The Surgical Significance of Phenotypic Variability in the Setting of Tetralogy of Fallot

Introduction

Described first 135 years ago as having “la maladie bleue,” ¹ individuals with tetralogy of Fallot still command substantial clinical interest. It is now accepted that the phenotypic feature of the lesion is anterocephalad deviation of the muscular outlet septum, or its fibrous remnant, coupled with abnormal septoparietal trabeculations; these features together producing the characteristic obstruction to pulmonary flow. ² As was commented long since, however, despite the various lesions being unified by this characteristic, no two cases are identical. ³ In terms of surgical correction, which is the only option subsequent to diagnosis, results achieved over the recent decades are an unequivocal success story. ⁴ In centers of excellence, attention is now turned to avoidance of morbidity, while still seeking to eradicate fatalities due to surgical intervention. It is, therefore, perhaps surprising that little attention thus far has been devoted by surgeons to the known phenotypic variability. Having conducted an extensive analysis of the literature devoted to surgical intervention, we found only one investigation that specifically focused on the phenotypic variability. ⁵ Even this study, however, focused largely on the extent of aortic override, and the influence of associated malformations, making no mention of the known variability in the make-up of the right ventricular margin of the interventricular communication, nor the variability in the substrates of subpulmonary obstruction. In this regard, it is also true that, as yet, there is no consensus on the terminologies best used to describe the interventricular communication, nor the substrates for subpulmonary obstruction.^2,6 There is still no general agreement as to whether tetralogy can be considered to coexist with a double outlet ventriculoarterial connection,^7–9 despite the fact that, in his initial account, Fallot described how the aortic root could be supported exclusively above the morphologically right ventricle. There is then no consensus as to whether the individuals with subpulmonary atresia rather than stenosis should be categorized as a subtype of tetralogy, rather than “pulmonary atresia with ventricular septal defect.” The cardiac surgeon, nonetheless, should now be well aware, in advance, of the specific anatomy with which he, or she, will be confronted in the operating room. This was certainly the case in our own institution. In this review, therefore, we have analyzed the significance, if any, of the phenotypic variations as observed in the operating room on the postoperative mortality and morbidities encountered in the experience of one surgeon undertaking surgical correction in over 1,000 individuals. Our aim was to establish whether lessons could be learned regarding the surgical treatment on the basis of the recognized phenotypic variations.

Patients and Methods

Some of us have provided extensive accounts of the phenotypic variability to be found in individuals diagnosed with tetralogy of Fallot on the basis of anterocephalad deviation of the muscular outlet septum or its fibrous remnant, combined with abnormal arrangement of the septoparietal trabeculations. ² These variations in phenotypic anatomy had been noted among 1,012 patients undergoing intracardiac repair at the All India Institute of Medical Science between January 2000 and June 2022. As did Fraser and colleagues in their analysis looking for phenotypic variability, ⁵ we excluded patients having pulmonary atresia with discontinuous pulmonary arteries, along with those having multifocal pulmonary arterial supply with major aortopulmonary collateral arteries requiring primary unifocalization and pulmonary arterial rehabilitation. ⁵ All the patients in our study had undergone cross-sectional echocardiographic interrogation, using also color flow and Doppler techniques. Cardiac catheterization and angiocardiography had been performed in 616 patients. Since 2013, we have used computed tomographic angiocardiography routinely on 396 patients to delineate the anatomic details. Statistical analysis was performed using STATA 16.0 software (Stata Corp). Categorical variables were expressed as in absolute number and percentages. Independent variables were expressed as mean plus or minus standard deviations or median values with the interquartile range. Mortality rates were calculated depending on the total number of years of follow-up. The probability of survival, with 95% confidence intervals was determined with Kaplan-Meier techniques (Figure 1). A P value of less than .05 was considered statistically significant.

OPEN IN VIEWER

Figure 1.

The graph shows the survival probability as calculated using the Kaplan-Meier curve for our cohort.

