**Restoration of Transposed Great Arteries With or Without Subpulmonary Obstruction to Nature**

Ing-Sh Chiu and Chi-Ren Hung

*Department of Surgery, National Taiwan University Hospital and National Taiwan University, College of Medicine, Taipei, Taiwan* 

*Surgical correction on transposition of the great arteries was proposed by many in the past half-century, and was claimed as the anatomical correction; but the treatment of choice was ever changing. The current technique usually included Lecompte maneuver to bring the pulmonary bifurcation in front of the aorta. Although the ventriculoarterial connection was corrected, it is not "normal" yet---------.* 

#### **1. Introduction**

Looking back into the evolution of surgical treatments on transposed great arteries (TGA), is full of fascinating and challenging stories. Many pondered the best option to correct this tricky, yet not the most complex congenital heart disease. Some operations that had been replaced by another were later revived. Senning once had been replaced by Mustard as the treatment of choice (Senning, 1959; Mustard, 1964), but was revived because autologous tissue was utilized, although it was more difficult (Quaegebeur, 1977). As people learned the functional implications of the ventricles, both atrial redirection procedures were replaced by the arterial switch operation (ASO) (Jatene et al., 1975; Lecompte et al., 1981; Castaneda et al., 1984). ASO was attempted initially, without transferring the larger right coronary artery, by Mustard (Mustard et al., 1954). Nikaidoh described aortic translocation, which in essence is an ASO including the arterial valve (Nikaidoh, 1984). In the past, few practiced this procedure because of its demanding techniques and potentially worse outcome. However, aortic translocation has recently gained popularity as an alternative to the Rastelli operation and Reparation l'etage ventriculaire (REV) as the treatment for TGA with a left ventricular outflow obstruction (Rastelli, 1969; Yeh et al., 2007; Emani et al., 2009). ASO has become the procedure of choice for TGA (Prêtre et al., 2001; Losay et al., 2002). However, TGA is considered to be a mere reversal of the great arteries anteroposteriorly (Shaher, 1964; Van Mierop, 1971); nonexistence of the normal spiral relationship of the great arteries in TGA has not been widely appreciated. Thus posterior pulmonary bifurcation is mobilised anteriorly to the aorta (the so-called Lecompte maneuver) in an effort simply to reverse transposed great arteries (Lecompte et al., 1981). We proposed an arterial Senning operation 13 years ago, to restore the spiral flow of nature in TGA (Chiu et al., 2000b; Chiu et al., 2002b, Chiu et al., 2010). The role of this operation is still uncertain and it requires continued refinement and development. The thinking process and evolving technique behind how we conceived our currrent technique was published (Chiu et al., 2001). Briefly, mobilization of the pulmonary arteries high above its original site to avoid compression the high take off coronary artey will result in supravalvular pulmonary stenosis (PS), similarly, Lecompte

Restoration of Transposed Great Arteries With or Without Subpulmonary Obstruction to Nature 171

since the outflow tract was not switched and did a spiral ASO offer any help by just restoring the spiral arterial trunks without altering outflow tracts below the valves? Although the ventricular outflow was not switched in spiral ASO, the facing commissure of the old aorta was mobilized along with a topmost outflow below it and fixed it to the direction of future MPA (Figure 1, solitus normal). In other words, by our technique the top portion of the right ventricular outflow just below the facing commissure is rotated toward the direction of future MPA (Chiu et al., 2000b; Chiu et al., 2002b; Chiu et al., 2010). As just clarified, when the aorta is left-sided with usual atrial arrangement, there CAN be spiraling of the outflow tracts; the outlet septum is vertical to the rest of the ventricular septum in the setting with a directly anterior or left anterior old aortic root (Chiu et al., 1984; Anderson & Weinberg, 2005), thus a slight rotation like our technique is adequate for spiral correction at the ventricular outlet in that group. Wheras in side-by-side gruop, Lecompte maneuver is usually not performed by all, the ventricular outflows are in parallel as they are. When the aortic root is right anterior to MPA, we fix the facing commissure to the right anterior aspect of old MPA (Chiu et al., 2000b). The result is as the inverted normal in Figure 1. We did not fix the facing commissure of the new MPA to directly anterior aspect of new aortic root or leave the new MPA alone without fixation, as in conventional ASO, which corrects the ventriculo-arterial discordance without modifying the outflows. *Restoring the natural curvature of both great vessels is the key to avoiding obstruction at the ventricular outflows,* 

