Acute-on-chronic: using magnetic resonance imaging to disentangle the haemodynamic responses to acute and chronic fetal hypoxaemia.

Introduction: The fetal haemodynamic response to acute episodes of hypoxaemia are well characterised. However, how these responses change when the hypoxaemia becomes more chronic in nature such as that associated with fetal growth restriction (FGR), is less well understood. Herein, we utilised a combination of clinically relevant MRI techniques to comprehensively characterize and differentiate the haemodynamic responses occurring during acute and chronic periods of fetal hypoxaemia. Methods: Prior to conception, carunclectomy surgery was performed on non-pregnant ewes to induce FGR. At 108-110 days (d) gestational age (GA), pregnant ewes bearing control (n = 12) and FGR (n = 9) fetuses underwent fetal catheterisation surgery. At 117-119 days GA, ewes underwent MRI sessions where phase-contrast (PC) and T2 oximetry were used to measure blood flow and oxygenation, respectively, throughout the fetal circulation during a normoxia and then an acute hypoxia state. Results: Fetal oxygen delivery (DO2) was lower in FGR fetuses than controls during the normoxia state but cerebral DO2 remained similar between fetal groups. Acute hypoxia reduced both overall fetal and cerebral DO2. FGR increased ductus venosus (DV) and foramen ovale (FO) blood flow during both the normoxia and acute hypoxia states. Pulmonary blood flow (PBF) was lower in FGR fetuses during the normoxia state but similar to controls during the acute hypoxia state when PBF in controls was decreased. Conclusion: Despite a prevailing level of chronic hypoxaemia, the FGR fetus upregulates the preferential streaming of oxygen-rich blood via the DV-FO pathway to maintain cerebral DO2. However, this upregulation is unable to maintain cerebral DO2 during further exposure to an acute episode of hypoxaemia. The haemodynamic alterations required at the level of the liver and lung to allow the DV-FO pathway to maintain cerebral DO2, may have lasting consequences on hepatic function and pulmonary vascular regulation after birth.

Echocardiographic assessment of cardiovascular physiology of preterm miniature piglets supported with a pumped artificial placenta system.

OBJECTIVES: We evaluated fetal cardiovascular physiology and mode of cardiac failure in premature miniature piglets on a pumped artificial placenta (AP) circuit. METHODS: Fetal pigs were cannulated via the umbilical vessels and transitioned to an AP circuit composed of a centrifugal pump and neonatal oxygenator and maintained in a fluid-filled biobag. Echocardiographic studies were conducted to measure ventricular function, umbilical blood flow, and fluid status. In utero scans were used as control data. RESULTS: AP fetuses (n = 13; 102±4d gestational age [term 115d]; 616 ± 139 g [g]; survival 46.4 ± 46.8 h) were tachycardic and hypertensive with initially supraphysiologic circuit flows. Increased myocardial wall thickness was observed. Signs of fetal hydrops were present in all piglets. Global longitudinal strain (GLS) measurements increased in the left ventricle (LV) after transition to the circuit. Right ventricle (RV) and LV strain rate decreased early during AP support compared with in utero measurements but recovered toward the end of the experiment. Fetuses supported for >24 h had similar RV GLS to in utero controls and significantly higher GLS compared to piglets surviving only up to 24 h. CONCLUSIONS: Fetuses on a pump-supported AP circuit experienced an increase in afterload, and redistribution of blood flow between the AP and systemic circulations, associated with elevated end-diastolic filling pressures. This resulted in heart failure and hydrops. These preterm fetuses were unable to tolerate the hemodynamic changes associated with connection to the current AP circuit. To better mimic the physiology of the native placenta and preserve normal fetal cardiovascular physiology, further optimization of the circuit will be required.

Pulmonary Vascular Regulation in the Fetal and Transitional Lung.

Fetal lungs have fewer and smaller arteries with higher pulmonary vascular resistance (PVR) than a newborn. As gestation advances, the pulmonary circulation becomes more sensitive to changes in pulmonary arterial oxygen tension, which prepares them for the dramatic drop in PVR and increase in pulmonary blood flow (PBF) that occur when the baby takes its first few breaths of air, thus driving the transition from fetal to postnatal circulation. Dynamic and intricate regulatory mechanisms control PBF throughout development and are essential in supporting gas exchange after birth. Understanding these concepts is crucial given the role the pulmonary vasculature plays in the development of complications with transition, such as in the setting of persistent pulmonary hypertension of the newborn and congenital heart disease. An improved understanding of pulmonary vascular regulation may reveal opportunities for better clinical management.

Reduced in utero substrate supply decreases mitochondrial abundance and alters the expression of metabolic signalling molecules in the fetal sheep heart.