Results

Study Population and Surgical Details

The age at operation varies from 6 months to 48 years with a median of 29 months, and interquartile ranges of 8 and 13 months. Of the overall group, almost two-fifths of the patients were younger than four years of age at the time of surgical correction, with a further fifth being older than 12 years. One-third of the study population were survivors of previous systemic-to-pulmonary shunt procedures. The characteristics of the overall group, and the details of the surgical procedures, are shown in Table 1.

OPEN IN VIEWER

Table 1.

Phenotypic Variations Noted at Surgery, and Surgical Results.^a

Patient-related variables	Number (%)
Right aortic arch—present	240 (23.7)
Dilated aortic root—present
Preoperative	650 (64.2)
Aortic override
≥50%	420 (41.5)
≤50%	592 (58.5)
Indexed aortic diameter (mm/m2)	27.80 ± 10.11
Aortic regurgitation
Yes	130 (12.8)
No	882 (87.2)
Concordant venticuloarterial connections	592 (58.5)
Double outlet venticuloarterial connections	420 (41.5)
Subtypes of tetralogy
with pulmonary stenosis	912 (90.1)
with absent pulmonary valve syndrome	30 (2.9)
with pulmonary atresia	62 (6.2)
with atrioventricular septal defect	8 (0.8)
Interventricular communication
Solitary	910 (89.9)
Multiple	102 (10.1)
Phenotypic varieties
Perimembranous solitary defect	518 (51.2)
Muscular outlet solitary defect	202 (19.9)
Juxta-arterial defect	182 (18)
Muscular posteroinferior rim (n = 100)
Perimembranous posteroinferior rim (n = 82)
Additional muscular septal defects	102 (10.1)
Inlet septal (n = 65)
Midseptal (n = 37)
In combination with
Perimembranous (n = 60)
Muscular (n = 12)
Juxta-arterial with perimembranous fibrous posteroinferior rim (n = 10)
Juxta-arterial with a muscular posteroinferior rim (n = 20)
Atrioventricular septal defect	8 (0.8)
Surgery
Transatrial, transpulmonary approach	738 (72.9)
Transatrial, transtricuspid approach	274 (27.1)
Transjunctional pulmonary artery patch	670 (66.2)
Intact pulmonary root	280 (27.7)
Right ventricle to pulmonary artery conduit	62 (6.1)
Hospital death
Massive pulmonary hemorrhage (n = 8) [Tetralogy with pulmonary stenosis (n = 2), Tetralogy with pulmonary atresia (n = 6)]	35 (3.5)
Low cardiac output syndrome (n = 16) [Aortic override >80% (n = 16)]
Intractable ventricular arrhythmias (n = 11)
Surgical complete heart block requiring permanent pacemaker	9 (0.9)
Atrioventricular septal defect (n = 1)
Juxta-arterial interventricular communication with perimembranous extension (n = 8)
Late death	10 (0.9)
Ventricular arrhythmias (n = 6)
Late complete heart block (n = 4)

^a

Data are presented as number and percentage.

We used the standard nomogram for measurement of the size of the aortic root at the sinotubular junction, indexed to body surface area and age. ¹⁰ We deemed the root to be dilated when its diameter at the sinotubular junction was 1.5 times greater than expected. Of the overall group, almost two-thirds had dilated roots. Within these individuals, 70 had evidence of trivial to mild aortic regurgitation and did not require surgical intervention. An additional 60 individuals had moderate aortic regurgitation. Of these, 14 underwent reconstruction of the root, while in 11 the aortic valve was replaced concomitant with intracardiac repair. In one patient who had previously undergone aortic valvular reconstruction using Carpentier's technique at the age of ten years, Bentall's procedure was subsequently required after 11 years because of progressive dilation of the root producing severe regurgitation. ¹¹

