of an incompletely fused vertebral body.

The incomplete form of common AV canal with mitral atresia occurred in 1 patient (3.7%; Case 154). The ostium primum type of AV septal defect (also known as an incomplete AV septal defect) was huge and almost constituted a common atrium.

A septum primum malposition ASD was found in 2 of these 27 patients (7.4%; Cases 169 and 170; see Fig. 14.3 ). The septum primum (the flap valve of the foramen ovale) was normally attached inferiorly. But the superior rim of the septum primum was displaced far to the left, into the LA. Consequently, the septum primum had an abnormally horizontal lie and leftward displacement into the LA. Because the superior limbic band of septum secundum was normally located between the “roof” of the right and left atria, and because the superior edge of the septum primum was displaced abnormally far to the left, the superior edge of the septum primum could not close against the superior limbic band of septum secundum in the normal way. This inability of the septum primum to occlude the interatrial communication and to prevent left-to-right atrial shunting we have called a septum primum malposition type of ASD (see Fig. 14.3 ).

It should be understood that a septum primum malposition ASD is different from the much more common secundum type of ASD. The usual type of ostium secundum ASD is caused by deficiencies of septum primum (the atrial septal “door”), deficiency of the superior limbic band of septum secundum (the “door jamb”), or both. However, in a secundum ASD, the septum primum, or its remnant, is normally located. Not so in a septum primum malposition ASD, in which septum primum is approximately horizontal, not approximately vertical (which, of course, is normal) (see Fig. 14.3 ). Septum primum malposition defect is often associated with HLHS, as in these two patients, whereas the typical secundum ASD can be associated with “anything.”

Case 169, an 11-day-old girl with a septum primum malposition ASD, had a large dysmorphic mass involving the inferior limbic band that measured 9 mm in length, 7 mm in width, and 11 mm in height. The tricuspid valve was dysmorphic and displayed tricuspid regurgitation. The tricuspid leaflets were myxomatous, and there were multiple blood cysts involving the tricuspid leaflets.

The other patient with septum primum malposition ASD, Case 170 (a 22½-month-old boy), did not have an additional anomaly of the inferior limbic band of septum secundum—from which the septum primum develops and normally grows upward —and no additional anomaly of the tricuspid valve. Consequently, we think that this abnormally leftward and horizontal lie of the septum primum (see Fig. 14.3 ), resulting in an uncloseable interatrial communication, is not necessarily associated with additional demonstrable malformations of the inferior limbic band and/or of the tricuspid valve.

Malposition of the septum primum into the LA is often better seen echocardiographically (see Fig. 14.3 ) than anatomically (see Fig. 14.2D ) because echocardiographically there is no spatial distortion that may be associated with organ removal and positioning for photography. In this apical four-chamber view (see Fig. 14.3 ), note that the inferior origin of the septum primum is normally located. But what is normally the more superior portion of the septum primum is instead displaced far to the left into the LA. The septum primum is almost horizontally oriented. It is malpositioned to the left of both the right and left pulmonary veins, resulting in totally anomalous pulmonary venous drainage into the RA and then into the right ventricle (see Fig. 14.3 ). The pulmonary veins are normally connected to the LA.

This patient did not have mitral atresia. The diagnosis in this 2½-year-old boy was tetralogy of Fallot (TOF) {S,D,S} with totally anomalous pulmonary venous drainage to the RA. The malpositioned and approximately horizontally oriented septum primum can simulate and be mistaken for a supramitral stenosing membrane with common atrium (see Fig. 14.3 ).

Right-sided juxtaposition of the atrial appendages has recently been described as juxtaposition of the atrial appendages to the right of both great arteries because of rightward malposition of the LAA ( Fig. 14.4 ), that occurred in only 1 of these 23 patients with mitral atresia (3.7%; Case 104). The latter designation specifies that it was malposition of the morphologically LAA to the right of the great arteries that resulted in juxtaposition of the atrial appendages. In this study we encountered no patients with the more common form of juxtaposition of the atrial appendages caused by malposition of the morphologically right atrial appendage (RAA) to the left of the great arteries.

I have very recently realized that the aforementioned descriptions ^, of the atrial appendages as malpositions relative to the great arteries in juxtaposition of the atrial appendages (JAA) is, in fact, wrong. The morphologically RAA or the morphologically LAA are not malpositioned, of necessity, in JAA. The appendages can be malpositioned, but that is not the essence of JAA. Instead, the atrial appendages in JAA are more like “innocent bystanders,” that is, essentially normal. The appearance of JAA is produced by abnormal ventricular morphogenetic movement. With solitus atria and D-loop ventricles, that is {S,D,-}, normal D-loop formation is “a two-step dance.” Step one: the straight heart tube loops to the right, forming a ventricular D-loop. Step two: the ventricles swing horizontally from right to left, passing from dextrocardia, through mesocardia, to normal definitive levocardia.

In the normal human embryo at 23 days of age in utero, the conotruncus lies to the right of both atrial appendages; that is, left-sided JAA is normally present. By 27 days of age, the ventricles have swung far enough leftward so that the conotruncus, mounted atop the ventricles, now runs between the RAA and the LAA, thereby “curing” the left-sided JAA. Thus, a subnormal step 2—a deficient right-to-left swing of the ventricles—results in left-sided JAA. A supernormal or excessive step 2 appears to result in right-sided JAA. Left-sided JAA typically has a hypoplastic right heart syndrome . ^, Right-sided JAA often has HLHS.

In the conventional description of JAA, the “malpositioned” atrial appendage is the atrial appendage that is adjacent to (closer to) the great arteries. In the conventional description of JAA, the adjacent atrial appendage is considered to be malpositioned relative to the great arteries, and relative malposition of the adjacent atrial appendage is indeed present.

However, as suggested earlier, a more basic understanding of JAA is now considered to be that the atrial appendages are essentially normal and can serve as markers of deficient or excessive ventricular movement during development.

Is HLHS, often with a hypertrophied and enlarged LV, responsible for subnormal ventricular right-to-left movement, resulting in left-sided JAA? Is HLHS, with a hypertrophied and enlarged RV, responsible for excessive ventricular right-to-left movement, resulting in right-sided JAA? These findings suggest that right ventricular development may be very important in ventricular right-to-left movement:

• 1.

  When there is a hypoplastic right heart syndrome with a normal or hypertrophied LV, right-to-left movement is deficient, indicated by left-sided JAA in which both great arteries remain to the right of both atrial appendages.

• 2.

  When HLHS is present, with a normal or hypertrophied RV, ventricular right-to-left movement is excessive, as is indicated by right-side JAA, in which both great arteries have been carried to the left of both atrial appendages.

These observations suggest that right ventricular development may be very important in normal right-to-left movement of the ventricles, that is, in step 2 of the normal ventricular D-loop “dance.”

In addition, both left-sided JAA and right-sided JAA are syndromes, because of their characteristic and opposite associations :

• 1.

  left-sided JAA and the hypoplastic right heart syndrome; and

• 2.

  right-sided JAA and HLHS.

Case 104 was a 19-day-old girl, a nonidentical twin, with MCAs, including epicanthal folds, hypertelorism, a cervical rib, and a bifid spleen. The other twin had a cleft palate and hare lip. As will become increasingly apparent, MCAs are an important reality in patients with this type of mitral atresia (see later).

Totally anomalous pulmonary venous connection (TAPVC) is potentially another way of avoiding the deleterious hemodynamic consequences of a normally formed and, hence, obstructive atrial septum leading to pulmonary venous congestion and low cardiac output postnatally. There were only 2 such cases in these 27 patients (7.4%; see Table 14.2 , anatomic type 3):

Case 144 was a 1-day-old boy with a TAPVC to the left innominate vein and then to the RSVC and RA. This, the classic “snowman” type of supracardiac TAPVC, did not have anatomic evidence of obstruction. This newborn boy also had a secundum type of ASD because of a deficient septum primum. The PDA was large. Why, then, did this patient die at only 1 day (30 hours) of age? He also had severe unicuspid aortic valvar stenosis with an aortic valvar orifice of only 1 to 1.5 mm in diameter.

Case 76, the second patient, had TAPVC to the coronary sinus, also without anatomically evident pulmonary venous stenosis. This 15-day-old boy had absence of the morphologically LA. To my knowledge, absence of the LA is a previously unknown and undescribed form of congenital heart disease. Using loupes, we searched unsuccessfully for any evidence of the LA. No patent foramen ovale or fossa ovalis was found.

In view of the rarity of this anomaly— absence of the LA —this heart merits detailed, individual description: mitral atresia; {S,D,S}; moderate left ventricular hypoplasia; two VSDs, a small membranous VSD and a large apical muscular VSD; a hypoplastic bicuspid aortic valve with a rudimentary right coronary–noncoronary (RC/NC) commissure; preductal coarctation of the aorta; right atrial hypertrophy and enlargement; a morphologically normally formed tricuspid valve, but with cardiac catheterization evidence of tricuspid regurgitation and with anatomic confirmation of regurgitation in the form of thickening and rolling of the free margins of the anterior and septal tricuspid leaflets; right ventricular hypertrophy and enlargement; and a PDA. (It should be understood that virtually all patients with mitral atresia and {S,D,S} have right atrial and right ventricular hypertrophy and enlargement.)

Therapeutically, in 1987, this patient (Case 76) at 4 days of age had a subclavian flap angioplasty for coarctation of the aorta, ligation of the PDA, and banding of the MPA. Because of a persisting coarctation gradient, at 15 days of age the patient underwent a modified Damus-Kaye-Stansel procedure in which a 6-mm Gore-Tex tube graft was placed between the proximal MPA and the descending thoracic aorta. Death occurred 6½ hours postoperatively.

Based on our experience with fetal and postnatal congenital heart disease, we now think that all cardiac chambers can be absent, except the morphologically RA. We have never seen or heard of a case in which the RA was absent. Because the RA is in part the systemic venous confluence forming the first cardiac chamber, we think that an RA is essential to the existence of a heart with a systemic circulation.

Partially anomalous pulmonary venous connection was found in 4 of these 27 patients (see Table 14.2 , anatomic type 3; Cases 18, 64, 121, and 122).

In Case 18, a 6-day-old girl, all of the left upper lobe drained anomalously to the RA via a “snowman” pathway (left vertical vein to left innominate vein to RSVC). The right lung drained normally to the LA and anomalously to the RSVC. There was incomplete incorporation of the common pulmonary vein into the LA, but without stenosis of the common pulmonary vein.

This patient also had MCAs with a mongoloid slant of the eyes, an increased carrying angle of the arms, a shield-shaped chest, widely spaced nipples, and pectus excavatum.

