Hypoxia: A teratogen underlying a range of congenital disruptions, dysplasias, and malformations

Abstract

In this review, we explore evidence that hypoxia in the developing human fetus can lead not only to the more commonly accepted disruptive-type defects, but also patterns of anomalies that suggest that hypoxia can exert a more classic teratogenic effect, using the brain as one example. We review neuropathology in the context of intrauterine hypoxia, particularly as it relates to carbon monoxide poisoning, in utero strokes, and homozygous alpha-thalassemia. In general, the associated brain injuries resemble those seen with other causes of hypoxic–ischemic injury. Fetal strokes during development usually lead to loss of brain tissue in areas that do not follow a typical embryologic pattern, and therefore are considered disruptions. However, there is also evidence that fetal brain ischemia can cause more classically recognized patterns of abnormal embryonic neuronal migration and organization such as polymicrogyria, cortical dysplasia, or dysgenesis, including select types of focal cortical dysplasia. This study summarizes available literature and evidence to raise clinicians' awareness regarding the association between hypoxia and congenital anomalies, including brain malformations.

1 ∣. INTRODUCTION

A birth defect is broadly defined as any structural difference present at birth that leads to abnormalities of function, appearance, or both (Centers for Disease Control and Prevention, 2019). In humans, birth defects have a variety of etiologies. Mechanisms leading to structural birth defects are often classified into categories including malformations, dysplasias, disruptions, and deformations (Jones et al., 2013). The term malformation is used to describe an inherently abnormal developmental process. Dysplasia refers more specifically to abnormalities of cell organization when attempting to form a particular tissue and/or organ and here will be included in the category of “malformation,” In contrast, a disruption is defined as a birth defect caused by destruction of a normally forming tissue, such as defects caused by disruption of normal blood flow to the developing embryo. Deformations refer to abnormalities that arise from prenatal mechanical forces acting on a structure that otherwise would have formed normally. As such, deformations will not be discussed further in this article.

Disruptions differ from malformations in that the affected organ or tissue would have formed normally if the developing fetus were not exposed to the disruptive event. Accordingly, disruptions usually result in physical differences that do not conform to typical embryologic developmental patterns, such as transverse limb defects involving the entire hand but no part of the forearm. Conversely, malformations often follow embryologic patterns of development, such as radial ray hypoplasia with absent thumb. As such, malformations are more likely to be driven by genetic insults than disruptions or deformations.

A teratogen can broadly be defined as any infection, physical, chemical, or environmental agent that can disrupt or disturb the development of a fetus or embryo (Adam, 2012). In order for a pregnancy exposure to be considered teratogenic, the exposure should produce a pattern of physical differences at greater than the background risk of birth defects, which is generally accepted to be about 3%–4% (CDC MMWR 2008). In general, the exposure is associated with an increased occurrence of a specific phenotypic effect in a dose–response relationship (Carey et al., 2009; Shepard, 1994).

Hypoxia, or low oxygen levels, can have multiple causes, including the inability of the tissues to utilize oxygen (i.e., due to carbon monoxide [CO] poisoning), a reduction or lack of blood flow itself, low oxygen levels in the blood (i.e., hypoxemia), and low levels of red blood cells and/or hemoglobin (Ferrer & Vidal, 2018). In this review, we will explore evidence that hypoxia in the developing fetus (using the fetal brain as one example) has the potential to lead not only to the more commonly accepted disruptive-type defects (such as porencephaly), but also to patterns of anomalies that suggest hypoxia can exert a more classic teratogenic effect.