Although there is a tendency in the current era to proceed directly to corrective surgery whenever possible, we had performed staged palliation in select instances, creating an initial systemic-to-pulmonary arterial shunt in individuals with severe cyanosis, with saturations of oxygen varying between 70% and 75%. ¹² We used standard cardiopulmonary bypass, myocardial preservation techniques, and modified ultrafiltration. The transatrial and transpulmonary route were used in almost three-quarters of our cohort. In the remaining patients, the repair was achieved by operating exclusively through the tricuspid valve. A patch across the ventriculopulmonary junction, the so-called “transannular patch” was required in three-fifths of these patients. We used an extracardiac conduit to reconstruct the pulmonary arterial pathways in 62 patients with pulmonary atresia, or when the right or left pulmonary arteries were severely hypoplastic or interrupted. In planning all these procedures, we measured the diameters of the subpulmonary outflow tract and the pulmonary root using Hegar dilators, calculating the Z-scores from accepted nomograms. ¹⁰

The Phenotypic Feature of Tetralogy of Fallot

The substrate for pulmonary obstruction, specifically the squeeze produced between the malaligned outlet septum and the septoparietal trabeculations, is the phenotypic feature of tetralogy of Fallot. It had previously been suggested that it would be inappropriate to diagnose tetralogy in the absence of a muscular outlet septum. ¹³ The outlet septum, however, is formed developmentally by myocardialization of the proximal parts of the outflow cushions (Figure 2A). If development proceeds normally, the entirety of the pulmonary component of the myocardializing cushions becomes the free-standing subpulmonary infundibulum (Figure 2B). ¹⁴ In the setting of tetralogy, in contrast, the proximal parts of the cushions become the malaligned outlet septum, with their distal parts producing the free-standing subpulmonary infundibular sleeve (Figure 2C). Should the cushions not myocardialize, then the malaligned septum can be fibrous, justifying this inclusion of those with juxta-arterial defects. Malalignment in itself, however, is insufficient to produce the overall phenotype, since comparable malalignment is to be found in the Eisenmenger ventricular septal defect. Malalignment combined with abnormal formation of the septoparietal trabeculations, which are usually hypertrophied, is required to produce the full phenotypic feature (Figures 3A), which is also found in those with the juxta-arterial defects (Figure 3B). The phenotypic features, along with the overriding of the aortic root, are now well displayed in the clinical setting (Figure 3C and D; Figure 4).

OPEN IN VIEWER

Figure 2.

The images show the date of the proximal parts of the outflow cushions in the developing heart. (A) The arrangement in a human embryo at approximately seven weeks of development. The proximal cushions have fused with each other, and with the muscular ventricular septum to commit the aortic root to the left ventricle. They are beginning to muscularize. In the normal heart, as shown in (B), the cushions form the subpulmonary infundibular sleeve. In tetralogy of Fallot, as shown in (C), their proximal part persists as the muscular outlet septum, while the distal part continues to produce an infundibular sleeve.

OPEN IN VIEWER

Figure 3.

The images show the detail of the subpulmonary squeeze (bracket). (A and C) The arrangement when the outlet septum is muscular, as seen in an autopsied specimen (A) and a computed tomographic dataset (C). (B and D) The features when the outlet septum is fibrous.

OPEN IN VIEWER

Figure 4.

The images show the triangular space subtended beneath the aortic root as seen in an autopsied specimen (A), and computed tomographic angiograms from patients with the aortic root mostly committed to the left ventricle (B) and supported predominantly by the right ventricle (C).