Embryologically, the coronary arteries (CA) develop after septation of the aortopulmonary trunk and pierce the aortic sinus at the nearest site after aortopulmonary rotation. We found that the pattern of the CA is dependent on aortopulmonary rotation, thus proposed a new categorization scheme based on the aortic root rotation (Chiu et al., 1995). In addition to the short-axis rotation, which is related to the juxtacommissural origin of the CA (JOCA) (Chiu et al., 1997), there is also a long-axis rotation, which is related to high takeoff of the CA (Chiu et al., 1996b). On the basis of these findings, we have also proposed appropriate diagnostic and surgical techniques to manage unusual CA patterns in TGA (Chiu et al., 1997; Chiu et al., 1996a). In short, for JOCA near the facing commissure, a superiorly based trapdoor (single-button technique) or lateral funnel (two-button technique), and near the nonfacing commissure, a medially based trapdoor, are vital for coronary redirection. The categorization of CA based on the aortopulmonary rotation can be applied to all congenital heart defects and normal hearts (Chiu et al., 2000a; Chiu et al., 2002a; Chiu et al., 2003; Chen

Paillole et al. documented the existence of central pulmonary artery (PA) hypoplasia before surgery and its persistence until after ASO (Paillole et al., 1988). Central PA hypoplasia, which is frequently seen in TGA, is related to posterior inclination of the proximal MPA in this setting (Figure 2, Chen et al., 2007). We have demonstrated that the same pathogenesis would lead to smaller PA size after a Senning procedure than ASO with Lecompte maneuver. In addition, Lecompte with Pacifico's modification (direct connection between the distal MPA and the aortic sinus defect), which mobilized the proximal MPA toward anterior aspect farther than by the patch-repair of the sinus defect, could facilitate better PA growth after ASO (Pacifico et al., 1983; Imoto et al., 1995). Spiral ASO would be even better

to block that pathogenesis than the above to permit PA growth (Chiu et al., 2010).

*instead of just reverse them anteroposteriorly.*

et al., 2007; Huang et al., 2011; Chiu et al., 2011).

**2.2 Coronary artery** 

**2.3 Pulmonary artery** 

maneuver mobilized pulmonary arteries away from its original position. To avoid the complications of unnecessary mobilization, an *in situ* transfer technique and the common wall concept to redirect the coronary arteries should also be applied to the pulmonary arteries (Chiu et al., 2000b; Chiu et al., 2001). Tissue deficiency of the pulmonary outflow tract in cyanotic cardiac defects could be recruited from the larger aorta to compensate for the smaller main pulmonary artery (MPA); or vice versa, from the big MPA to a small aorta. We called this the Robin Hood approach or, in the current vernacular, a redistributive approach (Chiu et al., 2009). In this review, we describe an innovative technique to reconstruct the great arteries in spiral fashion, which is the natural relationship of aorta and pulmonary artery. The structural and functional studies underlying the basics of natural spiral great arteries we published in the last two decades will be presented. We emphasize the surgical principles of nature and even distribution, using autologous tissues. Anatomical features will be discussed first, then the technique of the above two redistributive approaches in TGA and finally their three-dimensional CT follow-up results.

#### **2. Anatomical features**

Transposition of the great arteries is the consequence of distal fusion between the dorsal protrusion of the aortic sac and the outflow cushions during embryogenesis (Figure 1) (Van Praagh, 2010). The conal septum and great arteries remain straight and parallel instead of being torsaded around each other. As a consequence, the aorta emerges from the right ventricle and the MPA from the left ventricle.

Fig. 1. There are only two ways in which conotruncal rotation can be done correctly as in: (1) solitus normally related great arteries and (2) inversus normally related great arteries; in contrast to many other ways that it can be done wrong.