2 ∣. EMBRYONIC BRAIN DEVELOPMENT

Development of the human fetal brain is an extremely intricate and complex process that has been extensively reviewed elsewhere (Barkovich et al., 2012; Hoch et al., 2009; Sadler, 2015). Briefly, dorsal induction of the embryonic central nervous system begins with the notochord, which is present around 2 weeks post-fertilization. The neural tube forms around 3 weeks post-fertilization and closes around Day 23, with completion of neurulation around the 4th week post-fertilization. Failure of the dorsal process leads to severe embryonic fetal malformations, including anencephaly and craniorachischisis. Ventral induction at the rostral end of the early embryo leads to the development of the prosencephalon, mesencephalon, and rhombencephalon. The prosencephalon then gives rise to the telencephalon and diencephalon. Division of the telencephalon results in the formation of the two cerebral hemispheres and the lateral ventricles. Failure of ventral induction leads to holoprosencephaly, among other major brain malformations. Finally, between 2 and 5 months of gestation, the carefully orchestrated growth and modular development of the brain occurs, including neural proliferation, differentiation of cells (e.g., neuroepithelial cells differentiate into radial glial cells, which in turn differentiate into intermediate progenitor cells which form neurons), and subsequent cell migration (Barkovich, 2013; Sadler, 2015; van der Knaap & Valk, 1988). The estimated embryologic timing of onset of brain malformations of interest in this review includes: septooptic dysplasia and agenesis of the septum pellucidum around 6–7 weeks of gestation; schizencephaly or brain clefts typically extending from the ventricle to the pial surface starting at about 2 months of gestation; porencephaly (a fluid filled hole or a “cyst,” although it is not lined by an epithelial layer) around 3–4 months of gestation; pachygyria between 3 and 4 months of gestation; and lastly, polymicrogyria and neuronal heterotopias starting around 5 months of gestation (Barkovich, 2013; Guerrini & Dobyns, 2014; van der Knaap & Valk, 1988). Below, we use CO poisoning as a specific example of a teratogen with documented effects on fetal development.

3 ∣. CO POISONING

CO is an odorless and colorless gas that is among the most common causes of fatal poisoning in the United States (Kopelman & Plaut, 1998). Although reported CO poisoning during pregnancy is rare, it can have detrimental effects on the developing embryo or fetus. The three mechanisms by which CO poisoning causes embryonic perturbation all lead to a final common result of intrauterine hypoxia, which can cause fetal neurological harm and, in the most severe cases, maternal and fetal demise.

In the acute phase of maternal CO exposure, the first mechanism leading to fetal hypoxia is related to a decline in the maternal blood oxygen content (Greingor et al., 2001). The binding affinity of CO to the heme moiety of adult hemoglobin is so strong that it makes a covalent bond leading to fetal hypoxemia by lowering the partial pressure of available oxygen in the maternal blood. Even at this very early stage when CO has not yet entered the fetal circulation, danger to the fetus is particularly high (Buchelli Ramirez et al., 2014). Umbilical venous blood typically has an oxygen tension of about 20 mmHg, which is on a steep portion of the oxygen-hemoglobin dissociation curve, meaning that small changes in the oxygen tension in the umbilical vein can lead to substantial fetal hypoxia (Greingor et al., 2001; Kopelman & Plaut, 1998). However, once the CO concentration in maternal blood is high enough, it leads to a maternal-fetal CO concentration gradient and CO crosses the placenta. The mechanism for intrauterine CO intoxication can be through either facilitated or passive diffusion (Roderique et al., 2012). This leads to further fetal hypoxia because fetal hemoglobin has a CO affinity 200 times greater than that of adult hemoglobin (Friedman et al., 2015; Nowadly et al., 2018). Therefore, CO levels are often 10% to 15% higher in the fetus than in the mother, with maximum fetal CO concentrations being achieved about 4 hours after the maternal exposure, even if the mother is asymptomatic (Longo, 1977; Roderique et al., 2012). This phenomenon is characterized by a leftward shift of the oxygen-hemoglobin dissociation curve. In the normal state, when the partial pressure of oxygen is low, the percent saturation of hemoglobin is also low, meaning that hemoglobin is releasing oxygen into hypoxic tissues. However, when the oxygen-hemoglobin dissociation curve is shifted left, low oxygen tension is accompanied by higher hemoglobin saturations, meaning that hemoglobin does not release oxygen as efficiently to hypoxic tissues.

In addition to impaired oxygen delivery due to the strength of the CO-hemoglobin bond, CO can cause direct damage to the developing brain by impairing ATP biosynthesis (Ferrer & Vidal, 2018). CO binds to cytochromes, including cytochrome c oxidase, which interferes with mitochondrial function (Alonso et al., 2003; Weaver, 2009). The combination of hypoxia and mitochondrial dysfunction then leads to a cascade of cellular events, including inflammation, oxidative stress, and apoptosis (Piantadosi, 1996; Piantadosi et al., 1997). Iron-rich regions of the brain, such as the basal ganglia and substantia nigra, are particularly sensitive to CO damage.

Embryonically, the diffusion capacity for CO from mother to fetus increases over time, as there is a greater demand for oxygen as the fetus grows and develops (Greingor et al., 2001). Therefore, CO poisoning has more detrimental effects on the fetus later in pregnancy. Furthermore, while many fetal organs and tissues have developed by the end of the first trimester of pregnancy, the fetal brain continues to grow and develop during all phases of pregnancy and well after birth and therefore, the brain is one of the most sensitive organs to fetal disruption by hypoxia in the third trimester.

