The term double outlet right ventricle (DORV) is commonly applied to cardiac conditions where more than half of the aortic valve arises from the right ventricle (Kirklin et al., 1973; Zamora et al., 1975; Anderson et al., 1978; Restivo et al., 2006). Although this terminology is accurate in terms of ventriculo-arterial connection, it is too general. It mixes two distinct morphological conditions: the DORV and the aortic dextroposition (AD) defects.
The term “AD” indicates the anomalous position of the aortic root rightward above the right ventricle at various degrees, and includes both the tetralogy of Fallot and the Eisenmenger complex. It was described by Van Mierop et al. (1963b), and further elucidated by Goor et al. (1971). DORV was classified by Lev et al. (1972) in terms of the ventricular septal defect (VSD) position: 1) subaortic, 2) subpulmonary, 3) doubly committed, and 4) uncommitted. We think that this classification of DORV, although outstanding in terms of morphological details, mixes two different conditions: DORV and AD.
Thus, although in the majority of AD defects the predominant origin of both great arteries is from the right ventricle, we argue that DORV and AD defects represent two distinct morphological and embryological conditions, and they should be distinct in clinico-pathological diagnosis. We outline here what we think is the crucial morphological borderline between DORV and AD, and discuss the developmental aspects that appear to lead to these two different conditions. For the discussion, we use the definition of the conal cushions (or ridges), conus ridge 1 and conus ridge 3, introduced by Goor et al. (1972), rather than dextro-dorsal and sinistro-ventral conal cushions described by Van Mierop et al. (1963a).
DORV VERSUS AD DEFECTS
In early development, the outflow tract originates entirely from the right ventricle (Shaner, 1962; Van Mierop et al., 1963a; Goor et al., 1972; Anderson et al., 1974; de la Cruz et al., 1981; Lomonico et al., 1986). Counterclockwise conotruncal rotation (looking from the ventricular apex) and the leftward shift of the conoventricular flange normally allow the transfer of the right-sided aortic conotruncus backwards and leftwards above the left ventricle (Goor et al., 1972; Anderson et al., 1974; Lomonico et al., 1986).
In the context of the cardiac outflow tract development, it is essential to mention briefly the most important role of the secondary heart field (SHF) (de la Cruz et al., 1977; Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001). Indeed, according to these studies, precardiac mesodermic mesenchymal cells migrate from the SHF, located in the ventral pharyngeal mesoderm, towards the arterial pole of the cardiac tube, undergoing subsequently myocardial differentiation and consequently building up the conotruncal segment.
According to Goor and Edwards (1973), in DORV the distal conotruncus does not undergo normal rotatory inversion and, no leftward shift of the conoventricular flange occurs. This was confirmed and shown experimentally by Bajolle et al. (2006). Thus, in “true” DORV the infundibular septum is completely, or almost completely, dissociated from the ventricular septum. The infundibular septum occupies a more or less sagittal/oblique antero-posterior position and fuses posteriorly, through its septal end (conus ridge 1), with the back wall of the infundibular musculature (Fig. 1-a). Normally, septal conus ridge 1 would fuse distally with the left anterior division of the septomarginal trabeculation (SMT) (septal band) after the conotruncal rotation and proximal conal absorption. Instead, because of the lack of the conotruncal rotatory inversion in DORV, the parietal conus ridge 3 of the embryonic conus septum, remains attached to the anterior infundibular wall (Fig. 1-a) rather than moving postero-laterally to face the anterior leaflet of the tricuspid valve and to fuse with the right side of the conoventricular flange (Anderson et al., 1974).
As a result, the great arteries are straight parallel, the arterial valves are side by side or in oblique relationship with anterior aortic valve, and the infundibula are typically bilateral, although occasionally there is only one infundibulum. The VSD in DORV approximates the subpulmonary infundibulum although the infundibulum itself does not override the ventricular septum so that, strictly speaking, the defect is not subpulmonary but rather juxtapulmonary (Fig. 1-a) Freedom et al. (1997). In the Taussig–Bing defect (Taussig and Bing, 1949), a variant of DORV, the infundibular septum (septal end/conus ridge 1) fuses posteriorly with the ventricular septum near the right posterior division of the SMT. As a consequence, the subpulmonary infundibulum marginally overrides the ventricular septum, resulting in a subpulmonary VSD (Fig. 1-d). Other transitional forms of the Taussig–Bing anomaly, with various degrees of pulmonary valve override, extrapolate progressively into transposition of the great arteries (TGA) (Fig. 1-e) as a consequence of a progressive leftward shift of the conoventricular flange (Goor and Edwards, 1973).
