Fetal diagnosis of heart disease allows providers to better prepare and care for the family. One of the primary benefits for the family is the ability to learn about the condition and to decide where to seek treatment in a nonurgent manner compared to when the diagnosis is made postnatally. Families may discuss whether or not to continue the pregnancy, and this decision may be informed by additional testing including detailed extracardiac imaging and genetic studies (24). Parents not only have the opportunity for extensive counseling from both perinatology and pediatric cardiology, but also have the ability to meet with other specialists including cardiovascular surgeons, neonatologists, social workers, financial counselors, lactation specialists, nurses, and psychologists or psychiatrists if appropriate (25,26,27). Nonetheless, while prenatal diagnosis has many benefits for parental preparation, some studies suggest mothers with a prenatal diagnosis may have poorer bonding with the infant, possibly secondary to subconscious attempts to protect themselves emotionally in case of infant death (28). Fetal counseling for prenatally detected CHD is more extensively discussed later in the chapter.

A prenatal diagnosis allows the medical team to optimize care for the neonate based on the anticipated hemodynamic status. This includes the choice of treatment center, planned delivery close to this facility, optimal early management to prevent hemodynamic deterioration, and prompt intervention if necessary (29,30,31,32,33). Measures of morbidity including oxygenation, myocardial function, blood pH, renal function, and perioperative neurologic events have been shown to be more optimal in prenatally diagnosed infants compared to those postnatally diagnosed (34,35,36,37,38,39,40).

The question of whether prenatal diagnosis improves mortality has been elusive and controversial. Studies in which CHD lesions are grouped together suggest prenatal diagnosis is associated with increased mortality, but these findings are likely due to the lack of stratifying groups by lesion and the inclusion of higher mortality lesions in the prenatally diagnosed group (19,41). Most published studies also have the significant limitation of care at a single tertiary care facility that does not account for death prior to transfer or prior to surgery.

In hypoplastic left heart syndrome (HLHS), studies have demonstrated conflicting results regarding the potential mortality benefit from prenatal diagnosis (34,35,39). A large population-based study suggested that it is not the prenatal diagnosis itself, but delivery close to a surgical center that confers decreased preoperative mortality in HLHS (30). The authors showed more than a 6-fold greater preoperative mortality in neonates born more than 90 minutes from a cardiac surgical center, while postnatally diagnosed patients born close to a surgical center had preoperative mortality similar to those prenatally diagnosed. While several studies of transposition of the great arteries (TGA) have failed to show statistical significance in mortality among those prenatally and postnatally diagnosed, two large studies have demonstrated a significantly decreased mortality for infants who are prenatally diagnosed (32,39,42,43,44). The challenges in achieving statistical significance in studies of TGA are the low rate of prenatal diagnosis and low mortality. Although studies have suggested improved survival after prenatal diagnosis of coarctation of the aorta, a benefit has yet to be shown in right-sided lesions, complex single ventricle disease, and congenitally corrected TGA (36,38,45,46,47,48,49).

When a parent or sibling of the fetus has a history of CHD, the risk of CHD in the fetus is increased compared to the overall population, although this varies by lesion and relation. Overall, estimates of CHD when there is maternal heart disease are 5.7%, paternal heart disease 2.2%, and prior child with heart disease ∼2% to 6% (103,104).

When examining recurrence by lesion, a bicuspid aortic valve (BAV) is among the highest, with ∼9% of children having a BAV if either parent or a sibling has BAV (106). A history of AVSD in either parent is associated with a 7.7% to 7.9% birth incidence of CHD in the offspring (103). For TOF, there appears to be a higher recurrence rate if the mother is affected (∼4.5%) compared to the father (1.6%), and a similar pattern is seen in isolated pulmonary stenosis when the mother is affected (3.7%) compared to the father (1.7%) (107). Ventricular septal defects (VSDs) have a reported recurrence rate of 2% to 10%, and in some studies appear to be similar whether present in mother or father, but in others appear to occur at a higher frequency if the mother is affected (104,107).

Timing of the fetal echocardiogram depends on the specific lesion and the facilities available in an individual institution. While 18 to 22 weeks of gestation is generally considered to be optimal, earlier evaluation may be indicated if confirmation of fetal cardiac disease may affect the family’s decision about termination (29,108). However, it is important to keep in mind that fetal echocardiography at 18 to 22 weeks may miss cases in which disease is progressive or occurs late in gestation, for example, maternal diabetes-associated ventricular hypertrophy (109). If fetal heart disease is suspected later in pregnancy, fetal echocardiography should be performed promptly to aid decision-making about delivery and potentially to guide in utero therapy. This is especially true for fetal arrhythmias, which often do not manifest before 25 to 26 weeks of gestation and, in some cases, only in the third trimester (29).

