Ventricular Septal Defects

In 1989, Dr. John Kirklin asked me how I thought ventricular septal defects (VSDs) should be described, named, and classified. Soto, Ceballos, and Kirklin were just about to publish their best thoughts on this important topic, and Dr. Kirklin asked me and my colleagues to do the same.

Components of the Ventricular Septum

The four main anatomic components that make up the normal ventricular septum are as follows ( Fig. 16.1 ) :

  • 1.

    the septum of the atrioventricular (AV) canal (component 1);

  • 2.

    the muscular ventricular septum, or the ventricular sinus septum (component 2);

  • 3.

    the septal band, or the proximal conal septum (component 3); and

  • 4.

    the parietal band, or the distal conal septum (component 4).

Fig. 16.1

The Four Main Anatomic and Developmental Components of the Interventricular Septum of the Normal Heart.

(A) Morphologically right ventricle (RV). (B) Morphologically left ventricle (LV). Component 1, septum of the atrioventricular canal. Component 2, muscular ventricular septum, or sinus septum. Component 3 , septal band in A, or proximal infundibular (conal) septum; in B, component 3 is the proximal conal septum that forms the superior, nontrabeculated portion of the left ventricular septal surface. Component 4, the distal or subarterial infundibular septum. Component 1 is also known as the atrioventricular (AV) septum. Components 1 and 2 form the right ventricular inflow tract and the left ventricular inflow tract. Components 3 and 4 form the right ventricular outflow tract and the left ventricular outflow tract. The moderator band joins the septal band (component 3) to the anterior papillary muscle of the RV in A. The moderator band is also known as the septomarginal trabeculation because it runs from the right ventricular septal surface (septo) to the acute margin of the RV (marginalis). Some of our colleagues use the term septomarginal trabeculation to mean both the septal band and the moderator band. The lower portion of the distal conal septum (component 4) is also known as the crista suprventricularis because it forms a supraventricular crest above the right ventricular sinus (component 2).

Reproduced with permission from Van Praagh R, Geva T, Kreutzer J. Ventricular septal defects: how shall we describe, name and classify them? J Am Coll Cardiol. 1989;14:1298.

Anatomic Types of Ventricular Septal Defect

VSD of the AV Canal Type.

The septum of the AV canal (see Fig. 16.1 , component 1) is completely absent in the complete form of common AV canal.

In the incomplete form of common AV canal, the septum of the AV canal is absent above the leaflets of the AV valve(s) but present below the leaflets of the common AV valve. In other words, typically there is no VSD of the AV canal type in the incomplete form of common AV canal. Also typically, there is a cleft in the anterior leaflet of the mitral valve.

The incomplete form of common AV canal is also known as an ostium primum defect with a cleft anterior mitral leaflet. Accurately speaking, an ostium primum defect is an incomplete AV septal defect, not an atrial septal defect (ASD). The posterosuperior margin of an ostium primum defect is the anteroinferior margin of the atrial septum. Ostium primum defects have often been called, inaccurately, ostium primum atrial septal defects because the associated shunt (left to right, or right to left) is above the leaflets of the AV valve(s)—like an ASD. However, it is helpful to know that many different types of shunt occur above the leaflets of the AV valve(s). Ostium primum defects and ASDs are just two of many different anatomic types of abnormal communication that can occur above the AV valve(s).

Straddling tricuspid valve also has a VSD of the AV canal type. Typically, the right ventricular sinus (body or inflow tract) is underdeveloped compared with the left ventricular sinus. Consequently, the muscular ventricular septum (see Fig. 16.1 , component 2) is located well to the right of the normally located atrial septum. The muscular ventricular septum lies beneath approximately the middle of the normally located tricuspid orifice. This ventriculoatrial septal malalignment results in a straddling tricuspid valve with biventricular insertions of the tricuspid tensor apparatus, a VSD of the AV canal type between components 1 and 2 (see Fig. 16.1 ), and often with no septal defect above the level of the AV valves.

A defect in the septum of the AV canal (see Fig. 16.1 , component 1) is also known as an AV septal defect. AV septal defects occur both with and without common AV canal. An AV septal defect is an integral part of common AV canal, both complete and partial forms.

But an AV septal defect also occurs without common AV canal, as in a Gerbode defect. There is a left ventricle (LV)–to–right atrium (RA) shunt through a defect in the AV portion of the membranous ventricular septum, predominantly above the tricuspid valve. Megarity et al published a well-documented case of double-outlet right ventricle (DORV) with no VSD and with an LV-to-RA shunt. This patient did not have a common AV canal.

We avoid the term AV septal defect because of its confused usage. Some of our colleagues use this designation when they mean common AV canal. For clarity, we prefer the terms common AV canal and LV-to-RA shunt, because these terms make their different meanings clear.

We also avoid the designation inlet septal defect because this term is not specific. This term may apply to component 1, or to component 2, or to the component 1-2 junction (see Fig. 16.1 ). We prefer anatomically specific diagnoses in the interests of accuracy and clarity: defect in the septum of the AV canal, defect in the muscular ventricular septum, or defect at the junction of these two septal components. The latter diagnoses are anatomically accurate and their meanings are clear.

Muscular VSDs.

Muscular VSDs are openings in the muscular ventricular septum, also known as the ventricular sinus septum. Muscular VSDs, which is what these defects have long been called, involve component 2 (see Fig. 16.1 ).

Mid-muscular VSDs are often located at the junction of components 2 and 3 (see Fig. 16.1 ). From the right ventricular aspect, such apertures are seen slightly above or slightly below the septal band (see Fig. 16.1A , component 3). From the left ventricular aspect, mid-muscular VSDs typically are found at the junction of the smooth or nontrabeculated portion of the ventricular septum superiorly (see Fig. 16.1B , component 3) and the finely trabeculated part of the ventricular septum inferiorly (see Fig. 16.1B , component 2). These mid-muscular VSDs are sometimes called trabecular defects. Why? Because they are close to the septal band, as noted earlier. Some of our colleagues call the septal band the trabecular septomarginalis or the septomarginal trabeculation. The term trabecular septomarginalis was introduced in 1911 by Tandler for the structure that we now call the moderator band. Trabecula septomarginalis (Latin) means “little beam” ( trabecular ) that runs from the septum (septo) to the acute margin of the RV (marginalis). So, the trabecular septomarginalis , or septomarginal trabeculation, means the moderator band, not the septal band. But those who revived Tandler’s term , trabecular septomarginalis, being unfamiliar with Latin, did not understand what Tandler’s term really meant. So, they applied trabecular septomarginalis and its English equivalent septomarginal trabeculation to the septal band and the moderator band.

The foregoing is intended as an explanation, not as a criticism. The history of language is full of similar changes. Indeed, in terminology, the only constant we know is change. But it should be slow change, and necessary change, to minimize confusion and linguistic errors. What do we think is necessary terminologic change?

  • 1.

    If there is no extant term, the introduction of a designation may be necessary.

  • 2.

    If the extant term (or terms) is (are) inaccurate, a new and accurate designation may be necessary.

The foregoing explains why mid-muscular VSDs are called trabecular VSDs by some of our colleagues: trabecular comes from trabecula septomarginalis, or septomarginal trabeculations, that is, from the renaming of the septal band by some of our colleagues (see Fig. 16.1A ).

Muscular VSDs can be found in many different locations within component 2, and at its junctions with components 1 and/or 3 (see Fig. 16.1 ). Mid-muscular VSDs do not involve component 3, as seen from the right ventricular aspect (see Fig. 16.1A ). Instead, mid-muscular VSDs lie behind and to the left of the septal band.

Ironically, the muscular ventricular septum (see Fig. 16.1A–B , component 2) is trabeculated. The right ventricular septal surface is coarsely trabeculated (see Fig. 16.1A , component 2), whereas the left ventricular surface (see Fig. 16.1B , component 2) is finely trabeculated. Thus, although component 3 is not trabeculated per se (see Fig. 16.1A–B ), component 2 is coarsely trabeculated on the right ventricular side and finely trabeculated on the left ventricular side (see Fig. 16.1A–B ).

Infundibuloventricular Defects.

Infundibuloventricular defects lie between the infundibular (or conal) septum above (see Fig. 16.1A , component 4, and B) and the ventricular septum below (see Fig. 16.1A–B , components 1, 2, and 3). Infundibuloventricular defects are also known as conoventricular VSDs. The infundibular or conal septum (see Fig. 16.1A–B component 4) is the “lid” that normally fits on top of and seals the ventricular septal complex :

  • 1.

    the septum of the AV canal (see Fig. 16.1A–B , component 1), the top of which in the normally formed heart is called the membranous septum;

  • 2.

    the muscular ventricular sinus septum (see Fig. 16.1A–B component 2); and

  • 3.

    the septal band (see Fig. 16.1A , component 3), and the proximal infundibular or outflow tract septum (see Fig. 16.1B , component 3).

Thus, the ventricular septal complex is a tetralogy: components 1 to 4, inclusive (see Fig. 16.1A–B ). To achieve a normal heart, each of these four septal components must be normally formed, normally aligned, and normally connected.

Infundibular Septal Defects.

Infundibular septal defects result from an anomaly of, or within, component 4 (see Fig. 16.1A–B ). These malformations are also known as conal septal defects.

Relative Frequencies of the Four Anatomic Types of Ventricular Septal Defect

The four anatomic types of VSD are:

  • 1.

    AV canal type of VSD (Dr. Jesse Edwards’ term), also known as inlet septal defect (term of Soto et al);

  • 2.

    muscular VSDs, also known as inlet septal defects or trabecular defects (terms of Soto et al );

  • 3.

    infundibuloventricular (or conoventricular) VSDs (our terms), also known as infundibuloventricular (or conoventricular) VSD, (terms of Soto et al ); and

  • 4.

    infundibular septal (or conal septal) VSDs (our terms), also known as right ventricular outlet VSDs (term of Soto et al ).

The relative frequencies of these four anatomic types of VSD found in 76 patients of Soto, Ceballos, and Kirklin are presented in Table 16.1 .

Table 16.1

Frequencies of Four Anatomic Types of VSD (n = 76), Soto et al

Anatomic Type of VSD No. of Cases Percent of Series
Atrioventricular Canal Type 7 9
Muscular 23 30
Infundibuloventricular 25 33
Infundibular septal 21 28

Why is understanding the four main anatomic and developmental components that normally make up the ventricular septum (see Fig. 16.1 ) so important? Because VSDs occur within, and/or between, these four main ventricular septal components (see Fig. 16.1 ). This anatomic and developmental understanding is the key to understanding VSDs.

What is a membranous VSD? Membranous VSDs typically occur toward the top of the membranous septum of the AV canal (see Fig. 16.1A , component 1), just beneath a well-developed and normally located distal conal septum or parietal band (see Fig. 16.1A , component 4). For a VSD to be regarded as membranous, only the membranous septum should be abnormal. Everything else should appear to be normal. In particular, the distal conal septum (parietal band) should appear to be normal (see Fig. 16.1A , component 4).

Is the distal infundibular septum (see Fig. 16.1A , component 4) a little hypoplastic or mildly abnormally located? If it is, we do not make the diagnosis of membranous VSD, because such VSDs (with an abnormal parietal band, that is, Fig. 16.1A , component 4) are more than membranous. In such cases, we make the diagnosis of an infundibuloventricular (or conoventricular) VSD. The diagnosis of infundibuloventricular VSD does not imply that the distal conal septum (component 4) is normally formed, whereas the diagnosis of membranous VSD does have this connotation. Membranous VSD implies that the distal conal septum (component 4) is normally formed and that the defect is in the membranous septum only. Frequently, this is not the case. Infundibuloventricular VSDs often have hypoplastic and/or malaligned distal infundibular septa (components 4).