Fetuses who survive CO exposure in utero or who have undergone autopsy after death have displayed a variety of neuropathological abnormalities. In rare, documented cases of first trimester exposure, abnormalities of the development of the telencephalon, such as absence of the septum pellucidum, was accompanied by schizencephaly, heterotopias, and pachygyria. The authors concluded that hypoxia during brain development in the first trimester may have led to abnormalities in the formation of the primitive germinal matrices, which in turn led to subsequent abnormal cortical development as part of a brain dysgenesis sequence (Woody & Brewster, 1990). Brain anomalies such as polymicrogyria involving temporal, frontal, and anterior central regions have been reported in an infant whose mother experienced CO poisoning around 5 months gestation (Ginsberg & Myers, 1976). Another infant whose mother sustained CO poisoning at 6 months gestation had severe destruction of the cerebral hemispheres and symmetrical temporal microgyria (Ginsberg & Myers, 1976), demonstrating both a disruption and a malformation. Other neuropathologic findings in fetuses exposed between 6 and 7 months gestation include cystic degeneration of the white matter consistent with porencephaly and neuronal loss in the basal ganglia, medial globus pallidus, subthalamic nucleus, and putamen. In addition to the more typical destruction of brain tissue leading to severe loss of neural tissue (including microcephaly), brain injuries associated with fetal CO poisoning in the last 2 months of gestation resemble those seen in other causes of hypoxic–ischemic injury (Ginsberg & Myers, 1976; Gul et al., 2009; Nowadly et al., 2018).

Normal carboxyhemoglobin levels in nonsmoking pregnant women can be as high as 2.6%, whereas for smokers the levels may range from 2.0% to 8.3% (most typically around 6%) (Norman & Halton, 1990; Woody & Brewster, 1990). Women who experience CO poisoning often have nonspecific findings, such as headache, shortness of breath, nausea, and dizziness; loss of consciousness is seen in the most severe cases (Greingor et al., 2001). The in utero effects may be reported as nonspecific decreased fetal movements. As such, the possibility of CO poisoning is commonly overlooked because the effects overlap with common symptoms during pregnancy. Therefore, CO poisoning should be considered in the differential diagnosis of severe or unexplained headaches in pregnant women (Schoen et al., 2015). When interpreting maternal carboxyhemoglobin concentrations in the blood, maternal smoking history must be taken into account. In one instance, a carboxyhemoglobin concentration was taken 2 hours after an exposure to gas fumes for which the mother (who smoked one package per day of cigarettes) was symptomatic and had only mildly elevated levels at 6.9%. However, the infant, who was born at term 6 days later, had classic features of a hypoxic insult, including tricuspid regurgitation due to decreased right heart contractility, hepatomegaly, oliguria with hematuria, increased serum creatinine and BUN, lack of spontaneous movements, seizures, increased deep tendon reflexes, and evidence of cerebral edema on cranial ultrasound (Kopelman & Plaut, 1998).

While several case reports of in utero CO poisoning in humans mention birth defects outside of the brain, details about the specific findings are often lacking (Norman & Halton, 1990). Fetal CO exposure past the critical timing of organogenesis may not be sufficient to cause structural defects, as prolonged CO exposure leading to chronic fetal hypoxia is rare and may likely lead to fetal and/or maternal death as opposed to fetal disruptive defects, except in the organs that are most dependent on oxygen, such as the developing brain. A review of all known reported cases of first trimester CO exposure published in 1990 (Norman & Halton, 1990) found that 6 of 12 known cases had structural defects, although the specific anomalies were sometimes referred to only as “mongoloid-type.” In a fetus that experienced CO poisoning at 5 to 7 weeks of gestation, a limb reduction defect of the left arm with normal right arm was observed. The legs demonstrated absent left femur with right femoral hypoplasia, bilateral tibial hemimelia (absent tibia) and laterally missing toes (Ingalls & Philbrook, 1958). Another infant who was exposed to chronic CO from around 10 weeks to the 7th month of gestation was born with multiple congenital anomalies, including complex congenital heart defects, cleft lip and palate, bilateral retinal colobomas, hypoplastic external genitalia, and low set ears (Hennequin et al., 1993). While karyotype was normal, no further evaluations for a known single gene disorder were available at the time and the authors felt that the clinical features were consistent with CHARGE association. It is unclear what role, if any, chronic CO exposure played in the anomalies seen in this infant.