Variations in the Borders of the Interventricular Communication

As has been emphasized, no two cases of tetralogy of Fallot are identical. ³ The major variants are to be found in the borders and extent of the interventricular communication. It is these features, however, that have engendered most disagreements regarding terminology. The area in question was described by Fallot as the “communication interventriculaire.” ¹ It is now more frequent to find it described as a “ventricular septal defect.” ⁵ It is the entirety of the space subtended beneath the overriding aortic root, however, that provides the channel for interventricular shunting. Any chosen plane within this space (Figure 4A) can justifiably be considered to represent either a “ventricular septal defect” or an “interventricular communication.” From the stance of surgical repair, the leftward entrance to the space is the outflow tract for the left ventricle (Figure 4B and C). The essence of surgical correction, irrespective of the extent of aortic override, is to close the entrance to the aortic root from the right ventricle (Figure 4B and C). The surgeon, therefore, needs to appreciate the structures surrounding this right ventricular entrance. It is difficult for the echocardiographer to show the entirety of the border in the same image, but its margins can be demonstrated by careful sweeping across the aortic root. The view obtained by the morphologist (Figure 5A), however, can now be directly replicated using virtual dissections of computed tomographic datasets (Figure 5B). In the majority of individuals, part of the border is made up of fibrous continuity between the leaflets of the mitral and tricuspid valves, making the defect perimembranous (Figure 5A). In a smaller proportion, the posteroinferior border is myocardial, protecting the conduction axis (Figure 6A and B and D).The third, or the juxta-arterial, variant is found when the outlet septum is fibrous. In the majority of these defects, the posteroinferior border remains myocardial (Figure 6D). In a minority of cases, nonetheless, the juxta-arterial defect can extend to become perimembranous, placing the conduction axis at greater risk (Figure 6C).

OPEN IN VIEWER

Figure 5.

The images compare the view of the aorto-right ventricular communication (white dashed oval) as seen in an autopsied specimen and a computed tomographic dataset. The defect is perimembranous in both examples, with the red line showing the location of the atrioventricular conduction axis in its posteroinferior angle. The green lines show the location of the septomarginal trabeculation, or septal band. The bracket in (B) shows the squeeze at the mouth of the subpulmonary infundibulum.

OPEN IN VIEWER

Figure 6.

The images show the features of the muscular defect (A and C) and the juxta-arterial defect (C and D). The defect shown in (C) also extends to become perimembranous. (B and D) Virtual dissections of computed tomographic datasets. The red line shows the location of the atrioventricular conduction axis.

Short- and Long-Term Outcomes

Of our patients, 35 (3.5%) died prior to discharge from the hospital. The deaths were related to massive pulmonary hemorrhage in eight patients, intractable ventricular arrhythmias in 11, and low cardiac output syndrome with multiorgan failure in 16. Approximately three-fifths of our overall cohort had suffered postoperatively with low cardiac output. Those who died with this complication were all more than 12 years of age. They had all exhibited aortic override greater than 80%. Among those dying due to massive pulmonary hemorrhage, six individuals had atresia rather than pulmonary stenosis. All had undergone coil embolization of aortopulmonary collaterals. Transient heart block was seen in about one-tenth of the cohort. They had required temporary pacing for periods ranging from 24 to 120 h. In one patient with an atrioventricular septal defect, and eight individuals with juxta-arterial defects extending to become perimembranous, permanent pacing was needed due to complete heart block (Table 1). None of the individuals with a juxta-arterial defect and muscular posteroinferior rim exhibited any postoperative rhythm problems. Low cardiac output syndrome had been present in approximately three-fifths, who required standard inotropic support. Successful intra-aortic balloon counterpulsation was necessary in four adults. In eight patients, we needed to revise the transjunctional patch, while in an additional nine patients it was necessary to resect more aggressively the subpulmonary infundibulum.

An additional ten patients died later, due to ventricular arrhythmias in six, and late onset complete heart block in four between 70 and 194 months following corrective surgery. Patients dying of ventricular arrhythmias late postoperatively were 4 years, 12 years, 25 years, 28 years, 29 years, and 35 years of age, respectively, at the time of corrective surgery. All patients had a transjunctional patch between the right ventricle and pulmonary trunk, moderate pulmonary regurgitation and biventricular dysfunction with ejection fraction between 0.30 and 0.40. Although five out of six patients with repaired tetralogy dying of ventricular arrhythmias were late presenters, the relationship of older age and fatal arrhythmia did not achieve statistical significance (P = .41), possibly because of the fewer events. It may well be that these patients died because of the coexistence of both ventricular arrhythmias and ventricular dysfunction.

