Authors: Rachel L. Leon (1:Department of Pediatrics, University of Texas Southwestern Medical Center, Dallas, TX), Lynn Bitar (1:Department of Pediatrics, University of Texas Southwestern Medical Center, Dallas, TX), Vidya Rajagopalan (2:Department of Pediatrics, Children’s Hospital of Los Angeles and Keck School of Medicine University of Southern California, Los Angeles, CA), Catherine Y. Spong (3:Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, Dallas, TX)
Categories: Article
Source: Prenatal diagnosis
Doi: 10.1002/pd.6572
Authors: Rachel L. Leon, Lynn Bitar, Vidya Rajagopalan, Catherine Y. Spong
The placenta and fetal heart undergo development concurrently during early pregnancy, and, while human studies have reported associations between placental abnormalities and congenital heart disease (CHD), the nature of this relationship remains incompletely understood. Evidence from animal studies suggest a plausible cause and effect connection between placental abnormalities and fetal CHD. Biomechanical models demonstrate the influence of mechanical forces on cardiac development, while genetic models highlight the role of confined placental mutations that can cause some forms of CHD. Similar definitive studies in humans are lacking; however, placental pathologies such as maternal and fetal vascular malperfusion and chronic deciduitis are frequently observed in pregnancies complicated by CHD. Moreover, maternal conditions like diabetes and pre-eclampsia, which affect placental function, are associated with increased risk of CHD in offspring. Bridging the gap between animal models and human studies is crucial to understanding how placental abnormalities may contribute to human fetal CHD. Next steps will require new methodologies and multidisciplinary approaches combining innovative imaging modalities, comprehensive genomic testing, and histopathology. These studies may eventually lead to preventative strategies for some forms of CHD by targeting placental influences on fetal heart development.
The placenta and fetal heart develop concurrently in early pregnancy in a precisely choreographed series of events. Many studies have reported associations between placental abnormalities and fetal congenital heart disease (CHD) – both pathologic lesions in postpartum placentas (1–5) and dysfunction of the placenta in vivo (6–9). Although a causal link between abnormalities of the placenta and disrupted fetal cardiac development has not been fully described in humans, there is a plausible connection based on their concurrent development, data from animal models, and clinical association studies. Current evidence on the interdependence of the placenta and fetal heart development is best considered using the Bradford Hill Criteria (10), described in 1965 to evaluate causal relationships using epidemiologic data. These nine aspects of association are the most cited framework for causal inference in hypothesized relationships between exposures and disease occurrence; they include, 1) strength of association, 2) consistency, 3) specificity, 4) temporality, 5) biological gradient, 6) plausibility, 7) coherence, 8) experiment, and 9) analogy.
Modern day applicability of these criteria vary, and may have reduced capacity in some disease processes with a clear mechanistic knowledge base (11). However, our understanding of the human placenta – even the entire intrauterine environment – is significantly underdeveloped compared to other aspects of human health and disease. Given the lack of mechanistic insights into placental influences on fetal development, these criteria are highly applicable to a hypothesized cause and effect relationship between abnormal placental structure and/ or function and disrupted cardiac development. This review will assess the interdependence of the placenta and fetal heart development through an epidemiologic lens. We will highlight specific areas of knowledge deficits with suggested avenues for future research.
An in depth discussion on the formation of the placenta and fetal heart is beyond the scope of this review; interested readers are directed to comprehensive publications on these topics (12–15). The key principle of these processes relevant to the present discussion, however, is that placental and cardiac development are temporally linked (Figure 1). In brief, extraembryonic circulation can be divided into two distinct but overlapping vitelline circulation to the secondary yolk sac, and chorionic circulation to the placenta. Key aspects of cardiac formation take place while the embryo is supported by the secondary yolk sac and during establishment of placental circulation. This placental circulation has two compartments, 1) maternal blood flow that coalesces in the intervillous space and provides a low resistance and low afterload to the developing fetal heart that is accomplished by remodeling of maternal spiral arterioles, and 2) the fetal circulation involving a capillary network housed within villous structures that expand and arborize progressively within the intervillous space as gestation progresses. Throughout pregnancy, these villi exhibit a bilayer interface to maintain a separation between fetal and maternal circulations with an inner mononuclear cytotrophoblast layer and a thin, outer multinucleated syncytiotrophoblast layer; notably, these placental cell lines genetically diverge from the inner cell mass (which becomes the fetus), within the first few days after fertilization.