Malalignment infundibuloventricular VSDs often are associated with infundibuloarterial (conotruncal) malformations such as tetralogy of Fallot (TOF), truncus arteriosus, transposition of the great arteries (TGA), DORV, double-outlet left ventricle (DOLV), and anatomically corrected malposition of the great arteries (MPA).

A VSD may or may not be confluent with one or both semilunar valves and/or with one or both AV valves. For example, when a VSD is confluent with the tricuspid valve, such a VSD may be described as paratricuspid ( para = “beside,” in Greek), or as juxtatricuspid ( juxta = “beside,” in Latin). A VSD can be confluent with the membranous septum. Such VSDs are paramembranous (para = beside, or confluent with, in Greek ).

Some of our colleagues describe such VSDs as perimembranous . Unfortunately, this is a semantic error. Peri means “around” in Greek. An infundibuloventricular VSD can be confluent with the membranous septum (paramembranous). But such a VSD never surrounds the membranous septum. If the membranous septum were surrounded by a VSD, the membranous septum would be floating freely in space, attached to nothing. No such VSD has ever been documented.

By analogy, the para thyroid gland is correctly named, meaning beside the thyroid gland. The designation peri thyroid gland, meaning “around” the thyroid gland, would be absurd, because there is no such gland. Similarly, the designation perimembranous VSD is simply an error in terminology, because there is no such VSD.

Of the 3400 autopsied cases of heart disease in infants and children on which this book is based, 3216 had congenital heart disease (94.59%). The remaining 184 patients had acquired (not congenital) heart disease (5.41%). Of the 3216 patients with congenital heart disease, 1160 had a VSD (36.07%). Of these 1160 patients with a VSD, 128 had multiple VSDs (11.03%).

Table 16.2 needs a little clarification. Membranous VSDs (n = 169) are regarded as a subset of infundibuloventricular VSDs (with a normally developed distal conal septum; see Fig. 16.1A , component 4). Membranous VSDs accounted for 5.25% of the series of congenital heart disease as a whole (n = 3216). The other cases of infundibuloventricular VSD with an abnormally located component 4 (n = 682) composed 21.2% of the series as a whole. Thus, conoventricular or infundibuloventricular VSDs as a whole were found in 26.46% of this series of 3216 cases of congenital heart disease (see Table 16.2 ). VSDs that were hemodynamically disadvantageously restrictive (not big enough) were found in 4 patients (0.12%; see Table 16.2 , footnote).

Table 16.2

Frequencies of the Different Anatomic Types of Ventricular Septal Defect in the Present Study of 3216 Autopsied Cases of Congenital Heart Disease

Anatomic Type of VSD No. of Cases Percent of Series
Atrioventricular canal type 91 2.83
Muscular 203 6.31
Infundibuloventricular 682 (21.21) 851 26.46
Membranous 169 (5.25)
Infundibular septal 15 0.47

VSDs, multiple, 128; 3.98%; VSDs, restrictive, 4, 0.12%.

VSDs of all anatomic types were by far the most common form of congenital heart disease found in this series of 3216 cases of congenital heart disease: 1160 of 3216 = 0.36069, or 36.07%.

Limitations of These Statistics

Even though Table 16.2 is based on a large number of patients with congenital heart disease (3216), these statistics have limits to their accuracy that should be acknowledged:

  • 1.

    These patients are all autopsied cases, which is both good and bad. The good part is that the diagnoses are accurate, that is, with no clinical guesswork. The bad part is that these cases are all fatalities; thus, these statistics are skewed in the direction of badness (greater severity of disease than in the pediatric population of congenital heart disease as a whole).

  • 2.

    These statistics (see Table 16.2 ) are based on the number of patients in which VSD appears in the diagnosis. This means that TOF is omitted because a subaortic, malalignment infundibuloventricular type of VSD is understood to be present in TOF and hence is not stated.

If the conoventricular VSDs of our patients with TOF are included in these VSD statistics (see Table 16.2 ), as they certainly should be, the number of patients with VSD increases from 1160 (see Table 16.2 ) + 428 (TOF) to a new total of 1588, which is 49.38% of the whole series of congenital heart disease patients (n = 3216).

Similarly, if truncus arteriosus communis is included (only 1 of our 111 cases of truncus arteriosus had an intact ventricular septum), then 110 more cases of VSD should be included. This new final total of cases with a VSD equals 1698, of a total number of 3216 patients with congenital heart disease; that is, 52.8% of our patients had a VSD. This is a much higher percentage of patients with VSD than is usually found, because TOF and truncus arteriosus are often omitted. Thus, a VSD was present in more than half of our autopsied congenital heart patients: almost 53%.

Ventricular Septal Defect With and Without Aortic Insufficiency

One of the first questions that Dr. Robert E. Gross, our chief of the Department of Surgery, asked me to try to solve after I joined the staff of Boston Children’s Hospital in late 1965 was: Why do some patients with VSD have aortic insufficiency (AI; aortic regurgitation) whereas others do not? Dr. Gross suggested that I work with Dr. J. Judson McNamara, one of Dr. Gross’s Cardiac surgery residents, who later went on to have a distinguished surgical career in Hawaii.

Nadas et al found that of 756 patients with VSD studied at Boston Children’s Hospital between 1948 and 1962, only 34 developed AI (4.5%). Because clinical signs of AI were not discovered during the first year of life in 18 carefully followed patients (protodiastolic blowing murmur and wide pulse pressure), Nadas et al concluded that AI is an acquired complication of VSD, not a congenital anomaly per se. Similarly, Halloran et al reported 12 patients with VSD and AI; in only one of whom was a diastolic murmur heard within the first year of life, at 6 months of age. In 11 postmortem cases, we found 4 that there are essentially two anatomic types of VSD with AI ( Fig. 16.2 ):

  • 1.

    with an infundibuloventricular (subcristal) VSD in 7 of 11 (64%); and

  • 2.

    with an infundibular septal defect (subpulmonary VSD) in 4 of 11 (36%).

Fig. 16.2

I nfundibuloventricular (or Conoventricular) Type of Ventricular Septal Defects (VSDs) With Aortic Insufficiency (AI) (left and center columns).

The AI is caused by the coexistence of a bicommisural

aortic valve, with absence of the right coronary/noncoronary (RC/NC) commissure. The large combined right coronary leaflet (RCL) plus noncoronary leaflet (NCL) prolapses downward and does not coapt normally with the normal left coronary leaflet (LCL), resulting in AI. The subarterial infundibular (or conal) septum (CS) (also known as the infundibular septum and as the parietal band) may be well formed and approximately normally located (left column, top row). Or the CS may be hypoplastic and abnormally located, deviated anteriorly, superiorly, and leftward (CS, middle column, top row ). Depending on the severity of the subpulmonary infundibular stenosis, there may be right ventricular (RV) outflow tract obstruction as in tetralogy of Fallot: hemodynamically significant subpulmonary infundibular hypoplasia reflected as stenosis or atresia (middle column, top row). The hypoplastic subpulmonary infundibulum may be secondarily hypertrophied. Or there may be no right ventricular outflow tract obstruction, as in the Eisenmenger complex: hemodynamically insignificant subpulmonary infundibular hypoplasia (middle column, top row). An anatomically different type of VSD with AI is depicted in the right column. A conal or infundibular septal defect is present. The infundibular septum has failed to form, completely or in part. In the infundibular septal defect, the VSD is both subpulmonary and subaortic (right column). The infundibular septal type of VSD is doubly committed, subaortic and subpulmonary. In the infundibuloventricular type of defect, the VSD is subaortic only (left and middle columns). With infundibular septal defects, the aortic valve per se is normally formed: it is tricommissural. However, the infundibular septal defect is located completely beneath the RCL of the aortic valve. Consequently, the RCL of the aortic valve is unsupported below, facilitating downward prolapse of the RCL, with or without herniation of the RCL of the aortic valve into the right ventricular outflow tract. This results in AI, with or without a pulmonary outflow tract gradient. The RCL can reduce the size of the subsemilunar defect, or close it altogether. When an infundibuloventricular type of VSD is associated with a hypoplastic and antero-leftwardly malaligned CS (middle column), the VSD is larger than when there is little or no malalignment of the CS (left column). With CS malalignment (middle column), the VSD lies more beneath the RCL of the aortic valve than it does when the CS is normally located (left column). Thus, it is the right coronary leaflet component of a bicuspid aortic valve (RCL + NCL) that tends to prolapse most severely into the left ventricle (LV) and to herniate into the right ventricular outflow tract both with an infundibuloventricular VSD (middle column) and with an infundibular septal type of VSD (right column). Thus, no matter which type of VSD is present, it typically is the right coronary leaflet of the aortic valve that prolapses and herniates most severely. Ao, Ascending aorta; MB, moderator band; MV, mitral valve; PA, main pulmonary artery; RV, morphologically right ventricle; SB, septal band; TV, tricuspid valve; TSM, Trabecula septomarginalis, TSM was Tandler’s term for the moderator band; SMT, septomarginal trabeculation. TSM anglicized by Anderson, who applied SMT to both the septal band and the moderator band.

Reproduced with permission from Van Praagh R, McNamara JJ. Anatomic types of ventricular septal defect with aortic insufficiency. Am Heart J. 1968;75:604-619.

When the VSD was of the infundibuloventricular variety, 3 patients had no infundibular outflow tract stenosis (see Fig. 16.2 , left column; we called this type Ia), and 4 patients had infundibular outflow tract stenosis (see Fig. 16.2 , middle column, type Ib). When the VSD was of the subpulmonary infundibular septal defect type (see Fig. 16.2 , right column, type II) as in 4 patients, there was always pulmonary outflow tract stenosis.

The Essential Facts

When an infundibuloventricular VSD was present (see Fig. 16.2 , left and middle columns ), the aortic valve was often bicommissural (5/7 cases, 71.4%). This means that the aortic valve was often functionally bicuspid, because the number of well-formed aortic commissures equals the number of functional aortic leaflets. The deficient or absent aortic commissure in all 5 cases was the right coronary/noncoronary (RC/NC) commissure. This means that the functionally bicuspid aortic valve had just two functional leaflets: (1) a combined RC plus NC leaflet; and (2) a smaller left coronary (LC) leaflet (see Fig. 16.2 , lower row, left and middle columns ).

During the first year of postnatal life, the combined and poorly supported RC/NC leaflet got stretched (because it lacked RC/NC commissural support). The poorly supported RC/NC leaflet prolapsed downward, into the left ventricular outflow tract. When the large and prolapsing RC/NC leaflet could no longer coapt with the smaller, higher LC leaflet, AI appeared and progressively became more marked.

In the 4 patients with an infundibular (conal) septal type of subpulmonary VSD, the aortic valve was always tricommissural. None had a commissural deficiency or absence (see Fig. 16.2 , right column, lower row ). Instead, these 4 patients had a normally formed RC leaflet that prolapsed inferiorly and herniated partially through the subpulmonary conal septal defect into the right ventricular outflow tract, where it produced pulmonary outflow tract stenosis. Prolapse and herniation of the RC leaflet of the aortic valve also led to lack of aortic leaflet coaptation and AI.