Fetal echocardiography consists of a sequential segmental analysis of the cardiovascular structures (Figs. 5.3–5.6). Evaluation in standard planes can be useful, including the (1) Four-chamber view (Videos 5.1 and 5.2), (2) left ventricular outflow tract view, (3) right ventricular outflow tract (RVOT) view, (4) three vessel and trachea view, (5) bicaval view, (6) long axis of the aortic arch, (7) long axis of the ductal arch, (8) a high short axis view of the great arteries, and (9) a low short axis view of the ventricles. Views 1 to 4 can be acquired in a long superior sweep starting in an axial plane just superior to the fetal diaphragm (Fig. 5.3; Video 5.3). In addition to gray scale imaging in all views, color Doppler ultrasonography is required with the adjunctive use of pulsed Doppler ultrasonography as needed. Measuring cardiac structures is optional although this should be considered for suspected structural or functional anomalies. Reference values have been published for most fetal cardiac structures, and gestational age-based or size-based z-score calculators are available (OBSONO.org, fetal.parameterz.com) using published reference values (110,111,112,113,114). This detailed assessment may include an advanced assessment of cardiac function such as ventricular strain and myocardial performance index (MPI) measurements.

Medium to large VSDs can be detected in the fetus with careful imaging. However, VSDs are among the most commonly missed lesions by prenatal imaging, presumably due to the wide variation in size, and equal ventricular pressures precluding easy observation of shunting (144). Smaller defects may be visualized, but it can be much more challenging to distinguish these from artifact. In both early and later gestation, a combination of 2D ultrasonography and Doppler color imaging can help to identify VSDs and to minimize false positives (145). Although large defects can often easily be visualized by 2D imaging, the diagnosis should always be confirmed by color Doppler imaging, which usually demonstrates bidirectional flow across the defects.

Atrioventricular septal defects (AVSDs) are most easily demonstrated in the four-chamber view (Fig. 5.10; Video 5.6). This allows easy visibility of the atrioventricular valve attachments, primum
atrial septal defects, as well as any ventricular septal component. Short-axis imaging across the atrioventricular valves also allows for close inspection of the left atrioventricular valve for evidence of a common valve or zone of apposition/cleft (Fig. 5.11). Color Doppler should always be used, as this may allow better detection of the degree of ventricular level shunting, as this will significantly affect counseling. Color Doppler can also be used to assess the degree of fetal atrioventricular valvar regurgitation, which has been shown to correlate closely with the degree of postnatal regurgitation (150).

When AVSD is diagnosed in utero, there must be a high suspicion of genetic abnormalities. In one study, the frequency of genetic disorders in nonheterotaxy AVSD was 74% (151). The most common diagnosis was Trisomy 21 (34 of 53 tested fetuses with AVSD), but Trisomy 18 and Turner syndrome were also associated. One study has reported spontaneous in utero closure of the VSD component in four fetuses with complete AVSD (152).

Fetal TGA is a malformation often undetected due to a normal appearing fetal four-chamber view. Of the major congenital heart lesions, it is the least commonly detected in utero (153). Recently, the addition of outflow tract evaluation to routine screening has improved its detection (Fig. 5.12A–C; Videos 5.7, 5.8) (32).

The patent foramen ovale (PFO, Fig. 5.13) and the patent ductus arteriosus (PDA) are two very important structures to evaluate in the fetus with TGA and a functionally intact ventricular septum. Abnormal flow in either a PDA and/or an abnormal or a restrictive PFO may be predictive of severe neonatal hypoxemia associated with an increased morbidity and mortality. In the fetus with TGA the highly oxygenated blood from the placenta is directed into the left heart crossing through lungs and PDA, which results in pulmonary vasodilation, deceased pulmonary vascular resistance, increased pulmonary blood flow, and increased left atrial filling pressure that can shift the atrial septum to the right or restrict the PFO. In late gestation, abnormal constriction of the PDA may occur due to the higher oxygen content of the blood traversing it. A restrictive PDA is associated with an anatomic constriction or narrowing and high velocity, turbulent, continuous antegrade flow through it. Retrograde holodiastolic laminar flow into the PDA is abnormal and also associated with neonatal hypoxemia (154,155).

In the fetus with TGA, several features of the atrial septum are considered to be predictive of neonatal hypoxemia because of inadequate mixing of blood at the atrial level, necessitating urgent newborn balloon atrial septostomy. These concerning findings include a restrictive PFO that is flat or protruding toward the right atrium, the PFO bowing into the LA greater than 50% of its width, or a hypermobile PFO undulating between the left and right atrium. Appropriate parental counseling and multidisciplinary delivery planning is required when a fetus is diagnosed with TGA and may require an urgent neonatal intervention (154,155,156,157,158,159).