In addition, this family exemplified familial congenital heart disease . This patient’s older brother died at 4 9/12 years of age in 1970, and autopsy revealed an incomplete form of common AV canal with a large ostium primum type of ASD (the most common type of incomplete AV septal defect), no VSD of the AV canal type, congenital mitral stenosis with a rudimentary mitral valve with a diminutive cleft of the anterior mitral leaflet, thickening and rolling of the free margins of the anterior and septal leaflets of the tricuspid valve consistent with tricuspid regurgitation, right atrial hypertrophy and enlargement (marked), right ventricular hypertrophy and enlargement, and dilatation of the MPA. Thus, not only did this family display familial congenital heart disease but also the anatomic types of congenital heart disease were similar: mitral atresia in our 6-day-old female patient (Case 18) and a very severe congenital mitral stenosis in her 4 9/12-year-old brother.

Case 64 was a 9-day-old girl with scimitar syndrome . All of the right pulmonary venous blood coursed through and below the diaphragm in a vein that connected with the ductus venosus. An anomalous systemic artery passed upward through the diaphragm and supplied the lower right lung. And there was sequestration of the right lung. Once again there were MCAs, in addition to the right pulmonary sequestration, including accessory spleens or polysplenia (one large spleen and seven splenuli), bilaterally unilobed lungs, and a bicornuate uterus with a continuous vaginal septum.

Case 121 was a 2-day-old boy with a partially anomalous pulmonary venous connection: a very small pulmonary vein (internal diameter = 1 mm) ran from the hilum of the right lung to the RSVC, slightly above the connection of the azygos vein. This patient was premature (36 weeks estimated gestational age, birth weight 1870 g or 4.12 lb), and an “identical” twin. The co-twin was a normal boy. This family was also afflicted with familial congenital heart disease . An older brother, 2 years old, had a VSD with mild pulmonary valvar stenosis. This older brother was also a twin, the co-twin being stillborn.

Case 122 was a stillborn female fetus who underwent a spontaneous intrauterine death at an estimated gestational age of 32 to 34 weeks, with a weight of 1900 g (4.19 lb). This patient was also considered to have a partially anomalous pulmonary venous connection. The right pulmonary veins connected normally with the LA. The left lung was markedly hypoplastic, and no left pulmonary veins were found at autopsy. In other words, no left pulmonary veins connected with the LA or with any other site. The left pulmonary veins were considered to be absent. Note that absence of the left pulmonary veins is different from a partially anomalous pulmonary venous connection. Here we are dealing with a partially anomalous pulmonary venous nonconnection. The problem is that the finding of partial absence of the pulmonary veins is very rare and does not fit into the widely used classification of totally and partially anomalous pulmonary venous connections. I cannot remember ever having seen another such case. This finding is exceedingly rare and perhaps unique. However, this case of absence of the left pulmonary veins did not occur in isolation. Absence of the left pulmonary veins was associated with marked hypoplasia of the left lung, probably a very important additional fact.

This patient also had MCAs (marked hypoplasia of the left lung, as earlier). Left ventricular hypoplasia was the rule in association with mitral atresia and VSD (see Table 14.2 , anatomic type 3). Absence of tensor apparatus (no mitral chordae tendineae and no left ventricular papillary muscles) was specifically described in 8 of these 27 patients (29.63%). Hypoplastic left ventricular papillary muscles were described in 3 patients (11.11%). Therefore, the status of the mitral tensor apparatus was variable.

One or more VSDs were present in all 27 patients. A single VSD was present in 16 of 27 cases (59.26%), and multiple VSDs were found in 11 (40.74%). The anatomic types of VSD were as follows:

• 1.

  conoventricular, 15 of 27 (55.56%);

• 2.

  muscular, 11 of 27 (40.74%); and

• 3.

  AV canal type, 2 of 27 (7.41%).

The aortic valve, although patent, was frequently hypoplastic and bicuspid (15/27 patients, 55.56%); more accurately, it was often bicommissural because of the deficiency or absence of one commissure. Typically with bicommissural aortic valves, all three leaflets were present anatomically, but functionally such aortic valves have only two functioning leaflets.

The number of well-developed commissures equals the number of functional leaflets :

• 1.

  A bicommissural aortic valve is a functionally bicuspid aortic valve.

• 2.

  A unicommissural aortic valve is a functionally unicuspid aortic valve.

• 3.

  When no well-developed commissure is present, such an aortic valve typically is atretic, with no functional aortic valve leaflet(s). Rarely, however, it is possible to have an acommissural aortic valve with hypoplastic (“stenotic”) aortic leaflets and a patent aortic valvar orifice. In such a rare patient, as was presented previously, attempted balloon dilation of the aortic valve led to tearing of the unsupported aortic leaflets, followed by aortic valvar regurgitation.

Which aortic valve commissure was deficient or absent was recorded in 12 of these 15 cases of bicommissural aortic valve:

• 1.

  The RC/LC commissure was deficient or absent in 6 of 12 patients (50%).

• 2.

  The RC/NC commissure was deficient or absent in 3 of 12 cases (25%).

• 3.

  The LC/NC commissure was deficient or absent in 3 of 12 patients (25%).

A unicuspid aortic valve with only one well-developed commissure was described in only 2 of these 27 patients (7.41%; see Table 14.2 , anatomic type 3). The commissures were described in only one of these two patients (Case 10): The RC/LC commissure was well formed, whereas the LC/NC and the RC/NC commissures were rudimentary.

Note that surgical and pathology reports of abnormal aortic valves should always describe the status of the commissures specifically, as earlier, making it possible for others to understand the precise anatomic abnormality. Statements such as “the aortic valve is bicuspid,” or “the aortic valve is unicuspid” leave one wondering, What exactly is wrong with the aortic valve? Which commissure(s) is (are) absent? For clarity, anatomic descriptions must be specific. Specific description of the commissural anatomy revealed that deficiency or absence of the intercoronary commissure (RC/LC) was twice as common a cause of bicuspid aortic valve as were the other two possible types of commissural deficiency (RC/NC and LC/NC) in this subset of mitral atresia (see Table 14.2 , anatomic type 3).

In the area of improving diagnostic and surgical reports concerning congenital heart disease department, the anatomic type of conus (infundibulum) also always should be described specifically. Was it subpulmonary, with aortic–AV valvar fibrous continuity? Was it subaortic, with pulmonary–AV valvar fibrous continuity? Was it bilateral, that is, subaortic and subpulmonary, with no semilunar–AV valvar fibrous continuity? Or was it bilaterally absent or very deficient with bilateral semilunar–AV valvar fibrous continuity? The conus is the often “forgotten,” but very important connecting segment between the great arteries and the ventricles. The conus is part of the great arteries (not part of either ventricle), which is why it is known developmentally as the conotruncal segment. The conus plays a key role in determining the definitive ventriculoarterial (VA) alignments and connections, both normally and abnormally. This is why the conus merits careful study and specific description.

Subaortic stenosis was diagnosed in 5 of these 27 patients (18.52%). Further description in 2 of these patients indicated that posterior malalignment of the conal septum compressed or “squeezed” the immediately subaortic outflow tract. Not included in this estimate of subaortic stenosis (5 patients, 18.52%) is the probable role of a restrictively small VSD(s). Suffice it to say that we regard this statistic (18.52%) to be a significant underestimate of the true incidence of subaortic stenosis from all causes. This impression is heightened by the much higher prevalence of preductal coarctation.

Preductal coarctation of the aorta was found in 12 of these 27 patients (44.44%) with this variety of mitral atresia; see Table 14.2 , anatomic type 3). Preductal coarctation of the aorta (isthmic hypoplasia) is considered to result from a reduction in anterograde aortic blood flow, which in turn points to the hemodynamic importance of mitral atresia, restrictively small VSD(s), subaortic stenosis, and aortic valvar stenosis. There was also 1 patient with interruption of the aortic arch type B (between the left common carotid and the left subclavian arteries). If one includes this case, the incidence of preductal coarctation of the aorta and interruption of the aortic arch becomes 13 of those 27 patients (48.15%).

Hence, these data suggest that at least half of these patients with mitral atresia (see Table 14.2 , item 3) suffered from significantly reduced anterograde aortic blood flow. None had pulmonary outflow tract obstruction (stenosis or atresia). However, two patients (Cases 77 and 98) did have a bicuspid (bicommissural) pulmonary valve.

Case 77 was a 4-day-old boy with trisomy 18 (karyotype proved). He had typical polyvalvar disease with redundant pulmonary leaflets. Case 98, a 2-day-old boy, had MCAs, including a tracheo-esophageal fistula, bilateral cryptorchidism, malrotation of the intestines, Meckel diverticulum, dilatation of the ureters and hypertrophy of the bladder (posterior urethral valves or other obstruction not described), kernicterus (without jaundice), and intrauterine growth failure (birth weight at term = 4 lb 6 oz). Both semilunar valves were bicuspid. The aortic valve had a rudimentary LC/NC commissure. The pulmonary valve’s commissural anatomy was not described.

Although there was typically a large PDA, it was described as functionally closing in 7 of these 27 patients (25.93%). This, too, may be a significant underestimate. Narrowing or closure of a PDA in a patient with mitral atresia can be the immediate cause of death. This is true if the ductus arteriosus is the main, or the only, pathway for anterograde blood flow to the lower body or to the entire systemic arterial circulation. In this hemodynamic situation, narrowing or closure of the ductus arteriosus is equivalent to an acute coarctation of the aorta, or to complete interruption of the descending thoracic aorta. Normal ductal narrowing or closure in this situation is a hemodynamic disaster.

A persistent LSVC opened into the coronary sinus and then into the RA in 5 of these 27 patients (18.52%).

Absence of the right atrial ostium of the coronary sinus was found in Case 13, a 22¾-month-old boy. The coronary sinus blood flow passed into a left vertical vein and then into the left innominate (brachiocephalic) vein, the RSVC, and RA. This anomaly may be regarded as anomalous cardiac venous drainage of the supracardiac type via a snowman pathway, caused by atresia of the right atrial ostium of the coronary sinus. We do not know if this anomalous cardiac venous drainage was total or partial because we do not know if some of the cardiac venous blood returned to the heart via thebesian venous pathways, which, however, is thought to have been probable.

It is also noteworthy that we are talking about cardiac veins, not coronary veins. Anatomists prefer to reserve the term coronary for the arteries of the heart, not for the veins of the heart, because only the arteries form a crown ( corona, Latin) for the ventricles of the heart, whereas the cardiac veins do not. However, many widely used medical dictionaries do not follow this convention.

Anomalous cardiac venous drainage is not a widely recognized anomaly. Textbooks devoted to congenital heart disease often do not even mention anomalous cardiac venous drainage.

An aberrant subclavian artery (or arteries) was (were) found in 3 of these 27 patients (11.11%).