In Fig. 16.2 , left ventricular diagrams (lower row), note that in the left and middle drawings, the RC/NC commissures are absent, whereas in the right LV diagram , all three aortic commissures are well formed. Also note that in Fig. 16.2 diagrams, when a conoventricular VSD is present (left and middle left ventricular diagrams), the missing RC/NC commissure of the aortic valve is located above the middle of the conoventricular VSD. In other words, when a conoventricular VSD is present, the VSD lies approximately beneath half of the RC leaflet, and beneath approximately half of the NC leaflet.

But when a conal septal defect is present (see Fig. 16.2 , right diagrams ) the VSD lies beneath the RC leaflet only. This difference in aortic valve and VSD alignment appears to be important. This is why when a subpulmonary conal septal VSD is present, an RC aortic leaflet hernia into the RV outflow tract occurs (see Fig. 16.2 , right diagrams ). However, when a conoventricular VSD is present, an aortic leaflet hernia does not occur. Instead, if AI appears, the problem usually is due to RC/NC aortic commissural deficiency or absence and stretching of the combined RC + NC leaflets.

Is the foregoing explanation of AI with a VSD always true? No. Why not? We had 1 patient with a conoventricular VSD with pulmonary infundibular stenosis and no aortic commissural deficiency who, nonetheless, had prolapse of the RC/NC aortic leaflets with herniation of the RC aortic leaflet into the right ventricular outflow tract, resulting in a right ventricular outflow tract gradient of 13 mm Hg. How do we explain this exception? In our 4 cases with conoventricular VSD and deviation of the parietal band, none had clinical cyanosis. Instead, they resembled mild, acyanotic TOF. But in 1 patient (Case 7), there was enough anterior, superior, and leftward deviation of the parietal band (component 4) to permit prolapse and herniation of a normally formed RC aortic leaflet into the right ventricular outflow tract.


  • 1.

    When AI develops in a patient with a conoventricular VSD, the cause of the AI usually is an aortic valve anomaly: absence of an aortic valve commissure.

  • 2.

    When AI develops in a patient with a subpulmonary conal septal defect, the cause of the AI usually is an RC aortic leaflet hernia, the aortic valve being normally formed.

  • 3.

    When AI develops in a patient with a conoventricular VSD and a TOF type of infundibular stenosis, if the distal conal septum is deviated sufficiently anteriorly, superiorly, and to the left, a normally formed RC leaflet of the aortic valve occasionally can herniate into the right ventricular outflow tract.

  • 4.

    Careful diagnosis and early surgical management are necessary to prevent the aortic leaflets being stretched and deformed by prolapse and herniation.

  • 5.

    Heart specimen photographs are presented in Figs. 16.3 through 16.7 .

    Fig. 16.3

    Opened right ventricle (RV) showing a small infundibuloventricular type of ventricular septal defect (VSD) between a normally located conal septum (CS) or infundibular septum above and the septal band (SB) and ventricular septum below. There is no right ventricular outflow tract obstruction. The subpulmonary CS is normal, as in Fig. 16.2 (left column, top row). The patient was 3 8/12-year old boy. The VSD measured 4 × 2 mm. The aortic valve was bicommissural because of deficiency of the right coronary-noncoronary (RC/NC) commissure. The RC/NC leaflet prolapsed into the left ventricular cavity, causing aortic insufficiency (or regurgitation). There was no right ventricular outflow tract gradient. The VSD appeared too small to permit herniation of the combined RC leaflet plus the NC leaflet of the aortic valve. This type of defect has also been called a membranous VSD and a subcristal VSD . PV, Pulmonary valve.

    Reproduced from Van Praagh R, McNamara JJ. Anatomic types of ventricular septal defect with aortic insufficiency. Am Heart J. 1968;75:604-619.

    Fig. 16.4

    Opened right ventricle (RV) showing a small infundibuloventricular type of ventricular septal defect (VSD) between an approximately normally located infundibular septum above and a normally located septal band (SB) and ventricular septum below. The patient was a 20 7/12-year-old man. The VSD measured only 3 × 2 mm. Note the aneurysm of the membranous septum: see the VSD label. Aneurysms of the membranous septum are thought to be how such conoventricular VSDs can close. Note that the conal septum (CS) is unusually far from the tricuspid valve. The CS may be somewhat too short, predisposing to this VSD. “Membranous” VSDs often appear to be more than membranous. This new understanding has spurred the use of other terms such as paramembranous, juxtamembranous, and perimembranous (which also does not exist, accurately speaking). This patient also had a bicuspid (bicommissural) aortic valve—a rudimentary right coronary/left coronary (RC/LC) commissure. The combined RC/LC leaflet did not prolapse downward (as expected). Instead, the noncoronary aortic leaflet prolapsed, despite the fact that it was supported by two normal-appearing commissures. However, this young man also had a history of subacute bacterial endocarditis and cerebral infarction. We did not know whether subacute bacterial endocarditis played a role in this patient’s prolapse of his normally supported NC leaflet.

    Reproduced with permission from Van Praagh R, McNamara JJ. Anatomic types of ventricular septal defect with aortic insufficiency. Am Heart J. 1968;75:604-619.

    Fig. 16.5

    Opened right ventricle (RV) showing an anterosuperiorly deviated and shortened conal (infundibular) septum (CS). There is a large conoventricular type of ventricular septal defect (VSD) between the malaligned CS above and the normally located septal band (SB) below. The VSD measured 13 × 10 mm. The patient was a 7 11/2-year-old boy. The patient had a pulmonary outflow tract gradient of 40 mm Hg and a bicuspid pulmonary valve (PV); his diagnosis was acyanotic tetralogy of Fallot. The patient had a bicuspid aortic valve because of a rudimentary right coronary/noncoronary commissure that resulted in aortic insufficiency. This patient’s heart was virtually identical to the heart specimens diagrammed in Fig. 16.2 , middle column.

    Reproduced with permission from Van Praagh R, McNamara JJ. Anatomic types of ventricular septal defect with aortic insufficiency. Am Heart J. 1968;75:604-619.

    Fig. 16.6

    (A) Right ventricle (RV) viewed from the front and above. (B) Opened left ventricle (LV).

    This 13 11/12-year-old boy had a complete conal septal defect, also known as a complete infundibular septal defect. The subarterial ventricular septal defect (VSD) was large, measuring 20 × 10 mm. The VSD was both subpulmonary, as seem in A, and subaortic, as seen in B; that is, this VSD was subarterial and doubly committed. The fact that an infundibular septal defect lies beneath both great arteries is also seen in Fig. 16.2 , right column. In A, one can see that the space above the septal band, normally occupied by the parietal band or distal conal septum, is wide open. The right posterior division (Rt Post Div) and the left anterior division (Lt Ant Div) of the septal band form the inferior rim of the VSDs. But there is nothing above the septum band until one gets to the pulmonary valve (PV) and the main pulmonary artery (MPA). One can also see the right coronary leaflet (RCL) of the aortic valve. But the infundibular septum is absent. In B, showing the opened LV and ascending aorta, one can see the RCL and the noncoronary leaflet (NCL) of the aortic valve, and the subaortic VSD, but no distal infundibular or outflow tract septum. There was a gradient of 15 mm Hg across the right ventricular outflow tract, measured at cardiac catheterization. The RCL and NCL were both prolapsed at autopsy, and the aortic valvar commissures were normally developed. But we could not tell at autopsy whether the prolapsed RCL had herniated into the RV outflow tract to produce the 15 mm Hg outflow tract gradient.

    Reproduced with permission from Van Praagh R, McNamara JJ. Anatomic types of ventricular septal defect with aortic insufficiency. Am Heart J. 1968;75:604-619.

    Fig. 16.7

    (A) Opened right ventricle (RV). (B) Opened left ventricle (LV). The patient was an 18½-year-old man with a conal septal defect type of ventricular septal defect (VSD). The VSD measured 20 × 10 mm. The aortic valve was tricommissural and tricuspid. The right coronary leaflet of the aortic valve prolapsed into the LV and herniated into the RV outflow tract, resulting in a right ventricular outflow tract gradient of 50 mm Hg. This young man had a history of subacute bacterial endocarditis. He also had a Meckel diverticulum. Histology revealed extensive subendocardial left ventricular myocardial infarction. In A, note that just beneath the pulmonary valve (PV), the infundibular septal defect is filled by the right coronary leaflet (RCL) of the aortic valve that has herniated into the subpulmonary right ventricular outflow tract. The inferior rim of the infundibular septal defect is formed by the top of the septal band, that is, by the right posterior division (Rt Post Div) and the left anterior division (Lt Ant Div) of the septum band.

    In B, the RCL of the aortic valve can be seen bulging through the infundibular septal defect into the RV outflow tract. The prolapsed and herniated RCL of the aortic valve is essentially absent from its normal location, resulting in aortic regurgitation. The adjacent noncoronary leaflet (NCL) of the aortic valve appears uninvolved. SB, septal band.

    Reproduced with permission from Van Praagh R, McNamara JJ. Anatomic types of ventricular septal defect with aortic insufficiency. Am Heart J. 1968;75:604-619.

As noted earlier, when the infundibuloventricular VSD is present (see Fig. 16.2 , left and middle columns ), the VSD is beneath the posterior half of the RC leaflet and the anterior half of the NC aortic leaflet. However, when an infundibular septal VSD is present, the VSD is more anterior. The VSD typically lies entirely beneath the RC aortic leaflet, not just half beneath the RC leaflet. Consequently, when an infundibular septal VSD is present, RC aortic leaflet prolapse downward and herniation into the right ventricular outflow tract is more probable than when an infundibuloventricular VSD is present, because the VSD is more anterior with a conal septal type of VSD.

The exception to this statement is when an infundibuloventricular VSD is associated with marked deviation of the parietal band (component 4) in a leftward, anterior, and superior direction, making the VSD unusually anterior. This is also why a conal septal VSD may be accurately described as a doubly committed subarterial VSD . An infundibular septal VSD is both subpulmonary (see Fig. 16.2 , right column, top row ) and subaortic, with downward prolapse and rightward herniation of the RC aortic leaflet (see Fig. 16.2 , right column, bottom row ). However, an infundibuloventricular type of VSD typically is subaortic only (see Fig. 16.2 , left and middle columns ).

Apical Ventricular Septal Defects

Apical muscular VSDs are the most difficult to visualize through the tricuspid valve and the most difficult to close trans-atrially. Consequently, this has left closure of such defects to using a left apical ventriculotomy. The disadvantages of a systemic ventriculotomy and the possible postoperative complication of a left ventricular apical aneurysm led Dr. Stella Van Praagh to start thinking about a more risk-free approach. It was she who conceived of the right ventricular apical infundibulotomy approach. Dr. Giovanni Stellin, a friend and colleague from Padua, Italy, was so impressed that he tried it, liked it, and published it with us in 2000. Dr. Stella Van Praagh et al published a follow-up in 2002.

The background to this surgical success is as follows. Thanks to Dr. Maurice Lev, we had long understood that the RV has two apices ( Fig. 16.8A–B ): the right ventricular sinus or inflow tract apex, proximal to the moderator band, and the infundibular or outflow tract apex, distal to the moderator band. The infundibular or outflow tract apex is also known as the infundibular apical recess.

Fig. 16.8

Opened Normal Right Ventricle (RV).