Congenitally corrected transposition of the great arteries (CCTGA) is frequently not diagnosed prenatally, possibly because of the relatively normal chamber sizes on the four-chamber view (Fig. 5.14A) (47). CCTGA is most easily diagnosed in the fetus utilizing sweeps in multiple directions to confirm the discordant atrioventricular and ventriculoarterial relationships (Video 5.9). In the fetus with CCTGA, coexisting lesions are very common, and include VSD (62% to 79%), pulmonary outflow tract obstruction (35% to 50%), and tricuspid (systemic AV valve) valve abnormalities, for example, Ebstein anomaly and dysplasia (25% to 36%). Thus careful inspection of the ventricular septum, atrioventricular valves, and outflow tracts is important (Fig. 5.14 B and C) (47,160,161). CHB may be present in 6% to 10% of fetuses with CCTGA; thus the
fetal heart rhythm should be evaluated in detail (47,160,161,162). CCTGA in the fetus is also associated with cardiac malposition, heterotaxy, ventricular hypoplasia (Fig. 5.14B–D), or valvar atresia (47,161,163). Only 1% to 2% of CCTGA patients have no additional cardiac anomaly (164). In two fetal series of CCTGA, spontaneous intrauterine fetal demise (IUFD) was rare, occurring on 0/14 and 1/30 fetuses; however in one study of 34 fetuses, 3 had IUFD (9%), including one with CHB (47,160,161).

Conotruncal defects, reviewed in depth separately, can represent a challenge to prenatal diagnosis. TOF, double outlet right ventricle (DORV), PA with a VSD, and interrupted aortic arch are all thought to result from a disturbance in the development of the conotruncus. These defects may involve abnormal septation of the aorta and pulmonary outflow tracts, abnormal alignment between the outflow tracts and the ventricles, and the presence of a VSD. The link between these defects is further highlighted by the common association with 22q11.2 deletion syndrome, many with the DiGeorge phenotype (Table 5.2). In the long-axis view of the LV, the appearance of a VSD and overriding of a great vessel can signal the presence of any of the above lesions. In addition, even in a fetus with an isolated VSD and no conotruncal abnormalities, the aorta may appear to be overriding in certain views. Distinguishing these lesions and searching for known associated lesions is of the utmost importance.

TOF is the most common cyanotic cardiac defect, with a birth prevalence of approximately 0.5 per 1000 live births (176,177). Anterior and superior deviation of the infundibular septum is a pathognomonic feature (Fig. 5.15A) (178). An aorta that overrides the ventricular septum is present (Fig. 5.15B), although this same characteristic may be present in truncus arteriosus, PA with VSD, or DORV with subaortic VSD. The RVOT view, with attention paid to the relationship between the infundibular and trabecular septum, is crucial to making the diagnosis. Color and spectral Doppler interrogation of the PV (Videos 5.10, 5.11) and of the ductus arteriosus, as well as measurement of the PV and main pulmonary artery can aid in predicting the postnatal outcome, as retrograde flow in the ductus is associated with a need for postnatal patency of the ductus arteriosus to maintain adequate pulmonary flow (179,180,181). The presence of associated cardiac anomalies, such as additional VSDs, a right-sided aortic arch, or pulmonary artery hypoplasia should be excluded.

PA with a ventricular septal defect (PA-VSD) is viewed by many to be the most severe form of TOF, with complete obliteration of the RVOT. From the outflow tract views, an overriding aorta is visualized, and it can be indistinguishable from TOF (Image 16A). Determining whether the pulmonary arteries are confluent (image 16B), and the source of pulmonary flow are crucial. In the case of pulmonary

flow fed by a ductus arteriosus, prostaglandin initiation will be necessary at birth, as will neonatal surgery to establish a stable source of pulmonary blood flow (Video 5.12). Where pulmonary flow is provided by aortopulmonary collateral arteries, flow is typically stable at birth. In the case of pulmonary flow supplied by a ductus arteriosus, the ductus is often small and tortuous. A right-sided aortic arch is common in PA-VSD, and can be detected prenatally (Fig. 5.16C, Video 5.12) (165). Just as TOF may be associated with an AVSD, especially in Trisomy 21, PA-VSD may be associated with an AVSD (Video 5.13). In cases of PA-VSD with an AVSD, careful attention must be paid to systemic and pulmonary venous return, as this combination is common in heterotaxy (182).

Truncus arteriosus (also known as common arterial trunk) is a conotruncal defect characterized by a single outlet from the heart which gives risk to both the systemic and pulmonary blood flow (Fig. 5.17A and B). Truncus arteriosus can be mistaken for PA with a VSD given a similar LVOT view (Fig. 5.17C and D). Determining the source of pulmonary blood flow is essential to distinguishing these two lesions. Additionally, the identification of a dysplastic, regurgitant, truncal valve can aid in making this distinction. The ductus arteriosus is absent in the majority of cases (183). When the ductus arteriosus is present, arch interruption (Video 5.14) and ductal origin of a pulmonary artery should be excluded. Identifying aortic arch interruption is crucial, as this would signal the need for prostaglandin administration upon delivery and neonatal surgery to establish stable systemic blood flow. Follow up prenatal echocardiography is important, as progressive truncal regurgitation, fetal hydrops and in utero death may develop (184,185).