MCAs were prominent in patients with this type of mitral atresia (13/27 patients, 48.15%), and they were highly variable. The term MCAs means that not only was the cardiovascular system involved, but also that other systems of the body were malformed as well. These additional associated anomalies include a tracheo-esophageal fistula with esophageal atresia (5/27 patients, 18.52%); Meckel diverticulum (2/27, 7.41%); anomalous vertebral bodies with failure of normal fusion (2/27, 7.41%); duodenal atresia (1/27, 3.70%); cleft palate, low-set ears, midline skin tag of the neck anteriorly; bilateral simian transverse palmar creases, hypoplastic finger nails, absent thymus (DiGeorge syndrome?), horseshoe kidney, and hydroureter associated with a balanced translocation involving chromosomes 9 and 22, the mother being the carrier of this chromosomal anomaly (1/27, 3.7%); polycystic kidneys (1/27, 3.7%); shield-shaped chest, widely spaced nipples, pectus excavatum, increased carrying angle of the arms, and mongoloid slant of the eyes, associated with familial congenital heart disease. An older brother had incompletely common AV canal with a large ostium primum defect, severe congenital mitral stenosis, and tricuspid regurgitation (1/27, 3.7%); duodenal atresia (1/27, 3.7%); trisomy 18 (1/27, 3.7%); malrotation of the intestines (1/27, 3.7%); dilatation of the bladder and ureters (no obstruction such as posterior urethral valves being identified); kernicterus without jaundice in a patient with intrauterine growth restriction, birth weight at term being 4 lb 6 oz (1/27; 3.7%); epicanthal folds, hypertelorism, cervical rib, and bifed spleen in a nonidentical twin (1/27, 3.7%); marked hypoplasia of the left lung with absence of the left pulmonary veins in a stillborn 32- to 34-week fetus who weighed 1900 g (1/27, 3.7%); syndactyly of the left foot (1/27, 3.7%); and left diaphragmatic hernia of the foramen of Bochdalek type with stomach, spleen, pancreas, left lobe of liver, and a portion of small bowel, cecum, and appendix in the left hemithorax, with marked hypoplasia of a unilobed left lung (1/27, 3.7%).

Familial congenital heart disease was found in the families of 2 (Cases 18 and 121) of these 27 patients with this variety of mitral atresia (7.41%; see Table 14.2 , anatomic type 3). Both of these families have already been mentioned.

Case 18 was the 6-day-old girl with MCAs (mongoloid slant of the eyes, increased carrying angle of the arms, shield-shaped chest, widely spaced nipples, and pectus excavatum) whose 4 9/12-year-old brother had the incomplete form of common AV canal with severe congenital mitral stenosis and tricuspid regurgitation.

Case 121 was the 2-day-old boy—a twin—whose “identical” twin brother was normal. Their older brother, who was 2 years old, had a VSD with mild pulmonary stenosis. This older brother was also a twin, but his twin was a stillborn fetus.

Salient Anatomic Features

Segmental anatomy:

• 1.

  {S,D,S} in 2 of 7 (28.6%);

• 2.

  TGA in 3 to 7 (42.9%):

  • TGA {S,D,D}, 1;

  • TGA {S,D,A}, 1; and

  • TGA {S,L,D}, 1 in which the mitral atresia (or atrial outlet atresia) was right-sided;

• 3.

  DORV in 1 of 7 (14.3%):

  • DORV {S,D,A}, 1; and

• 4.

  double-outlet infundibular outlet chamber (DOIOC):

  • DOIOC {S,L,L} in 1 of 7 (14.3%), in which the mitral atresia (or atrial outlet atresia) was right-sided. The identity of the AV valves corresponds to that of the ventricular loop of entry, not to that of the atria of exit. In a ventricular L-loop, the tricuspid valve is left-sided and the mitral valve is right-sided.

What does double-outlet from the infundibulum outlet chamber (DOIOC) mean? Both great arteries arise from the infundibular outlet chamber or conus arteriosus. This designation also means that neither great artery originates above a right or a left ventricular sinus or inflow tract. A description of the associated ventricular anatomy is necessary for full comprehension.

Ventricular Anatomy

A single LV was present in 4 of these 7 patients (57.1%) (Cases 46, 57, 72, and 93) because the RV (the right ventricular sinus, body, or inflow tract) was absent. A large LV and a small RV (sinus) were found in 3 of these 7 patients (42.9%) (Cases 23, 111, and 148). The presence of a small or hypoplastic RV in a few of these cases made it possible to have DORV, as mentioned earlier.

Other important anatomic findings in this rare form of mitral atresia with a single LV and no right ventricular sinus or with a large LV and a small right ventricular sinus were as follows:

• obstructive atrial septum in 2 of 7 patients;

• coronary sinus ostial atresia with an associated persistent LSVC in 1;

• cor triatriatum in 1;

• straddling tricuspid valve in the 3 patients with a hypoplastic right ventricular sinus; preductal coarctation of the aorta in 3 patients, and interrupted aortic arch (type B, distal to the left common carotid artery) in 1, associated with aortic outflow tract obstruction caused by a restrictive bulboventricular foramen;

• pulmonary infundibular outflow tract stenosis in 1;

• an aberrant right subclavian artery in 1;

• a single left coronary artery (absence of the right coronary arterial ostium) in 1; high ostium of the left coronary artery in 1; and hypoplasia of the left coronary arterial ostium in 1 patient, with enlargement of the right coronary arterial ostium; and closing PDA in 2 of these 7 patients.

Now it is time to present this infrequent form of mitral atresia with a large morphologically LV and a small or absent morphologically right ventricular sinus photographically ( Fig. 14.6 to Fig. 14.14 , inclusive). These figures and their legends merit careful consideration.

Developmental Hypothesis

Saying that mitral atresia with a large morphologically LV is a “developmental impossibility,” as I did by way of introduction to this rare form of mitral atresia is, of course, itself an impossibility. Anything that in fact occurs cannot be developmentally impossible. (But it can be mind boggling.)

How is it possible to have a large LV, or even a single LV, in association with mitral atresia? Almost always with mitral atresia, the LV is diminutive. But occasionally, the LV can be large. Why? Our hypothesis is that the small or absent right ventricular sinus is the key. When the RV inflow tract is hypoplastic or absent, the ventricular septum, or its remnant, is displaced toward the side of the hypoplastic or absent RV—to the right with a ventricular D-loop or to the left with a ventricular L-loop. This ventricular septal displacement is one type of ventriculoatrial malalignment. The geometry of ventriculoatrial malalignment in this type of mitral atresia is presented in Fig. 14.15 .

With D-loop ventricles, for example, rightward displacement of the ventricular septum places the LV abnormally to the right of the atrial septum, beneath the right-sided RA and the tricuspid valve. Consequently, the blood from the left and right atria flows through the tricuspid valve and into the abnormally right-sided LV, permitting the LV to grow and become large.

Usually with typical mitral atresia, when the right ventricular sinus or inflow tract develops well and the LV is diminutive, the ventricular septum comes to underlie the atrial septum. Because of typical mitral atresia, blood cannot flow from the embryonic and fetal LA into the LV; hence, the LV, deprived of pressure work and flow work, remains diminutive. The RV grows and develops and becomes functionally the systemic ventricle of the embryo and fetus in typical cases of mitral atresia, such as those considered previously.

Ventricular septal displacement toward the side of the absent ventricle is what one sees in hearts with single LV, ^, which 4 of these 7 patients with mitral atresia, large LV, and absent right ventricular sinus had. Displacement of the ventricular septum toward the side of the hypoplastic RV is what one sees anatomically in straddling tricuspid valve, which is what 3 of these 7 patients had. When a lung is hypoplastic or absent, a very similar phenomenon occurs: the mediastinum is displaced toward the side of the hypoplastic or absent lung.

Summary

Mitral atresia with a large LV and a small or absent right ventricular sinus and with the resulting ventriculoatrial malalignment occurred with normally related great arteries and normal {S,D,S} segmental anatomy (2/7, 28.6%), with TGA and D-loop TGA {S,D,D} and TGA {S,D,A} (2/7, 28.6%), with L-loop TGA (S,L,D} (1/7, 14.3%), with D-loop DORV {S,D,A} (1/7, 14.3%), and with L-loop DOIOC {S,L,L} (1/7, 14.3%). Absence of the right ventricular sinus occurred in 4 of 7 patients (57.1%), resulting in single LV, and hypoplasia of the right ventricular sinus was found in 3 of 7 (42.9%), resulting in a large LV and a small RV.

To the best of our knowledge, this anomaly was first reported by Dr. Manuel Quero of Madrid, Spain in 1970. Dr. Quero, a good friend, was trying to show us that a single ventricle could occur with AV valvar atresia, that is, that the then accepted definition of single ventricle was wrong. The classic premorphologic definition was single ventricle is present if both AV valves or a common AV valve open entirely or predominantly into one ventricular chamber. Quero’s point, to quote Cole Porter, was “It ain’t necessarily so.” Look at this case of mine, he was saying. It has a single LV with absence of the right ventricular sinus and normally related great arteries. The segmental anatomy is normal {S,D,S}, but it also has mitral atresia. In fact, this is a Holmes heart with mitral atresia.

Our Case 93 is Dr. Manuel Quero’s case, which he very kindly sent to us, just in case we had any lingering doubts. We did not. We fully agreed that Quero was right. We also knew that the classic premorphologic definition of single ventricle cited earlier was wrong for several additional reasons:

• 1.

  Double-inlet or common-inlet RV typically has a small LV. An anatomically single RV usually is not present; instead, one finds morphologically that there is a large RV and a diminutive LV (not a single RV).

• 2.

  The Lambert heart ^, also violates the classic definition of single ventricle because double-inlet LV (DILV) is not present. Instead, the tricuspid valve opens predominantly or entirely into the infundibular outlet chamber.

We refer to mitral atresia with a single LV, an absent right ventricular sinus, an infundibular outlet chamber, and normally related great arteries with normal {S,D,S} segmental anatomy as the Quero heart, in his honor.

One may ask why we put all 7 of these cases of mitral atresia with a large LV and a small or absent right ventricular sinus together into anatomic type 6 (see Table 14.2 and Fig. 14.1 ), despite the fact that there are several very different segmental combinations in these 7 patients. The answer is that we think of mitral atresia with a large LV and a small or absent right ventricular sinus as one anatomic type of mitral atresia, which occurred in five different segmental anatomic sets: (1) {S,D,S} in 2; TGA {S,D,D/A} in 2; TGA {S,L,D} in 1; DORV {S,D,A} in 1; and DOIOC {S,L,L} in 1 patient.

An alternative way of classifying mitral atresia with a large LV and a small or absent right ventricular sinus would be to regard each of the five foregoing segmental sets as an anatomic type of mitral atresia. The disadvantage of the latter approach is that it would increase the anatomic types of mitral atresia from 18 (see Table 14.2 and Fig. 14.1 ) to 23. As a practice matter, our preference has been to keep the number of anatomic types of mitral atresia (and of all other anomalies) as low as possible, without oversimplifying the data.

