Modern day surgical correction of congenital heart disease leads to improvement in hemodynamics in most cases, yet outcomes and quality of life for many survivors is limited. For example, the neonatal arterial switch operation for TGA leads to resumption of a normal cardiovascular state in the newborn; however, neurodevelopmental outcomes are impaired in a substantial proportion of children and adolescents [20, 21]. One explanation is that this may be due to alteration in organ development and newborn biological substrate, related to factors that are present prior to birth and are not modified by cardiac surgery. Fetal “programming” in CHD may be taking place within an altered maternal-fetal environment, in which the placenta plays a key role.

Despite its importance in influencing fetal and postnatal life, there is a relative paucity of knowledge concerning the structure and function of the placenta in congenital heart malformations. Based on a large population-based study from Denmark, lower placental weight at birth is associated with tetralogy of Fallot, double-outlet right ventricle and “major” ventricular septal defects [22]. In a small series of 16 subjects with hypoplastic left heart syndrome, the placenta is reported smaller than normal in weight with a marked increased in fibrin deposition, decreased terminal villi and increased expression of leptin, an angiogenic and mitogenic hormone produced by the placenta, which may indicate an attempt to compensate for vascular abnormalities [23]. Placental dysfunction-related complications such as preeclampsia are associated with CHD [7, 8]. Umbilical cord insertion abnormalities of eccentric, marginal or velamentous cord insertions are also noted with increased frequency in CHD [24]. These findings suggest the concept of an early, common origin to both abnormal placentation and developmental abnormalities of the fetal cardiovascular system.

In our study, we found abnormally low placental weight-to-birth weight ratios for newborns with CHD as well as abnormalities of placental thrombosis, infarction, chorangiosis, and hypomature villi. When present, thrombosis and infarction were most notable in the periphery of the placenta in “watershed” regions in which the circulation may be most vulnerable to compromise. Chorangiosis is of particular interest as it indicates an increase in capillary density per region of villous tissue and is proposed to reflect long-standing, low-grade placental hypoxia [25]. Chorangiosis is reported in a number of clinical complications of pregnancy such as maternal diabetes and preeclampsia and is associated with increased risk of placental thrombosis [26]. Chorangiosis is present in nearly 1 out of 5 overall in our cohort, with placental thrombosis in 2 out of 5. It is conceivable that selected fetuses with CHD may exhibit a thrombotic placental vasculopathy, and if so, these may benefit from anti-thrombotic therapies like aspirin, as is commonly used in other conditions of high-risk pregnancies [27].

In order to better understand the possible influence of fetal cardiovascular physiology on placental characteristics at birth, we analyzed our findings based on physiological categories of CHD. Although placental abnormalities are present in all physiological groups, our data provide a “signal” for one type in particular, with the most profound differences noted in TGA. Newborns with TGA have the smallest placenta relative to birth weight with all TGA subjects in our study falling below the 10th percentile. In addition, those with TGA also had the highest frequency of chorangiosis and villi hypomaturity (present in 1/3, twice as common in comparison to the other categories) potentially reflecting placental vascular derangement.

Why are these findings present to a greater degree in newborns with TGA? It is possible that there may be a primary early co-incident derangement in placental formation occurring very early in gestation at the same time as a defect in septation of the great arteries from the common arterial trunk. Alternatively, the patterns of fetal blood flow unique to TGA may be a contributor. In the fetus with TGA, umbilical arterial oxygen delivery to the placenta is lower than normal. Streaming of highly oxygenated blood exiting the placenta through the umbilical vein is directed via the ductus venosus to the left side of the heart, which is connected to the pulmonary artery, while poorly oxygenated blood is directed towards the right side of the heart, which is connected to the aorta. Streaming phenomena and mixing result in blood oxygen content in the descending aorta and the umbilical arteries that is lower than normal, and thus oxygen delivery to the developing placenta throughout gestation may be reduced. Novel fetal MRI techniques are now able to quantify blood oxygen content in utero and have recently confirmed in vivo low arterial saturation in various forms of CHD [10]. These investigators also discovered low umbilical venous saturation levels in cases of CHD suggesting dysfunction of the placenta in its role as the organ of fetal oxygenation. Depending upon the type of cardiac structural abnormality present, blood oxygen content to, as well as blood oxygen content exiting from the placenta, may be different than normal.