To assess progress in our own center, we compared the results in the cohort of 563 individuals undergoing surgery in the initial decade of the experience of the operating surgeon with the 449 individuals corrected in the second decade. We found no difference in the phenotypic features of the individuals coming forward for correction in the two decades. Although there was reduction of mortality from 3.5% to 3.3% in the recent cohort, this did not prove to be statistically significant (P = .86).

Of our overall cohort, 967 survivors underwent follow-up by clinical examination, electrocardiography, and echocardiography every six months. We noted their functional class, their need for cardiac medications, and any development of arrhythmias, ventricular dysfunction, congestive heart failure, or late complications. If six-monthly evaluation was not possible after repeated attempts to contact the patients, the data were deemed to be missing. If two consecutive evaluations were missing, the patient was considered lost to follow-up. In all, we lost 8 (0.8%) patients to follow-up. Thus, the follow-up has been virtually complete, yielding 12,392.91 patients-years of data. At a median follow up of 156 months, with an interquartile range of 95 to 214 months, the probability of 22-years survival was 95.49% ± 0.006% with 95% confidence interval from 0.94 to 0.96 (Figure 1). At their last follow up, nine-tenths of the patients were in the first class of the New York Heart Association, with the others in the second class.

Left ventricular ejection fraction was greater than 0.5 in 748 patients, with mild tricuspid regurgitation found in 40. In those patients requiring a transjunctional patch, 602 had mild, and 62 had moderate pulmonary regurgitation according to the prescribed echocardiographic criteria.^21,22 During late follow-up, five patients have needed additional surgery to relieve right ventricular outflow tract obstruction in three, and occlusion of the left pulmonary artery in two. There were no reoperations for residual ventricular septal defect. The pulmonary valve has been replaced with a bioprosthesis in 15 patients, all initially with pulmonary stenosis. Revision of their conduit was needed in 32 patients with pulmonary atresia. Another nine patients are awaiting conduit revision. An additional individual underwent a successful Bentall's procedure. ¹¹

Comment

The essence of tetralogy is the separation of the three building blocks of the normal subpulmonary outflow tract, each retaining its own individuality. ²³ Anterocephalad deviation of the outlet septum, or its fibrous remnant, the most characteristic phenotypic feature, is insufficient in itself to produce the abnormal pathology. It is also seen in the Eisenmenger ventricular septal defect. ²⁴ The other essential feature is hypertrophy of the abnormal septoparietal trabeculations. Marked additional individual variations of potential surgical significance are then found in the majority of patients. Of these, perhaps the most significant is the area closed by the surgeon to restore septal integrity. This area is the right ventricular entrance to the overriding aortic root (Figure 4). A significant number of individuals with tetralogy, however, have the larger part of their aortic root supported by the right ventricle, in other words a double outlet ventriculoarterial connection. Fallot had described such an arrangement in his original account. ¹ In this setting, however, it is the channel between the ventricles that currently is described as the “ventricular septal defect.” There can be no question but that it is a “ventricular septal defect.” It is also the outflow tract for the left ventricle. The area, therefore, rather than being closed during surgical repair, is baffled to the aortic root. In all other circumstances, however, it is the right ventricular entrance to the aortic root that is usually described as the “ventricular septal defect.” To avoid any ambiguity, we have distinguished between the two areas as the aorto-right and aorto-left ventricular communications. It is the aorto-right ventricular communication that is closed to restore septal integrity. As we have suggested previously, ⁹ the procedure used by the surgeon during the operative procedure provides a means of resolving the ongoing conundrum of how to define double outlet ventriculoarterial connections when there is overriding of an arterial root. If the surgeon considers that he, or she, has “closed the defect,” then the patient must initially have had concordant ventriculoarterial connections. If, in contrast, the procedure is considered to be “baffling to the aortic root,” then the ventriculoarterial connection must have been one of double outlet.