(A) Photograph. (B) Explanatory diagram. The infundibulum or conus arteriosus

has two parts: a distal or subarterial or parietal band part; and a proximal, or apical, or septal band (SB) and moderator band (MB) part. Consequently, the RV has two apices. The apex of the right ventricular sinus, body, or inflow tract lies proximal and inferior to the MB. The apex of the infundibulum is distal and superior to the MB and is also known as the infundibular apical recess. Apical muscular ventricular septal defects often run between the left ventricular apex and the infundibular apex. The infundibular apex above and the right ventricular sinus apex below are frequently separated by an infundibulosinus partition. In single left ventricle (LV) with an infundibular outflow chamber, the outlet chamber is formed by the infundibular apical recess. This infundibular myocardium has often been mistaken for RV sinus myocardium, causing confusion regarding single LV, in which the right ventricular sinus typically is absent. The problem is that the apical part of the infundibulum has not been widely understood, and consequently this infundibular apical myocardium has been mistaken for right ventricular sinus myocardium. This, in turn, has led to the mistaken notion that single LV is not really a single LV because a small RV is present (the outlet chamber). However, once one understands that the outlet chamber typically is the proximal part of the infundibulum, and that the right ventricular sinus is absent, then one understands that the LV really is single, because the right ventricular sinus beneath the SB is absent. The “small RV” is really the infundibulum only, without the right ventricular sinus. PB, Parietal band; PV, pulmonary valve; TV, tricuspid valve.

Reproduced with permission from Van Praagh S, Mayer JE, Berman NB, Flanagan ME, Geva T, Van Praagh R. Apical ventricular septal defects: follow-up concerning anatomic and surgical considerations. Ann Thorac Surg. 2002;73:48.

Double-chambered RV shows these two component parts of the RV with great clarity. Many people do not understand that the septal band and the moderator band are apical infundibular structures. They think that the infundibulum begins with the parietal band (component 4). Typically, the apical muscular VSD opens from the left ventricular apex into the infundibular apical recess (the infundibular apex).

The surgery of the right ventricular apical infundibulotomy is presented diagrammatically in Fig. 16.9 . The surgical incision is parallel to, and to the right of the distal part of the anterior descending coronary artery. The length of this incision, which extends close to the apex of the heart, varies from 1.5 to 2.5 cm. The apical muscular VSD can extend above and below the moderator band, opening into both the infundibular and the right ventricular sinus apices.

Fig. 16.9

Diagrammatic Presentation of Apical Infundibulotomy.

The incision (dotted line) is parallel to and to the right of the distal left anterior descending (LAD) coronary artery. The incision extends close to the ventricular apex and varies from 15 to 2.5 cm in length. The inset shows that the apical muscular ventricular septal defect (VSD) opens into the infundibular apical recess distal to the moderator band (MB) and in some cases the VSD also opened proximal to the MB into the apex of the right ventricular sinus. Ao, Ascending aorta; LV, left ventricle; MPA, main pulmonary artery; RV, right ventricle; SB, septal band.

Reproduced with permission from Van Praagh S, Mayer JE, Berman NB, Flanagan ME, Geva T, Van Praagh R. Apical ventricular septal defects: follow-up concerning anatomic and surgical considerations. Ann Thorac Surg. 2002;73:48.

In Fig. 16.10 , a heart specimen with an apical muscular VSD is presented. The patient was a 2.5-month-old girl who also had a coarctation of the aorta and a patent ductus arteriosus (PDA). In Fig, 16.10A , the opened right ventricle (RV) is shown. The apical VSD is located between the left ventricular apex and the infundibular apical recess. Fig. 16.10B shows the opened LV and the apical VSD, which is large (25% of the LV’s septal length).

Fig. 16.10

The heart of a 2.5-month-old girl with an apical muscular ventricular septal defect (VSD) that measured 40 × 10 mm. The VSD lay between the left ventricular sinus apex and the infundibular apex. The VSD was considered to be relatively large because its maximal dimension equaled 25% of the length of the left ventricular septal surface. (A) Opened right ventricle (RV). (B) Opened left ventricle (LV). (C) Selective left ventricular angiocardiogram, left anterior oblique projection. (A) Apical muscular VSD opens into the infundibular apical recess (Inf Apex). A muscular partition separates the infundibular apex from the right ventricular sinus apex, but the exit from the infundibular apex is nonrestrictive superiorly. (B) Apical muscular VSD from the left ventricular aspect. (C) Documents the apical muscular VSD’s superiorinferior dimension ( rightward arrows with superoinferior orientation). The unobstructed exit from the infundibular apical recess is demarcated by the leftward, horizontally oriented arrows. This patient had normal segmental anatomy, that is {S,D,S}, a secundum type of atrial septal defect (ASD), double orifice mitral valve with mild congenital mitral stenosis, coarctation of the aorta, and a patent ductus arteriosus. Before this angiocardiogram was done at 17 days of age, she had the coarctation repair, patent ductus arteriosum ligation, and banding of the main pulmonary artery.

Reproduced with permission from Van Praagh S, Mayer JE, Berman NB, Flanagan ME, Geva T, Van Praagh R. Apical ventricular septal defects: follow-up concerning anatomic and surgical considerations. Ann Thorac Surg. 2002;73:48.

We concluded that large apical VSDs can be successfully closed through small right ventricular apical infundibulotomies. This approach, applicable even in small infants, can avoid both pulmonary artery banding and left ventriculotomy. However, extremely large apical VSDs with severe biventricular dysplasia and dysfunction may require cardiac transplantation.

The Story of Ventricular Septal Defects

If you do not know the history of something, you do not really understand it. So this is a story I have to tell you. No one knows the whole story of anything. But I have been reading the medical literature for my whole professional life, taking notes on 5 inch by 8 inch white cards, and keeping them in chronologic order. No one can read everything, of course. But now, more than 60 years later, these white cards have an amazing story to tell: the birth and development of our field, how and when it happened, and who the major players were.

The year 1784 saw the posthumous publication of Three Cases of Mal-conformation in the Heart, written by William Hunter (1718–1783), a Scot who was a leading obstetrician of his time in London, England. It should be added that William Hunter was the elder brother of John Hunter, the leading surgeon of his time and the founder of surgical pathology. In William Hunter’s paper, Case 3 was a stillborn infant of 6 months gestation who had a VSD and a secundum type of ASD with multiple perforations of the septum primum. This is the earliest report of a VSD that I have in my notes. This report is also remarkable because it contains a clearcut statement of the concepts of evolution, natural selection, and survival of the fittest published 74 years before papers by Darwin and Wallace (1858 to 1860). On page 308, William Hunter wrote:

As in vegetables too, the parent generally produces a species very like itself; but sometimes a different constitution, whether better or worse. Whatever may happen in a particular instance, or with regard to an individual, the most perfect and sound animal upon the whole, will have the best chance of living to procreate others of his kind: in other words, the best breed will prevail: and the monstrous constitution, and that which is defective, or of such a fabric as necessarily to breed disease, will be cut off. The most perfect constitution will be preserved; it will be most susceptible of love, and the most likely to meet with a warm return of that passion: so that, in every way, the sound constitution will have the preference in procreation, and the defective, weak, or diseased line will be wearing out.

So there it is: evolution by natural selection and survival of the fittest. Am I surprised? No. Why not? Because we have known since 1985 that the concept of evolution by natural selection was understood by Empedocles (495–435 BC), an ancient Greek pre-Socratic physicist from Agragas (now Agrigento), Sicily, which at that time was part of Magna Graecia (Great Greece).

To return to Dr. William Hunter, he continued:

If this doctrine were as well known, and as much attended to in our country, with regard to the human species, as it is in breeding up horses and many other animals, personal qualifications would be more attended to in match-making then they generally are. We every day see preference given to rank or birth, or weight of possessions, at the expense of entailing disease of body and mind upon a devoted race doomed to early extinction.

Then, speaking directly to the Society of Physicians in London, Dr. William Hunter wrote:

Your plan, Gentlemen, professes to keep to practical subjects. Some of your readers may look upon this communication as rather philosophical and curious, than practical and useful. But though the cure of diseases be the first object of our profession, the knowledge of incurable complaints [such as congenital heart disease] is of much importance to humanity; particularly in restraining us from bleeding, blistering, vomiting, purging, cutting issues, applying caustics: in a word, torturing a miserable and incurable human creature. (Bold words are the author’s.)

This was William Hunter’s postmortem suggestion to his fellow London medical practitioners in 1784 concerning their medical therapy that he understood often amounted to the torture of patients with incurable disease (such as VSD). Today, few would disagree with William Hunter. How is it possible that Hunter’s understanding of evolution by natural selection and survival of the fittest (1784), 74 years before Darwin and Wallace (1858), has been forgotten? Or Empedocles, whose primacy in this regard is proved by Aristotle’s disapproving citation of Empedocles in the Physics. ,

Now, astronomers and astrophysicists know that the concept of evolution applies to the chemistry of our universe, which is about 13.8 billion years old. Shortly after the Big Bang, about 400,000 years later when the universe had cooled enough for the elementary particles to form atoms, these atoms were mostly hydrogen (1 proton in the nucleus), with some helium (2 protons), and a little lithium (3 protons).

What were (and still are) the stars doing, apart from twinkling (an effect of our atmosphere) and adorning our night sky? Large stars, with enormous heat and pressure were, and are, producing the heavier element, in a process known as nuclear fusion or nuclear synthesis. Then, some of these large stars exploded (i.e., they go supernova), seeding interstellar space with the heavy elements found in the Periodic Table. Another method of spreading heavier elements into the interstellar gas is by the stellar wind. The many neutrinos that are passing through you right now are from our solar wind.

Our solar system is now thought to have formed about 4.6 billion years ago, that is, about 9.2 billion years after the beginning of our universe (the so-called Big Bang): 13.8 – 4.6 = 9.2. This 9.2 billion years was a long enough time that enough heavy elements had been formed in the stars and distributed into the interstellar space of our universe so that life as we know it could evolve. We are star dust, the result of the chemical evolution of our universe over the past 13.8 billion years.

When an astronomer or astrophysicist looks at the human body, what does he or she see?

Oxygen 61% 8 protons (in the nucleus)
Carbon 23% 6 protons
Hydrogen 10% 1 proton
Nitrogen 3% 7 protons
Trace elements 3% Variable

The number of protons in the nucleus is known as the atomic number. Life as we know it on Earth is all carbon-based. Are we capable of assessing objectively the overall effects of successful modern treatment of congenital heart disease?

What is the result of successful therapy on our human gene pool, given that molecular genetic abnormalities appear to be the cause of much congenital heart disease? Are we treating individual patients increasingly more successfully but also polluting the human gene pool ? Sir Brian Barratt-Boyes asked me to consider this question. In his introduction, Sir Brian said, “Now Dr. Van Praagh is going to tell us why we should not be doing, what we are doing.” Perhaps, fortunately, I could not do that because we do not have enough relevant data.

Do the children and grandchildren of parents with congenital heart disease have a significantly increased prevalence of congenital heart disease? We still do not know the answer to this very important question. But we must continue to do such studies in order to find out.

Now back to the story of VSDs, as it unfolds over the years. I shall omit papers that have little or nothing “new” to say, in the interests of attempted brevity.

In 1897, Victor Eisenmenger described what became known as the Eisenmanger complex, that is, TOF without pulmonary outflow tract stenosis. There was subpulmonary infundibular underdevelopment, with leftward, anterior, and superior deviation of the parietal band (see Fig. 16.1 distal conal septum, component 4). However, the hypoplasia and deviation of the conal septum were enough to produce a large malalignment type of conoventricular VSD, but not severe enough to produce significant infundibular stenosis.