While our newborns with TGA were relatively larger in weight, their placentas were the smallest in absolute weight size compared to the other physiological categories, although neither newborn nor placental weight alone reached statistical significance. Nevertheless, the ratio of birth weight-to-placental weight was the highest for any of our physiological categories. This raises some provocative questions: are fetuses with TGA more “efficient” at utilization of placental tissue, with good somatic growth taking place with a smaller placenta? Is it possible that somatic growth in TGA is taking place at the expense of neural maturation and growth, in the presence of limited placental mass? Although conjecture at this time, it is worth considering the possibility that the neurodevelopmental delay seen in TGA and other forms of CHD such as hypoplastic left heart syndrome may in part have its origins in fetal life due to placental abnormalities such as small placental size, poor vascularity or thrombosis [28]. Further study looking at placental characteristics in-vivo during gestation with longitudinal follow-up and clinical outcomes is necessary in order to answer these questions.

We found no relationship between placental characteristics and fetal echocardiography Doppler derived flow resistance indices of the umbilical artery, or as ratios of middle cerebral artery-to-umbilical artery. As anticipated there is an association between type of CHD and distinctions in middle cerebral artery pulsatility indices, as previously reported [29]. These likely reflect local changes in cerebrovascular impedance based on cardiac blood flow patterns, with the lowest middle cerebral artery impedance noted in those with single ventricle and aortic obstruction (e.g., hypoplastic left heart syndrome). The condition of fetal growth restriction when no CHD is present is typically due to placental “insufficiency,” with elevated placental vascular resistance manifesting as a high umbilical artery Doppler pulsatility index. The fact that we found no association between umbilical artery Doppler data and any of our placental pathology findings suggests a different mechanism of placental pathophysiology in the fetus with CHD compared to the fetus with classic growth restriction. For example, chorangiosis—an increase in capillary density per placental tissue—may in fact be a compensatory response that may decrease overall placental vascular resistance by increasing the vascular cross-sectional area. Fetal echocardiography-derived Doppler flow measures such as pulsatility index calculations are a measure of vascular resistance, not flow. Doppler-derived volumetric calculations of umbilical venous flow as an absolute measure of placental blood flow, or perhaps as a relative fraction of overall fetal combined cardiac output may be more informative [30]. Other methodologies such as MRI techniques will add to the capacity to better gauge the relationship between blood flow and placental structure and function during pregnancy, which will complement fetal echocardiography and add important insights to our understanding [10, 31].

Our study is limited in its conclusions in large part by the absence of postnatal outcome data to correlate with our placental findings. Nevertheless, our findings of placental abnormality are hypothesis generating as we look to identify worthy elements to incorporate into larger longitudinal studies. We report on static postnatal placental observations at birth, and do not provide information on the temporal progression of these findings during pregnancy. Development of safe and non-invasive surveillance techniques for characterizing the human placenta in a longitudinal manner from the earliest point of first trimester and onward is necessary and is currently the focus of a multicenter effort under the NIH sponsored Human Placenta Project [http://​nichd.​nih.​gov/​hpp] [2]. Furthermore, our study is limited by the absence of direct comparison to a contemporaneous normal population analyzed in the same manner as the CHD group. A wide spectrum of placental changes and weights can be seen in the healthy newborn population, thereby limiting the conclusions that can be made from the prevalence of findings we identify.

Our findings support the value of further investigation of the placenta in CHD, in particular in patients with TGA. Prospective investigation looking at details of maternal factors influencing placental health (i.e., overall health, nutrition, preeclampsia, diabetes) will be important to incorporate into future study analysis. Knowledge gained may identify modifiable variables during pregnancy leading to treatment such as maternal supplemental oxygenation, or maternal anti-thrombotic therapy in select pregnant women, thus ushering in an era of fetal therapy that may improve outcomes for patients with CHD.