Preclinical models of CHD are created from either genetic manipulation or anatomic disruption that affects biomechanical forces. These models provide strong experimental evidence to support a causal inference between placental abnormalities and cardiac defects and will be reviewed in the context of examining the plausibility of placental abnormalities leading to human forms of CHD.
Although precisely timed gene expression and molecular signaling guide cardiac development, there is also an important role of biomechanical forces and fluid mechanics. These mechanical forces include intracardiac pressure, wall shear stress (WSS), and strain. Computational models have been used to quantify the wall shear stress and strain that drive chamber growth in animal embryos with a hypothesized model of circulatory flow being essential for ventricular chamber development, primarily composed of trabecular cardiomyocytes initially with smaller populations of compact cardiomyocytes as gestation progresses (16,17). Pressure gradients created by the circulatory flow exert a circumferential strain on the myocardium that affects growth of trabeculae and compact myocardium differentially through unique gene expression profiles (18). Insights from zebrafish have demonstrated the importance of precisely localized and temporally exerted WSS in the growth of cardiac valves and chambers. Disruptions in this organization reliably result in cardiac defects of outflow tracts (19), chamber size (20), and trabeculation (21). As maturation of cardiomyocytes occurs throughout gestation (22), placental perturbations even later in pregnancy may impact cardiac development.
Additional insights into the effects of mechanical influences come from animal models in which anatomic disruption leads to changes in preload or afterload early in cardiac development. One of the best known of these models is the left atrial ligation (LAL) model of hypoplastic left heart syndrome (HLHS) in the chick embryo wherein the left atria is partially ligated in ovo resulting in obstructed flow to the left ventricle. Days later, these embryos show underdevelopment of the left sided structures along with higher right ventricle (RV) inflow that leads to RV overdevelopment (23). This model has recently been replicated in fetal lambs with similar results (24). Interestingly, a variation of this model to reduce the preload to the RV through right atrial ligation did not result in the expected pattern of underdevelopment of the right heart (25) emphasizing that biomechanical forces that shape each ventricle during cardiac development are unique. A second model of mechanical disruption known as conotruncal banding (CTB) leads to immediate afterload increase and results in dilation, thickening of the compact myocardium and subsequently a ventricular septal defect (VSD) with either double outlet right ventricle or persistent truncus arteriosus (26). To our knowledge, no models exist yet that physically alter the placenta with resultant disruption to cardiac development in mammals. A comprehensive review of animal models of CHD by Jalil Rufaihah was published in 2021 (27).
These animal models of mechanical disruption of fluid dynamics, preload, and afterload on the developing heart that evolve into CHD, provide an important analogy to the human condition. In human fetuses, we can suspect that similar disruptions to developmental biomechanical forces will likely result in CHD. With the placenta providing the primary source of left ventricle preload and right ventricle afterload, alterations in the placental structure or function that subsequently disrupt these fluid dynamics may contribute to CHD in the fetus; significant histopathologic lesions of the placenta are one potential source of disruption to fetal placental blood flow. However, due to limitations in our current methodologies to study the in vivo human placenta, we do not yet have clinical evidence of this relationship. Future studies that may provide such evidence will involve innovative methods and multidisciplinary teams that can leverage computational, imaging, machine learning, and other technologies to further our understanding human placental function.