In geometric Fig. 14.15 and in those preceding this figure, note that in this infrequent type of so-called mitral atresia with a large LV and with a small or absent right ventricular sinus, ventriculoatrial malalignment often resulted in blocking from below what should have been the mitral orifice. This occurred both with D-loop ventricles (see Fig. 14.15 A–F , inclusive) and L-loop ventricles (see Fig. 14.15G–H ). Often, ventriculoatrial malalignment placed the left ventricular free wall beneath what should have been the mitral orifice (see Fig. 14.15A–E, H ). Occasionally, a prominent muscular ridge in the LV underlay appeared to block what should have been the mitral orifice (see Fig. 14.15 F–G ).

Thus, in this infrequent anatomic type of mitral atresia (see Fig. 14.1 type 6), the primary problem appears to be ventriculoatrial malalignment, not atresia of the mitral valve.

Mitral atresia, that is, atrial outlet atresia with a large LV and a small or absent RV, can be diagnosed angiocardiographically ( Fig. 14.16 ) and echocardiographically ( Fig. 14.17 ).

Anatomic Features

The salient anatomic features of these 28 patients with mitral atresia, DORV{S,D,D/“S”}, and VSD (see Table 14.2 , anatomic type 8) are summarized in Table 14.7 .

In the interests of attempted brevity, I must let Table 14.7 largely speak for itself. However, there are several noteworthy aspects that do require some explanation.

When DORV occurs in association with HLHS—of which mitral atresia is an excellent example, the conus (infundibulum) is often unilateral—subpulmonary (only), or subaortic (only) —rather than bilateral, as is typical of DORV with two well-developed ventricles.

Note that a subpulmonary conus with aortic-to-tricuspid direct fibrous continuity occurred in 18 of these 28 cases (64%; see Table 14.7 and Figs. 14.3D–E and 14.18B–C ). Of these 18 patients, 4 had a TOF-like infundibulum and great arteries, with pulmonary outflow tract stenosis. Because the conotruncus was basically of the normal type, on external and internal inspection, the great arteries looked almost solitus normally related. But DORV was present, so the great arteries were not entirely normally related; hence, the designation DORV {S,D,D/“S”}. “S” indicates that the great arteries were almost solitus normally related, as mentioned earlier. So, in this sense, these cases displayed the almost normally related great arteries type of DORV . And when the subpulmonary infundibulum and the pulmonary valve were obstructive, such cases exemplified what may be called the TOF type of DORV.

When the pulmonary outflow tract was widely patent with a subpulmonary infundibulum, this situation was often associated with aortic outflow tract obstruction. Because the subaortic infundibular free wall was resorbed, permitting aortic-tricuspid fibrous continuity, the subaortic outflow tract was often “squeezed” between the conal septum anteriorly and the tricuspid valve posteriorly, resulting in subaortic stenosis. Note the high incidence of subaortic stenosis (13 patients, 46%; see Table 14.7 ). This was not the only mechanism of subaortic stenosis but also was a frequent and important one.

When the subpulmonary infundibulum was underdeveloped, resulting in a TOF-like conotruncus, the infundibular septum was deviated anteriorly, superiorly, and leftward—away from the tricuspid valve, which opened up the aortic outflow tract.

Note also the prevalences of bicuspid aortic valve (5 patients, 18%), aortic valvar stenosis (4 patients, 14%), preductal coarctation of the aorta (8 patients, 29%), and interrupted or atretic aortic arch (type B) (2 patients, 7%), all indicating aortic outflow tract obstruction.

The other type of unilateral conus, subaortic only, with pulmonary-tricuspid fibrous continuity, resulted in the transposition type of DORV associated with mitral atresia. On external and internal inspection, the great arteries looked very much like TGA, except that DORV was in fact present. The transposition-like type of DORV occurred in only 2 patients (7%; see Table 14.7 ).

A bilateral conus (subaortic and subpulmonary) was found in only 3 of these cases of DORV with mitral atresia (11%; see Table 14.7 and Fig. 14.18 ). Although a bilateral conus is the rule in association with DORV and two well-developed ventricles, it was the exception in those patients with HLHS.

Although all 28 of these patients with mitral atresia had a VSD, multiple VSDs were found in 8 patients (29%; see Table 14.7 ). When VSDs were multiple (typically two), there was always a high conoventricular type of VSD and a lower muscular type of VSD.

Other noteworthy findings in Table 14.7 include:

• absence of the LA (in 1, 4%);

• absence of the LAA (in 2, 7%);

• leftward malalignment of the septum primum (in 2, 7%) and right-sided juxtaposition of the atrial appendages (in 1, 4%).

Anatomic Features

First, let’s answer the question posed by the very existence of this anatomic type of mitral atresia: how is it possible to have DORV {S,D,D} without a VSD? There are two logical possibilities, and both were found to occur.

First, the ventricular septum can indeed be intact, as in 4 of these 7 rare cases (57%): Cases 157, 158, 167, and 173 ( Fig. 14.19 ). In 3 of these 4 patients, the LV was diminutive or tiny (Cases 157, 167, and 173). However, in 1 patient with membranous mitral atresia (not the more common muscular mitral atresia in which there may be a dimple in the floor of the LA but no sign of membranous mitral valve tissue), the ventricular septum was intact (no VSD), but the LV was small-chambered and thick-walled, that is, not definitely hypoplastic (Case 158).

Thus, it is interesting to realize that in HLHS, not only is the LA often hypertrophied (not at all hypoplastic), but also the LV may not be definitely hypoplastic. When the LV is thick-walled but small-chambered (because it is doing pressure work, but no flow work), whether the LV is truly hypoplastic is uncertain.

Hypoplastic (adjective) and hypoplasia (noun) in pathology mean weighing significantly less than is normal for the age, often qualified as meaning more than 2 standard deviations below the mean that is normal for the age. The reason this has been a difficult problem to solve in HLHS when a thick-walled but small-chambered LV is present is as follows. It is easy to weigh the left ventricular free wall. But what about the left ventricular component of the interventricular septum? How much of the muscular interventricular septum belongs to the LV (as opposed to the RV), and how much septal myocardium should be weighed to give an accurate weight of the LV as a whole, meaning the left ventricular free wall and the LV component of the muscular ventricular septum? It is because of uncertainty concerning how much of the ventricular septal myocardium to include in the weight of the whole LV that this question has thus far eluded a definite answer. However, it is suspected that such a thick-walled and small-chambered LV, as in Case 158, my well not be significantly hypoplastic.

Another perhaps more important problem associated with the diagnosis of HLHS is its nonspecificity. What are we talking about—mitral atresia, aortic atresia, both, or neither? In individual cases, we prefer specific diagnoses, such as mitral atresia. However, we agree that HLHS may be a useful general heading covering the many different anomalies that can be associated with hypoplasia of the LV.

Now, to return to the question, How can DORV {S, D, D} have no VSD? The first mechanism, as earlier, is yes, indeed, the ventricular septum can be intact with DORV, both when the LV is tiny and when the LV is thick-walled and small-chambered (and only questionably hypoplastic). These are anatomic facts. The real question is How is it possible for such a patient to survive prenatal life? We speculate that the via sinistra is minimal to nonexistent, with little or no inferior vena cava (IVC) blood flow going into the LA because of the presence of mitral atresia. In addition, the pulmonary venous return to the LA late in pregnancy must be able to shunt left-to-right from the LA into the RA.

The second mechanism is when the LV is absent, resulting in an anatomically single RV. When the LV is absent, there can be no VSD. It should be understood that VSD is a short form for IVSD. There can be no IVSD between the RV and the LV unless both ventricles are present.

With single RV, there is no bulboventricular foramen. Again, it should be understood that bulboventricular foramen is a short form for infundibulo-left ventricular foramen, when the right ventricular sinus is absent, as in single LV with an infundibular outlet chamber. In single LV, one cannot talk accurately about a VSD, because the RV (meaning the right ventricular sinus, body, or inflow tract) is missing. There can be no IVSD (or VSD) when the RV is absent and only the LV is present; hence, the term bulboventricular foramen is used, not VSD, in the interests of anatomic accuracy.

This is one reason why it is important to understand that the infundibulum or conus arteriosus is not a ventricle. The conal or infundibular outlet chamber is often mistakenly regarded as a small, deformed, or hypoplastic RV. In fact, the conus arteriosus and the ventricles belong to different cardiac segments. The conus is the crucial connecting segment between the great arteries and both ventricles (not just the RV). The RV is part of a main cardiac segment (the ventricular loop) that normally provides the systemic pump (the left ventricular sinus) and the pulmonary pump (the right ventricular sinus). The infundibulum or conus is not a good pump; its function normally is to perform an aortic switch from above the RV to above the LV during cardiogenesis to achieve concordant VA alignments. The conus is an embryonic “architect” but not a good pump.

The LV was absent, resulting in single RV, in 3 of these 7 patients (43%): Cases 39, 153, and 164 . The anatomic type of mitral atresia (9) (see Table 14.2 and Fig. 14.1 ) is called mitral atresia with no VSD and DORV {S,D,D}, not mitral atresia with intact ventricular septum and DORV {S,D,D}, because when the LV is absent , there can be no VSD. With single RV and absent LV, it is difficult to identify the ventricular septum with certainty when there is no left ventricular septal surface. The designation “intact ventricular septum” suggests that there could be a VSD, but the ventricular septum has been identified, examined, and found to be intact. However, when the LV is absent, none of the foregoing connotations is correct. There cannot be a VSD, and the ventricular septum has not been identified precisely and examined. Hence, “no VSD” is preferred to “intact ventricular septum” in Fig. 14.1 , mitral atresia type 9.

It is also noteworthy that single RV is not included in conventional classifications of DORV. Other important findings in these 7 cases of mitral atresia with DORV {S,D,D}, single RV (absent LV), and no VSD are summarized in Table 14.8 . In Table 14.8 , the atrial septum was highly obstructive in 3 of these 7 patients, totally obstructive in 1, and not obstructive in 3.

The patient with a large ostium primum type of defect had common AV canal with atresia of the mitral component of the common AV valve. Because this patient also had absence of the LV, resulting in a single RV, the patient had no ventricular septum. Consequently, it was impossible to say whether a VSD of the AV canal type was present or would have been present had a ventricular septum been present anatomically. This is why our diagnosis was common AV canal, not otherwise qualified . Was this a complete form of common AV canal or a partial form? We do not know.

This is an instructive problem. Conventionally, we classify common AV canal (AV septal defect) not in terms of itself, but in terms of another variable, the ventricular septum, to decide whether the common AV canal is complete or partial. This approach works well, except when the other variable is abnormal, or nonexistent as in Case 164. Then one understands the weakness of classifying one variable (such as the anatomic status of the AV junction) in terms of another variable (such as the ventricular septum).

