lable at ScienceDirect

Seminars in Fetal & Neonatal Medicine 20 (2015) 210e216

Contents lists avai

Seminars in Fetal & Neonatal Medicine

journal homepage: www.elsevier .com/locate/s iny

Physiology of the fetal and transitional circulation

Anna Finnemore a, *, Alan Groves b

a Department of Perinatal Imaging and Health, King's College, London, UKb Department of Pediatrics, Weill Cornell Medical College, New York, NY, USA

Keywords:Fetal physiologyNeonatal critical carePerinatal management

* Corresponding author. Address: Department of P1st Floor, South Wing, St Thomas' Hospital, London7188 9145; fax: þ44 (0)20 7188 9154.

E-mail address: anna.finnemore@kcl.ac.uk (A. Finn

http://dx.doi.org/10.1016/j.siny.2015.04.0031744-165X/© 2015 Elsevier Ltd. All rights reserved.

s u m m a r y

The fetal circulation is an entirely transient event, not replicated at any point in later life, and functionallydistinct from the pediatric and adult circulations. Understanding of the physiology of the fetal circulationis vital for accurate interpretation of hemodynamic assessments in utero, but also for management ofcirculatory compromise in premature infants, who begin extrauterine life before the fetal circulation hasfinished its maturation. This review summarizes the key classical components of circulatory physiology,as well as some of the newer concepts of physiology that have been appreciated in recent years. Theimmature circulation has significantly altered function in all aspects of circulatory physiology. Themechanisms and significance of these differences are also discussed, as is the impact of these alterationson the circulatory transition of infants born prematurely.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

The fetal circulation is an entirely transient event, not replicatedat any point in later life, and functionally distinct from the pediatricand adult circulations. An understanding of the physiology of thefetal circulation is core to accurate interpretation of hemodynamicassessments in utero.

Given that premature infants begin extrauterine life before thefetal circulation has finished its maturation, infants manifest manyof the functional and structural characteristics of the fetal vascu-lature. Therefore awareness of the fetal physiology is also relevantfor many aspects of postnatal circulatory care. Indeed in theabsence of conclusive clinical trials determining the optimal clinicalmanagement of many common clinical presentations of the pre-term infant (hypotension, patent ductus arteriosus, pulmonaryhypertension), treatment decisions in the neonatal intensive careunit (NICU) frequently rely on applying knowledge of physiologyand pathophysiology to identify optimal therapy.

This review aims to summarize the distinct features of the fetaland transitional circulations and to put forward the best currentunderstanding of the functional significance of each of thesefactors.

erinatal Imaging and Health,SE1 7EH, UK. Tel.: þ44 (0)20

emore).

2. Brief summary of cardiac anatomy

The workhorses of the cardiac musculature are car-diomyocytes ‒ individual muscle units that split and recombineinto a complex network, which is then encased in a collagenousextracellular matrix. The extracellular matrix both withstandsand dissipates the forces created by the muscle fibers [1]. Thecardiomyocytes are made up of myofibrils, themselves formedfrom chains of sarcomeres. Within each sarcomere, muscle fila-ments cross-link and relax to produce muscular contraction.Surrounding these fibers and linking them to the extracellularmatrix is a cytoskeleton of protein or endomysium [1], the con-stituents of which have a crucial role in development andremodeling of the myocytes in response to mechanical andchemical signals [2].

The sarcoplasmic reticulum (SR) is a network of sleeve-likestructures surrounding the myofibrils within a muscle fiber. Itsrole is the storage, release and reuptake of calcium that is crucial tothe conversion of an action potential to a muscular contraction.When an action potential arrives at the plasma membrane of themuscle fiber, it is conducted via the T tubules to the sarcoplasmicreticulum, which releases calcium into the cytosol [3]. This binds totroponin, a complex of three regulatory proteins, I, C, and T, whichcontrol contraction and relaxation of the muscle in the presence ofcalcium (Fig. 1) [4]. When calcium levels increase, some binds totroponin, altering its shape and allowing actin‒myosin cross-bridges to form and contraction to occur. Removal of the calcium

Fig. 1. Troponin (Tn) complex structure at low (a) and high (b) Ca2þ concentrations. TnI, TnC, and TnT are shown in blue, red, and yellow, respectively. Reproduced with permissionfrom Katrukha [3].

A. Finnemore, A. Groves / Seminars in Fetal & Neonatal Medicine 20 (2015) 210e216 211

from troponin then restores the tropomyosin blocking action andthe muscle relaxes.