Some studies suggest that more than 300 genes collectively contribute to heart defects in humans (28) but only 9–18% of human CHD can be attributed to genetic mutations in contemporary investigations (29). Research using knockout animal models has shown that the disruption of genes from many different biologic pathways result in cardiac defects (30), but the role of the placenta is largely missing in those investigations. The placenta represents a distinct tissue with divergent genetic makeup from that of the heart because it arises from trophoblast cells that differentiate from the embryo-forming inner cell mass within days after conception. Specifically, in a study of whole-genome sequencing study of multiple samples from each of 37 term placentas, investigators found an abundance of somatic mutations primarily comprised of single nucleotide variants (average of 192 variants per micro-dissected trophoblast cluster), with at least one copy number change in 41 of 86 bulk placenta samples assessed (31). This degree of mosaicism of the placenta is greater than any other human organ studied (31–33), and animal studies suggest placental mutations alone may account for some forms of human CHD that could escape detection through genetic testing of the patient (32). This also infers that current sampling methods for human placental pathologic examination could also miss tissue-level changes driven by localized mutations. Animal models linking placental dysfunction with fetal CHD have been difficult to study, owing to the complexity of genetic manipulation of placental genes in conditional knockout models and the high rate of embryo demise in many of the mutations affecting the placenta. The ones that do exist provide experimental evidence of causal inference and, in some cases, offer a semblance of coherence between clinical association studies and laboratory findings.
In a comprehensive study of mouse gene knockout models that are embryonic lethal in the first stages of embryogenesis, the cause of lethality almost uniformly arises from placental abnormalities, emphasizing the importance of early placental function (34). Furthermore, phenotyping of mutant mice in the Deciphering the Mechanisms of Developmental Disorders (DMDD) project revealed a strong association between embryos with placental defects and comorbidities involving the heart, brain, and vascular system, totaling 44% of 41 mutant lines assessed at E14.5 (34). Using that data, investigators examined the isolated placental effects of three of the mutant lines studied in the DMDD project (35). They developed embryo-specific and trophoblast-specific conditional knockouts for Atp11a, Smg9, and Ssr2, chosen because each of their constitutional knockouts showed full penetrance of both cardiac defects and placental abnormalities. They found that disruption of trophoblast genes alone was at least partially responsible for the abnormal cardiac phenotype in Smg9 knockout mice and fully accounted for the severe cardiac phenotype in Atp11a mutants. The investigators further specified the syncytiotrophoblast-I layer as the primary site of placental abnormalities affected by the trophoblast-specific gene knockouts (35).
Additional studies disrupting genes involved in normal placental development further underscore the essential role of placental function in heart development. Cardiac defects emerge when p38α-mitogen-activated protein (MAP) kinase (36) or its upstream activator Mekk3 (37) are knocked out in mice models. Rescue of the defective placental function completely ameliorates the deleterious effects on cardiogenesis in p38α-MAP kinase null mice. Likewise, peroxisome proliferation-activated receptor gamma (PPAR) missing solely in the placenta leads to cardiac defects in mice embryos and supplementing PPAR gamma knockouts with a normally functioning placenta similarly rescues the cardiac phenotype (36,38). Placental insufficiency-driven cardiac defects also arise from extraembryonic disruption of small ubiquitin-related modifier (SUMO)-specific protease 2 (SENP2) with its absence leading to marked abnormalities in all trophoblast layers. Conditional knockout of SENP2 in only placenta results in embryonic lethality and a complex CHD characterized by hypoplastic ventricles, pericardial effusion, thinning myocardium, and absence of atrioventricular endocardial cushions; however, isolated embryonic deletion of SNEP2 results in a normal cardiac phenotype and viability at birth (39).
Together, these studies highlight specific genes that disrupt placental structure and function resulting in abnormal cardiac phenotype in animal models. These research advances are slowly shifting the prevailing dogma of embryonic mutations as a leading etiology for CHD. Yet, research in the human placenta to bolster this paradigm shift is still lacking. No studies to date have demonstrated a placental-driven form of CHD in the human fetus. In fact, placental investigations in pregnancies complicated by fetal CHD continue to focus largely on associations between heart defects and histopathologic findings in post-delivery placentas. New longitudinal studies are needed that measure in vivo placental hemodynamics over the course of pregnancy, correlate those metrics with traditional histopathologies, and determine the genetic drivers of hemodynamic disturbances.
