Due to the heterogeneous nature of DORV, there is no pathognomonic electrocardiogram (EKG) pattern. In fact, the EKG may often be normal or demonstrate only nonspecific findings. However, in the most common subtype, the TOF subtype or sub-aortic VSD with PS, the EKG may show right axis deviation with right ventricular hypertrophy (denoted by rR’, rsR’, or qR pattern in V1) similar to classic TOF (61). Similarly, DORV in the setting of AVSD may show superior QRS axis in the frontal plane with prolongation of the PR interval due to posterior displacement of the conduction system as seen in classic AVSD. The presence of heterotaxy with atrial isomerism may result in rare findings of twin AV nodes in right atrial isomerism denoted by two distinct QRS morphologies. Left atrial isomerism may reveal a superior P wave axis or AV conduction deficits. In general, atrial isomerism is a rare finding; however, DORV is not an uncommonly described ventriculo-arterial arrangement in this population. This was originally described as the Van Praagh class III with right/left imbalance and heterotaxy (9). There is also a higher incidence in Asian populations related to higher incidence of right atrial isomerism (62). Congenital AV block is a rare finding in common subtypes of DORV but may have a higher incidence in the setting of additional cardiac malformations including AVSD (61) and left atrial isomerism (63).

Computed tomography (CT) stands as a powerful diagnostic instrument for both preoperative and postoperative evaluation of DORV. It offers certain advantages over echocardiography and magnetic resonance imaging (MRI), such as the absence of limitations related to a small acoustic window or the need for anesthesia (64). Furthermore, CT imaging can also be applied to patients with metallic implants. With the current generation of CT scanners providing high spatial and temporal resolution, broad detectors, high-pitch scanning mode, dose-reduction algorithms, and advanced three-dimensional post-processing tools, CT serves as a safer and high-quality alternative to diagnostic cardiac catheterization or MRI. Consequently, CT imaging has seen increasing use in the diagnosis and management of virtually every form of congenital heart defect, but it is particularly useful in DORV. CT can also identify the coronary artery origins. Three-dimensional displays and printing of computed tomography reconstructions can assist in the detailed analysis of the ventricles’ connection and the pathway from each ventricle to an arterial trunk. This kind of detailed imaging has proven to aid in determining the optimal surgical treatment for complex cases (Figure 6).

The purpose of additional studies in addition to an echocardiogram would be to clarify the details of the anatomy such as branch pulmonary arteries, arch anatomies, and coronaries; it may also assist in clarifying the location of the VSD.

Pre-operative management is highly dependent on the patient’s hemodynamics. For instance, patients with subaortic VSD or subpulmonary VSD can have pulmonary over-circulation requiring aggressive diuresis. At times, this may lead to the need for respiratory support and intubation. If there is too little pulmonary blood flow, prostaglandin E1 infusion and/or atrial septostomy should be considered. A subset of patients will have adequate pulmonary blood flow initially and they can be sent home to grow and await later surgical intervention. These patients may develop increasing cyanosis over time and may even develop hyper-cyanotic spells like patients with TOF. These patients will need their patent ductus arteriosus (PDA) to be kept patent, either with prostaglandin E1 administration or PDA stent placement; or sometimes they may need the placement of a central shunt. Consideration should be given to the urgent need for atrial septostomy in a patient who is desaturated and has evidence of a poor cardiac output state. If the patient has significant aortic stenosis or coarctation of the aorta (9, 64), then they may present with cardiogenic shock, which may require inotropy, cardiac catheter-based intervention, and even veno-arterial extracorporeal membrane oxygenation support.

DORV represents an extremely heterogeneous group that requires a variety of surgical approaches, considerations, and techniques for proper repair. The spectrum of surgical techniques spans from biventricular repairs, which include intraventricular rerouting patch, Rastelli-type repair (with RV to PA conduit), root translocations and the arterial switch operation (ASO), or univentricular staged palliation procedures (Table 2).

Some important anatomical features to considered when planning the surgical repair of DORV include the following: 1.

The relationship of the atrioventricular conduction axis to the margin of the VSD

In addition, some patients may require palliative procedures before a definitive surgical repair, especially in premature neonates; a syndromic infant with heterotaxy; a baby who is small for gestational age/with intrauterine growth restriction (IUGR); and/or infants with significant extracardiac abnormalities such as renal or intracranial issues. Such palliative procedures include PA band placement for babies with pulmonary over-circulation, PDA stent, or operative central shunt creation for those with inadequate pulmonary blood flow. A consideration in the modern era is the utilization of the hybrid procedure to optimize the Qp/Qs.

Surgical management of these patients commonly involves the creation of an intraventricular baffle. It is recommended to undergo repair within the first 6 months of life to limit the long-term effects of pulmonary over-circulation. When a delay is foreseen, pulmonary band placement can be pursued before a complete repair occurs at a later time.