Maternal treatment of CO exposure must be undertaken rapidly to avoid detrimental effects to the fetus. Furthermore, resolution of maternal symptoms alone should not be misinterpreted as meaning that the fetus is no longer at risk for continued damage. Treatment typically begins with supplemental 100% oxygen therapy for the mother; however, this does not address the issues of CO clearance from the fetal circulation. Maternal hyperbaric oxygen treatment is now recommended, as this expedites dissociation of CO from both hemoglobin and from the mitochondrial respiratory chain, in addition to raising the amount of dissolved oxygen in maternal blood that the fetus can then utilize (Arslan, 2021; Greingor et al., 2001; Mathieu et al., 2017). It is unclear if hyperbaric oxygen therapy itself poses a separate risk to the fetus, although normal birth after hyperbaric oxygen therapy has been reported (Arslan, 2021; Elkharrat et al., 1991; Greingor et al., 2001; Koren et al., 1991).

4 ∣. NONGENETIC DISRUPTION OF BLOOD SUPPLY IN THE BRAIN

Strokes fall into two broad categories, hemorrhagic and nonhemorrhagic (ischemic) strokes (Ferrer & Vidal, 2018). Hemorrhagic strokes are characterized by a blood vessel rupture or leak in the brain parenchyma, the periventricular area, intraventricularly, or in other spaces (Ferrer & Vidal, 2018; Kirkham et al., 2018). A hemorrhagic stroke involving brain parenchyma typically causes a void of brain tissue around the rupture or leakage in which the brain tissue cannot maintain conditions for survival. Nonhemorrhagic, or ischemic, strokes are caused by a narrowing or blockage of a blood vessel in the brain. The loss of blood flow causes cell death in the affected area.

Fetal stoke is typically defined as a stroke that occurs anytime between 14 weeks of gestation until the beginning of labor (Ozduman et al., 2004). Fetuses who experience a stroke during development usually have loss of brain tissue in an area that does not follow a typical embryologic developmental pattern, and thus is therefore considered a disruption. Fetal strokes of both types can have similar causes, including preeclampsia, premature rupture of membranes, chorioamnionitis, gestational diabetes, placental abruption, or placental thrombosis (Ferrer & Vidal, 2018).

Hemorrhagic strokes are often acquired from maternal vascular conditions such as autoimmune disorders or congenital heart disease. Examples of brain abnormalities that can be found alongside fetal hemorrhagic stroke include porencephaly and schizencephaly (Takanashi et al., 2003). Schizencephaly is usually distinguished from porencephaly by the presence of gray matter heterotopias lining the cleft (Granata et al., 2005). While schizencephaly was originally considered a primary defect in brain development, the association of schizencephaly with maternal alloimmune thrombocytopenia, death of a co-twin, and maternal trauma (among other factors) strongly suggests that it may be due to a disruptive process (Howe et al., 2012). Furthermore, the paucity of studies identifying genetic mutations in children with schizencephaly (including the later disproven studies of the EMX2 gene) further strengthens the evidence of nongenetic insults as probably causative (Merello et al., 2008; Tietjen et al., 2007). Schizencephaly has also been reported to be associated with optic nerve hypoplasia, both of which are associated with absence of the septum pellucidum. It has been hypothesized that ischemia in the watershed area of the transcallosal branches of the medial artery of the corpus callosum during the 7th week of gestation can lead to all of these findings and could account for their associated cooccurrences (Barkovich & Norman, 1988; Fernandez-Bouzas et al., 2006; Howe et al., 2012; Lubinsky, 1997; Raybaud et al., 2001).

Typical congenital brain abnormalities caused by disruption of blood flow (ischemia) include the well-recognized findings of encephalomalacia, schizencephaly, and/or cyst formation, such as porencephaly discussed above (Kirkham et al., 2018). However, there is also evidence that fetal brain ischemia later in gestation can cause more classically recognized patterns of abnormal embryonic neuronal migration such as polymicrogyria, focal cortical dysgenesis, and low grade focal cortical dysplasia (Diamandis et al., 2017; Durrani-Kolarik et al., 2017; Kirkham et al., 2018; Krsek et al., 2010). This is not surprising given that incomplete or focal cortical dysplasia has been proposed to be on the mildest end of the schizencephaly spectrum (Fernandez-Bouzas et al., 2006). In some instances, porencephaly and schizencephaly can also be accompanied by neuronal migration defects as a result of continued abnormal neuronal migration after the inciting insult (Takanashi et al., 2003).