Babies born with fetal growth restriction (FGR) are at higher risk of developing cardiometabolic diseases across the life course. The reduction in substrate supply to the developing fetus that causes FGR not only alters cardiac growth and structure but may have deleterious effects on metabolism and function. Using a sheep model of placental restriction to induce FGR, we investigated key cardiac metabolic and functional markers that may be altered in FGR. We also employed phase-contrast magnetic resonance imaging MRI to assess left ventricular cardiac output (LVCO) as a measure of cardiac function. We hypothesized that signalling molecules involved in cardiac fatty acid utilisation and contractility would be impaired by FGR and that this would have a negative impact on LVCO in the late gestation fetus. Key glucose (GLUT4 protein) and fatty acid (FATP, CD36 gene expression) substrate transporters were significantly reduced in the hearts of FGR fetuses. We also found reduced mitochondrial numbers as well as abundance of electron transport chain complexes (complexes II and IV). These data suggest that FGR diminishes metabolic and mitochondrial capacity in the fetal heart; however, alterations were not correlated with fetal LVCO. Overall, these data show that FGR alters fetal cardiac metabolism in late gestation. If sustained ex utero, this altered metabolic profile may contribute to poor cardiac outcomes in FGR-born individuals after birth. KEY POINTS: Around the time of birth, substrate utilisation in the fetal heart switches from carbohydrates to fatty acids. However, the effect of fetal growth restriction (FGR) on this switch, and thus the ability of the fetal heart to effectively metabolise fatty acids, is not fully understood. Using a sheep model of early onset FGR, we observed significant downregulation in mRNA expression of fatty acid receptors CD36 and FABP in the fetal heart. FGR fetuses also had significantly lower cardiac mitochondrial abundance than controls. There was a reduction in abundance of complexes II and IV within the electron transport chain of the FGR fetal heart, suggesting altered ATP production. This indicates reduced fatty acid metabolism and mitochondrial function in the heart of the FGR fetus, which may have detrimental long-term implications and contribute to increased risk of cardiovascular disease later in life.

Feasibility of MRI assessment of maternal-fetal oxygen transport and consumption relative to maternal position in healthy late gestational pregnancies.

Late gestational supine positioning reduces maternal cardiac output due to inferior vena caval (IVC) compression, despite increased collateral venous return. However, little is known about the impact of maternal position on oxygen (O2 ) delivery and consumption of the gravid uterus, fetus, placenta and lower limbs. We studied the effects of maternal positioning on these parameters in 20 healthy pregnant subjects at 36 ± 2 weeks using magnetic resonance imaging (MRI); a follow-up MRI was performed 6-months postpartum (n = 16/20). MRI techniques included phase-contrast and T1/T2 relaxometry for blood flow and oximetry imaging, respectively. O2 transport was measured in the following vessels (bilateral where appropriate): maternal abdominal descending aorta (DAoabdo ), IVC, ovarian, paraspinal veins (PSV), uterine artery (UtA) and external iliacs, and umbilical. Maternal cardiac output was measured by summing DAothoracic and superior vena cava flows. Supine mothers (n = 6) had lower cardiac output and O2 delivery in the DAoabdo , UtA and external iliac arteries, and higher PSV flow than those in either the left (n = 8) or right (n = 6) lateral positions during MRI. However, O2 consumption in the gravid uterus, fetus, placenta and lower limbs was unaffected by maternal positioning. The ratio of IVC/PSV flow decreased in supine mothers while ovarian venous flow and O2 saturation were unaltered, suggesting a major route of pelvic venous return unaffected by maternal position. Placental-fetal O2 transport and consumption were similar between left and right lateral maternal positions. In comparison to non-pregnant findings, DAoabdo and UtA O2 delivery and pelvic O2 consumption increased, while lower-limb consumption remained constant , despite reduced external iliac artery O2 delivery in late gestation. KEY POINTS: Though sleeping supine during the third trimester is associated with an increased risk of antepartum stillbirth, the underlying biological mechanisms are not fully understood. Maternal cardiac output and uteroplacental flow are reduced in supine mothers due to inferior vena caval compression from the weight of the gravid uterus. This MRI study provides a comprehensive circulatory assessment, demonstrating reduced maternal cardiac output and O2 delivery (uteroplacental, lower body) in supine compared to lateral positioning; however, O2 consumption (gravid uterus, fetus, placenta, lower limbs) was preserved. Unlike other mammalian species, the ovarian veins conduct substantial venous return from the human pregnant uterus that is unaffected by maternal positioning. Lumbar paraspinal venous flow increased in supine mothers. These observations may have important considerations during major pelvic surgery in pregnancy (i.e. placenta percreta). Future studies should address the importance of maternal positioning as a potential tool to deliver improved perinatal outcomes in pregnancies with compromised uteroplacental O2 delivery.

MRI Characterization of Blood Flow and Oxygen Delivery in the Fetal Sheep whilst Exposed to Sildenafil Citrate.