3. The classical triad of circulatory physiology: preload,contractility, and afterload

3.1. Preload

This is the initial stretching of the cardiac muscle fibers prior tocontraction. As preload increases and sarcomeres stretch, there isan increase in the number of actin‒myosin bridges formed (up to acertain point) and therefore of the scope for force generation [5].The association between increasing preload and increasing cardiacoutput is known as the Frank‒Starling mechanism. However,whereas numbers of actin‒myosin bridges were previouslyconsidered the central component of the Frank‒Starling mecha-nism, awareness of cardiac anatomy and physiology continues toevolve. It is now clear that titin (also known as connectin), thelargest known mammalian protein, plays a key role in myocytefunction [5,6]. Titin links the Z-disc region of the thin filament tothe myosin thick filament (Fig. 2) [7], and has an elastic structurewhich allows it to stretch. This makes it a key component ofmyocardial stiffness, but also enhances its role in force generation[5]. Although this is partly due to release of the kinetic energystored as potential energy during stretch of the molecule duringdiastole, titin also mediates force generation by altering Ca2þ

release. At the time of writing, many aspects of titin's functionremain unknown or contentious. The properties of titin are com-plex, and highly variable across species, over time, within differentlayers of the heart, with altering isoforms (over days/weeks) andchanging calcium concentrations (over milliseconds during thecardiac cycle). The characteristics of this complex molecule havebeen the subject of a number of excellent reviews [8,9].

3.2. Contractility

This is the ability of sarcomeres to change their inherent con-tractile force (independent of preload), a trait unique to cardiacmuscle. Increased inotropy (or contractility) manifests as increasedvelocity of muscle fiber shortening and therefore increased ejectionvelocity of blood. As the time available for ejection of blood isrelatively constant, increased ejection velocity over a constanttime-period will result in increased ejection volume [10].Contractility is determined largely by the interaction of the sym-pathetic and parasympathetic autonomic nervous systems [10].Circulating catecholamines also have an impact, as do poorly un-derstood interactions with heart rate (the Treppe effect) [10] andafterload (the Anrep effect) [11]. Most changes in cardiac contrac-tility appear to be mediated by changes in intracellular calciumlevels [10].

3.3. Afterload

This is the load against which the ventricle must eject blood andis probably best thought of in terms of vascular resistance, althoughit is not synonymous with this. As afterload increases, so does thewall stress in the ventricle, decreasing the rate at which musclefibers contract and (with ejection time being relatively constant)decreasing the ejection volume. Changes in vessel diameter(particularly small arteries and arterioles) regulate blood flowwithin individual organs, and the total resistance created by each ofthese vascular beds is systemic vascular resistance (SVR) [10].Vasoregulation is determined by intrinsic myogenic mechanisms,local vasoactive mediators and extrinsic factors including theautonomic innervation of blood vessels and circulating vasoactivehormones [10]. Blood pressure is the product of vascular resistanceand blood flow and is therefore frequently closely associated with

Fig. 2. Schematic arrangement of myofilaments in an extended and contracted sarcomere. The upper, middle, and lower panels illustrate the highly extended, shortening, andshortened sarcomere, respectively. Reproduced with permission from Nagy et al. [7].

Fig. 3. Predicted force developed by single titin molecule as a function of sarcomerelength (SL). Titin develops restoring force below the slack SL (1.9 mm) and passive forceabove the slack SL. Reproduced with permission from Lahmers et al. [16].

A. Finnemore, A. Groves / Seminars in Fetal & Neonatal Medicine 20 (2015) 210e216212

afterload. Particularly in the transitional circulation it is thoughtthat systemic vascular resistance has a dominant influence overblood flow in determining blood pressure [12,13].

The separation of the effects of preload, contractility and after-load in hemodynamic physiology is somewhat artificial since theclosed nature of the cardiovascular system means that any changein one component of the circulation will have ‘knock on’ effectselsewhere. For example, an increase in afterload will cause a reflexincrease in myocardial inotropy, though this will only partly pre-serve stroke volume. Residual blood not ejected from the leftventricle will act to increase preload, also enhancing contractileforce [10].