Recently, in a study of brain malformations in primarily discordant monozygotic and dizygotic twins for whom no known genetic or infectious etiology was identified, disruption of vascular perfusion was hypothesized to be the most likely cause (Park et al., 2020). These included malformations of cortical development such as polymicrogyria, periventricular nodular heterotopia, and absent septum pellucidum. Proposed mechanisms for the fetal vascular perfusion defects included fetal hypotension from hemorrhaging into a demised co-twin or the placenta, which also can cause concurrent severe fetal anemia or twin-twin transfusion from fetal vascular anastomoses (Kirkham et al., 2018). Interestingly, a case of chronic placental abruption led to the birth of a term male who was identified to have bilateral perisylvian polymicrogyria, thickened and dysplastic cortex characterized by shallow sulci and broad gyri, and limb defects consisting of hypoplasia and syndactyly of fingers 2 through 4 (with the third finger most severely affected) and hypoplasia of the left great toe (Yamanouchi et al., 2002). The authors proposed that the findings were due to fetal circulatory disturbance from placental abruption, and given the types of malformations identified, likely occurred during the second trimester of pregnancy. Although the authors attributed the limb defects to amniotic bands, the defects were strikingly similar to what has been described in fetuses with homozygous alpha-thalassemia due to chronic hypoxia primarily during the second and third trimesters (see Homozygous Alpha-Thalassemia below).

5 ∣. DISRUPTION OF BLOOD SUPPLY DUE TO GENETIC CONDITIONS

Fetal strokes due to known single gene disorders are uncommon, and include fetal von Willebrand disease, fetal protein C deficiency, pyruvate carboxylase deficiency, and mutation of specific subsets of Type IV collagen (Ozduman et al., 2004). Von Willebrand disease predisposes affected individuals to abnormal bleeding while protein C deficiency promotes abnormal clotting. The mechanisms by which pyruvate carboxylase deficiency causes fetal stroke are not very well-understood but one hypothesis is chronic ischemia of the germinal matrix during neurogenesis due to the high metabolic rate of germinal matrix tissue (Brun et al., 1999). The vascular effects commonly associated with heterozygous pathogenic variants in COL4A1 and COL4A2 are intraparenchymal hemorrhage (Kirkham et al., 2018), lacunar infarcts, microbleeds, carotid artery aneurysms, and hemolytic anemias (Yoneda et al., 2013). Prenatal vascular disruptions associated with heterozygous pathogenic variants in COL4A1 and COL4A2 are associated with bilateral asymmetric porencephaly, schizencephaly, polymicrogyria near the schizencephalic cleft or in the opposite hemisphere, pontocerebellar atrophy, focal cortical dysplasia, cerebellar hypoplasia, unilateral cerebral atrophy, and periventricular leukoencephalopathy (Cavallin et al., 2018; Verbeek et al., 2012; Yoneda et al., 2013). This strikingly similar repertoire of brain anomalies compared to fetal strokes from any cause is consistent with the hypothesis that loss of blood flow or oxygen delivery to the developing brain leads to the same or similar constellation of features. However, it should be noted that the pathways and processes that drive normal human brain development are extremely conserved, and susceptible to a wide variety of genetic and nongenetic insults that may converge mechanistically to lead to similar developmental abnormalities. This in part may explain the high degree of overlap or similarities in these associated defects (Barkovich et al., 2012; Guerrini & Dobyns, 2014; van der Knaap & Valk, 1988).

6 ∣. HOMOZYGOUS ALPHA-THALASSEMIA

A recurrent cause of severe anemia and fetal hydrops is homozygous alpha-thalassemia due a common 20.5 kilobase biallelic deletion of HBA1 and HBA2, also referred to as the Southeast Asian deletion (Ko et al., 1991). This results in the complete absence of the alphaglobin chain of hemoglobin, such that affected fetuses produce a majority of hemoglobin Barts (a combination of four gamma chains). Hemoglobin Barts has a high affinity for oxygen and results in fetal hypoxemia (Hsieh et al., 1989). Additionally, hemoglobin Barts can lead to ineffective hematopoiesis and shortened red blood cell survival due to its relative instability and tendency to precipitate (Chui & Way, 1998; Ng et al., 1998). Affected fetuses typically demonstrate hydrops fetalis in the mid to late second trimester of pregnancy.