In Western countries, in about four-fifths of individuals, the aorto-right ventricular communication is perimembranous. In our cohort, however, only about three-quarters had perimembranous defects. Often times in these patients, the borders of the defect are reinforced by a membranous fibrous flap. We noted such membranous flaps in approximately three-fifths of our patients. In the remaining one-fifth of Western cohorts, but in about one quarter of our patients, the presence of a myocardial posteroinferior margin makes the defect muscular. Both the perimembranous and muscular outlet variants are described by others as being conal malalignment defects. ²⁵ The defects can be difficult to close surgically when approaching through the tricuspid valve, particularly if the septoparietal trabeculations are not fully mobilized and resected. The atrioventricular conduction axis penetrates in their posteroinferior margin and is at greater risk when the defect is perimembranous. ²⁶ Having penetrated, the bundle is usually carried away from the septal crest. This is not always the case. The branching bundle may be prone to surgical injury from sutures placed directly on the septal crest.^26,27 When the posteroinferior margin is muscular rather than fibrous, nonetheless, it can be safe to take superficial bites along with all the borders of the defect. It is safer, however, to place interrupted and pledgeted mattress sutures around the base of the septal leaflet of the tricuspid valve. Apart from the crucial posteroinferior region, the remainder of the margins of the defect is free from conduction tissue. The sutures placed through the ventriculoinfundibular fold should be shallow so as to avoid damage to the branches of the right coronary artery running within the inner heart curvature. By the same token, extreme care should be exercised to avoid excessive trimming of the fold to prevent iatrogenic aortic regurgitation. ²⁸

It is unusual to find the juxta-arterial defect in Western populations, although such defects are frequent in the Far East and Latin America, especially Central America.^29–31 We found these defects in almost one-quarter of our cohort. Their phenotypic feature is the presence of a fibrous, rather than a muscular, outlet septum. At one stage, we had argued that such defects should not be considered as part of tetralogy of Fallot. ¹³ We now know that it is failure of muscularization of the primordium of the outlet septum that produces the juxta-arterial defect. ¹⁴ The defects usually have a myocardial posteroinferior rim but can extend to become perimembranous. In most instances, those with a muscular posteroinferior rim can be closed through the pulmonary trunk. In those with perimembranous extension, sutures are more readily placed using the transtricuspid approach. In either instance, particular care is needed at the posteroinferior angle. The superior portion of the patch should be secured at the base of the pulmonary root, thus avoiding retraction to the leaflets of both arterial valves. The presence of an additional muscular defect is concerning. When such a defect opens to the right ventricular inlet, only a small muscle bundle remains to carry the conduction axis. Closure using separate patches during infancy then runs the risk of producing complete heart block. Because of this, closure has been recommended using a single all-encompassing patch. ²⁸ Neonates, infants, and children with hemodynamically significant inlet multiple septal defects with less than 4 mm of space between the defect and the atrioventricular or arterial valves are not recommended for device closure by the American Heart Association.^32,33 The inlet extension of the malaligned muscular septum found with straddling of the tricuspid valve is also a formidable surgical challenge. In the solitary patient in our cohort with this variant, we were able to retract the straddling tension apparatus into the right ventricle and close the defect. The patient, nonetheless, did require a salvage superior cavopulmonary connection to ensure survival. Another challenge is to find the phenotypic features of tetralogy in the setting of a common atrioventricular junction. This was found in less than 1% of our patients, all of whom survived, although one required permanent pacing because of complete heart block.