DORV occurs when both of the great arteries arise from the RV (186). The ventriculoarterial alignment results in both the great arteries positioned at least 50% over the RV. There is usually a loss of fibrous continuity between the mitral valve and the posterior semilunar valve. DORV is associated with a broad spectrum of intracardiac defects and great artery relationships, all of which impact on the clinical presentation and surgical management. In DORV, the outlet from the LV is via the VSD. DORV types are classified based upon VSD location. In subaortic VSD with normally related great arteries, LV output is directed to the aorta and often there is associated subpulmonary obstruction. The physiology is either that of pulmonary over circulation due to the VSD or depending on the degree of pulmonary outflow tract and pulmonary arterial obstruction, physiology akin to TOF. When the VSD is subpulmonary, the LV output is directed to the PA and the physiology is usually similar to TGA with a VSD. A subpulmonary VSD may be associated with a subaortic obstruction, coarctation of the aorta, aortic valvar stenosis or arch hypoplasia. Typically when the defect is subpulmonary the great arteries are side-by-side (Fig. 5.18, Video 5.15). DORV with an uncommitted or doubly committed VSD are the least common types. A perimembranous inlet VSD (atrioventricular canal type VSD in Van Praagh nomenclature), which is typically uncommitted is commonly associated with heterotaxy.

DORV can be undetected in the fetus when the four-chamber view appears normal (63,187). Recent inclusion of outflow track visualization to routine screening has significant potential to improve detection rates (62,80). The location of the VSD and the relationships of the great arteries can usually be accurately determined on fetal echocardiography. Genetic abnormalities are found in 14% to 20% of fetuses with DORV, although 22q11 deletion syndrome is rare (188,189). Fetuses with DORV and multiple extracardiac anomalies have the worst outcomes (190). Given this, prenatal counseling for DORV is highly variable and is based on the intracardiac anomalies, the great artery relationships and the presence of extracardiac or chromosomal abnormalities (191).

Aortopulmonary window is a rare condition in which there is a communication between the ascending aorta and pulmonary trunk proximal to its bifurcation (Fig. 5.19). Common associations include interrupted aortic arch, arch hypoplasia, and VSD. Prenatal diagnosis of aortopulmonary window is uncommon, but has been reported (192,193). Two-dimensional sweeps and use of color Doppler allow visualization of an aortopulmonary window in utero (Videos 5.16, 5.17).

Pulmonary atresia with an intact ventricular septum (PA-IVS) is a morphologically heterogeneous lesion characterized in which there is right ventricular outflow tract obstruction (RVOTO) either due to an atretic PV or muscular infundibular atresia (Fig. 5.20A). It is often accompanied by hypoplasia of the right ventricular cavity, due in part to increased muscularization of the RV (Fig. 5.20B). Fetal critical PV stenosis may evolve into atresia in utero (194). RV size and growth potential is variable and is related to the size of the tricuspid (TV) annulus (195). Fetal PA-IVS with TV/RV hypoplasia exhibits impaired growth as the pregnancy progresses (195,196). At mid-gestation the RV may be tripartite with a well-developed infundibulum. Where the RV is tripartite, the TV z score >−3 and the RVOT is patent, a biventricular repair is anticipated (195). A bipartite RV is a hypoplastic, hypertrophied RV which is nonapex forming. Fetus with PA-IVS and TV z score ≤−3 are more likely to undergo univentricular repair. This group with TV annular hypoplasia also have a high prevalence of RV to coronary artery fistulae (Fig. 5.20C) (196). Fetuses with coronary artery fistulae have a worse long-term prognosis due to possible RV-dependent coronary flow (195,196). In another study, a fetal score using four measurements in PA-IVS were predictive for a univentricular repair: TV/MV annulus ratio ≤0.83; PV/AV annulus ratio ≤0.75 RV/LV length ≤0.64, and TV inflow duration/cardiac cycle length ≤36.5% (197). A univentricular repair was performed in all fetuses in whom four criteria were present and in 92% of those with three criteria. Another study found that while TV annular size was a good predictor of outcome at all gestational ages, the best echocardiographic predictors varied by gestational age (198).

Fetal counseling is based upon RV and TV size and the absence or presence of coronary artery fistulae. When fistulae are present, prenatal counseling should include increased risk of fetal death, coronary artery insufficiency after delivery, and possible need for cardiac transplantation. Fetal pulmonary valvuloplasty has been shown to result in improved growth of the RV in selected fetuses with PA-IVS, and is discussed further in Section XIII (199,200,201,202).