Dr. Bill Rashkind translated Eisenmenger’s paper from German into English. Eisenmenger’s patient was a 32-year-old man who had been cyanotic since early childhood. He had marked digital clubbing and a buzzing systolic murmur over the cardiac apex. S 2 was loud, but P 2 was described as not increased. Autopsy revealed a large subaortic VSD, and what I thought was unrecognized pulmonary vascular obstructive (PVO) disease. The PVO disease was not recognized as such in 1897. On page 180, Dr. Rashkind’s translation was as follows: “The slightly dilated pulmonary artery shows endoarteritic thickening on its inner surface, which continues into the main branches of the vessel.” I interpreted the latter as lipid streaking and pulmonary atherosclerosis. No pulmonary histology was described.

In 1952, Drs. Muller and Dammann described and proposed pulmonary artery banding to treat malformations of the heart that are characterized by increased pulmonary blood flow. The idea of pulmonary artery banding was to create pulmonary stenosis, thereby reducing pulmonary hypertension and excessive pulmonary blood flow. Their patient was a 3½-month-old boy. Pulmonary artery banding was suggested for single ventricle, VSD, and truncus arteriosus.

In 1955, Drs. Lillehei, Cohen, Warden, and Varco —the famous Minneapolis–St. Paul team—introduced direct-vision open-heart surgery by means of controlled cross circulation. They reported their surgical results in 32 patients with VSDs, TOF, and common AV canal. This report was a bombshell. Open heart surgery had arrived.

In 1955, Dr. John W. Kirklin et al from the Mayo Clinic in Rochester, Minnesota reported intracardiac surgery in 8 patients with the aid of a mechanical pump oxygenator of the Gibbon type. This changed everything. Not only did Kirklin confirm that open heart surgery is possible. He and his team did it in a way, using a mechanical pump-oxygenator, that would prove clinically feasible. Clinical open heart surgery had begun.

In 1956, Dr. Kirklin et al reported 20 patients with large VSDs who had undergone intracardiac repair with a pump oxygenator. Of these 20 patients, 4 died (a 20% mortality), which was considered low in that era.

In 1960, Lev did a detailed study of the AV conduction system in hearts with a VSD.

In 1963, Kirklin, McGoon, et al. used Lev’s study to develop surgical techniques to avoid creating heart block during VSD repair.

In 1960, Morgan, Griffiths, and Blumenthal from Babies’ Hospital in New York City published a large study of infants with VSD. Of 125 patients with VSD, 17 developed congestive heart failure (14%). Of the 17 infants with congestive heart failure, 10 died (59%). If congestive heart failure was going to develop, it always did so in the first 6 months of postnatal life.

In 1961, Neufeld, Titus, DuShane, Burchell, and Edwards from the Mayo Clinic described and named VSDs of the AV canal type , based on a study of 15 postmortem cases, because this anatomic type of VSD “usually occupied the position of the ventricular component of the defect in persistent common atrioventricular canal.”

They also noted that “the electrocardiographic features were striking. In all cases the mean electrical axis of the QRS lay above the isoelectric point; the vector loop obtained in the frontal plane from the scalar electrocardiogram was directed counterclockwise, and its main mass was above the zero line. In addition, in all cases there were signs of right ventricular overload and in some cases of left ventricular overload as well.” These authors proposed that the characteristic counterclockwise and superior frontal plane QRS loop “is the result of congenital displacement of the bundle of His in its relation to the VSD.” These authors also noted that this characteristic counterclockwise and superior frontal QRS loop occurred with the common AV canal ( not present in these cases of isolated VSD of the AV canal type). These patients did not have an ostium primum type of septal defect, and they did not have common AV valves. But the presence of an isolated VSD of the AV canal type indicated that this anatomic type of VSD was related to, but different from, common AV canal. This is what these authors meant by calling this anomaly isolated VSD of the AV canal type.

I understand and agree with this distinction. We have always regarded VSD of the AV canal type as a partial form of common AV canal, often caused by ventriculoatrial malalignment. For example, if the RV is underdeveloped, the muscular ventricular septum may underlie the middle of the tricuspid orifice and the tricuspid valve. This results in a VSD of the AV canal type that is not related to an anomaly of the AV canal. The real anomaly is at the ventricular level, not at the AV canal level. A VSD of the AV canal type often appears to be due to ventriculoatrial malalignment, not to an anomaly of the AV canal per se. This is why in patients with a VSD of the AV canal type, there is no ostium primum defect and no common AV valve. Ironically, in VSD of the AV canal type, the basic diagnosis often goes unnoticed and unmentioned: ventriculoatrial malalignment.

In 1965, Kidd, Rose, Collins, and Keith from the Hospital For Sick Children in Toronto, Ontario, Canada published a study of the hemodynamics of VSDs in infancy. Infancy is defined as the first 12 months of postnatal life. In 151 infants with the diagnosis of isolated VSD who underwent cardiac catheterization before 1 year of age, the salient findings were as follows. A pulmonary-to-systemic flow ratio (Qp/Qs) was greater than 2/1 in 78%. In those with a low flow ratio (Qp/Qs <2/1), the low flow ratio was due to the presence of a small VSD. A low or normal total pulmonary vascular resistance was found in 59%. The total pulmonary vascular resistance was greater than normal in only 25% of patients. Of these 151 patients, 25 were recatheterized subsequently. Of these 25 patients, 3 achieved functional closure (12%).

In the group with low flow and high resistance in the first study, the follow-up study subsequently revealed lower pulmonary resistance. However, 14 of these 25 patients revealed increasing pulmonary vascular resistance at follow-up study (56%). Of these 14 patients, 10 had low or normal pulmonary resistance earlier (71%).

Kidd et al reached the following conclusions. “It is suggested that progressive pulmonary vascular obstruction is a sequel to high flow and low vascular resistance in the pulmonary circulation and is not present from the time of birth. It is proposed that such patients be subjected to serial study in order to define those who are reacting to the pressure-flow stimulus in this way.”

In 1965, Friedman, Mehrizi, and Pusch published a study of multiple muscular VSDs. It was estimated that approximately 10% of VSDs are muscular and that of isolated VSDs, 4% are muscular. Friedman et al reported 7 patients with muscular VSDs; 6 of the 7 died (86% mortality). Their recommendations in 1965 were that if a VSD is suspected, do a left ventricular angiocardiogram, and if multiple VSDs are present, band the main pulmonary artery (MPA).

In 1965, Barnard and Kennedy published a case of postinfarction VSD. It was apical and muscular, below the septal band. They suggested waiting 5 weeks after the infarction before closing the VSD surgically.

In 1965, Hoffman and Rudolph published a study of the natural history of VSDs in infancy, based on 62 infants. They estimated the incidence of VSD to be approximately 0.94 in 1000 full-term live births, 4.51 in 1000 premature live births, and 1.35 in 1000 full-term and premature live births. Congestive heart failure occurred in 31 of 62 infants (50%), before 6 months of age in all, much earlier in premature infants than in full-term babies. Spontaneous functional closure occurred in 13 of 36 local infants (36%), but in only 2 of 26 referred infants (8%). For the whole group (local and referred infants), 52% of VSDs closed or became smaller. Only 16% of these infants were seriously affected. Complete closure of VSDs occurred between 7 and 12 months, usually when the VSD was small. But spontaneous closure also “could occur with large defects.”

What about pulmonary vascular resistance? “Rise in pulmonary vascular resistance in infancy is not rare when the VSD is big and that, even in these children, the pulmonary vascular resistance does not usually persist at the high level present at birth, but first falls postnatally to normal levels before rising.”

In 1965, Ritter, Feldt, Weidman, and DuShane found that congestive heart failure occurred in infants with VSD between 6 weeks and 6 months in 53 of their 78 cases (68%).

In 1996, Simmons, Moller, and Edwards published a paper on spontaneous closure of VSD. Assisted by a personal communication from Hoffman and Rudolph, Simmons et al summarized the findings in 40 infants with VSD:

No change in VSD size 11/40, 27.5%
VSD became smaller 19/40, 47.5%
VSD closed 10/40, 25.0%

Simmons et al observed two anatomic types of muscular VSD closure: (1) closure by the septal leaflet of the tricuspid valve and (2) closure by fibrous tissue within the ventricular septum toward the right ventricular side, but not by tricuspid valvar tissue.

In 1966 , Osborn, Hall, Winn, Capper, and Blake reported a fatal complication of pulmonary artery banding. The patient was a male infant who was banded at 5 months of age because of an unsatisfactory response to medical treatment. Cardiac catheterization had shown a 20% left-to-right shunt. Twenty-two months after banding, the infant was found to have developed a thrombus at the band site. The thrombus almost occluded the MPA. The patient died postoperatively. These authors noted that banding removes much of the adventitial circulation at the band site and that banding impedes reestablishment of adventitial circulation to the pulmonary artery wall at the band site. Internally, intimal sclerosis results from the focal narrowing of the MPA. The authors thought that ischemia had occurred involving the inner one-third of the main pulmonary arterial wall at the band site, leading to the occlusive intima thrombus. Osborn et al recommended wide, nonreactive Teflon tape for pulmonary artery banding.

In 1966, Hoffman and Rudolph published a paper in the general pediatric literature (Pediatrics) warning of the dangers of VSDs. Their paper was entitled, “Increasing Pulmonary Vascular Resistance During Infancy in Association With Ventricular Septal Defect.” They reported 3 infants with VSD and increasing pulmonary vascular resistance. The point was that elevated pulmonary vascular resistance would make such patients inoperable, and it was therefore essential to have such patients diagnosed accurately and operated on successfully before elevated pulmonary vascular resistance (Rp) would make such patients inoperable. Note also that Hoffman and Rudolph were talking about infants needing open-heart surgery in 1966 . These were still early days in open-heart surgery.

Also in 1966, Rose, Collins, Kidd, and Keith were spreading the word about the importance of VSD in infancy and childhood. Dr. John Keith, the Canadian father of Pediatric Cardiology, and colleagues were publishing another huge and well-documented study based on 407 infants and children with VSD in the general pediatric literature (Journal of Pediatrics). Noncardiologists—pediatricians, neonatologists, obstetricians and gynecologists, radiologists, and pathologists—all had to learn about VSDs in infants and children so that referrals would be made soon enough, diagnoses would be made soon enough, and surgical repair would be done soon enough, so that these otherwise normal infants and children would survive and thrive. That was the challenge in 1966.

In 1966, Girod, Raghib, Adams, Anderson, Wang, and Edwards published a study of 46 autopsied cases with VSD, focusing on associated cardiac malformations. They studied VSDs that are “not part of a recognized complex” (e.g., TOF or DORV). Why? Because such “associated anomalies” can be very important surgically. Significant surprises at surgery can be fatal.

In 1966, Gould and Lyon reported the prolapse of a pulmonary leaflet through a VSD; they thought theirs was the first report of this anomaly. The patient was a 37-year-old white man. He had a subpulmonary conal septal defect (their Figure 16.5 ). The left septal leaflet of the pulmonary valve prolapsed inferiorly and herniated through the subpulmonary conal septal defect, resulting in marked pulmonary insufficiency (regurgitation).