In some cases in which the affected fetus has not been terminated or evaluation of the brain after death has been undertaken, several brain abnormalities have been found (Chan et al., 2019). Neuropathological findings have included the more common signs of brain hypoxia, such as diffuse white matter gliosis, but also neuronal migrational defects, including subependymal nodular heterotopias (Adam et al., 2005; Chan et al., 2019). Given that the fetus begins to generate hemoglobin Barts around 8 weeks gestation, fetal hypoxia likely is present for the end of the first trimester and, without treatment, continues throughout the remainder of gestation. If aggressive in utero therapy with fetal exchange transfusion is undertaken, the risk of hypoxic brain injury is likely reduced, although data are limited. In one case, a head CT was normal after birth of an affected infant who had the classic limb defects associated with homozygous alpha-thalassemia but cognitive development was still mildly delayed (Dwinnell et al., 2011).

Interestingly, affected fetuses have also been identified to have extracranial congenital anomalies, most frequently limb defects, that are often of a disruptive type (Adam et al., 2005; Chen et al., 2006). Although frank limb reduction defects have been observed, a pattern of digital anomalies has also been described, with digits 2 to 4 typically being the most severely affected and the feet being more severely affected than the hands (Adam et al., 2005; Chan et al., 2019; Dwinnell et al., 2011). In rats exposed to hypoxia through uterine blood vessel clamping for 45 min during gestational days 14 to 16, sacrificed fetuses had hemorrhages of distal structures including the tail, forelimbs, and genital tubercle (Webster et al., 1987). It was hypothesized that if these fetuses were allowed to deliver at term, they would have shown hypoplasia, syndactyly, and amputation of the limbs and digits, with digits 2 to 4 most severely affected. In the most extreme cases, complete amputation of the foot plate would be observed. This closely resembles the pattern of limb anomalies seen in humans with homozygous alpha-thalassemia causing in utero hypoxia (Golden et al., 2003) and in the reported case of Yamanouchi et al. (2002) of chronic placental abruption in a term male who was identified to have bilateral perisylvian polymicrogyria, thickened and dysplastic cortex, and limb defects. Although the limb defects associated with hypoxia in these instances suggest a pattern, likely the limb defects are due to a disruption of previously formed tissue, as the hypoxia seen in homozygous alpha-thalassemia occurs just after the completion of limb formation at 8 weeks of gestation. Interestingly, the embryonic development of the lower limbs lags that of the upper limbs by several days, such that hypoxia may have an arresting affect that is more severe in the lower limbs compared to the upper limbs (Adam et al., 2005; Sadler, 2015).

7 ∣. SUMMARY

Hypoxia is a known cause of tissue destruction both before and after the birth of a human fetus. Very few fetuses have been reported in which hypoxia is known to have occurred in the first trimester of pregnancy. As most tissues are fully formed at the end of the first trimester, hypoxia that occurs outside of this developmental window is more likely to cause disruptive-type birth defects as opposed to malformation-type anomalies. However, the human brain continues to grow and develop during all stages of pregnancy. Brain development and growth involves a complex and carefully orchestrated process of proliferation, migration, and organization of neurons and glia to form the final layers of gyri or a fully developed neonatal brain. Therefore, hypoxia during the second and third trimesters of pregnancy has the potential to cause not only destruction of already formed components of the brain (e.g., porencephaly) but also to impact the typical migration and organization of brain structures, leading to more subtle findings such as polymicrogyria and other types of cortical dysplasia (Diamandis et al., 2017; Durrani-Kolarik et al., 2017; Kirkham et al., 2018; Krsek et al., 2010), which sometimes can lead clinicians to consider a genetic etiology for the brain anomalies. Human and animal data also suggest that hypoxia can cause a pattern of anomalies typical of a teratogen. Animal models of hypoxia during early stages of organ development also include patterns of limb defects with more severe destruction of the feet compared to the hands and specific digits (2 through 4, particularly the 3rd digit) being affected. In humans, some of these same patterns of anomalies have been identified as well. Although this pattern is recognizable, we hypothesize it may be due to different susceptibilities of tissues to destruction by hypoxia. As is true for most teratogens, the ultimate fetal risk and final outcome may be dependent on as-yet unidentified genetic factors that may predispose certain individuals to more severe and/or extensive anomalies compared to those who do not have the genetic susceptibility.

DATA AVAILABILITY STATEMENT

This manuscript was based on an extensive review of previously published literature, so no new data was generated.