INTRODUCTION: Newborns exposed to sildenafil citrate (SC) in utero have increased rates of persistent pulmonary hypertension. The mechanism behind this has not yet been fully elucidated. We aimed to utilize a combination of clinically relevant MRI techniques to comprehensively characterize the haemodynamics of the fetal sheep whilst under the influence of SC. We hypothesized that these MRI techniques would detect SC-induced increases in pulmonary blood flow and oxygen delivery prior to birth. METHODS: At 116-117 days gestational age (term, 150 days), pregnant Merino ewes (n = 9) underwent fetal catheterization surgery. MRI scans were performed during a basal state and then repeated during a constant umbilical vein infusion of SC to measure blood flow and oxygenation within the major vessels of the fetal circulation using phase-contrast-MRI and T2 oximetry. RESULTS: Right and left ventricular cardiac outputs were not different between states. Pulmonary blood flow increased during the SC state resulting in elevated pulmonary oxygen delivery. Right to left heart shunting through the foramen ovale was reduced without reducing cerebral oxygen delivery. CONCLUSION: SC induces alterations to pulmonary haemodynamics in utero; a characteristic that if maintained may underlie or act as a precursor towards the elevated rates of poor pulmonary outcomes after birth. These MRI techniques are the first to comprehensively characterize sildenafil's direct impact on the pulmonary vasculature and its indirect detriment to the flow of oxygen-rich blood through the foramen ovale.

Impact of the Addition of a Centrifugal Pump in a Preterm Miniature Pig Model of the Artificial Placenta.

The recent demonstration of normal development of preterm sheep in an artificial extrauterine environment has renewed interest in artificial placenta (AP) systems as a potential treatment strategy for extremely preterm human infants. However, the feasibility of translating this technology to the human preterm infant remains unknown. Here we report the support of 13 preterm fetal pigs delivered at 102 ± 4 days (d) gestation, weighing 616 ± 139 g with a circuit consisting of an oxygenator and a centrifugal pump, comparing these results with our previously reported pumpless circuit (n = 12; 98 ± 4 days; 743 ± 350 g). The umbilical vessels were cannulated, and fetuses were supported for 46.4 ± 46.8 h using the pumped AP versus 11 ± 13 h on the pumpless AP circuit. Upon initiation of AP support on the pumped system, we observed supraphysiologic circuit flows, tachycardia, and hypertension, while animals maintained on a pumpless AP circuit exhibited subphysiologic flows. On the pumped AP circuit, there was a progressive decline in umbilical vein (UV) flow and oxygen delivery. We conclude that the addition of a centrifugal pump to the AP circuit improves survival of preterm pigs by augmenting UV flow through the reduction of right ventricular afterload. However, we continued to observe the development of heart failure within a matter of days.

Intrauterine growth restriction alters the activity of drug metabolising enzymes in the maternal-placental-fetal unit.

PURPOSE: Ten percent of pregnancies are affected by intrauterine growth restriction (IUGR), and evidence suggests that affected neonates have reduced activity of hepatic cytochrome P450 (CYP) drug metabolising enzymes. Given that almost all pregnant individuals take medications and additional medications are often required during an IUGR pregnancy, we aimed to determine the impact of IUGR on hepatic CYP activity in sheep fetuses and pregnant ewes. METHODS: Specific probes were used to determine the impact of IUGR on the activity of several CYP isoenzymes (CYP1A2, CYP2C19, CYP2D6 and CYP3A) in sheep fetuses and pregnant ewes. Probes were administered intravenously to the ewe at 132 days (d) gestation (term 150 d), followed by blood sampling from the maternal and fetal circulation over 24 h. Maternal and fetal liver tissue was collected at 139-140 d gestation, from which microsomes were isolated and incubated with probes. Metabolite and maternal plasma cortisol concentrations were measured using Liquid Chromatography - tandem mass spectrometry (LC-MS/MS). RESULTS: Maternal plasma cortisol concentration and maternal hepatic CYP1A2 and CYP3A activity was significantly higher in IUGR pregnancies. Maternal hepatic CYP activity was higher than fetal hepatic CYP activity for all CYPs tested, and there was minimal CYP1A2 or CYP3A activity in the late gestation fetus when assessed using in vitro methods. CONCLUSIONS: The physiological changes to the maternal-placental-fetal unit in an IUGR pregnancy have significant effects on maternal drug metabolism, suggesting changes in medications and/or doses may be required to optimise maternal and fetal health.

Impact of maternal late gestation undernutrition on surfactant maturation, pulmonary blood flow and oxygen delivery measured by magnetic resonance imaging in the sheep fetus.