In 1965, Schrire, Vogelpoel, Beck, Nellen, and Swanepoel from Cape Town, South Africa, presented a paper concerning the clinical spectrum of VSD. Their hemodynamic classification of VSDs was as follows :

<SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='Type1PAsystolicpressureSystemicsystolicpressure&lt;40%’>Type1PAsystolicpressureSystemicsystolicpressure<40%Type1PAsystolicpressureSystemicsystolicpressure<40%

<SPAN role=presentation tabIndex=0 id=MathJax-Element-2-Frame class=MathJax style="POSITION: relative" data-mathml='(almostnormalPApressure)’>(almostnormalPApressure)(almostnormalPApressure)

<SPAN role=presentation tabIndex=0 id=MathJax-Element-3-Frame class=MathJax style="POSITION: relative" data-mathml='Type2PAsystolicpressureSystemicsystolicpressure=40%−795%’>Type2PAsystolicpressureSystemicsystolicpressure=40%795%Type2PAsystolicpressureSystemicsystolicpressure=40%−795%

<SPAN role=presentation tabIndex=0 id=MathJax-Element-4-Frame class=MathJax style="POSITION: relative" data-mathml='(moderatepulmonaryhypertension)’>(moderatepulmonaryhypertension)(moderatepulmonaryhypertension)

<SPAN role=presentation tabIndex=0 id=MathJax-Element-5-Frame class=MathJax style="POSITION: relative" data-mathml='(PA,Pulmonaryartery.)’>(𝑃𝐴,Pulmonaryartery.)(PA,Pulmonaryartery.)

In 1966, Reynolds published a case with a subpulmonary conal septal defect (that was called a supracristal VSD), without aortic or pulmonary insufficiency. “The defect was closed by direct suture.” This is a very important sentence. The subpulmonary conal septal VSD was closed by direct suture, not with a patch. It will be recalled that this type of VSD can be complicated by herniation of a normally formed RC aortic leaflet through a subsemilunar conal septal defect, resulting in significant AI. The way this subsemilunar VSD was closed in Dr. James Reynolds’ patient—by direct suture (not with a patch)—is the best way to make this repair, if technically feasible, to prevent the RC leaflet of the aortic valve from herniating through the subsemilunar conal septal defect (see Fig. 16.2 , right column, lower row ). Direct suture closure repair of the conal (infundibular) septal hernia is what the normally formed aortic leaflet needs (normal support from below) to support the aortic valve, before an RC aortic leaflet hernia and stretching occur, resulting in AI.

In 1966, Williams, Hara, and Bulloch reported a traumatic VSD.

In 1966, Saab, Burchell, DuShane, and Titus published a paper that focused on muscular VSDs.

In 1967, Mitchell, Berendes, and Clark suggested the following: “The hypothesis is proposed that the normal time of VSD closure may not be limited to the fourth and fifth postconceptive weeks, but rather may extend, for a minority of patients, throughout pregnancy and into the postpartum period.” This is a reasonable way of viewing the rightmost end (the long end) of this distribution curve. This view suggests that spontaneous (medically unassisted) postnatal VSD closure is the rightmost end of the normal VSD closure curve.

In 1967, Grainger et al published a paper about pulmonary artery banding as treatment for VSD.

In 1967, Dr. Jesse E. Edwards (one of the great congenital and acquired cardiovascular pathologists of the 20th century; Dr. Maurice Lev was another) published a paper about “unresolved problems” concerning VSDs:

  • 1.

    Should subpulmonary or subarterial VSDs be called “supracristal” VSDs (his Fig. 16.5 )?

  • 2.

    How could a VSD have subaortic stenosis (see Fig. 16.7 )?

  • 3.

    Should the ventricles be designated hemodynamically, such as “venous” and “arterial” ventricles? Or should the ventricles and the AV valves be named in terms of their morphologic anatomy, as Lev and I were advocating?

  • 4.

    In classic congenital physiologically corrected TGA, is it permissible to call regurgitation of the left-sided AV valve “mitral” insufficiency? Or should it be called “left-sided tricuspid” insufficiency?

At that time, we (my colleagues and I) thought that the correct answers to these questions were as follows:

  • 1.

    No. The anatomic meaning of crista supraventricularis was confused and erroneous. Initially, it was thought that the crista supraventicularis consisted of the parietal band (component 4) and the septal band (see Fig. 16.1 , component 3). Later, study of the conotruncal (infundibuloarterial) anomalies such as TOF, TGA, and DORV made it obvious that the parietal band (component 4) and the septal band (component 3) are two separate structures, not the parietal and septal extensions of one structure, the crista supraventicularis as classically defined and understood. Only the parietal band (component 4) formed a supraventricular crest. The septal band (component 3) did not form a supraventricular crest because, typically, the septal band was attached to the superior right ventricular septal surface. Briefly, the meaning of the crista supraventricularis changed and then collapsed when the crista supraventicularis was understood to be a misconception. So other more accurate anatomic and developmental names were preferred (see Fig. 16.1 ).

  • 2.

    How could a VSD have subaortic stenosis? By having a conal septal left shift.

  • 3.

    Should the ventricles be designated hemodynamically (“venous” ventricle and “arterial” ventricle)? No. We thought that the cardiac chambers (atria and ventricles) should be diagnosed and designated morphologically. The type of blood conveyed by a morphologically RV could be either venous or arterial; the hemodynamics were a variable, whereas a chamber’s morphology was a constant.

  • 4.

    In TGA {S,L,L}, the left-sided AV valve is the tricuspid valve (not the mitral valve). Consequently, we are talking about left-sided tricuspid regurgitation (not about mitral regurgitation).

In 1967, Anderson, Levy, Naeye, and Tabakin published a case report concerning a 4-year-old girl with a VSD who had developed what the authors regarded as rapidly progressive PVO disease. The authors were shocked. The dangers of a VSD and the risk for becoming inoperable at an early age (4 years) were becoming apparent.

In 1967, Levin, Spach, Canent, Boineau, Capp, Jain, and Barr published a study of intracardiac pressure-flow dynamics in isolated VSD:

  • 1.

    When right ventricular pressure is low, shunting is left-to-right throughout the cardiac cycle.

  • 2.

    But when right ventricular pressure is systemic, the shunt is right-to-left during isovolumic relaxation.

In 1967, Sigmann, Stern, and Sloan published their results of early surgical correction of large VSDs. In 45 infants who underwent surgery, mortality was 10%. PVO disease was found pathologically as early as 8 months of age, suggesting that surgery in infancy might be required for successful management of patients with VSD.

In 1967, Tikoff, Schmidt, and Tsagaris published a study of an adult with a large VSD. They found that isoproterenol induced right ventricular infundibular obstruction. They thought their findings suggested that right ventricular infundibular obstruction could be acquired at any time in the history of a patient with VSD. They thought that such obstruction may be intermittent. They also suggested that infundibular obstruction could result in syncope.

In 1968, Sakakibara and Konno published a paper about the pathologic anatomy of patients with VSD and congenital aneurysm of the sinus of Valsalva, based on 70 cases (55 surgical cases and 15 autopsied cases).

Instead of talking about VSD with AI, as we and others have done, Sakakibara and Konno proposed that these conditions should be understood as VSDs with aneurysms of the sinuses of Valsalva. (Readers should be aware that this was a famous Japanese cardiovascular team. Dr. Shigeru Sakakibara was Professor of Cardiac Surgery and Director of the Heart Institute of Japan. Sakakibara discovered and described double-outlet left ventricle . Dr. S. Konno was Chief of Staff of the Heart Institute of Japan and Head of the Catheterization Laboratory.) They classified aneurysms of the sinus of Valsalva with VSD into four anatomic types.

In 1968, Dr. J. I. E. Hoffman published a study of the natural history of VSD. This study found estimated incidence ≈ 20 in 10,000 live births (0.2%) and death before 15 years of age ≈ 10% of VSDs, that is, 2 in 10,000 live births (0.02%). There “should” be about 18 in 10,000 school children with VSD. However, intensive school surveys show only about half of this number, that is, about 9 in 10,000 (0.009%). These findings support the suggestion that there is a high incidence of spontaneous closure of VSDs—perhaps 50%.

In 1968, Baron, Wolf, Steinfeld, and Van Mierop published a paper concerning the angiocardiographic diagnosis of subpulmonary VSD. In addition, this paper contains many helpful insights concerning the anatomy, embryology, and surgery of subpulmonary and conoventricular VSDs, which comes as no surprise because Ludwig (“Bob”) Van Mierop was the senior coauthor. Van Mierop, who was both a cardiac surgeon and an embryologist, says that membranous VSDs can be readily closed by the RA, whereas surgical closure of a subpulmonary VSD requires a right ventriculotomy. Despite his surgical training, Van Mierop spent most of his professional life as a pediatric cardiologist. He was also an excellent artist, as this paper demonstrates. His developmental description merits quotation:

As development continues, both the truncus and the conus become divided into two separate channels. The truncus swellings first join in their midportion and with continued growth their line of fusion extends distally toward the aorticopulmonary septum and proximally toward the conus. . . .

At first the conus swellings grow slower and lag behind those in the truncus. Fusion begins in the mid region of the conus and extends in both directions. . . .

. . . In the adult heart, the line of junction of the two septa is located in the ventricular septum, 1 to 3 mm below the pulmonic valve, but is not demarcated by any gross anatomic landmark.

Distally, if the conal septal cushions do not fuse with each other, or with truncal septum that extends 1 to 3 mm below the pulmonary valve leaflets, the result is a subpulmonary and a subaortic conal septal defect, that is, a subsemilunar conal septal defect, that typically lies directly beneath the RC leaflet of the normally located aortic valve (see Fig. 16.2 ).

Proximally, if the conal septal cushions do not fuse completely, in a normal way, the proximal conal septum (the parietal band) is hypoplastic, higher than normal, and also can be deviated anterosuperiorly or in some other direction. This results in a conoventricular type of VSD between the abnormally located conal septum above and the normally located ventricular septum below. We do not call such defects membranous VSDs because there often is much more wrong than an anomaly of the membranous septum. Typically, there is an anomaly of the inferior part of the conal septum that forms the roof of the conoventricular VSD, as seen from the right ventricular perspective.

  • Conoventricular VSDs and subpulmonary conal septal defects are both conal septal anomalies:

  • 1.

    Conoventricular VSDs typically are defects of the proximal or inferior conal septum.

  • 2.

    Subpulmonary VSDs are defects of the distal or superior conal septum.

Van Mierop’s diagrams show this beautifully.

In 1968, Clarkson, Frye, DuShane, Burchell, Wood, and Weidman published a paper on the prognosis of patients with VSD and severe PVO disease.

In 1968, Rastelli, Ongley, and Titus published a paper about 4 patients with VSD of the AV canal type with straddling tricuspid valve and a mitral valve deformity.

In 1968, Edgett et al published a case of spontaneous VSD closure after pulmonary artery banding. The patient was a 6-year-old white boy. Banding may have helped promote spontaneous VSD closure.

In 1918, Weber is the first person known to have suspected spontaneous VSD closure. Beyond infancy, 25% of small VSDs will close spontaneously.

In 1968, Coleman et al from the Royal Hospital for Sick Children in Glascow, Scotland reported their results concerning VSD repair in childhood. In 91 children from 1 to 12 years of age, the mortality rate was 15%. Infundibular stenosis was also present in 18%; I wondered if these patients with infundibular stenosis had TOF, but I could not be sure.

In 1968, Leachman and Pereyo reported the case of a 2-year-old girl who had a VSD with elevated pulmonary capillary wedge pressure and severe pulmonary arterial hypertension. After surgical VSD closure, her pulmonary vascular resistance decreased significantly, indicating that this is possible. In other words, severe pulmonary arterial hypertension and elevated pulmonary vascular resistance do not always contraindicate VSD closure—a surprising and hopeful finding.

In 1968, Tandon reported chylothorax after the surgical repair of a VSD in an 8-year-old girl. Treatment involved repeated aspiration and a low-fat, high-protein diet.