#### **2.1 Outlet septum**

The ventricular outflows are in parallel when there is a side-by-side great arteries. When the aortic root is directly anterior or even left anterior, the outlet septum is no longer in parallel, it is vertical to the rest of the ventricular septum (Chiu et al., 1984). It is often asked that since the outflow tract was not switched and did a spiral ASO offer any help by just restoring the spiral arterial trunks without altering outflow tracts below the valves? Although the ventricular outflow was not switched in spiral ASO, the facing commissure of the old aorta was mobilized along with a topmost outflow below it and fixed it to the direction of future MPA (Figure 1, solitus normal). In other words, by our technique the top portion of the right ventricular outflow just below the facing commissure is rotated toward the direction of future MPA (Chiu et al., 2000b; Chiu et al., 2002b; Chiu et al., 2010). As just clarified, when the aorta is left-sided with usual atrial arrangement, there CAN be spiraling of the outflow tracts; the outlet septum is vertical to the rest of the ventricular septum in the setting with a directly anterior or left anterior old aortic root (Chiu et al., 1984; Anderson & Weinberg, 2005), thus a slight rotation like our technique is adequate for spiral correction at the ventricular outlet in that group. Wheras in side-by-side gruop, Lecompte maneuver is usually not performed by all, the ventricular outflows are in parallel as they are. When the aortic root is right anterior to MPA, we fix the facing commissure to the right anterior aspect of old MPA (Chiu et al., 2000b). The result is as the inverted normal in Figure 1. We did not fix the facing commissure of the new MPA to directly anterior aspect of new aortic root or leave the new MPA alone without fixation, as in conventional ASO, which corrects the ventriculo-arterial discordance without modifying the outflows. *Restoring the natural curvature of both great vessels is the key to avoiding obstruction at the ventricular outflows, instead of just reverse them anteroposteriorly.*

#### **2.2 Coronary artery**

170 Front Lines of Thoracic Surgery

maneuver mobilized pulmonary arteries away from its original position. To avoid the complications of unnecessary mobilization, an *in situ* transfer technique and the common wall concept to redirect the coronary arteries should also be applied to the pulmonary arteries (Chiu et al., 2000b; Chiu et al., 2001). Tissue deficiency of the pulmonary outflow tract in cyanotic cardiac defects could be recruited from the larger aorta to compensate for the smaller main pulmonary artery (MPA); or vice versa, from the big MPA to a small aorta. We called this the Robin Hood approach or, in the current vernacular, a redistributive approach (Chiu et al., 2009). In this review, we describe an innovative technique to reconstruct the great arteries in spiral fashion, which is the natural relationship of aorta and pulmonary artery. The structural and functional studies underlying the basics of natural spiral great arteries we published in the last two decades will be presented. We emphasize the surgical principles of nature and even distribution, using autologous tissues. Anatomical features will be discussed first, then the technique of the above two redistributive

Transposition of the great arteries is the consequence of distal fusion between the dorsal protrusion of the aortic sac and the outflow cushions during embryogenesis (Figure 1) (Van Praagh, 2010). The conal septum and great arteries remain straight and parallel instead of being torsaded around each other. As a consequence, the aorta emerges from the right

Fig. 1. There are only two ways in which conotruncal rotation can be done correctly as in: (1) solitus normally related great arteries and (2) inversus normally related great arteries; in

The ventricular outflows are in parallel when there is a side-by-side great arteries. When the aortic root is directly anterior or even left anterior, the outlet septum is no longer in parallel, it is vertical to the rest of the ventricular septum (Chiu et al., 1984). It is often asked that

approaches in TGA and finally their three-dimensional CT follow-up results.

**2. Anatomical features** 

ventricle and the MPA from the left ventricle.

contrast to many other ways that it can be done wrong.