Restriction of fetal substrate supply has an adverse effect on surfactant maturation in the lung and thus affects the transition from in utero placental oxygenation to pulmonary ventilation ex utero. The effects on surfactant maturation are mediated by alteration in mechanisms regulating surfactant protein and phospholipid synthesis. This study aimed to determine the effects of late gestation maternal undernutrition (LGUN) and LGUN plus fetal glucose infusion (LGUN+G) compared to Control on surfactant maturation and lung development, and the relationship with pulmonary blood flow and oxygen delivery ( DO2 ) measured by magnetic resonance imaging (MRI) with molecules that regulate lung development. LGUN from 115 to 140 days' gestation significantly decreased fetal body weight, which was normalized by glucose infusion. LGUN and LGUN+G resulted in decreased fetal plasma glucose concentration, with no change in fetal arterial PO2 compared to control. There was no effect of LGUN and LGUN+G on the mRNA expression of surfactant proteins (SFTP) and genes regulating surfactant maturation in the fetal lung. However, blood flow in the main pulmonary artery was significantly increased in LGUN, despite no change in blood flow in the left or right pulmonary artery and DO2 to the fetal lung. There was a negative relationship between left pulmonary artery flow and DO2 to the left lung with SFTP-B and GLUT1 mRNA expression, while their relationship with VEGFR2 was positive. These results suggest that increased pulmonary blood flow measured by MRI may have an adverse effect on surfactant maturation during fetal lung development. KEY POINTS: Maternal undernutrition during gestation alters fetal lung development by impacting surfactant maturation. However, the direction of change remains controversial. We examined the effects of maternal late gestation maternal undernutrition (LGUN) on maternal and fetal outcomes, signalling pathways involved in fetal lung development, pulmonary haemodynamics and oxygen delivery in sheep using a combination of molecular and magnetic resonance imaging (MRI) techniques. LGUN decreased fetal plasma glucose concentration without affecting arterial PO2 . Surfactant maturation was not affected; however, main pulmonary artery blood flow was significantly increased in the LGUN fetuses. This is the first study to explore the relationship between in utero MRI measures of pulmonary haemodynamics and lung development. Across all treatment groups, left pulmonary artery blood flow and oxygen delivery were negatively correlated with surfactant protein B mRNA and protein expression in late gestation.

Redox ratio in the left ventricle of the growth restricted fetus is positively correlated with cardiac output.

Intrauterine growth restriction (IUGR) is a result of limited substrate supply to the developing fetus in utero, and can be caused by either placental, genetic or environmental factors. Babies born IUGR can have poor long-term health outcomes, including being at higher risk of developing cardiovascular disease. Limited substrate supply in the IUGR fetus not only changes the structure of the heart but may also affect metabolism and function of the developing heart. We have utilised two imaging modalities, two-photon microscopy and phase-contrast MRI (PC-MRI), to assess alterations in cardiac metabolism and function using a sheep model of IUGR. Two-photon imaging revealed that the left ventricle of IUGR fetuses (at 140-141 d GA) had a reduced optical redox ratio, suggesting a reliance on glycolysis for ATP production. Concurrently, the use of PC-MRI to measure foetal left ventricular cardiac output (LVCO) revealed a positive correlation between LVCO and redox ratio in IUGR, but not control fetuses. These data suggest that altered heart metabolism in IUGR fetuses is indicative of reduced cardiac output, which may contribute to poor cardiac outcomes in adulthood.

Open or closed: Changes in ductus arteriosus flow patterns at birth using 4D flow MRI in newborn piglets.

The ductus arteriosus (DA) functionally closes during the transition from fetal to postnatal life because of lung aeration and corresponding cardiovascular changes. The thorough and explicit measurement and visualization of the right and left heart output during this transition has not been previously accomplished. We combined 4D flow MRI (dynamic volumetric blood flow measurements) and T2 relaxometry (oxygen delivery quantification) in surgically instrumented newborn piglets to assess the DA. This was performed in Large White-Landrace-Duroc piglets (n = 34) spanning four age groups: term-9 days, term-3, term+1, and term+5. Subject's DA status was classified using 4D flow: closed DA, forward DA flow, reversed DA flow, and bidirectional DA flow. Over all states, vessel diameters and flows normalized to body weight increased with age (for example in the ascending aorta flow-term-9: 126.6 ± 45.4; term+5: 260.2 ± 80.0 ml/min per kg; p = 0.0005; ascending aorta diameter-term-9: 5.2 ± 0.8; term+5: 7.7 ± 0.4 mm; p = 0.0004). In subjects with reversed DA blood flow there was lower common carotid artery blood flow (forward: 37.5 ± 9.0; reversed: 20.0 ± 7.4 ml/min per kg; p = 0.032). Linear regression revealed that as net DA flow decreases, common carotid artery flow decreases (R2  = 0.32, p = 0.004), and left (R2  = 0.33, p = 0.003) and right (R2  = 0.34, p = 0.003) pulmonary artery flow increases. Bidirectional DA blood flow changed oxygen saturation as determined by MRI between the ascending and descending aorta (ascending aorta: 90.1% ± 8.4%; descending aorta: 75.6% ± 14.2%; p < 0.05). Expanded use of these techniques in preterm animal models will aid in providing new understandings of normal versus abnormal DA transition, as well as to test the effectiveness of related clinical interventions.