In 1969, Gotsman, Beck, Barnard, and Schrire reported their hemodynamic studies after surgical VSD closure.

In 1969, Miller et al studied the relationship between hemodynamics in VSD patients and their height and weight. They found “a close association between pulmonary hypertension and poor weight gain.”

In 1969, Hallidie-Smith et al assessed the effects of surgical VSD closure on pulmonary vascular disease. If there were no complicating lesions, surgical VSD closure reduced their mortality rate from 22% to 15%.

In 1969, Yang et al reported 110 cases of membranous septal aneurysm. Only 2 had a VSD (1.8%). Why do aneurysms of pars membranacea septi (the membranous part of the ventricular septum) matter?

In 1969, a paper by Varghese et al answers this question. They regard a membranous septal aneurysm as “a method of spontaneous closure of a small [membranous] VSD.” The VSD almost did not close. Why not? Often it looks as though the inferior rim of the conal septum (the parietal band) was a bit too far away from the membranous septum; that is, mild hypoplasia of the inferior part of the conal (infundibular) septum. Hence, an aneurysm of the membranous septum suggests that the ventricular septum just barely managed to close and one should look for a small, residual membranous VSD.

In 1969, Morton et al also reported an aneurysm of the ventricular septum with an aortic valvar malformation in an infant. Thus, aneurysms of the membranous septum and their relationship to membranous VSDs were now being recognized diagnostically and understood as part of the VSD closure process.

In 1969, Herbert pointed out that some VSDs could be visualized angiocardiographically only in diastole. These VSDs were thought to be muscular. Systole resulted in approximation of the margins of the VSD and functional closure, whereas diastolic relaxation permitted separation of the margins of such VSDs, facilitating their diagnosis.

In 1969, Scheinman et al reported early repair of an apical muscular VSD caused by nonpenetrating trauma in an 8-year-old boy.

In 1969, Gonzales-Lavin and Barratt-Boyes reported their surgical experience in patients with VSD and aortic valvar incompetence. In a series of 7 cases, they described VSD closure using the RC leaflet of the aortic valve and homograft aortic valve replacement.

In 1969, Meng reported a 28-month-old girl with spontaneous closure of a VSD in tricuspid atresia. Hence, VSD closure can be physiologically disadvantageous or disastrous.

In 1969, Nghiem et al reported spontaneous VSD closure after pulmonary artery banding. Because it was being recognized that this can happen, these authors suggested doing an angiocardiogram before debanding to see if the VSD has closed, to avoid an unnecessary right ventriculotomy.

In 1970, Pombo et al published 4 cases of aneurysm of the membranous ventricular septum and a VSD. They suggested that the best method of diagnosing these interrelated defects was by selective left ventricular angiocardiography.

In 1970, Dr. Jane Somerville, Dr. A. Brandao, and Mr. Ronald Ross reported their surgical experience and clinical findings in 20 patients with VSD and aortic regurgitation. They preferred pulmonary homograft replacement of the aortic valve, rather than repair. Of their 20 patients, 2 died (10%), 18 survived (90%), and 17 had a satisfactory result (85%).

In 1970, Stark, Tyman, and Aberdeen reported their experience with spontaneous VSD closure after pulmonary artery constriction (banding). They recommended reconstruction of the MPA with a pericardial patch.

In 1970, Ebert, Canent, Spach, and Sabiston reported their experience with VSD closure and prior pulmonary artery banding. The resistance ratio, pulmonary arterial resistance/systemic resistance (Rp/Rs), did not change: after total correction, it was the same as before banding.

In 1970, Wagner, Ankeney, and Liebman tried to answer an important question: Can one predict how well a patient with a large VSD is likely to do after surgical closure? Their patients had “nonrestrictive” VSDs. “Nonrestrictive” means that they had systemic pressure in the RV. Salient findings are:

  • 1.

    If the pulmonary arterial mean pressure preoperatively was less than 95 mm Hg, they did well.

  • 2.

    If the pulmonary arterial pressure preoperatively was greater than 65 mm Hg, mortality was high.

Flows and resistances did not correlate well with surgical results, but mean pulmonary arterial pressures did. This is what the data in their 30 patients showed.

In 1970, Yokoyama, Takao, and Sakakibara published their findings concerning the natural history and surgical indications in patients with VSD.

In 1970, Hunt, Formanek, Castañeda, and Moller published a paper concerning 8 patients who had had pulmonary artery banding followed by surgical VSD closure. Pulmonary vascular resistance was normal in 3 (37.5%), moderately increased in 4 (50%), and markedly increased in 1 (12.5%). Mild, acquired pulmonary valvar stenosis was found in 7 of 8 (87.5%). Mild acquired (secondary) infundibular stenosis was observed in 3 of 8 (37.5%). Thus, it seems reasonable to conclude that pulmonary artery constriction (banding) can result in mild secondary (acquired) pulmonary valvar and infundibular stenosis.

In 1970, Li and Keith found that the incidence of spontaneous VSD closure was about 25%.

In 1971 , Rao and Sissman published a paper about the spontaneous closure of physiologically advantageous VSDs, as in tricuspid atresia and DORV.

In 1970, Daicoff and Miller proposed that the way to treat patients with VSD and congestive heart failure in infancy was by early repair of the VSD. The ages of their patients were 6 months, 10 months, 11 months, 15 months, 22 months, and 33 months, and all survived. My comment in my private notes was “Bravo!”

In 1970, Goor, Lillehei, Rees, and Edwards discussed the embryology and anatomic classification of VSDs based on a study of 101 heart specimens with 112 VSDs.

In 1971, Jarmakani et al found that after VSD closure, left ventricular volume and mass regressed considerably. However, there may be some partially irreversible changes associated with long-term left ventricular hypertrophy.

In 1971, Lev et al presented the quantitative anatomy of isolated VSD based on 53 cases older than 3 months of age.

In 1971, Brammel et al presented a clinical and physiologic study of 91 cases of the Eisenmenger syndrome. The Eisenmenger syndrome is defined as a VSD with PVO disease and a right-to-left shunt at the atrial, ventricular, or great arterial level (e.g., with a PDA).

In 1971, Young and Mark published a paper concerning the fate of the patient with the Eisenmenger syndrome. In 57 patients from 21 months to 58 years, congestive heart failure occurred in 22 of 57 (39%), causing death in 5 patients. Of these 57 patients, 17 died of cardiac causes (30%).

In 1971, Dr. John W. Kirklin wrote an editorial about pulmonary arterial banding in babies with large VSDs in which he stated that he prefers open repair early. In 1965, Kirklin reported VSD repair between 6 and 12 months of age, with a hospital mortality of 5%.

In 1971, Campbell presented the natural history of VSD. Salient features are:

  • 1.

    Spontaneous VSD closure occurs up to the 5th decade of life, that is, 50 years of age;

  • 2.

    20% of VSDs have closed by 30 years of age; and

  • 3.

    24% of VSDs have closed by 60 years of age.

Closure occurs mainly with small VSDs, which occur in approximately one-third of the total. Hence, about two-thirds of VSDs are large. The overall mortality in large VSDs was 25 of 117 (21.4%). In the first two decades (up to 20 years of age), mortality is 2.2% to 2.9%/year. In the fourth and later decades, mortality is double that.

In 1972, Coleman et al reported their results of VSD repair after pulmonary artery banding. First they banded the MPA, and then they closed the VSD 3 to 5 years later. The mortality of the second operation was 4 of 15 patients (27%).

In 1972, Dr. Aldo R. Castañeda became the Chief of Cardiac Surgery at the Boston Children’s Hospital, and I can almost hear him saying, “One operation is better than two. No more palliative surgery. Let’s fix them whenever they need it, no matter how young they are. No problem.” This soon became what we called the Castañeda doctrine.

In 1972, Parisi, Holden, Plauth, and Nadas described the syndrome of VSD with AI based on 72 patients (1950–1971). This was the Boston Children’s Hospital experience when Dr. Alex Nadas was our Chief of Cardiology and Dr. Robert E. Gross was our Chief of Cardiac Surgery—great pioneers both. The median age at onset of AI was 6.9 years, ranging from 0.1 to 20.5 years. Males were 61%, and females were 39%. The median Qp/Qs was 2.1/1.0, ranging from 1 to 5. AI was moderate or severe in 61%, and there was significant pulmonary outflow tract stenosis in 19%. Three patients developed AI 1.4 to 4.2 years after VSD closure. Surgery was performed in 57% of patients, at an average of 4.3 years after the onset of AI. Mortality was 32% (23/72). Surgical deaths were 16; deaths from congestive heart failure occurred in 6 patients; and the cause of death was unknown in 1. Subacute bacterial endocarditis (SBE) occurred in 12.5% (9/72); 7 of 9 underwent surgery (78%), and 5 of 7 died (71%). This study by Dr. Lucy Parisi et al delineates the natural history of VSD with AI and our therapeutic results in this early time period (1950 to 1971).

In 1972, Glasser, Cheitlin, McCarty, Haas, Hall, and Mullins published their experience with VSD and AI in 32 cases. Surgery was performed in 62.5% (20/32). In general, the AI was slowly progressive. A pulmonary outflow tract gradient was found in 41% (13/32). This gradient was due to herniation of the RC leaflet of the aortic valve in 6 of 32 patients (19%) (see Fig. 16.2 , right column ). Of the 20 patients who underwent surgery, none had aortic valve replacement. Annuloplasties were done in 6 of 20 (30%), with VSD closure only in 14 of 20 (70%). Of the 14 patients who had VSD closure only, the AI completely disappeared in 3 (21%). The AI was significantly reduced in 4 of 14 (29%). Hence, the AI was significantly reduced in half (7/14, 50%). These authors recommended early VSD closure to prevent progressive AI. These were the same conclusions that we had reached in 1968.

In 1972, Alpert and Rowe estimated the probability that small VSDs would close spontaneously in the first 5 years of postnatal life. In this clinical study of 52 patients, 30 closed spontaneously (58%).

In 1972, Steinfeld et al published a paper on the diagnosis of subpulmonary (supracristal) VSD. Elective left ventricular angiocardiography was the diagnostic method of choice at that time.

In 1972, Collins, Calder, Rose, Kidd, and Keith published a paper on patients with VSD in the first 5 years of life, based on 200 cases. When the Qp/Qs was less than 2/1, patients were asymptomatic and the rate of spontaneous VSD closure was 20%. Most had a Qp/Qs ratio greater than 2/1. At least 15% die in infancy.

In 1973, Fisher, Brawley, Neill, Donahoo, Haller, Rowe, and Gott published a paper describing severe tricuspid regurgitation after VSD closure.

In 1972, Rothfeld et al described a patient who had a postmyocardial infarction VSD. Then 17 years later, the patient developed the Eisenmenger syndrome, that is, severe pulmonary hypertension with shunt reversal (right-to-left shunt).

In 1973, Tandon and Edwards described “aneurysm-like formations” of the membranous ventricular septum.

In 1973, Kawashima et al from Osaka, Japan described their surgical experience with VSD and AI, based on 35 patients. They had 7 hospital deaths (20%) and 3 late deaths (9%). Since 1968, with 18 consecutive cases, they have had no deaths.

In 1973, Momma, Mimori, and Takao studied the natural history of pulmonary hypertension with VSD in infancy and childhood. They found progressive pulmonary vascular obstruction in 9 of 33 cases (27%).