Existing human CHD placenta studies have had significant limitations including small sample sizes, heterogenous CHD groups (often no subcategorization beyond “complex”), no separation of patients with syndromic CHD versus isolated CHD, and a lack of adherence to standardized placental pathologic sampling guidelines and diagnostic definition as described in the Amsterdam Placenta Workshop Group Consensus Statement (40). Nevertheless, clinical investigations of the placenta have increased significantly in recent years with many focusing on pathologic investigation after delivery, often without control groups. These types of observational association studies require all the rigor and internal controls of interventional studies in order to lay a solid foundation for the next investigational steps. These studies provide specific clues to how human placental disruption manifests and evolves during pregnancies complicated by fetal CHD.
Placental abnormalities are common, and range from abnormalities in location, depth of invasion, shape, and function (Figure 2). In pregnancies complicated by fetal CHD, placental abnormalities occur in approximately three-fourths with chronic placental histopathologic lesions more common than acute inflammation (1,4,41). Placenta to birthweight ratios are also much less in fetal CHD (3). This association also has a degree of dose-dependence – in complex forms of fetal CHD abnormalities of the placenta occur at higher rates than in pregnancies complicated by less severe forms of fetal CHD (3). Placental histopathologies contribute to our understanding of a potential role for the placenta in development of CHD due to the microstructural changes that characterize these lesions and their anticipated effect on hemodynamics. The exact timing of these lesions and mechanisms by which they alter normal placental blood flow remain unknown. Histopathologic lesions of greatest interest are maternal vascular malperfusion (MVM), fetal vascular malperfusion (FVM), and chronic deciduitis which occur in 25%, 19%, and 16% of cases, respectively, in our study of over 380 CHD placentas (1).
MVM represents a constellation of histologic findings characterized by inadequate spiral arteriole remodeling in early gestation; it is a hallmark of the preeclamptic placenta and a common finding in fetuses with intrauterine growth restriction. MVM results in a hypoxic microenvironment with smaller chorionic villi, accelerated villous maturation, intervillous fibrin deposition, and often areas of infarcted villi (42). Smaller caliber arterioles with decreased vascular remodeling result in higher velocity and turbulent flow within the intervillous space which disrupts the normal branching morphogenesis of chorionic villi. Gross pathology of the placenta with MVM is hypoplastic and may contain retroplacental hemorrhage and/ or hematomas. Placental MVM is associated with stillbirth with odds ratio of 1.48 (95% CI, 1.30–1.69) in one recent report (43), with similar results in prior studies (44–46); however, associated fetal anomalies are not described. Similarly, pregnancies complicated by fetal CHD have a higher risk of stillbirth with an adjusted OR 4.09 (95% CI 1.62–10.33) and one comprehensive study of stillbirths found CHD in 9.4% of the cohort (47); unfortunately, placental abnormalities are not reported in these studies, underscoring the disconnect in much of the research in CHD populations and that involving the placenta.
Fetal vascular malperfusion arises in the presence of obstructed fetal blood flow to or from the placenta which can be complete or partial and acute or intermittent. This obstruction can arise from structural abnormalities in the umbilical cord as well, with umbilical cord hypercoiling, true knots, and shortened or elongated umbilical cords representing the cause of FVM in some cases. Most commonly, obstruction occurs in the umbilical venous circulation, which is more easily compressed. In severe cases, FVM leads to stillbirth and it is the placental pathology most closely associated with cerebral palsy (48). Other maternal morbidities also demonstrate higher incidence of FVM including hypertension with or without preeclampsia (49), maternal diabetes mellitus (DM) (50), and hypercoagulable conditions (51). FVM placentas have histologic findings of thrombosis that can be found in umbilical vessels or the chorionic plate itself, avascular chorionic villi, intramural fibrin, and/ or stromal-vascular karyorrhexis, which is characterized by red blood cell extravasation through degenerating villous endothelium (40). FVM as a consequence of abnormal fetal hemodynamics, versus an etiology of hemodynamic disturbance, remains unclear. This point, however, is key in deciphering the role of FVM in how CHD evolves in the fetus.