Classifying one variable primarily in terms of a different variable is a fundamental error in logic and mathematics. Then why do we physicians do this so often? Because these other variables are clinically and surgically so important. Despite being illogical, this method of classification works well as long as the other variable (or variables) is (are) relatively normal. But when the other variable—such as the ventricular septum in the classification of common AV canal—is very abnormal, this practical, if illogical, approach to classification breaks down, as in Case 164 (earlier). The point is that when such breakdowns in classification occur, we should not be too surprised because classifying one variable in terms of another variable, instead of in terms of itself, is fundamentally flawed, an error in logic.

Rightward malposition of the septum primum, that is, when the septum primum is on the right atrial side (instead of being on the left atrial side, which is normal) is a noteworthy finding (see Table 14.8 ). Heretofore, we have considered leftward malposition of the septum primum in association with various types of HLHS. In rightward malposition of the septum primum, this septum (the flap valve of the foramen ovale) lies to the right of the superior limbic band of septum secundum, thereby placing the septum primum on the right atrial side. Rightward malalignment of the septum primum is often associated with the heterotaxy syndrome and polysplenia with interruption of the IVC. Was the polysplenia syndrome present in our patient (Case 167)? Unfortunately we do not know; this autopsy was limited to the heart and lungs and did not include the abdomen. However, because Case 167 had bilaterally hyparterial bronchi, bilaterally bilobed lungs, and rightward malposition of the septum primum, we strongly suspect that the heterotaxy syndrome with polysplenia was indeed present, but we shall never know for sure.

Rightward malposition of the septum primum may well be related to the abnormal hemodynamics associated with interruption of the IVC, which is common with visceral heterotaxy and polysplenia. When the IVC is interrupted, the via sinistra of the IVC’s normal blood flow into the RA and through the patent foramen ovale into the LA and then to the left heart and up to the brain is greatly reduced because the venous blood flow from the lower body is conveyed to the SVC (right, left, or both) by a necessarily enlarged azygos vein(s), which is often called an “azygos extension.”

Normally, the via sinistra “pushes” the septum primum into the LA, to the left of the superior limbic band of the septum secundum. But when the IVC is interrupted, as is frequent with the polysplenia syndrome, there is much less blood flowing into the RA via the hepatic and suprahepatic portions of the IVC, and, hence, the hemodynamic forces “pushing” the septum primum into the LA are considerably reduced. Consequently, with interruption of the IVC in the polysplenia syndrome, the septum primum can be startlingly right-sided—on the right atrial side, instead of on the LA side, which is normal. This is our best hypothesis to explain rightward malposition of septum primum, which we have repeatedly found in association with the heterotaxy syndrome and polysplenia.

It also should be understood that it is almost always possible to diagnose the atrial situs (solitus, or inversus) in the heterotaxy syndrome with polysplenia ^, because the concept of atrial isomerism in the heterotaxy syndromes with asplenia, polysplenia, or occasionally with a normally formed but right-sided spleen, is anatomically erroneous (see Chapter 29 for further discussion).

Anatomic Features

The atrial septum was highly restrictive in 2 of these 5 patients (Cases 16 and 41) (40%). TAPVC of the supracardiac type occurred in 1 of these 5 cases (20%) (Case 60). The anomalous venous pathway included a stenotic vertical vein to the RSVC. A persistent LSVC that drained into the coronary sinus was not part of the anomalous pulmonary venous pathway. A persistent LSVC connected with the coronary sinus and then to the RA in 2 patients (Cases 60 and 91) (40%). Rightward displacement of the septum primum that lay to the right of the superior limbic band of the septum secundum was present in 1 patient (Case 60) (20%); that is, the septum primum was on the right atrial side, instead of being on the left atrial side, which is normal. A straddling tricuspid valve, through a VSD of the AV canal type, was associated with a small right ventricular sinus (body, or inflow tract) in 1 patient (Case 14) (20%). Absence of the RV (the right ventricular sinus) resulted in single LV with an infundibular outlet chamber and D-TGA in 1 patient (Case 16) (20%). An intact ventricular septum was found in 1 of these 5 patients (Case 91) (20%). This patient also had stenosis of the proximal left pulmonary artery at the ductal insertion site (20%). It is well to remember that a closing ductus arteriosus can cause important obstruction at both of its ends: superiorly, ductal medial musculature can encircle and constrict the aorta, causing discrete coarctation of the aorta; and inferiorly, ductal medial musculature can encircle and constrict a proximal pulmonary artery branch, as in this patient.

The transposed great arterial outflow tracts were obstructed in all cases, which was a striking finding; pulmonary outflow tract atresia was present in 4 of these 5 patients (Cases 14, 41, 60, and 91) (80%). This is what one would expect in association with mitral atresia and D-TGA in which the pulmonary artery arises from a typically diminutive LV. However, in the infrequent form of mitral atresia with a single LV (Case 16) (20%), the transposed pulmonary artery arises from a large LV, and in this situation the pulmonary outflow tract typically is widely patent, and so it was in this patient. When a D-transposed aorta arises from an infundibular outlet chamber, the patient is vulnerable to subaortic stenosis from a restrictively small bulboventricular foramen, if the conal septum above the bulboventricular foramen is aligned too well with the ventricular septal remnant below the bulboventricular foramen. Then the foramen can be restrictively small, as was the case in Case 16. Proof of the significant reduction in anterograde aortic blood flow was provided by hypoplasia of the ascending aorta and aortic arch and preductal coarctation of the aorta (internal diameter 1 mm). The coup de grâce in this 7-week-old girl was that the PDA was also closing, depriving her lower body of adequate systemic arterial blood flow.

The ductus arteriosus was found to be closing in 3 of these 5 patients (Cases 16, 41, and 60) (60%). In view of the very high prevalence of great arterial outflow tract obstruction in this anatomic type of mitral atresia (100%)—pulmonary outflow tract atresia in 80% and aortic outflow tract obstruction in 20%—normal or even delayed closure of the ductus arteriosus, although a normal phenomenon, is in these patients a hemodynamic catastrophe, as noted earlier.

Anatomic Features

The atrial septum was obstructive in all 4 of these patients (Cases 7, 59, 125, and 166). In one patient (Case 7), a finely fenestrated but obstructive septum primum bulged far into the LA; this aneurysm of septum primum constituted supratricuspid stenosis above the left-sided tricuspid valve ( Fig. 14.20 ). All four patients had visceroatrial situs discordance.

Why is the atrial septum often restrictive with right-sided mitral atresia? When the atrial septum is normally formed, with a well-formed septum primum (the “door” that normally opens into the LA prenatally) and a well-formed superior limbic band of septum secundum (the “door jamb” against which the septum primum normally closes postnatally) and when the pulmonary veins are normally connected with the LA, after birth and expansion of the lungs, the bright red pulmonary venous blood comes flooding into the LA, tending to close the septum primum against the septum secundum in the normal way.

However, this normal narrowing of the interatrial communication makes the atrial septum obstructive when all of the right atrial systemic venous return must shunt right-to-left because of right-sided mitral atresia. Consequently, if the atrial septum is well formed, it usually is obstructive regardless of whether the mitral atresia is left-sided or right-sided.

Dextrocardia was present in 2 of these 4 patients (Cases 7 and 125) (50%) (see Fig. 14.20 ). Why? Because L-loop ventricles “belong” in the right hemithorax, just as D-loop ventricles “belong” in the left hemithorax. Ventricular loop formation may be regarded as a two-step “dance.” Normally, the straight heart tube first loops to the right (forming a D-loop); and then later the ventricular apex swings from right to left, ending up predominantly in the left hemithorax, resulting in levocardia. Abnormally, the straight heart tube first loops to the left (L-loop formation), and later the ventricular apex may then swing in the opposite direction, to the right, resulting in dextrocardia, as in situs inversus totalis. When L-loop ventricles are present in a left-sided heart, levocardia is an additional (often unrecognized) abnormality.

The anatomic type of conus was recorded in 2 of these 4 patients: it was bilateral in 1 (Case 7) and subaortic (only) in 1 (Case 59). Thus, a unilateral conus, subaortic only , with pulmonary-to–AV valve direct fibrous continuity, also occurs with discordant L-loop ventricles with right-sided HLHS.

Case 59, a 2 10/12-year-old boy, has several other lessons to teach us. He had an infrequent form of incompletely common AV canal with no ostium primum type of ASD, a common AV valve with right-sided mitral atresia, and an undivided and unattached anterior leaflet of the common AV valve (type C–like), and a large VSD of the AV canal type. This patient had a mildly small right ventricular sinus (left-sided), a double-chambered RV (left-sided), and a stenotic subaortic os infundibuli that was associated with moderate hypoplasia of the aortic isthmus (but the hypoplasia was not severe enough to warrant the diagnosis of preductal coarctation of the aorta). But the remarkable finding was that this patient with right-sided mitral atresia had hypertrophy and enlargement of the right-sided LV . So this patient with mitral atresia did not have HLHS (meaning the hypoplastic morphologically LV syndrome). How is this possible? We thought that the presence of a large VSD beneath the anterior and posterior leaflets of the common AV valve may have made it possible for the LV to perform enough pressure and flow work, in combination with the mild hypoplasia of the RV, for the LV to become hypertrophied and enlarged, despite the coexistence of mitral atresia. We interpreted the VSD as a confluent conoventricular plus AV canal type of VSD. One may also conclude that this patient demonstrates that it is possible for right-sided mitral atresia to be associated with a large LV (right-sided) and a mildly small RV (left-sided). At the AV valvar and ventricular levels, these findings are a mirror-image of the findings presented earlier in mitral atresia type 6 with left-sided mitral atresia and a ventricular D-loop (see Fig. 14.1 ). This same patient also had an obstructive atrial septum that resulted in right atrial congestive heart failure with hepatomegaly, ascites, and prominent right atrial pulsations in the venae cavae. At cardiac catherization, the “a” wave measured 26 mm Hg. It will be recalled that this patient’s incomplete form of common AV canal had no ostium primum type of ASD.

This patient illustrates yet another important point, namely, t hat mitral atresia can occur in association with common AV canal . In other words, mitral atresia with common AV canal is one of the anatomic types of mitral atresia; conversely, common AV canal with mitral atresia is one of the anatomic types of common AV canal. Although they are two different diagnoses, mitral atresia and common AV canal are not mutually exclusive malformations and can occur together. The same is true of tricuspid atresia and common AV canal; they too can coexist.

The great arterial outflow tracts in these 4 patients with right-sided mitral atresia and DORV {S,L,L} were very suboptimal. There was pulmonary outflow tract obstruction in 3 of these 4 patients (75%): pulmonary outflow tract stenosis in 2 (Cases 7 and 166), and pulmonary outflow tract atresia in 1 (Case 125).