When the VSD is perimembranous or subaortic, a relatively straight intraventricular tunnel can be created between the VSD and the aortic valve (Figure 7). In this scenario, the relationship between the tricuspid and semilunar valve plays an important role predicting the risk of pulmonary outflow tract obstruction (POTO) (65, 66). This relationship is described by four grades: 0: fibrous continuity, 1: slight discontinuity, 2: discontinuity but distance < aortic orifice diameter, and 3: Discontinuity but distance > aortic orifice diameter. Since the tunnel will pass inferior and to the right of the pulmonary valve, it is accepted that an intraventricular repair is suitable without POTO when the distance between the tricuspid and pulmonary valve is graded as a 3 or the tricuspid to pulmonary valve distance is greater than the tricuspid to aortic distance (Figure 8). Unobstructed intraventricular baffle is usually achievable beyond the age of 3 months and the repair can be performed through the right atrium. If the VSD is restrictive, which is defined as less than the diameter of aortic annulus, enlargement of the communication is advisable to avoid the development of left ventricular outflow tract obstruction. This is carried out by resecting the antero-superior portion of the ventriculo-infundibular fold and the conal septum.

When the VSD is doubly committed, the creation of an intraventricular tunnel will likely need to be performed through a right ventriculotomy or through the pulmonary artery (67, 68). This is because the intraventricular channel is near the infundibular septum, underneath the semilunar valves, with absence or near absence of the conal septum. Resection of this remnant tissue should be done. We must keep in mind that the baffle is anchored to the fibrous tissue between the pulmonary and aortic annulus and thus there is a risk for semilunar valve distortion. Patch enlargement for closure of the ventriculotomy is necessary to prevent POTO.

For this subset, the same principles for creating an unobstructed intraventricular tunnel from the LV to the aorta are utilized in addition to the strategies used to relieve general pulmonary outflow tract obstruction. Similar to TOF, the associated POTO can be sub-valvar, valvar, or supra-valvar. In scenarios where the obstruction is at the sub-valvar level, a right ventricular infundibular patch enlargement should be considered, whereas if the right ventricular outflow tract (RVOT) obstruction is instead at the valvar level, a Rastelli-type repair with an RV-to-PA conduit is indicated. Assessment of the annulus size on echocardiography as well as visually through the pulmonary trunk to determine if the z score is <−2 will help determine whether a transannular incision is warranted. In cases where a valve-sparing repair cannot be performed, a transannular incision is extended from the prior ventriculotomy. Transannular repair can be performed using a polytetrafluoroethylene (PTFE) monocusp valve, a valved conduit (Rastelli-type), or a valveless reconstruction (Figure 9).

Any additional obstruction produced by septomarginal and septoparietal trabeculations should be resected out. Invariably, this type of DORV is better approached directly through the ventricle due to enhanced visualization. Importantly, the patient’s acuity is determined by the degree of POTO. In this sense, symptomatic neonates might require prompt palliation with a systemic to PA shunt or a PDA stent before complete repair, whereas asymptomatic patients could undergo an elective repair at 3–6 months of age.

Often these patients will require a palliative procedure initially, either with one of three options: a systemic shunt, a PDA stent if POTO is present, or a pulmonary band placement to prevent pulmonary over-circulation. In patients with DORV non-committed VSD (nc-VSD) without POTO, biventricular repair could be accomplished with the creation of an intraventricular tunnel (VSD to aorta) if the tricuspid valve apparatus was anatomically amenable to rerouting. The main principles previously discussed regarding intracardiac tunnel are also applicable here. The main difference is that multiple patches may be needed to construct the intracardiac baffle (69) (Figure 10). Nevertheless, the route of the tunnel must be as short as possible to decrease long non-contractile areas within the left ventricular outflow tract and to avoid impinging upon the tricuspid valve apparatus. Alternatively, in patients with DORV remote VSD without valvar POTO, a VSD to PA baffle creation and ASO is preferred. The creation of a baffle between the VSD and the PA in conjunction with an arterial switch operation provides a shorter tunnel and is not impacted by the tricuspid valve (70). The VSD is always enlarged by resecting out the superior and the anterior rim. If the tunnel obstructs the right ventricular outflow tract, an infundibular patch should be placed.

However, in the presence of pulmonary obstruction, or when the intraventricular tunneling is not desirable, double root translocation can be considered for DORV remote VSD.

In DORV-TGA, the vessel and the ventricular communication are closest to the PA, and this anatomical relationship determines the method of anatomic repair. This subgroup is frequently associated with coronary artery anomalies, aortic arch hypoplasia, and coarctation, thereby increasing the complexity of surgical repair. Different surgical repairs have been used for this subgroup of patients, some of them of historical value in the current times. For example, tunneling the VSD to the pulmonary trunk combined with an atrial switch (Mustard/Senning) is rarely performed in the modern era. Instead, depending on the relationship of the great arteries and the associated cardiac or great vessel anomalies, the following procedures can be performed: 1.