Maternal and Fetal Hemodynamic Adaptations to Pregnancy and Clinical Outcomes in Maternal Cardiac Disease.

BACKGROUND: Although insufficient maternal cardiac output (CO) has been implicated in poor outcomes in mothers with heart disease (HD), maternal-fetal interactions remain incompletely understood. We sought to quantify maternal-fetal hemodynamics with the use of magnetic resonance imaging (MRI) and explore their relationship with adverse events. METHODS: Pregnant women with moderate or severe HD (n = 22; mean age 32 ± 5 years) were compared with healthy control women (n = 21; 34 ± 3 years). An MRI was performed during the third trimester at peak output (maternal-fetal) and 6 months postpartum with return of maternal hemodynamics to baseline (reference). Phase-contrast MRI was used for flow quantification and was combined with T1/T2 relaxometry for derivation of fetal oxygen delivery/consumption. RESULTS: Third-trimester CO and cardiac index (CI) measurements were similar in HD and control groups (CO 7.2 ± 1.5 vs 7.3 ± 1.6 L/min, P = 0.79; CI 4.0 ± 0.7 vs 4.3 ± 0.7 L/min/m,2P = 0.28). However, the magnitude of CO/CI increase (Δ, peak pregnancy - reference) in the HD group exceeded that in the control group (CO 46 ± 24% vs 27 ± 16% [P = 0.007]; CI 51 ± 28% vs 28 ± 17% [P = 0.005]). Fetal growth and oxygen delivery/consumption were similar between groups. Adverse cardiovascular outcomes (nonmutually exclusive) in 6 HD women included arrhythmia (n = 4), heart failure (n = 2), and hypertensive disorder of pregnancy (n = 1); premature delivery was observed in 2 of these women. The odds of a maternal cardiovascular event were inversely associated with peak CI (odds ratio 0.10, 95% confidence interval 0.001-0.86; P = 0.04) and Δ,CI (0.02, 0.001-0.71; P = 0.03). CONCLUSIONS: Maternal-fetal hemodynamics can be well characterised in pregnancy with the use of MRI. Impaired adaptation to pregnancy in women with HD appears to be associated with development of adverse outcomes of pregnancy.

An MRI approach to assess placental function in healthy humans and sheep.

KEY POINTS: Human placental function is evaluated using non-invasive Doppler ultrasound of umbilical and uterine artery pulsatility indices as measures of resistance in placental vascular beds, while measurement of placental oxygen consumption ( VO2 ) is only possible during Caesarean delivery. This study shows the feasibility of using magnetic resonance imaging (MRI) in utero to measure blood flow and oxygen content in uterine and umbilical vessels to calculate oxygen delivery to and VO2 by the gravid uterus, uteroplacenta and fetus. Normal late gestational human uteroplacental VO2 by MRI was ∼4 ml min-1  kg-1 fetal weight, which was similar to our MRI measurements in sheep and to those previously measured using invasive techniques. Our MRI approach can quantify uteroplacental VO2 , which involves the quantification of maternal- and fetal-placental blood flows, fetal oxygen delivery and VO2 , and the oxygen gradient between uterine- and umbilical-venous blood, providing a comprehensive assessment of placental function with clinical potential. ABSTRACT: It has not been feasible to perform routine clinical measurement of human placental oxygen consumption ( VO2 ) and in vitro studies do not reflect true metabolism in utero. Here we propose an MRI method to non-invasively quantify in utero placental and fetal oxygen delivery ( DO2 ) and VO2 in healthy humans and sheep. Women (n = 20) and Merino sheep (n = 10; 23 sets of measurements) with singleton pregnancies underwent an MRI in late gestation (36 ± 2 weeks and 128 ± 9 days, respectively; mean ± SD). Blood flow (phase-contrast) and oxygen content (T1 and T2 relaxometry) were measured in the major uterine- and umbilical-placental vessels, allowing calculation of uteroplacental and fetal DO2 and VO2 . Maternal DO2 (ml min-1  kg-1 fetus) to the gravid uterus was similar in humans and sheep (human = 54 ± 15, sheep = 53 ± 21, P = 0.854), while fetal DO2 (human = 25 ± 4, sheep = 22 ± 5, P = 0.049) was slightly lower in sheep. Uteroplacental and fetal VO2 (ml min-1  kg-1 fetus; uteroplacental: human = 4.1 ± 1.5, sheep = 3.5 ± 1.9, P = 0.281; fetus: human = 6.8 ± 1.3, sheep = 7.2 ± 1.7, P = 0.426) were similar between species. Late gestational uteroplacental:fetal VO2 ratio did not change with age (human, P = 0.256; sheep, P = 0.121). Human umbilical blood flow (ml min-1  kg-1 fetus) decreased with advancing age (P = 0.008), while fetal VO2 was preserved through an increase in oxygen extraction (P = 0.046). By contrast, sheep fetal VO2 was preserved through stable umbilical flow (ml min-1  kg-1 ; P = 0.443) and oxygen extraction (P = 0.582). MRI derived measurements of uteroplacental and fetal VO2 between humans and sheep were similar and in keeping with prior data obtained using invasive techniques. Taken together, these data confirm the reliability of our approach, which offers a novel clinical 'placental function test'.