It is the extent of aortic override that remains the most contentious point with regard to the description of tetralogy. As shown by our experience, a spectrum exists with the aorta almost exclusively connected either to the left or the right ventricle. Previous arguments in this respect have devolved upon the definition used for double outlet right ventricle. ⁸ For quite some time, the presence of bilateral infundibulums was considered the phenotypic feature. ⁷ It is now accepted that such an approach abrogates the “morphological method,” which states that one variable feature should not be used to define another feature that is itself variable. The question then remains as to whether more than half, or more than nine-tenths, of the aortic root should be supported by the right ventricle to justify the diagnosis of double outlet? As we have indicated, this debate can now be defused by the surgeon determining whether he, or she, has “closed the septal defect” or “baffled it to the aortic root.” ⁹

It is the “squeeze” produced by the deviated muscular outlet septum, or its fibrous remnant, and the hypertrophic septoparietal trabeculations that are the phenotypic feature of tetralogy (Figure 3). A clear understanding of these structures requires equally clear descriptive terminology. The attachments of the outlet septum itself can simply be considered parietal and septal. The various types of subpulmonary obstruction found in tetralogy are readily described using these terms. The outlet septum is usually of good size, but is hypoplastic and fibrous when the defect is juxta-arterial. Hypertrophy of its septal attachment, and the septoparietal trabeculations, is almost always present, with the obstruction often exacerbated by formation of a subvalvar fibrous shelf. The pulmonary root itself is also usually hypoplastic, frequently with a bicuspid valve. Further narrowing can be found at the sinotubular junction, or within the pulmonary arteries, particularly at the site of origin of the arterial duct. The finding of rudimentary leaflets, as observed in 30 of our patients, produces the so-called “absent valve syndrome.” The arterial duct is usually absent in this setting, but the main feature of this complication is the aneurysmal dilation of the pulmonary arteries, producing compression of the tracheobronchial tree. Those presenting with this feature as neonates are usually ventilator-dependent and require early intervention. Later surgical intervention, as in our cohort, is possible on those with less severe symptoms.

Just as no two individuals with tetralogy have identical hearts, ³ no two surgeons perform the corrective surgery in an identical fashion. Despite surgical experiences now spanning over seven decades, controversies continue regarding optimal management. The surgical community is still debating which operation to do, for what sort of patient, and when. Questions remain, particularly in centers such as our own, as to whether neonates and infants with recurrent cyanotic spells below a certain age require palliation prior to complete repair, or should have primary repair regardless of size or age? And should those not undergoing cyanotic spells be repaired at diagnosis, or when aging and weight gain may lead to improved outcomes? Further questions relate to the optimal approach to rehabilitation of the right ventricular outflow tract, and the acceptable upper limit of postoperative ratio of peak systolic right-to-left ventricular pressures. We make no claims to have answered these questions. Surgical management, nonetheless, has evolved with time. ⁴ It is customary in the current era to proceed directly to surgical correction. Staged palliation, nonetheless, remains a requirement in select instances, proving effective in our hands. With regard to rehabilitation of the pulmonary outflow tract, we performed a limited transjunctional patch in about two-thirds of the patients. In about half of these, postoperative pressure ratios were greater than 0.7. We deemed this acceptable, since our myocardial resection and patching had produced an outflow tract of the recommended diameter. Most of our patients, furthermore, had presented late, with hypertrophied myocardium. Within the current era of elective neonatal or infant intervention, excellent results are reported for the transatrial–transpulmonary approach and complete repair. ⁵ Some use the valve-sparing options in up to four-fifths of their patients. Although this protocol may decrease ventricular arrhythmias, and improve right ventricular function, higher incidences of residual outflow tract obstruction have also been reported, requiring increased rates of reintervention.

Computed tomographic angiography, especially when electrocardiographically gated, has an excellent spatial resolution with isometric voxels, which allows for three-dimensional visualization of the intracardiac defects using surface rendered and endoluminal reconstructions in addition to the routine multiplanar reformatted images. Echocardiography is frequently limited by availability of optimal acoustic windows, and by its known blind areas. Evaluation of the coronary arteries is also suboptimal when using echocardiography, as is, to some extent, the evaluation of pulmonary and systemic venous anomalies. Computed tomography allows for excellent visualization of all these features, as well as identifying major systemic-to-pulmonary collateral arteries. The technique additionally gives information about the airways and lung parenchyma. The advantages obtained far outweigh the cost of the investigation.