In 1973, under the leadership of Dr. Pierre Corone and et al, an international meeting was held in France that focused on VSD. Numerous excellent papers were published in Coeur (NS) in 1973. NS is an abbreviation for Numéro Special (Special Number). The papers were published in French, which I, a native Canadian, can read without difficulty. Unfortunately, Coeur (meaning Heart) is no longer being published. In the commentary that follows, I will translate everything into English for the convenience of readers.

Dr. Pierre Corone et al published a paper about the spontaneous evolution of VSDs after the age of 2 years, based on 275 cases. By “spontaneous evolution,” these authors mean what we would call natural history. It is also relevant to mention that Dr. Pierre Corone (a good friend) was basically an adult cardiologist very interested in pediatric cardiology. So his understanding of congenital heart disease in the adult was much better than that of most pediatric cardiologists. In his series, no deaths were related to VSDs. In a series of 181 cases, complete VSD closure occurred in 26 patients (14%). In 181 cases, incomplete closure was observed in 23 patients (13%). The average age of delayed spontaneous closure of VSD was 13½ years of age. More than half of these spontaneous VSD closures occurred after the age of 10 years. Large VSDs can also close spontaneously. Infundibular pulmonary outflow tract stenosis was found in 30% of their large VSDs. An increase in pulmonary resistance was found in only 1 patient.

Dr. Gordon Danielson of the Mayo Clinic in Rochester, Minnesota reported that they were not able to reproduce pulmonary hypertension in dogs experimentally.

In 1973, Dr. Alex Nadas’ presentation at a French meeting had to do with operative indications for VSD surgery and the role of pulmonary arterial hypertension. In 170 infants with a large VSD treated medically, cyanosis and a pulmonary artery pressure greater than 75% systemic were thought to be key factors leading to PVO disease. Among 553 infants treated medically and followed for 4 to 8 years, there were 3 deaths (mortality = 0.5%).

Among 46 patients with VSD and pulmonary hypertension, only those who were operated on at younger than 2 years of age had normal pulmonary arterial resistance (Rp). Operation is thought to be imperative when the Qp/Qs is greater than 2/1, and when the pulmonary artery/systemic artery pressure ratio is greater than 1/3. Operation is thought to be debatable when the Qp/Qs ratio is greater than 2/1, and when the pulmonary artery/systemic artery pressure ratio is less than 1/3. Dr. Nadas said to look for coarctation of the aorta, ASD, or common AV canal. “Debatable” means that the operative indications are unclear, not definite.

Operation for VSD closure is thought to be useless or dangerous when the Qp/Qs ratio is less than 2/1, and the pulmonary artery/systemic artery pressure ratio is greater than 1/3. Operation for VSD closure is unnecessary when the Qp/Qs flow ratio is less than 2/1 and when the pulmonary artery/systemic artery pressure ratio is less than 1/3.

Dr. Richard Bonham-Carter’s view from Great Ormond Street Hospital, London, England (p. 372) was the trial of survival . Operate only if intensive medical therapy has failed and if the infant’s life is in jeopardy. Otherwise, wait and see.

My own view is to remember that there really are no absolute surgical indications for anything. It is all relative. What matters is what one’s surgeon can do successfully. Both physicians are right. Each is trying to express what he thinks his surgeon can do successfully at that time.

Dr. Alex Nadas, surgical indications for VSD closure in infants in 1973:

  • 1.


    • Intractable congestive heart failure

    • Qp/Qs >3/1

  • Pressure in the pulmonary artery/pressure in a systemic artery (P pa /P ps ) >0.75

  • 2.

    With delay of 6 to 12 months

    • Controlled congestive heart failure

    • Failure to thrive

    • Repeated pulmonary infection

    • Persistent hyperkinetic pulmonary artery hypertension

    • PPA, pressure in the pulmonary artery

    • PSA, pressure in a systemic artery

Dr. Nadas’ criteria for inoperability of VSDs in 1973:

    • Arterial oxygen saturations <90%

    • Qp/Qs <1.5/1 and PPA/PSA >0.75

    • Peripheral pulmonary stenoses

    • Inoperable, associated lesions, for example, mitral atresia

Dr. Dupon’s suggestions in 1973 for helping the cardiac surgeon to decide whether to close the VSD:

  • 1.

    Record the pulmonary artery pressure with the patient breathing room air and with the patient breathing oxygen. If the pulmonary artery pressure falls with oxygen, that is a good sign, suggesting that the patient’s elevated pulmonary resistance is labile, not fixed.

  • 2.

    Test band the MPA, and then measure the distal MPA pressure. If the distal pulmonary artery pressure falls, this suggests that the MPA band will do what it is designed to do, namely, to lower the distal pulmonary artery pressure and flow, thereby lowering the pulmonary arterial resistance (Rp).

  • 3.

    Digitally obstruct the VSD. If the RV does not “blow up” and start to fail, this is good, suggesting that the pulmonary arterial resistance is low enough that all the right ventricular blood can flow through the lungs, that is, that a right-to-left shunt through the VSD is not physiologically essential. If a right-to-left shunt through the VSD is essential because of a high and fixed pulmonary arterial resistance, then patch closure of the VSD would be fatal.

Mercier et al presented their experience with 160 VSD closures in 1973 at this meeting in France. The pulmonary artery pressure was greater than 60 mm Hg in 91 of these 160 patients (57%). The operative mortality was 13 of 91 (14%). There were 78 survivors, of whom 33 were restudied. A residual shunt was found in one-fourth of these patients, and the pulmonary artery pressure remained high in 10%.

In the discussion that followed Dr. Mercier’s presentation, Dr. d’Allaines said that if digital test closure of the VSD is negative, that is, if the pulmonary artery pressure does not fall by at least 20%, and if the systemic pressure does not rise significantly, then do NOT close the VSD.

In 42 of 52 patients (81%), the digital test was positive, that is, the pulmonary artery pressure did fall and the systemic pressure did rise. Moreover, after surgical closure of the VSD, the pulmonary artery pressure was the same as it had been with digital test closure of the VSD; in other words, the pulmonary artery pressure obtained after test digital VSD occlusion was the same as the pulmonary artery pressure after surgical VSD closure. So test intraoperative VSD digital occlusion was an accurate predictive test of what the pulmonary artery pressure would be after surgical VSD occlusion. Dr. d’Allaines added that in 4 patients, digital test VSD occlusion was negative (the pulmonary pressure did not fall). These VSDs were closed, and all 4 patients died. In 4 other patients, test digital occlusion was negative (the pulmonary pressure did not fall). The VSDs in these 4 patients were not closed, and all 4 patients survived surgery.

Then Dr. Alex Nadas said he thinks that when the pulmonary arterial resistance (Rp) is high (>10 units), it will fall after operation only in those patients who are younger than 1 year of age at the time of surgery (not older than 1 year of age at the time of surgery).

Dr. Gordon Danielson from the (Mayo Clinic) agreed. He added that this is why banding of the pulmonary artery is still being done at major centers, even though it makes mortality at corrective surgery higher. His final sentence was, “Correction is preferable to banding.” Dr. Danielson’s comment that “this is why banding of the pulmonary artery is still being done at major centers” may require brief explanation. The hope of banding the MPA is that it will reduce or prevent the progressive increase in Pr, thereby making it possible to delay surgical VSD occlusion to later than 1 year of age. Then Dr. Danielson added another important consideration to the conversation: Banding “makes mortality at corrective surgery higher.” This is why banding of the pulmonary artery really is not a good idea. As per Dr. Danielson, “Correction is preferable to banding.” (Now, in 2021, we know that for VSDs, early correction is the best solution, unless the VSD is likely to close spontaneously.)

Then Dr. Neveux proposed that we should rely on hemodynamic studies, not on surgical explorations with digital tests and test bandings. Neveux dismissed these approaches as “acrobatics” that artifact the circulation. Dr. Danielson said that he agrees entirely with Dr. Neveux. Regarding whether to operate, at the Mayo Clinic they rely on the Rp/Rs. When the Rp/Rs is less than 0.7, they operate. “In effect, we close all whose shunt is predominantly left-to-right. But we have established that when the Rp/Rs > 0.7, 25% will die at surgery, 25% will stay unchanged, 25% will get worse, and 25% are improved. Since we can’t tell in advance which 1 of 4 will be benefited, we prefer not to operate on this group at all.”

Pernot described the medical management of infants with VSD.

Galey presented a surgical review of VSDs in infancy. He began by saying that VSDs constitute 20% to 30% of all congenital heart disease, that 1 in 1000 live births has a VSD, and that about half are large VSDs with high flows (a large Qp/Qs ratio). His table of banding mortality, based on published data, varied from a high of 23% (Cooley) to a low of 7% (Binet-Langlois), with a mean of 14%. His table of VSD closure mortality in 1973 revealed the following :

Using Extracorporeal Circulation

McGoon, 94 cases 21%
McGoon, last 42 cases 10%
Sloan, 45 cases 20%

Using Profound Hypothermic (Surface )

Horiuschi 61 cases
Okamoto 62 cases

Using Extracorporeal Circulation and Profound Surface Hypothermia

Barratt-Boyes, 20 cases, 5%
(<1 year) (1 death)
Subramanian 14 cases 7%
(1 death)

Mortality resulting from prior banding at VSD closure was 6% (Subramanian).

Dr. Galey’s Conclusions

The only way to avoid the development of PVO disease is to close all VSDs that need closing at younger than 2 years of age.

Guilmet et al presented their experience with VSD and AI. He began by saying that the first description of VSD with AI was by Pezzi and Laubry in 1921 . This is why the French often call VSD with AI the syndrome of Pezzi and Laubry. Guilmet et al presented 2 cases, both treated successfully. They advocated VSD closure and commissuroplasty, not aortic valve replacement.

Cabral pointed out that the VSD can be closed through the aortic valve. Danielson said that his experience involved approximately 30 cases of VSD with AI: one-third had TOF and two-thirds had a simple VSD only. If the AI is minor, they just close the VSD. As far as aortic annuloplasty is concerned, they have obtained a good result in only about half of their cases. Dr. Danielson concluded that if the AI is severe, they presently close the VSD and replace the aortic valve. Dr. Jean Kachaner et al in 1973 presented their experience with 341 infants with VSD, studied between 1957 and 1965. The methods of diagnosis were:

Cardiac catheterization 154 (45%)
Postmortem examination 20 (6%)
To clinical 167 (49%)

The duration of follow-up was: average, 54/12 years, and range, 6 months to 16 5/12 years.

The results were summarized as follows :

Dead 45/341 13%
Unchanged 161/341 47%
Improved 72/341 22%
Worse 7/341 2%
Closed 39/341 11%
Infundibular stenosis 17/341 5%

Surgery was performed in 154 patients; 67 were operated on twice. Of the 45 infants who died, 44 had pulmonary hypertension (98%). Dr. Kachaner said that the natural history is essentially a function of the type of VSD.

Groups Ia and Ib: Normal Pulmonary Artery Pressure

  • History: the same, or improved

Ia: Maladie de Roger

77 cases 6 Operated (8%)
Mortality 0
Unchanged 55/77 (71%)
Closed 21/77 (27%)
Worse 0
Infundibular stenosis 1/77 (1%)

Group Ib

  • Left-to-right shunt >1.5

  • Pulmonary arterial pressure: Normal

  • n = 51 cases

  • Surgery in 13 of 51 (25%)

  • Surgery twice in 4 of 51 (8%)

Aug 8, 2022 | Posted by in CARDIOLOGY | Comments Off on Ventricular Septal Defects

Full access? Get Clinical Tree

Get Clinical Tree app for offline access