The Kawashima procedure: The VSD is baffled to the aortic valve with a patch enlargement of the ventriculotomy. This alternative suits patients with side-by-side great arteries associated with anomalous coronary anatomy that precludes root translocation. Furthermore, a Kawashima procedure can be appropriate for patients when the pulmonary valve is not of adequate size to support the systemic circulation. In this case, the conus should be resected completely to allow rerouting the VSD to the aortic valve (Figure 11). In cases where the great vessel is anterior–posterior relationship related, the distance between the tricuspid valve and pulmonary valve tends to be insufficient for an intraventricular patch. As such, the ventriculotomy patch enlargement is needed to increase the size of the RVOT.

The decision to perform a Nikaidoh procedure or a double root translocation is determined by the suitability of the pulmonary valve to being translocated. Irrespective of this, some anatomic factors we need to address when considering this procedure are 1.

size of the pulmonary annulus size,

The double root translocation was initially described for patients with DORV remote VSD, since the construction of a long prosthetic intracardiac tunnel significantly augments the risk of LVOTO and unfavorable hemodynamics due to the akinetic area. The most favorable anatomy for double root translocation according to Hu et al. (72) is DORV non-committed VSD and DORV-TGA with bilateral conus, with or without pulmonary stenosis.

Contraindications to biventricular repair include severe ventricular hypoplasia, abnormal tricuspid chordae, multiple VSDs, and a straddling mitral valve. This subgroup is the most likely to undergo single ventricle palliation. There are stepwise surgical palliations in the single ventricular pathway. In certain non-committed VSD patients, creation of a non-obstructed intraventricular tunnel may not be felt to be feasible. In other circumstances, there may be severe ventricular hypoplasia, severe atrioventricular valve straddling, or Swiss-cheese VSDs. Thus, the univentricular pathway may be the best management strategy for this subset of patients.

If a single ventricle pathway is elected, significant cyanosis or over-circulation is initially palliated with a systemic to PA shunt (or PDA stent) or PA band placement until 3–6 months of age. Following these first approaches, a bidirectional cavopulmonary connection is created (73). Third stage palliation is completed with a total cavopulmonary connection at 2–3 years of age. It is important that any restrictive VSD be enlarged at or ideally prior to the time of Fontan completion. Although many of these patients require single ventricle approaches, new methods are evolving. For example, a report originating from Japan recently described a successful biventricular repair of DORV with transposition of the great arteries, pulmonary stenosis, and straddling mitral valve (73).

The clinical outcomes in patients with DORV have improved significantly over time with the degree of success depending on the complexity of the underlying lesions and the institution’s experience. For example, outcomes among patients with DORV-VSD are excellent. Li et al. (74) published their outcomes, including 129 patients with DORV-VSD, with in-hospital mortality of 1.6%. The surgical outcomes of DORV-Fallot are like those achieved in the TOF, with operative mortality between 0% and 5%. Although classically reported due to the location of the VSD, heart block was only noted in less than 1%. Shim et al. (75) reported their 20-year experience with arterial switch operation for patients with Taussig–Bing anomaly, and there were only 2.2% of patients succumbing to in-hospital death and only one patient (0.7%) undergoing early reoperation due to coronary insufficiency. According to a recent analysis of the STS database by Seese et al. (76) in patients who underwent Nikaidoh, Rastelli, or REV procedure in the US from 2010 to 2019, the operative mortality was 3.1% (95% CI 1.0%–7.0%; n = 5) for Rastelli, 4.4% (95% CI 1.4%–9.9%; n = 5) for the Nikaidoh, and 11.1% (95% CI 0.3%–48.3%, n = 1) for REV patients. A major postoperative morbidity was experienced in 18.8% (95% CI 13.0%–25.7%, n = 30) in the Rastelli group, 25.0% (95% CI 17.4%–33.9%; n = 29) in the Nikaidoh group, and 22.2% (95% CI 2.8–60.0%; n = 2) in the REV group. Unplanned reoperation was reported in 11.9% (95% CI 7.3%–17.9%) among Rastelli/RV to PA conduit placement patients and in 19.8% (95% CI 13%–28.9%) among Nikaidoh patients.

In the Rastelli procedure, the tunnel can be long and can become obstructed over time and the child can also have somatic outgrowth of the conduit. This will ultimately result in the need for RV to PA conduit replacement but can be temporized with catheter-based ballooning and stenting prior to surgery. In these cases of obstruction, there may be RV dysfunction over time. Coronary anomalies increase the risk of unfavorable outcomes in the Nikaidoh procedure.

Kreutzer et al. (77) has reported that the Rastelli procedure can be performed with very low early mortality. However, late mortality occurred in patients with RVOTO and LVOTO. Moreover, the European Congenital Heart Surgeons Association reported a high reoperation rate after the Rastelli procedure for the expected reasons of somatic outgrowth of the conduit over time. From the European Congenital Heart Surgeons Association data, freedom from death, reoperation rate, or reintervention rate of the REV are 90%, 78%, and 64% at 1, 5, and 10 years, respectively (78).

While any of these repairs can result in good outcomes, patient selection and determining the best operation for the patho-anatomy and the patient’s condition are essential. Furthermore, procedural selection also must align with surgeon experience and comfort level or consultation at another institution can be considered if needed.