**2.1 Outlet septum** 

Embryologically, the coronary arteries (CA) develop after septation of the aortopulmonary trunk and pierce the aortic sinus at the nearest site after aortopulmonary rotation. We found that the pattern of the CA is dependent on aortopulmonary rotation, thus proposed a new categorization scheme based on the aortic root rotation (Chiu et al., 1995). In addition to the short-axis rotation, which is related to the juxtacommissural origin of the CA (JOCA) (Chiu et al., 1997), there is also a long-axis rotation, which is related to high takeoff of the CA (Chiu et al., 1996b). On the basis of these findings, we have also proposed appropriate diagnostic and surgical techniques to manage unusual CA patterns in TGA (Chiu et al., 1997; Chiu et al., 1996a). In short, for JOCA near the facing commissure, a superiorly based trapdoor (single-button technique) or lateral funnel (two-button technique), and near the nonfacing commissure, a medially based trapdoor, are vital for coronary redirection. The categorization of CA based on the aortopulmonary rotation can be applied to all congenital heart defects and normal hearts (Chiu et al., 2000a; Chiu et al., 2002a; Chiu et al., 2003; Chen et al., 2007; Huang et al., 2011; Chiu et al., 2011).

#### **2.3 Pulmonary artery**

Paillole et al. documented the existence of central pulmonary artery (PA) hypoplasia before surgery and its persistence until after ASO (Paillole et al., 1988). Central PA hypoplasia, which is frequently seen in TGA, is related to posterior inclination of the proximal MPA in this setting (Figure 2, Chen et al., 2007). We have demonstrated that the same pathogenesis would lead to smaller PA size after a Senning procedure than ASO with Lecompte maneuver. In addition, Lecompte with Pacifico's modification (direct connection between the distal MPA and the aortic sinus defect), which mobilized the proximal MPA toward anterior aspect farther than by the patch-repair of the sinus defect, could facilitate better PA growth after ASO (Pacifico et al., 1983; Imoto et al., 1995). Spiral ASO would be even better to block that pathogenesis than the above to permit PA growth (Chiu et al., 2010).

Restoration of Transposed Great Arteries With or Without Subpulmonary Obstruction to Nature 173

We have to take down the "Lecompte maneuver" in a patient eight years after conventional

The aorta is not the only structure behind the MPA bifurcation after Lecompte maneuver, the left main bronchus (airway-pressure with cartilage) is also present. Bronchial compression or even atelectasis has been reported after Lecompte maneuver (Robotin et al., 1996; Toker et al., 2000), although in the majority of cases the bcronchial patency was not so

Fig. 4. After conventional ASO with Lecompte maneuver, not only supra-aortic stenosis occurs at the site posterior to the anteriorly mobilized pulmonary bifurcation (\*), but also the left main brobchus, that is cartilage stented, is ccompreesed; its lumen becomes pin hole like on brochoscope, both virtual and real. In the right lower panel, the aortic arch window

becomes slit-like at lesser curvature site.

ASO to restore the spiral arterial trunks (Chen et al., 2010).

**2.5 Bronchus** 

severely compromised.

Fig. 2. Our studies on 101 angiograms showed the wider the main to branch PA angle, the better PA size. Narrow L/R PA angle resulted in bilateral PA hypoplasia in the upper panel. Thus wider angle facilitates bilateral PA growth before ASO. Any maneuver that did not correct this inborn error of ventriculoarterial discordance or make even narrower angle like Lecompte maneuver will compromise PA growth.

#### **2.4 Aorta**

The aortic arch is kept wide open by the presence of the MPA bifurcation below it (Chiu et al., 2000b; Chiu et al., 2002b). Neo-aortic kinking is the narrowing of the aortic arch window and may become slit-like, can also be called aortic neocoarctation (Figure 3, Chiu et al., 2010; Muster et al., 1987). One group reported an incidence of 0.54% (Serraf et al., 1995). A coarctation is not present and the arch is wide open before conventional ASO with Lecompte maneuver. A pressure gradient across it is rare, less seen than is supravalvular PS because of systemic-pressure in aorta.

Fig. 3. After conventional arterial switch with Lecompte maneuver, a slit like arch window can be seen in the right panel, which was wide open before switch.

We have to take down the "Lecompte maneuver" in a patient eight years after conventional ASO to restore the spiral arterial trunks (Chen et al., 2010).