Achieving sustained extrauterine life: Challenges of an artificial placenta in fetal pigs as a model of the preterm human fetus.

Artificial placenta (AP) technology aims to maintain fetal circulation, while promoting the physiologic development of organs. Recent reports of experiments performed in sheep indicate the intrauterine environment can be recreated through the cannulation of umbilical vessels, replacement of the placenta with a low-resistance membrane oxygenator, and incubation of the fetus in fluid. However, it remains to be seen whether animal fetuses similar in size to the extremely preterm human infant that have been proposed as a potential target for this technology can be supported in this way. Preterm Yucatan miniature piglets are similar in size to extremely preterm human infants and share similar umbilical cord anatomy, raising the possibility to serve as a good model to investigate the AP. To characterize fetal cardiovascular physiology, the carotid artery (n = 24) was cannulated in utero and umbilical vein (UV) and umbilical artery were sampled. Fetal UV flow was measured by MRI (n = 16). Piglets were delivered at 98 ± 4 days gestation (term = 115 days), cannulated, and supported on the AP (n = 12) for 684 ± 228 min (range 195-3077 min). UV flow was subphysiologic (p = .002), while heart rate was elevated on the AP compared with in utero controls (p = .0007). We observed an inverse relationship between heart rate and UV flow (r2  = .4527; p < .001) with progressive right ventricular enlargement that was associated with reduced contractility and ultimately hydrops and circulatory collapse. We attribute this to excessive afterload imposed by supraphysiologic circuit resistance and augmented sympathetic activity. We conclude that short-term support of the preterm piglet on the AP is feasible, although we have not been able to attain normal fetal physiology. In the future, we propose to investigate the feasibility of an AP circuit that incorporates a centrifugal pump in our miniature pig model.

Impact of resveratrol-mediated increase in uterine artery blood flow on fetal haemodynamics, blood pressure and oxygenation in sheep.

NEW FINDINGS: What is the central question of this study? Uterine artery blood flow helps to maintain fetal oxygen and nutrient delivery. We investigated the effects of increased uterine artery blood flow mediated by resveratrol on fetal growth, haemodynamics, blood pressure regulation and oxygenation in pregnant sheep. What is the main finding and its importance? Fetuses from resveratrol-treated ewes were significantly larger and exhibited a haemodynamic profile that might promote peripheral growth. Absolute uterine artery blood flow was positively correlated with umbilical vein oxygen saturation, absolute fetal oxygen delivery and fetal growth. Increasing uterine artery blood flow with compounds such as resveratrol might have clinical significance for pregnancy conditions in which fetal growth and oxygenation are compromised. ABSTRACT: High placental vascular resistance hinders uterine artery (UtA) blood flow and fetal substrate delivery. In the same group of animals as the present study, we have previously shown that resveratrol (RSV) increases UtA blood flow, fetal weight and oxygenation in an ovine model of human pregnancy. However, the mechanisms behind changes in growth and the effects of increases in UtA blood flow on fetal circulatory physiology have yet to be investigated. Twin-bearing ewes received s.c. vehicle (VEH, n = 5) or RSV (n = 6) delivery systems at 113 days of gestation (term = 150 days). Magnetic resonance imaging was performed at 123-124 days to quantify fetal volume, blood flow and oxygen saturation of major fetal vessels. At 128 days, i.v. infusions of sodium nitroprusside and phenylephrine were administered to study the vascular tone of the fetal descending aorta. Maternal RSV increased fetal body volume (P = 0.0075) and weight (P = 0.0358), with no change in brain volume or brain weight. There was a positive relationship between absolute UtA blood flow and umbilical vein oxygen saturation, absolute fetal oxygen delivery and combined fetal twin volume (all P ≤ 0.05). There were no differences between groups in fetal haemodynamics or blood pressure regulation except for higher blood flow to the lower body in RSV fetuses (P = 0.0170). The observed increase in fetal weight might be helpful in pregnancy conditions in which fetal growth and oxygen delivery are compromised. Further preclinical investigations on the mechanism(s) accounting for these changes and the potential to improve growth in complicated pregnancies are warranted.

Umbilical vein infusion of prostaglandin I2 increases ductus venosus shunting of oxygen-rich blood but does not increase cerebral oxygen delivery in the fetal sheep.