Chronic deciduitis is characterized by an abundance of lymphocytes and plasma cells in the basal plate of the placenta (52). It commonly occurs in the presence of other forms of chronic inflammation, specifically villitis of unknown etiology (53) and chronic chorioamnionitis, both of which are suspected to arise from maternal immune activation. Inhibition of maternal immune function is essential to the successful invasion of fetal trophoblast cells either from failure of maternal tolerance to fetal antigens or undetected infectious agents at the feto-maternal interface. This area has been previously reviewed thoroughly (54,55). However, chronic deciduitis may be an extension of chronic endometritis that predates conception, and which has been linked with recurrent miscarriage (56) as well as repeated unexplained implantation failure with in vitro fertilization (57). Given the high rate of heart defects in miscarriage, this connection between chronic deciduitis and unsuccessful pregnancy raises the question of its impact on fetal heart development. In an experimentally-induced neutrophil-driven placental inflammation model, investigators found that the resultant irregular placental development was associated with migration of maternal inflammatory monocytes to the embryonic heart, which caused a shift in the composition of the cardiac macrophage population and ultimately led to disrupted myocardial structure (58). This study definitively linked maternal inflammation with disrupted cardiogenesis in a mouse model.
The pathophysiologic connection between placental histopathologies and fetal cardiac development in humans remain unclear. The gap in knowledge is determining how placental histopathologies affect placental function, particularly in terms of the effects on LV preload and RV afterload, known factors to impact fetal cardiac formation in animal studies. Advanced imaging techniques initially developed to evaluate other organ systems are slowly being adapted to study the intrauterine environment. Next steps will require combining these in vivo placental hemodynamic metrics with investigations of underlying tissue-level underpinnings.
Pregnancies complicated by fetal CHD often have maternal comorbidities that can provide clues for the role of the placenta in fetal CHD. Specifically, studies have shown an increased incidence of fetal CHD in pregnancies complicated by diabetes mellitus (DM) (59–62), pre-eclampsia (63–66), uncontrolled hypertension (67,68), alcohol or substance use (69), smoking (70), infections (REF), and other maternal diseases as well. Many of these morbidities also affect the placenta in characteristic ways, with some animal studies providing clues to the mechanistic link between these maternal morbidities and fetal CHD. However, the inconsistency of these conditions to result in fetal CHD argues against a placental-CHD cause and effect relationship according to our causal inference criteria. Clearly, more translational research is needed that can bridge the gap between animal models showing mechanistic links to fetal CHD and human studies. Without known mechanisms linking these disorders of pregnancy with fetal CHD in humans, there continue to be no targeted interventions to prevent fetal CHD in women with risk factors.
Both pre-gestational and gestational diabetes mellitus are known to cause an oversimplification of the placental villous structure as well as overabundance and crowding of capillaries within the villi known as chorangiosis (71). Maternal pre-pregnancy hemoglobin-A1c directly correlates with degree of chorangiosis (72) as well as the risk for fetal CHD (60,62,73). This relationship meets the criteria of a biological gradient, yet the mechanistic evidence for this association is still an active area of investigation. Animal models demonstrate that the relationship between maternal DM, placental structural abnormalities, and fetal CHD is multifactorial. Preclinical data suggests that hyperglycemia exerts negative effects on cardiac development through reactive oxygen species-mediated effects, inhibition of myocyte migration, differentiation, proliferation, and maturation, transcriptional alterations of key genes in cardiac development pathways, and in late gestation cardiac hypertrophy (59).