Moving to AD, we basically deal with two anomalies: tetralogy of Fallot and Eisenmenger defect (Fig. 1-b). The morphological essence of the AD defects is the connection of the septal end (distal conus ridge 1) of the infundibular septum with the left anterior division of the SMT. This is essentially the same arrangement as the normal heart (Fig. 1-c), except that the infundibular septum is anteriorly displaced and deviated from the plane of the ventricular septum base. Both the tetralogy of Fallot and Eisenmenger defect share various degrees of AD (Saphir and Lev, 1941). In extreme cases, the aortic valve is almost entirely connected to the right ventricle. The basic difference between the two is that in tetralogy of Fallot there is stenosis of the subpulmonary infundibulum, whereas the infundibulum is large in the Eisenmenger defect (Oppenheimer-Dekker et al., 19851985).
Developmentally, we speculate that the variable degree of aortic override in AD is not the result of an incomplete conotruncal rotation but rather, the result of various degrees of incomplete leftward shift of the conoventricular flange. Once the rotated infundibular septum (septal conus ridge 1) fuses with the left anterior division of SMT, even if the aortic valve is almost completely connected to the right ventricle, the aortic root itself remains almost connected with the left ventricle (Fig. 1-b). In the minority of AD cases where the subaortic infundibulum is present, the infundibulum is close not only to the tricuspid valve but also to the mitral valve. In contrast, in DORV the subaortic infundibulum is close only to the right ventricle and to the tricuspid valve but not the mitral valve and the left ventricle. This occurs because the infundibular septum does not rotate, maintaining itself in a sagittal position away from the left anterior division of the SMT (Fig. 1-a).
Within the context of DORV, it is important to note that in a minority of cases the aorta is left sided even though ventricular position is normal. In other words, the ventricular loop which turns to the right (d-loop) and the conotruncus which twists to the left are dissociated, with the consequence that the aorta is positioned on the contralateral side (L) of the right ventricle (D) (Fig. 1-f) (Van Praagh et al., 1975).
We wish to make it clear that the distinction between DORV and AD defects that we are emphasizing refers specifically to the DORV with aortic D-position. The fascinating dissociation that occurs when the ventricular d-loop and the conotruncus twist in opposite directions, leading to conotruncal defects such as DORV with aortic L position, remains beyond the scope of the present discussion.
Based on the concepts outlined here, in DORV the aortic valve is side by side to the pulmonary valve or anterior oblique, and the VSD is only associated with the subpulmonary infundibulum (directly in Taussig–Bing or in approximation to it in straight typical DORV). In contrast, subaortic VSD is found only in AD defects. It is noteworthy that in a minority of cases the aortic valve can be directly anterior to the pulmonary valve.
The high incidence of straight parallel great arteries conotruncal defects TGA and DORV) associated with asplenia syndrome is worth mention in this discussion. Among conotruncal defects found in a large series of patients with asplenia syndrome, 92% exhibited TGA (Phoon et al., 1994). In the same study, the reported incidence of DORV was 13% and pulmonary atresia was 37%.
Considering that the subaortic VSD, which implies spiraliform great arteries, is part of an AD defect and not of DORV, it is interesting that among the many DORV cases reported by Van Praagh et al. (1982) in association with the asplenia syndrome, none exhibited subaortic VSD. These epidemiological data support the concept proposed by us (Marino et al., 2002) that a high percentage of straight parallel great arteries conotruncal anomalies, TGA particularly, fit into the spectrum of the heterotaxy laterality defects in terms of genetic control and development. Similar views were expressed by Bajolle et al. (2006) with respect to the developmental interrelation between straight parallel great arteries cardiac conotruncal defects and anomalies of left-right asymmetry. More recent studies by our group (De Luca et al., 2010; Oliverio et al., 2010) have reemphasized this concept, investigating the link between TGA and the asplenia type of heterotaxy by testing genetic mutations in families, with special emphasis on the Nodal gene.
We have described some of what we believe are the most significant embryological and morphological aspects of DORV and AD defects. We view it as important to consider DORV not just a generic ventriculo-arterial connection disorder, but rather a specific morphological entity with at least two primary fixed features: the dissociation versus association of the infundibular septum with the left anterior division of the SMT and the straight parallel versus spiraliform course of the great arteries. Thus, we argue that any degree of aortic override of the posterior aortic valve, anterior connection of the infundibular septum to the SMT, subaortic VSD and concomitant spiraliform great arteries, should be considered an AD defect. Consequently, presumptive DORV in aortic D-position with subaortic VSD should, in our opinion, be excluded from DORV diagnosis and it should instead be considered an Eisenmenger type of AD.
Developmentally, we hypothesize that in AD defects, distal conotruncal rotation occurs, allowing the septal conus ridge 1 of the embryonic conus infundibular septum to join and fuse with the left anterior division of the SMT. We speculate that it is likely that the AD malformations are not the result of a rotatory failure, but rather the consequence of a deficient passive migration of various degrees of the conoventricular flange to the left above the left ventricle.
Finally, we point out the association of straight parallel great arteries conotruncal anomalies with the asplenia type of heterotaxy. The almost total absence of association of spiraliform great arteries with asplenia syndrome further reinforces our belief that this type of conotruncal pattern should not be defined only in terms of ventriculo-arterial connection, but rather in terms of intrinsic morphology distinguishing genuine DORV from AD defects.