KEY POINTS: The ductus venosus (DV) is a dynamic fetal shunt that allows substrate-rich blood from the umbilical vein to bypass the hepatic circulation. In vitro studies suggest a direct role of prostaglandin I2 (PGI2 ) in the regulation of DV tone; however, the extent of this regulation has not been determined in utero. 4D flow and T2 oximetry magnetic resonance imaging can be combined to determine blood flow and oxygen delivery within the fetal circulation. PGI2 increases DV shunting of substrate-rich blood but this does not increase cerebral oxygen delivery. ABSTRACT: During fetal development, the maintenance of adequate oxygen and nutrient supply to vital organs is regulated through specialized fetal shunts. One of these shunts, the ductus venosus (DV), allows oxygen-rich blood to preferentially stream from the placenta toward the heart and brain. Herein, we combine magnetic resonance imaging (MRI) techniques that measure blood flow (4D flow) and oxygen saturation (T2 oximetry) in the fetal circuit to determine whether umbilical vein infusion of prostaglandin I2 (PGI2 , regulator of DV tone ex utero) directly dilates the DV and thus increases the preferential streaming of oxygen-rich blood toward the brain. At 114-115 days gestational age (dGA; term = 150 days), fetal sheep (n = 6) underwent surgery to implant vascular catheters in the fetal femoral artery, femoral vein, amniotic cavity and umbilical vein. Fetal MRI scans were performed at 119-124 dGA. 4D flow and T2 oximetry were performed to measure blood flow and oxygen saturation across the fetal circulation in both a basal state and whilst the fetus was receiving a continuous infusion of PGI2 . The proportion of oxygenated blood that passed through the DV from the umbilical vein was increased by PGI2 . Cerebral oxygen delivery was unchanged in the PGI2 state. This may be a result of decreased flow from the right to left side of the heart as blood flow through the foramen ovale was decreased by PGI2 . We have shown that although PGI2 acts on the DV to increase the proportion of oxygen-rich blood that bypasses the liver, this does not increase cerebral oxygen delivery in the fetal sheep.

Technique for comprehensive fetal hepatic blood flow assessment in sheep using 4D flow MRI.

KEY POINTS: The comprehensive visualization and quantification of in vivo fetal hepatic haemodynamics, particularly the shunting of ductus venosus blood, has been elusive and is not yet fully understood. We introduce the combination of chronically instrumented fetal sheep and 4D flow MRI of the whole fetal liver, which allows retrospective blood flow measurement in all visible vessels as well as qualitative assessment. The applicability and usefulness of this technique is exhibited in normally grown fetal Merino sheep in mid- and late-gestation with detailed dynamic distribution of hepatic blood flow presented. The feasibility of this approach in clinical pathology is demonstrated in two growth-restricted fetuses at mid-gestation. Further exemplification of blood flow quantification is performed over major hepatic vessels. ABSTRACT: Although the fetal vasculature has been demarcated and well understood for several decades, the corresponding haemodynamics permitting oxygen- and nutrient-rich blood delivery to the fetal organs has been comparatively difficult to study. We married two well-established methods: 4D flow MRI, a volumetric and dynamic blood-flow measurement technique, and chronically instrumented sheep to broadly assess fetal hepatic circulation. We performed this technique in mid- and late-gestation fetal Merino sheep under normoxemic conditions and major hepatic vasculature was segmented to quantify blood flow and related parameters. Dynamic blood flow was visualized, exhibiting an acceleration of umbilical vein blood through the ductus venosus as well as spiralling into the inferior vena cava where its stream remained separate from that of the hepatic veins and lower body. Ductus venosus changes from mid- to late-gestation included larger diameter (mid: 5.8 ± 0.9 vs. late: 7.1 ± 1.1 mm; P = 0.003) and cross-sectional area (mid: 27.1 ± 8.6 vs. late: 40.4 ± 11.8 mm2 ; P = 0.003), and lower velocity averaged over the cardiac cycle (mid: 15.7 ± 5.4 vs. late: 9.8 ± 7.0 cm s-1 ; P = 0.020). This resulted in higher magnitude blood flow (indexed to umbilical vein input) at mid-gestation in the ductus venosus (mid: 0.73 ± 0.21; late: 0.46 ± 0.21; P = 0.008). The visualization and quantification results support the further use of this technique to better understand regional blood flow changes during normal or abnormal fetal growth, as well as to observe acute haemodynamic responses to physiological challenges or drug interventions.

Feasibility of ventricular volumetry by cardiovascular MRI to assess cardiac function in the fetal sheep.