Similarly, pre-eclampsia has all the hallmarks of a plausible placental-mediated etiology of fetal CHD given the temporality of its occurrence, but does not provide a consistent nor specific association. Pre-eclampsia is defined as hypertension and proteinuria that arise in the second half of gestation in previously normotensive women. It is a primary disorder of the placenta arising from inadequate invasion of the trophoblast into the endometrium with deficient remodeling of the maternal spiral arterioles. This leads to reduced blood flow for placental expansion, oxidative stress and release of placental factors into the maternal circulation that contribute to intravascular damage of maternal endothelium. Newer evidence has emerged implicating poor maternal cardiovascular health as a necessary backdrop to the placental issues that arise in preeclampsia. Recent reviews provide a more comprehensive discussion of this complex disease process (74–76). Pre-eclampsia is often associated with a small placenta and intrauterine growth restriction, in addition to conferring a higher risk of fetal CHD (65,77). Clinical studies have found that timing of the development of preeclampsia plays a significant role in its association with fetal CHD. In a population based Canadian study overall adjusted prevalence ratios (PR) for CHD in preeclamptic women was 1.57 (95% CI, 1.48–1.67) which increased to PR of 5.53 (95% CI, 4.98–6.15) in those diagnosed with early onset preeclampsia (63). Animal models have suggested hypoxia and its downstream effects to be the mediators of pre-eclampsia-associated CHD. Specifically, in animal models, hypoxia affects the proliferation of the multipotent second heart field progenitor cells and leads to higher rates of heart defects, particularly in genetically-predisposed mutant lines (78). These studies provide experimental evidence for causal inference. These studies have not yet yielded a targeted treatment for preeclampsia-associated fetal CHD, but hundreds of interventional clinical trials are underway to prevent preeclampsia (clinialtrials.gov); their effects on incidence of fetal CHD in those pregnancy may be difficult to study given the higher sample size needed to detect that outcome.
There are currently no established clinical standards for direct prenatal assessment of placental function and health. Fetal Doppler, commonly employed in clinical practice, is used to infer placental function through secondary measurements of umbilical artery and fetal hemodynamics (cerebroplacental ratio). However, Doppler measures have shown limited effectiveness in distinguishing abnormal placental function in CHD compared to healthy controls (9,79,80). Furthermore, Doppler measurements of placental function have demonstrated poor predictability regarding fetal biometric and neurodevelopmental outcomes (81,82). Associations between fetal Doppler measures of placental function and fetal outcomes also appear to vary with gestational age, indicating a an evolution of abnormal placental function in CHD (8,9).
Fetal magnetic resonance imaging (MRI) is increasingly acknowledged as an adjunct that provides additional information beyond the capabilities of Doppler alone. Application of MRI to the placenta has increased our knowledge about its normal structure, including its typical growth throughout gestation (83), the textural analysis as it matures (84,85), and new perfusion sequences are providing information on maternal perfusion of the placenta (6,86). In the CHD placenta, recent MRI studies not only confirm morphometric and structural abnormalities (4,7), but also reveal impaired functional characteristics (6,87). While predominantly utilized for research purposes, current MRI modalities exploit placental vascularity and oxygenation as surrogates for overall placental function. CHD placentas have been observed to exhibit reduced baseline oxygenation (88,89), perfusion (90), and altered hemodynamic responses to maternal hyperoxia stimuli (91). Although limited in number, these MRI studies have demonstrated associations between abnormalities in placental functional characteristics and fetal hemodynamics, particularly in the brain (88,91–93). While these MRI studies are cross-sectional and involve small sample sizes, they establish the proof of concept that placental functional impairments can be measured prenatally in CHD. This underscores the critical need for larger longitudinal studies aimed at characterizing the ontogeny and progression of placental functional impairments in CHD, ultimately aiming to translate placental MRI from a research tool to a clinical diagnostic standard that will drive the development of targeted therapeutics.
Both animal studies and clinical investigations support the hypothesis that placental development and cardiogenesis are strongly linked. The concurrent timing of their formation and the experimental models that alter placental genetics with resultant CHD in the offspring are the most convincing evidence to date that placental abnormalities may be responsible for some forms of CHD in the human fetus. Clinical studies to demonstrate this placenta-heart connection are needed. Advanced imaging techniques coupled with genomic testing and traditional histopathologic analyses will be key tools to bridge this gap. With this type of multidisciplinary approach, the development of treatments to prevent at least some forms of CHD is a potential reality.