Case 166 had congenital absence of pulmonary valve leaflets in association with severe subvalvar pulmonary outflow tract stenosis (3 to 4 mm in internal diameter). Congenital absence of the pulmonary valve (leaflets) is usually associated with TOF, and is seen only rarely in other settings, such as in Case 166. Case 166 also had marked left-sided tricuspid regurgitation, with thickening and rolling of the tricuspid leaflet free margins.

Finally, Case 166 had one other noteworthy finding: membranous mitral atresia. Typically, when one examines the atrial “floor” where the mitral valve is expected to be, one often sees nothing, that is, no fibrous mitral valve–like tissue but rather only muscle. This we call the muscular type of mitral (or tricuspid) atresia. But in this patient, there was membranous tissue where the mitral valve was expected to be. Perhaps in cases of membranous mitral atresia, the concept of mitral atresia may be accurately valid. Perhaps a mitral valve was present but failed to develop an orifice. The etymology of atresia is: a, without, and tresis, hole, Greek. By contrast, in the so-called muscular form of mitral atresia, the mitral valve may well be absent; hence, the concept of mitral atresia may not be accurate in muscular mitral atresia . The alternative concept of absent AV connection has the advantage of, at the least, not being developmentally and/or anatomically wrong. We clearly have much to learn about the cause and morphogenesis of what is generally known as mitral atresia.

Absence of the right-sided morphologic LV resulted in a single morphologically RV in a 50-day-old girl with right-sided mitral atresia, DORV {S,L,L}, and pulmonary outflow tract atresia. Accurately speaking, this would be a case of the aplastic LV syndrome (not HLHS).

Again it should be remembered that I am not trying to change conventional diagnostic terminology. Instead, my hope is to deepen anatomic and developmental understanding. We continue to use the conventional designations mitral atresia (and tricuspid atresia ), even though we fully agree that we have much to learn concerning their etiology, morphogenesis, and anatomy.

Anatomic Features

The cardiac position in the thorax was noteworthy in two patients. Case 62 had dextrocardia, which is normal for visceroatrial situs inversus but abnormal for D-loop ventricles. Case 25 had mesocardia, which is always abnormal. Mesocardia may be regarded as incomplete cardiac morphogenetic movement in the following sense. Ventricular looping to the right (D-loop formation) has occurred, but the subsequent swing of the ventricular apex from right to left was incomplete. In this patient, mesocardia is regarded as half way to levocardia (which is normal for D-loop ventricles, but abnormal for viscerotrial situs inversus). Perhaps mesocardia in DORV {I,D,D} may be regarded as a “compromise” between what is normal for visceroatrial situs inversus (i.e., dextrocardia) and what is normal for D-loop ventricles (i.e., levocardia); however, this is speculation.

The atrial septum was restrictive in 2 of these 3 patients (Cases 25 and 160). In the third patient (Case 62), an incompletely common AV canal with mitral atresia had a large typical ostium primum defect that did not restrict left-to-right shunting at the atrial level necessitated by left-sided mitral atresia. This same patient (Case 62, a 7 4/12-year-old boy) had visceral heterotaxy but with a normal spleen. In other words, this rare patient had the asplenia syndrome, but with a normally formed left-sided spleen, and with atrial situs inversus (not atrial situs ambiguus, and not “right atrial isomerism”). The right-sided IVC veered to the left at the level of the liver, received the hepatic veins, and then drained into the left-sided RA. A right-sided SVC drained into the coronary sinus and then to the left-sided RA. A LSVC connected with the left-sided RA; hence, there were bilateral SVCs. There was TAPVC to the junction of the LSVC with the left-sided RA, without obstruction of the pulmonary venous return. This relatively “old” patient (7 4/12 years) had significant and hemodynamically disadvantageous tricuspid regurgitation .

Case 62’s segmental anatomy was DORV { AI ,D,D}; A indicates situs ambiguus of the viscera; and I means situs inversus of the atria. Note that A and I are not separated by a comma because the viscera and the atria are regarded as different parts of the same cardiac segment, that is, the visceroatrial segment.

This kind of case has taught us that it is definitely possible to have visceral heterotaxy, very similar if not identical to that of the asplenia syndrome, but with a normally formed spleen. This is why we prefer the heterotaxy syndrome as the main diagnosis (rather than the asplenia syndrome); but we always describe the anatomic state of the spleen with care (in view of its importance in the body’s defenses against infectious diseases). Also, the spleen can be present, but hypoplastic in mitral atresia with DORV {I,D,D}, as in Case 25, in which the spleen weighed only 60% of the normal weight. The LV was hypoplastic in all 3 patients. In 2, absence of the mitral tensor apparatus—absence of chordae tendineae and papillary muscles—was documented. The ventricular septum was intact in 2 of these 3 patients: in Case 160, ^a

aAlthough it is impossible to thank all of our many friends and colleagues from all over the world, I would like to acknowledge Dr. Walter Silver and Dr. Nakul Chandra from Maimonides Medical Center in Brooklyn, NY, who kindly sent Case 160 to Dr. Stella Van Praagh and me for consultation in 1987. Because of the kindness and generosity of so many friends, we have learned much that we have never previously been able to publish. This work is an attempt to say thank you and to make amends.

and in Case 62 with an incomplete form of common AV canal. Case 25 had a high, small conoventricular type of VSD.

The anatomic type of conus was recorded in 2 of these 3 patients: bilateral in 1 (Case 25), and subaortic (only) in 1 (Case 160). Pulmonary outflow tract obstruction was found in 2 of these 3 patients: pulmonary infundibular and valvular stenosis in Case 25, who had a bilateral conus, and pulmonary outflow tract atresia in Case 160, who had a subaortic conus. Case 25 had a right aortic arch, a closing right-sided ductus arteriosus, stenosis (“coarctation”) of the proximal right pulmonary artery at the ductal insertion site, and a vascular ring that compressed the morphologically left mainstem bronchus that was positionally right-sided. The vascular ring about the right-sided bronchus was formed by the right aortic arch superiorly and posteriorly, the right pulmonary artery branch anteriorly, and the closing right ductus arteriosus superiorly and anteriorly where it inserted into the proximal right pulmonary artery. The proximal left pulmonary artery also played a role because it lay immediately to the left of the right-sided bronchus that was compressed by its neighboring structures ( Fig. 14.21 ). When Dr. Louise Calder and I studied this 42-day-old boy in 1970, we thought that ligation and division of the right-sided ductus arteriosus would have decompressed the entrapped right-sided left mainstem bronchus.

Anatomic Features

The atrial septum was obstructive in both cases, despite the fact that the atrial septum in visceroatrial situs inversus is designed to open from left to right, that is, from the left-sided RA into the right-sided LA.

An interesting mechanism for decompressing the left-sided RA was found in Case 17 . There was a large coronary sinus septal defect between the coronary sinus and the right-sided LA. Thus, in addition to shunting left-to-right through the restrictive atrial septum, systemic venous blood from the left-sided RA could flow retrogradely through the coronary sinus ostium and then through the large coronary sinus septal defect into the right-sided LA and then into the right-sided RV.

The other patient, Case 65, had an aneurysm of septum primum that bulged into the right-sided LA and down into the right-sided tricuspid orifice, producing supratricuspid stenosis .

• The left-sided LV was markedly hypoplastic in both patients.

• Pulmonary outflow tract atresia was present in both cases.

• A closing ductus arteriosus was found in both patients.

• Isolated levocardia was considered to be present in both patients; that is, the heart was left-sided but the viscera and atria were in situs inversus.

• A right aortic arch was present in both patients, which is usual for visceroatrial situs inversus.

The mother of Case 65 with TGA {I,D,A} and mitral atresia (left-sided) was addicted to cocaine. Her son, who died at 16 days of age (Case 65), had signs of methadone withdrawal. The presence of maternal cocaine addiction is recorded here, in case cocaine may prove relevant to the cause and/or morphogenesis of mitral atresia with TGA {I,D,A} or a similar anomaly.

Heterotaxy With Polysplenia

Case 82 was a 7½-month-old boy who died in 1951. He had DORV {A(S),D,D} with mitral atresia and polysplenia. The A (S) part of his segmental anatomy means he had situs ambiguus (A) of the viscera and we thought the atrial situs very probably was solitus (S). DORV {A(S),D,D} indicates that he definitely had visceral situs ambiguus and that we thought that he probably had atrial situs solitus. The S symbol is placed in parentheses to indicate some doubt: we thought the atria very probably were in situs solitus, but we could not be entirely certain.

The patient had two spleens; in that sense polysplenia was present. The liver was bilaterally symmetrical. The lungs were bilaterally trilobed (more like the asplenia syndrome than the polysplenia syndrome). There was TAPVC to the RSVC-atrial junction. A LSVC also connected with the poorly septated atrial segment. The left innominate vein was absent. The IVC was midline, not interrupted, and connected with the atrial segment. There was an ostium primum defect. To the left of the atrial septal remnant, there was no patent AV valve or valvar component. Mitral atresia was considered to be present. No vestige of the morphologically LV was found. The patient had a single morphologically RV. The RV was of the right-handed or solitus type in terms of its chirality. A ventricular D-loop with a single RV and an absent LV was our diagnosis. The tricuspid valve lay to the right of the atrial septal remnant and opened into the RV. The segmental anatomy, as earlier, was DORV {A(S),D,D}. A common AV canal was thought to be present because of the ostium primum defect (the most common incomplete form of AV septal defect).

As is usual with single RV, because of absence of the LV, no ventricular septum or ventricular septal remnant could be identified. Consequently, we could not say whether this was a complete form or a partial form of common AV canal. In other words, was a VSD of the AV canal type present or not? When a ventricular septal remnant cannot be identified, and the morphologically LV is absent, this question cannot be answered (to the best of our present understanding).

A bilateral conus was present (subaortic and subpulmonary). Pulmonary outflow tract stenosis (infundibular and valvar) was present, with no aortic outflow tract obstruction. The pulmonary valve was bicuspid (bicommissural). The ductus arteriosus was absent. There was a right aortic arch. The small and large bowel had a common mesentery.

In summary, the mitral valve and the morphologically LV were both absent; in this sense, mitral atresia was diagnosed. It is noteworthy that we still do not know how to distinguish between mitral atresia (mitral valve present, but with no orifice) and mitral valve absence . At present, the diagnosis of mitral atresia appears to include both potential anomalies. The same applies to the tricuspid valve: our uncertain ability to distinguish between tricuspid valvar atresia (which definitely occurs with the Ebstein anomaly, that is, with imperforate Ebstein malformation) and tricuspid valvar absence. Typical tricuspid and mitral atresia, that is, so-called muscular tricuspid atresia and muscular mitral atresia, may well represent tricuspid valvar absence and mitral valvar absence, not a lack of an opening in either valve. Imperforate Ebstein anomaly may be the only real form of tricuspid valvar atresia because the tricuspid leaflets are definitely present, but with no valvar opening (imperforate). In other words, “typical” tricuspid atresia and typical mitral atresia (the “muscular” forms) may well be erroneous concepts, accurately speaking. Perhaps membranous atresia and muscular atresia (both mitral and tricuspid) may distinguish between AV valvar atresia, as opposed to AV valvar absence, respectively. Much remains to be learned.