KEY POINTS: The application of fetal cardiovascular magnetic resonance imaging (CMR) to assess fetal cardiovascular physiology and cardiac function through the quantification of ventricular volumes has previously been investigated, but the approach has not yet been fully validated. Ventricular output measurements calculated from heart rate and stroke volumes (SV) of the right and left ventricles measured by ventricular volumetry (VV) exhibited a high level of agreement with phase-contrast (PC) blood flow measurements in the main pulmonary artery and ascending aorta, respectively. Ejection fraction of the right ventricle, which is lower than that of the left ventricle in postnatal subjects, was similar to the left ventricular ejection fraction in the fetus; probably due to the different loading conditions present in the fetal circulation. This study provides evidence to support the reliability of VV in the sheep fetus, providing evidence for its use in animal models of human diseases affecting the fetal circulation. ABSTRACT: The application of ventricular volumetry (VV) by cardiovascular magnetic resonance imaging (CMR) in the fetus remains challenging due to the small size of the fetal heart and high heart rate. The reliability of this technique in utero has not yet been established. The aim of this study was to assess the feasibility and reliability of VV in a fetal sheep model of human pregnancy. Right and left ventricular outputs by stroke volume (SV) measured using VV were compared with 2D phase-contrast (PC) CMR measurements of blood flow in the main pulmonary artery (MPA) and ascending aorta (AAo). At 124-140 days (d) gestation, singleton bearing Merino ewes underwent CMR under general anaesthesia using fetal femoral artery catheters, implanted at 109-117d, to trigger cine steady state free precession acquisitions of ventricular short-axis stacks. The short-axis cine stacks were segmented at end-systole and end-diastole, yielding right and left ventricular SV, ejection fraction, and cardiac outputs (SV × heart rate). PC cine acquisitions of MPA and AAo were analysed to measure blood flow, which served as comparators for the right and left cardiac outputs by VV. There was good correlation and agreement between VV and PC measures of ventricular outputs with no significant bias (r2  = 0.926; P < 0.0001; Bias = -4.7 ± 10.5 ml min-1  kg-1 ; 95% limits of agreement: -15.9 to 25.2 ml min-1  kg-1 ). This study validates fetal VV by CMR in a large animal model of human pregnancy and provides preliminary reference values of fetal sheep right and left ventricles in late gestation.

Normal human and sheep fetal vessel oxygen saturations by T2 magnetic resonance imaging.

KEY POINTS: Human fetal Doppler ultrasound and invasive blood gas measurements obtained by cordocentesis or at the time of delivery reveal similarities with sheep (an extensively used model for human fetal cardiovascular physiology). Oxygen saturation (SO2 ) measurements in human fetuses have been limited to the umbilical and scalp vessels, providing little information about normal regional SO2 differences in the fetus. Blood T2 MRI relaxometry presents a non-invasive measure of SO2 in the major fetal vessels. This study presents the first in vivo validation of fetal vessel T2 oximetry against the in vitro T2-SO2 relationship using catheterized sheep fetuses and compares the normal SO2 in the major vessels between the human and sheep fetal circulations. Human fetal vessel SO2 by T2 MRI confirms many similarities with the sheep fetal circulation and is able to demonstrate regional differences in SO2 ; in particular the significantly higher SO2 in the left versus right heart. ABSTRACT: Blood T2 magnetic resonance imaging (MRI) relaxometry non-invasively measures oxygen saturation (SO2 ) in major vessels but has not been validated in fetuses in vivo. We compared the blood T2-SO2 relationship in vitro (tubes) and in vivo (vessels) in sheep, and measured SO2 across the normal human and sheep fetal circulations by T2. Singleton pregnant ewes underwent surgery to implant vascular catheters. In vitro and in vivo sheep blood T2 measurements were related to corresponding SO2 measured using a blood gas analyser, as well as relating T2 and SO2 of human fetal blood in vitro. MRI oximetry was performed in the major vessels of 30 human fetuses at 36 weeks (term, 40 weeks) and 10 fetal sheep (125 days; term, 150 days). The fidelity of in vivo fetal T2 oximetry was confirmed through comparison of in vitro and in vivo sheep blood T2-SO2 relationships (P = 0.1). SO2 was similar between human and sheep fetuses, as was the fetal oxygen extraction fraction (human, 33 ± 11%; sheep, 34 ± 7%; P = 0.798). The presence of streaming in the human fetal circulation was demonstrated by the SO2 gradient between the ascending aorta (68 ± 10%) and the main pulmonary artery (49 ± 9%; P < 0.001). Human and sheep fetal vessel MRI oximetry based on T2 is a validated approach that confirms the presence of streaming of umbilical venous blood towards the heart and brain. Streaming is important in ensuring oxygen delivery to these organs and its disruption may have important implications for organ development, especially in conditions such as congenital heart disease and fetal growth restriction.

MR imaging of the fetal heart.

In the last decade, technological advances have enabled the acquisition of high spatial and temporal resolution cardiac magnetic resonance imaging (MRI) in the fetus. Fetal cardiac MRI has emerged as an alternative to ultrasound, which may be helpful to confirm a diagnosis of congenital heart disease when ultrasound assessment is hampered, for example in late gestation or in the setting of oligohydramnios. MRI also provides unique physiologic information, including vessel blood flow, oxygen saturation and hematocrit, which may be helpful to investigate cardiac and placental diseases. In this review, we summarize some of the main techniques and significant advances in the field to date. Level of Evidence: 5 Technical Efficacy: Stage 5 J. Magn. Reson. Imaging 2020;51:1030-1044.