Limitations of the Study

We recognize that our review is a retrospective evaluation and has not addressed all the phenotypic variations. For example, we considered only those with pulmonary atresia with arterial supply via the persistently patent arterial duct. None of these patients required unifocalization of systemic-to-pulmonary arteries. We have also not discussed the significance of abnormalities of the coronary arteries, although we took care to identify such problems.

In this investigation, furthermore, we recognize that we did not specifically correlate the morphological findings with the age of our patients. It is noteworthy that two-thirds of our patients exhibited aortic dilation and regurgitation. Since we had previously examined the relationship of variables, such as older age, the degree of aortic override, previous palliative shunts, low lamellar count, abnormal aortic histopathology, and the occurrence of Fibrillin-1 gene polymorphisms/mutations to aortic dilation and postoperative remodeling, we have not repeated those analyses. In a retrospective evaluation with very late term follow-up, it is difficult to control other variables for logistic regression analysis that could impact mortality and/or heart block. Again, since we had validated the diagnostic accuracy of cardiac magnetic resonance imaging in evaluating biventricular ejection fraction and quantified pulmonary regurgitant fraction in 280 survivors of repaired tetralogy within our studied population, we have not repeated that analysis. ²² Due to the limitations of space, we have chosen to focus on the major intracardiac variations.

Conclusions

Although our institution is a tertiary-level center, we recognize that our experience is not necessarily representative of that now expected for specialist centers in the Western world. The major evolution in the “Western world” took place over the last two decades of the 20^th century and essentially involved a shift to primary repair during infancy. The median age of our patients was 29 months, with a range between 6 months and 48 years. Of the cohort, however, one-fifth of the patients were older than 12 years. We created systemic-to-pulmonary arterial shunts in one-third. Late presentation is not unusual in the developing world. Indeed, it is not uncommon to be confronted by adult patients with tetralogy physiology who have not undergone previous palliation. Such patients are cyanotic and polycythemic and are different from the typical populations undergoing surgery in the Western world. It is the socioeconomic profile of the patients, and their lack of health insurance that usually leads to delayed referral, and hence late surgical intervention. By their very survival, it is arguable that our patients represented the most favorable spectrum of the phenotypic variants. Their myocardium, nonetheless, had been subjected to long-standing cyanosis and pressure overload. In previous investigations, furthermore, we had described a preponderance of degenerative changes, such as myocytolysis and perivascular fibrosis in these late presenters, who have higher postoperative peak ratios of systolic right-to-left ventricular pressures, and perioperatively suffer low cardiac output.^20,34 We speculate that the higher incidence of ventricular arrhythmias and myocardial dysfunction may be due to late presentation at the time of corrective surgery. The importance of coexisting ventricular arrhythmias and heart block with ventricular dysfunction as a cause of death has been highlighted by previous investigators and the American Heart Association Scientific Statement.^35,36

Despite these constraints, we have been able to identify all the phenotypic variations and our results are encouraging. Of the phenotypic features assessed, origin of more than half of the aortic root arising from the right ventricle was associated with increased risk of death, whereas the finding of a juxta-arterial defect with perimembranous extension was more likely to be associated with rhythm problems. Other than these features, we did not find other phenotypic features to be associated with increased mortality or morbidity, but they did require attention to the particular fashion in which the subpulmonary obstruction was relieved, and the aorto-right ventricular communication was closed. When using multimodality imaging, it is now possible to characterize and delineate the variations in phenotypic morphology that previously were the province of the anatomist. These details can all be ascertained by the observant surgeon. We predict that it will be attention to the significance of such variations that will guide optimal future surgical management, not only in developing countries but also in the Western world.

Supplemental Material