Summary and Discussion: Pathologic Anatomy of Mitral Atresia

Table 14.2 and diagrammatic Figs. 14.1 to 14.21 summarize the salient features of the pathologic anatomy of mitral atresia that were found in 177 postmortem cases. The findings were classified into six anatomic groups, and into 18 anatomic types of mitral atresia. The median ages at death (or cardiac transplantation) in the 18 anatomic types of mitral atresia are summarized in Table 14.10 . Table 14.10 is also a convenient summary of the salient features of the 18 different anatomic types of mitral atresia.

The mean age at death or cardiac transplantation (only 1 patient had a transplant) for the series as a whole (all 18 anatomic types) was 231 • 24 ± 431 • 42 days, or 7.7 months ± 1.18 years, ranging from 1 hour to 3.72 years, with a mean of 28 days.

The classic forms of mitral atresia had normal segmental anatomy {S,D,S} in 118 of these 177 cases (66.67%, types 1 to 7 inclusive). In type 6 with a single LV and absence of the right ventricular sinus, that is, the right ventricular inflow tract, or a large LV and a very small RV, only 2 of these 7 cases had normal {S,D,S} segmental anatomy (see Table 14.2 and Fig. 14.1 ). Patients with an intact ventricular septum and aortic valvar atresia predominated (45%). Infrequently, patients with mitral atresia and an intact ventricular septum had a patent but hypoplastic aortic valve (1%). But when the aortic valve was patent, it was more usual for a VSD to coexist (15%). Nonetheless, despite the presence of a VSD, aortic valvar atresia was found in 3% of these patients with mitral atresia and normal segmental anatomy (see Table 14.2 and Fig. 14.1 ).

Rare and noteworthy cases were encountered in group 1 with mitral atresia and normal {S,D,S} segmental anatomy:

• 1.

  A large morphologically LV and a small or absent morphologically right ventricular sinus were found in 2 of these 118 patients with mitral atresia {S,D,S} (1 • 69%); these 2 cases constitute 1 • 13% of all patients with mitral atresia in this study (see Table 14.2 and Fig. 14.1 ). Single LV with absence of the right ventricular sinus and an infundibular outflow chamber with mitral atresia {S,D,S} is the Quero heart.

• 2.

  Truncus arteriosus occurred in a rare patient with mitral atresia {S,D,S} (0.56%; see Table 14.2 and Fig. 14.1 ).

• 3.

  One patient was a functionally acardiac thoracopagus conjoined twin with atresia of all four cardiac valves.

• 4.

  Mitral atresia occurred in hearts with normal segmental anatomy, {S,D,S}, in 118 of 177 patients (66.67%) (see Table 14.2 and Fig. 14.1 ), as earlier. Mitral atresia also occurred in hearts with abnormal segmental anatomy in 59 of 177 patients (33.33%). To simplify and make this easier to remember, in mitral atresia the segmental anatomy was normal in two-thirds of cases and was abnormal in one-third of cases.

In the one-third of patients with mitral atresia and abnormal segmental anatomy (n = 59), the infundibuloarterial (conotruncal) anomalies were DORV in 47 of 177 patients (26.55%), TGA in 11 of 177 patients (6.21%), and DOIOC with absence of the right ventricular sinus in 1 of 177 patients (0.56%).

When examining the statistics in Table 14.2 and Fig. 14.1 , remember that type 6 mitral atresia contains only 2 patients with normal {S,D,S} segmental anatomy. Type 6 also contains TGA {S,D,D/A} in 2, TGA {S,L,D} in 1, DORV {S,D,A} in 1, and DOIOC {S,L,L} in 1. It is necessary to remember this for all the members to add up correctly.

Most textbook accounts of mitral atresia assume that the segmental anatomy is normal. This assumption does not apply accurately to about one-third of patients with mitral atresia (see Table 14.2 and Fig. 14.1 ).

Absence of the morphologically LA was another exceedingly rare finding. To my knowledge, absence of the LA is a previously unknown anomaly. This patient was Case 135, autopsy number A76-8, a 10-day-old boy with mitral atresia, anatomic type 8, that is, mitral atresia with VSD and DORV {S,D,D/“S”} (see Table 14.2 and Fig. 14.1 ).

DORV in association with mitral atresia often had a unilateral conus, that is, a subpulmonary conus with aortic-tricuspid fibrous continuity (made possible by resorption of the subaortic conal free wall) or a subaortic conus with pulmonary valve-tricuspid valve direct fibrous continuity (made possible by resorption of the subpulmonary conal free wall). DORV with a unilateral conus—subpulmonary only or subaortic only—is very different from the bilateral conus (with subaortic and subpulmonary infundibular musculature) that is usual when DORV is associated with two well-developed ventricles.

Although not widely recognized, DORV with a unilateral conus (subpulmonary or subaortic) occurring in association with HLHS is a specific anatomic type of DORV. DORV with a unilateral conus (subpulmonary only or subaortic only) is as specific an anatomic type of DORV as is the Taussig-Bing malformation with two well-developed ventricles and ^, a bilateral (subpulmonary and subaortic) conus. DORV with a unilateral conus merits wider recognition because of its importance in the understanding of mitral atresia with DORV.

What the Literature Teaches Concerning the Pathologic Anatomy of Mitral Atresia

If one consults many of the excellent contemporary textbooks that deal with pediatric cardiology and/or congenital heart surgery, there is a remarkable dearth of detailed anatomic information concerning mitral atresia, absent left AV connection, imperforate left AV connection, or atresia of the left AV connection, these being synonyms that some of our colleagues currently favor. However, the journal literature remains a rich source of information. Pioneering studies include those of Edwards and Rodgers in 1947 and of Edwards and DuShane in 1950 in which these authors also describe premature closure of the foramen ovale and a decompressing levo-atrio-cardinal vein between the LA and the left innominate vein. In 1955, Friedman, Murphy, and Ash emphasized the hypoplastic nonfunctioning left heart associated with mitral atresia. Three years later, in 1958, Noonan and Nadas focused on the same aspect and named it HLHS. In 1960, Watson, Rowe, Conen, and Duckworth reported 11 cases of mitral atresia in which the aortic valve was of normal size, not hypoplastic. In 1962, Lucas, Lester, Lillehei, and Edwards again described the combination of mitral atresia with a levoatrial cardinal vein; note the updated spelling of levoatrial.

Classification

In 1965, Dr. Jesse Edwards and colleagues attempted to classify mitral atresia, based on a study of 32 postmortem cases. Eliot, Shone, Kanjuh, Ruttenberg, Carey, and Edwards proposed the following anatomic classification of mitral atresia, using three variables to classify mitral atresia:

• 1.

  the presence or absence of TGA;

• 2.

  the presence or absence of hypoplasia of left-sided cardiac structures; and

• 3.

  the presence or absence of a VSD.

How did they do? As well as was possible at that time, I thought. Their group I was clear: 24 of 32 cases (75%) had normally related great arteries; 14 of these patients (43.75%) had aortic valvular atresia with marked left ventricular hypoplasia. Of these 14 with aortic valvular atresia, 13 (92.86%) had an intact ventricular septum and only 1 had a VSD. The aortic valve was patent but hypoplastic in 10 patients (31.25%). In group II, with “transposition” of the great arteries (TGA in 8 of 32 patients [25%] and with “common” ventricle in 7 of these 8 patients [87.5%]), the diagnostic analysis becomes much less clear, for now very understandable reasons.

In 1965, TGA was still being used—by “everybody”—in Abbott’s sense, meaning any abnormal relationship between the great arteries themselves, and/or between the great arteries and the underlying ventricles, ventricular septum, and AV valves. It was not until the latter half of the 1960s and the early 1970s that it was appreciated that for diagnostic clarity, it is essential that the various VA alignments—normal and abnormal— must be defined with literal accuracy. For example, transposition of the great arteries must mean that both great arteries are “placed across” ( trans = across, and ponere = to place, Latin) the ventricular septum; hence, each arises above the morphologically inappropriate ventricle: aorta above the RV and pulmonary artery above the LV. ^, In other words, transposition must be used with literal anatomic accuracy. It does not matter if the aortic valve is anterior to the pulmonary valve, beside the pulmonary valve, or posterior to the pulmonary valve. ^, What matters clinically, hemodynamically, and surgically is which ventricle each great artery arises above, not the anteroposterior relationship between the great arteries. We had introduced the concepts of concordance and discordance of the AV alignments in 1964. Dr. John Kirklin extended the concept of concordance and discordance to the VA alignments in 1973 concerning ACM of the great arteries. In many subsequent publications, Dr. Robert H. Anderson of London, England and his colleagues did much to popularize the use of AV and VA concordance or discordance in the description and classification of congenital heart disease. Hence, by the late 1960s and early 1970’s, TGA was redefined, with literal anatomic accuracy, as VA discordance, meaning that the RV ejected (inappropriately or discordantly) into the ascending aorta, and the LV ejected inappropriately into the MPA. Similarly, DORV, double-outlet LV, and ACM of the great arteries quickly came to be used with literal anatomic accuracy. Before this time, any and all VA malalignments were regarded as some type of TGA, with this term then being used in a very broad and imprecise way. So, what did Edwards and his colleagues mean by “transposition” of the great arteries in 1965? Because of the broad general usage of transposition then current, it is now difficult to be certain what the precise VA alignments were in this pioneering paper. For example, almost certainly some of these “transpositions” would today be called DORV. But as I said at the outset, these authors did as well as could be done in 1965. Whether we realize it or not, we are all “prisoners” of our time.

In 1965, nonmorphologic terminology was still being used to designate the various cardiac chambers: “venous” atrium and arterial “atrium” and “venous” ventricle and “arterial” ventricles. These hemodynamic terms (“venous” and “arterial”) lead to the questions of, anatomically, which is the “arterial” ventricle? Is it the morphologically LV, or is it the morphologically RV? The answers to these questions can be very unclear, particularly when “transposition” of the great arteries—used in the old way —coexisted. In other words, one of Dr. Maurice Lev’s most important papers, published in 1954, was concerned with how to diagnose the morphologic anatomic identity of the various cardiac chambers, no matter what the spatial location of the cardiac chambers may be, by examining the morphologic anatomic features of the septal surfaces of the atria and ventricles. This was an essential key to the understanding and accurate anatomic diagnosis of complex congenital heart disease.

