Transposition of the great arteries (TGA) may be either complete or corrected. Complete TGA is also known as d-transposition, simple transposition, or atrioventricular concordance with ventriculoarterial discordance. This anomaly probably occurs because of failure of the aorticopulmonary septum to follow a spiral course during embryogenesis, resulting in the aorta arising from the right ventricle and the pulmonary artery arising from the left ventricle (de la Cruz et al., 1981). Atrial septal defect (ASD) and ventricular septal defect (VSD) are commonly seen with complete TGA and may also be associated with pulmonary artery obstruction. This is a potentially critical abnormality because, without a persistent communication between right and lefts sides of the heart, both the systemic and pulmonary circulations will run in parallel thereby preventing adequate oxygenation. Complete TGA usually causes no significant hemodynamic compromise when the fetus is in utero, but rapid deterioration occurs soon after birth in those cases without sufficient mixing of the right- and left-sided circulations.

In a cohort of 130 neonates delivered with TGA, Jouannic et al. (2004) noted that 13 of 130 neonates (10%) with TGA had profound hypoxemia (defined as a Pao₂ < 25 mmHg) and metabolic acidosis (pH < 7.15) in the first 30 minutes of life (Jouannic et al., 2004). Fortunately, concurrent cardiac structural anomalies such as VSDs may allow for communication between right and left sides of the heart to allow adequate mixing of blood. Even in infants with an intact ventricular septum, a nonrestrictive foramen ovale and patent ductus arteriosus may also allow for adequate mixing of systemic circulation and oxygenated venous return to avoid significant hypoxemia early in the neonatal course. However, neonates with an intact ventricular septum and either a restrictive foramen ovale or constricted ductus arteriosus may experience significant preoperative morbidity due to inadequate intracardiac mixing. In these cases, severe hypoxemia may necessitate immediate balloon atrioseptostomy in order to prevent end organ damage from ongoing hypoxemia.

In order to optimally plan for delivery and neonatal management, efforts have been directed at identifying fetal echocardiographic predictors of fetuses who will require emergent atrial septoplasty at the time of birth. Jouannic et al. (2004) retrospectively assessed the degree of antenatal restriction of the foramen ovale and/or constriction of the ductus arteriosus in 119 fetuses (mean GA ~ 36 weeks’ gestation). Twenty-four of 119 (20%) fetuses were noted to have at least one abnormal shunt on prenatal echocardiography, and 4 (3.4%) had restriction of both the foramen ovale and the ductus arteriosus. Although the specificity of prenatal echocardiography in predicting neonatal emergency was high (84%), the sensitivity was only 54% and the authors concluded that prenatal echocardiography cannot detect all fetuses that will require required immediate balloon atrioseptostomy as neonates (Jouannic et al., 2004).

By contrast, corrected TGA refers to the connection of the right atrium to the morphologic left ventricle, which connects to the pulmonary artery, while the left atrium connects to the morphologic right ventricle and then to the aorta. This anomaly is also known as l-transposition, or atrioventricular discordance with ventriculoarterial discordance. Because both right and left cardiac blood flow follow their intended paths from systemic to pulmonary and back to systemic circulations, this anomaly effectively cancels itself out (Romero et al., 1988). While the right atrium empties into the anatomic right ventricle, this ventricle is in fact the morphologic left ventricle with a transposed pulmonary artery as its outflow tract. Similarly the left atrium empties into the anatomic left ventricle, which is in fact the morphologic right ventricle, connected to a transposed aorta. Physiologically it might be expected that this anomaly would not lead to hemodynamic compromise; however, because of a frequent association with other abnormalities such as VSD, pulmonic stenosis, conduction defects, and atrioventricular valve abnormalities, significant neonatal morbidity and mortality may still occur.

Earlier studies suggested that TGA may account for up to 10% of all infants born with congenital cardiac defects, representing approximately 2 to 3 per 10,000 livebirths (Fyler et al., 1980; Hoffman and Kaplan, 2002). In a more recent birth defects registry from Australia covering 631,209 births from 1993 to 2002, there were 4897 cases of congenital heart disease, of which 47 were TGA (Chew et al., 2007). This represents a birth prevalence of 0.7 per 10,000 births, or 1% of all cases of congenital heart disease.

Complete TGA is recognized on prenatal sonography by careful visualization of the cardiac outflow tracts. Sonographic diagnosis relies on the demonstration of parallel outflow tracts, with absence of the normal crossover (Morelli et al., 1996). The normal crossover of the pulmonary artery and aorta is not seen, and the outflow tracts appear to run parallel to each other. Each outflow tract should be followed to its branches to positively differentiate the pulmonary artery from the aorta (McGahan et al., 2007) (Figures 55-1 and 55-2). The pulmonary trunk should be visible arising from the posterior ventricle, bifurcating into the left and right pulmonary arteries and ductus arteriosus. The aorta is visible arising anterior to the pulmonary artery and connecting to the aortic arch with brachiocephalic vessels. One of the most reliable prenatal diagnostic signs of complete TGA is the visualization of a straight vessel arising from the left ventricle, and giving off lateral branches, which represents the pulmonary artery bifurcation (Vinals et al., 2006).

Prenatal sonographic diagnosis of corrected TGA is extremely difficult because the ventricular outflow tracts may appear to arise correctly from the anatomic right and left ventricles. In addition, the morphologic appearance of the anatomic left ventricle is more suggestive of a right ventricle, with a moderator band and triangular-shaped ventricular cavity. A normal left ventricular cavity should not demonstrate a moderator band and should be elliptical, rather than triangular, in shape.

Prenatal sonography is an important means of diagnosing or excluding additional coexisting cardiac malformations. VSD may be present in 50% of cases of TGA and, when found, is commonly accompanied by subvalvular pulmonic stenosis (Schiebler et al., 1961). As a result of interference with the conduction apparatus of the heart, arrhythmias are common, with complete heart block found in 4 of 21 fetuses with corrected TGA (Gembruch et al., 1989). Other cardiac malformations that may be detectable include ventricular hypoplasia and coarctation of the aorta (Santoro et al., 1997). The presence of situs inversus may also be found in the presence of corrected TGA (Abossolo et al., 1996). In a series of nine neonates with TGA, five were diagnosed prenatally and four of these five had complex cardiac malformations (McGahan et al., 2007). These complex cardiac malformations that were amenable to prenatal diagnosis included abnormal cardiac axis, abnormal ventricular size, and VSDs.

The precision of sonography as a tool in the prenatal diagnosis of TGA is unclear. In one series of screening sono-grams of 11,894 fetuses, none of the 4 cases of TGA were detected prenatally (Tegnander et al., 1995). In another series of 50 infants operated on for TGA, 17 (34%) were successfully diagnosed prenatally through the use of obstetric sonography (Lupoglazoff et al., 1997). In addition, in another recent series of 111 congenital cardiac abnormalities, 5 of the 8 cases (63%) of TGA were detected on prenatal screening sonography (Kirk et al., 1997). In a review of all congenital cardiac malformations in one geographical area, only 5 of the 80 cases (6%) of TGA were detected prenatally (Montana et al., 1996).

It would appear that in the hands of expert sonologists in a referral population, the accurate prenatal detection of TGA is relatively high, but in general population screening, the vast majority of such cases go undiagnosed. This is supported by a review of all livebirths in Victoria, Australia, from 1993 to 2002, when only 17% of TGA cases were correctly identified prenatally (Chew et al., 2007). In fact, TGA was the least likely major congenital cardiac malformation to be diagnosed prenatally.