Ten years later, in 1964, Van Praagh, Ongley, and Swan used a modification of Lev’s morphologic method to understand the pathologic anatomy of what was then called “single” or “common” ventricles. Van Praagh et al used not only the morphologic anatomic features of the ventricular septal surfaces, as Lev had done, but also extended the morphologic anatomic analysis to include the ventricular free walls . Anatomic and embryologic study in humans had convinced Van Praagh et al ^, that the morphologic features of the ventricular free walls were every bit as specific, different, and diagnostically informative as the morphologic features of the ventricular septal surfaces. Because the morphologic anatomy of the ventricular septal surfaces can be altered or absent in single ventricle, the additional use of the morphologic anatomic features of the ventricular free walls made it possible to understand the pathologic anatomy of the single ventricle. ^,

What did Eliot et al mean in 1965 by common ventricle ? The premorphologic meaning of single or common ventricle was that both AV valves or a common AV valve open entirely or predominantly into one ventricular chamber. The designation common ventricle, that was used by Edwards et al in this first attempted classification of mitral atresia ( Table 14.11 ), suggests that the ventricles may be in common, or undivided by a ventricular septum. The systemic and pulmonary venous blood streams certainly appeared to be in common, or undivided. Hence, the question became morphologically or anatomically, what is a single or common ventricle? This is the question that Van Praagh et al had tried to solve in 1964, presumably when the study of Eliot et al was being written.

So, 1965 was the very early days. To summarize, the problems that Eliot et al faced unknowingly in 1965 were:

• 1.

  TGA was still being used in a very broad and anatomically imprecise way ; hence, the meaning of “transposition” of the great arteries in their classification (see Table 14.11 ) is imprecise.

• 2.

  The initial morphologic method of cardiac chamber identification, pioneered by Dr. Maurice Lev in 1954, still was not widely understood. It should be added that Dr. Lev himself thought his morphologic method could not be used in single or common ventricle anomalies because he thought that the ventricular septal surface morphologies were either abnormal or absent in these malformations.

• 3.

  The extended or full morphologic method using both the septal surface and the free-wall morphologies, pioneered by Van Praagh et al, ^, had revealed that there are two anatomic types of single ventricle: (1) single (unpaired) morphologically LV caused by absence of the morphologically right ventricular sinus in about 75% of cases and (2) single (unpaired) RV, caused by absence of the LV, in about 25% of cases. However, in 1965, Eliot et al could not be expected to understand this, because it was still in the early days—hence, the common ventricle of their classification (see Table 14.11 ).

• 4.

  The year 1965 was also too early for the adoption of the segmental approach to the diagnostic understanding of congenital heart disease pioneered by Van Praagh et al.

This diagnostic method, based on both morphologic anatomy and embryology, was used by Van Praagh et al since 1964, but was not formally named and presented until 1972. The segmental approach was developed and tested repeatedly both clinically and pathologically from 1960 to 1972. We were trying to develop a diagnostic method that always worked, no matter how complex the congenital heart disease. No “exceptions” or other excuses were accepted. Only when we could not find anything wrong with it did we present it.

Thus, by 1965, the necessary understanding and methods for the accurate anatomic diagnosis of large series of complex cases of congenital heart disease had not been completely worked out. This is why I said at the outset that these outstanding investigators did the best that they could do in 1965 (see Table 14.11 ).

In 1968, Summerell et al expanded the classification of Eliot et al by reporting that it is possible for mitral atresia with normally related great arteries and left ventricular hypoplasia to have a normal-sized aortic valve, thereby confirming the findings of Watson et al in 1960.

In 1969, Navarro-Lopez et al described a patient with mitral atresia who had an occlusive left atrial thrombus and who survived for 11 years. We have not seen a patient with a left atrial thrombus. The oldest patient found in their literature review was 22 years of age. Of the approximately 160 cases of mitral atresia studied at autopsy, they also found that 10 had lived for more than 1 year (6.25%). Their literature review found that only 25% of patients had normally related great arteries, whereas 75% had “transposition” of the great arteries; these findings are the opposite of ours (see Table 14.2 and Fig. 14.1 ) and the opposite of those of Eliot and Edwards et al, perhaps explaining the relative longevity of the patients in the literature review of Navarro-Lopez et al.

In 1970, Dr. Manuel Quero from Madrid, Spain reported a patient with a Holmes heart {S,D,S} with single LV and an infundibular outlet chamber who had mitral atresia, that is, mitral atresia with a large LV, an absent right ventricular sinus, and normally related great arteries. This type of mitral atresia was thought at the time to be unique; hence, we called this anomaly the Quero heart, in his honor. In 1972, Quero did it again. He published the case of a 3-month-old girl with TGA {S,L,D} with right-sided mitral atresia, a large right-sided morphologically LV, and a left-sided infundibular outlet chamber. In other words, a single LV with an infundibular outlet chamber, but this time with right-sided mitral atresia, a ventricular L-loop, and D-TGA. Quero demonstrated that mitral atresia could occur with a single LV both in a ventricular D-loop (1970) and a ventricular L-loop (1972). It should be recalled that the old premorphologic definition of single (or common) ventricle was as follows: Single or common ventricle is present if both AV valves or a common AV valve open entirely or predominantly into one ventricular chamber. In other words, the old definition of single ventricle excluded mitral or tricuspid atresia. Dr. Manuel Quero was proving that single LV and mitral atresia could indeed coexist, suggesting that tricuspid atresia and mitral atresia should not be excluded from the category of single LV.

Edwards et al had described and classified both tricuspid and mitral atresia, but nobody yet understood double-inlet single ventricle. In 1974, Dr. Glen Rosenquist published 3 cases of mitral atresia with a normal-sized LV, a straddling tricuspid valve, and a small RV; 2 of these patients had TGA. Also in 1974, Cabrera et al published the case of a 14-day-old boy with TGA {S,D,D} with left-sided mitral atresia, single LV with infundibular outlet chamber, and coarctation of the aorta. This case was similar to Quero’s 1972 report, except that the patient of Cabrera et al had a ventricular D-loop (not a ventricular L-loop). The case of Cabrera et al was thought to be the first published case of its type. In 1974, Bjørnstad and Michalsen published a case of mitral atresia with aortic atresia and intact ventricular septum in siblings, suggesting the possibility of a heritable genetic etiology in this anatomic type of mitral atresia with HLHS. In 1976, Moreno, Quero, and Perez Diaz reported 18 cases of mitral atresia with a normal aortic valve, confirming earlier reports. ^, Also in 1976, Friedman et al reported a rare hemodynamic cause for failure of a Blalock-Taussig anastomosis to relieve inadequate pulmonary blood flow: mitral atresia with premature closure of the foramen ovale, resulting in pulmonary venous and pulmonary arterial hypertension. Creation of an ASD helped. In 1979, Ostermeyer et al reported the case of a 6 8/12-year-old girl with mitral atresia, VSD, normal-sized ventricles, and D-TGA. Rarely, mitral atresia can be associated with two well-developed ventricles— not with hypoplastic LV syndrome—if a good-sized VSD coexists. In 1980, Mickell, Mathews, Park, Lenox, Fricker, Neches, and Zuberbuhler reported 40 cases of left AV valvar atresia. They understood that 31 of these patients had mitral atresia (77.5%) and 9 had left-sided tricuspid atresia (22.5%). In other words, they were aware that the identity of the left-sided AV valve is ventricular-loop dependent: mitral, when a ventricular D-loop is present and tricuspid, when a ventricular L-loop coexists. The study of Mickell et al was focused primarily on clinical management. In 1981, Thiene, Daliento, Frescura, DeTommasi, Macartney, and Anderson reported an anatomic study and classification of 62 postmortem cases of mitral atresia, that they preferred to call atresia of the left AV orifice. When Thiene et al used the conventional term mitral atresia, they put mitral between quotation marks: “mitral” atresia. However, they used the conventional diagnosis tricuspid atresia, unadorned with quotation marks. This study merits careful consideration because the first author and the senior author are both world-class congenital heart pathologists. By 1981, Dr. Gaetano Thiene and Dr. Robert H. Anderson and their colleagues had learned many significant lessons, including the importance of morphologic anatomy for cardiac chamber identification, and the utility of the segmental approach for diagnostic data analysis. Consequently their study and classification represent an advance over earlier pioneering investigations.

AV and VA “Connections”

Thiene et al write about AV and VA connections because they do not regard the AV junction or canal and the infundibulum or conus arteriosus as separate connecting cardiac segments between the atria and the ventricles and between the ventricles and the great arteries, respectively, as we do.

In the interests of anatomic accuracy, as noted earlier, we prefer the concepts of AV alignments and VA alignments (not connections). Why, exactly? Because the atria and the ventricles normally do not connect tissue-to-tissue, because of the interposition of the AV canal and valves. Similarly, the ventricles and the great arteries normally do not connect tissue-to-tissue, because of the interposition of the conus arteriosus or infundibulum.

Their nonconnection explains why the main cardiac segments can be aligned in so many different ways, accounting for much of the complexity of congenital heart disease. For example, if the morphologically RA really were connected with the morphologically RV muscle-to-muscle, and if the morphologically LA really were connected with the morphologically LV muscle-to-muscle, then discordant L-loop ventricles would be developmentally impossible. It is because the atria are not connected muscle-to-muscle with the ventricles that the AV alignments can be so variable: concordant, discordant, double-inlet, right atrial outlet atresia, left atrial outlet atresia, etc. Similarly, if the RV really were connected to the MPA, and if the LV really were connected with the aorta, conotruncal anomalies such as TGA, DORV, double-outlet left ventricle (DOLV), and ACM would be developmentally impossible. It is because the ventricles and the great arteries are not connected tissue-to-tissue that virtually any conceivable VA alignment can and does exist. Note that our concern has nothing to do with terminologic preference; rather, it has everything to do with anatomic and embryologic accuracy.

Because so many of our friends and colleagues like to talk about AV and VA connections, one may well wonder why we cannot somehow make connections anatomically correct. I would say that of course we can. All it takes is a little grammatical “subterfuge,” as follows: It is anatomically accurate to say that the atria are connected to the ventricles by the AV canal or junction. Similarly, the ventricles and the great arteries are connected by the infundibulum or conus arteriosus. In these two sentences, the verb are connected is in the passive voice, the meaning of which is anatomically correct. By contrast, in the sentence the atria connect with the ventricles, the verb connect is in the active voice, the meaning of which is anatomically incorrect. So, use of the passive voice solves the problem; only the active voice is anatomically wrong. But if grammar was never your strong suite, don’t worry. What really matters is meaning . Grammar is the slave of meaning. When some of our colleagues say “the atria connect with the ventricles” what they mean is “ the atria are connected with the ventricles ” in this or that way.