4. Additional components of circulatory physiology: recoil,torsion, and rotation

4.1. Recoil

The classical description of preload given in Section 3.1 conjuresan entirely passive process. However, it is now known that diastolicfunction and cardiac filling are much more dynamic processes. Inthe early stages of diastole there is a rapid filling phase known asrecoil, in which blood is effectively suctioned into the ventricles[14]. Titin again seems to play a key role here. Isolated sarcomereshave a resting length, or “slack length,” of around 1.9 mm. Whenstretched beyond that length, sarcomeres naturally contract downto their slack length in an energy-independent fashion. This processis now known to be due to stretching of subregions of titin. Simi-larly, when compressed to below 1.9 mm in length, sarcomeres willnaturally lengthen back to their slack length when the contractingforce is removed [15]. Although the exact mechanism of the processis uncertain, it may be that titin proteins are anchored such thattitin has its shortest length when the sarcomeres measure 1.9 mm.Alternatively, titin might be further shortened at sarcomere lengths<1.9 mm, expanding back to its own resting length when the

compressing force is removed, effectively acting as a molecularcoiled spring [6]. In either case, titin is the principal determinant ofthis recoil [15], and the extent of the expansile force at lengths<1.9 mm is equivalent to the contractile force at length >1.9 mm(Fig. 3) [16].

4.2. Torsion

The contraction of the myocardium in the healthy human heartis not simply a circumferential movement. Innate to the efficiencyof the myocardium is the structure of differently oriented fibers tomaximize contractile efficiency. The role of these oblique layers hasbecome more clearly defined in recent years and the concept of awringing motion of contraction has been introduced. In adulthearts, the apex rotates in a counter-clockwise direction to the base,so twisting the ventricle and squeezing blood up towards theoutflow tract [17]. Sub-epicardial and sub-endocardial fibers move

A. Finnemore, A. Groves / Seminars in Fetal & Neonatal Medicine 20 (2015) 210e216 213

in opposite directions to increase the efficiency of this wringingmotion [1]. The systolic twist not only assists with the ejection ofblood, but also stores further energy, augmenting recoil asdescribed above [14].

4.3. Rotation

Another contributor to this early ventricular filling phasemay bethe complex intracardiac rotational flow blood patterns that areconsistently seen, at least in the adult heart [18]. As blood entersthe left atrium from the pulmonary veins, or the right atrium fromthe superior and inferior venae cavae, it does not spill randomlyinto the chamber, and then wait idly for opening of the atrioven-tricular valves. Instead it forms highly reproducible rotational flowpatterns. These patterns were initially proposed to maintain thekinetic energy of inflowing blood, allowing it to ‘slingshot’ into theventricles in early diastole [18]. However, the degree of kineticenergy conserved by this process has been debated [19], and therole of the process in the healthy or sick heart is currently difficultto quantify [20].

5. Unique aspects of the fetal and transitional circulation

5.1. Structural/flow path differences in the fetal circulation

The central unique feature of the fetal circulation is that gasexchange occurs not in the lungs, but via the placenta. Relativelywell-oxygenated blood returns to the fetus from the placenta via asingle umbilical vein. This blood has a partial pressure of oxygen ofaround 3.7 kPa (27 mmHg), at which 70e80% of fetal hemoglobin issaturated [21e23]. Blood flows from the umbilical vein into theumbilical recess where it mixes with a small volume of blood fromthe portal vein. Around 50% of this blood passes through the ductusvenosus to bypass the liver and join with the inferior vena cava andleft and right hepatic veins immediately before entering the rightatrium [24].

Oxygenated blood from the ductus venosus is preferentiallydiverted into the left atrium by the Eustachian valve. Passage intothe left atrium is possible since higher pressure in the right atrium(due to high volume venous return) stents open the valve over theforamen ovale [25]. This relatively well-oxygenated blood joinswith a small volume of pulmonary venous return to flow into theleft ventricle before being ejected into the ascending aorta. Thusthe brain, heart, and upper body are preferentially suppliedwith relatively highly oxygenated blood (oxygen saturation around65%) [25].

The remainder of the venous return to the right atrium, from theinferior vena cava, hepatic veins and superior vena cava, flows intothe right ventricle before being ejected into the pulmonary artery.However, only a small proportion (10e25%) of the right ventricularoutput reaches the lungs to supply the lung tissue's basic metabolicneeds [26], with the proportion increasing at higher gestations [27].The remainder of the right ventricular output is diverted throughthe ductus arteriosus into the descending aorta. In addition toperfusing the abdominal organs and lower limbs of the fetus, thisdeoxygenated blood (oxygen saturation around 35%) flows to thelow-resistance placenta to collect oxygen and dispose of carbondioxide and other waste products.

5.2. The placental circulation

The development of an adequate placental circulation,providing a low-resistance vascular bed for efficient gas andnutrient exchange between maternal and fetal blood, is vital to thedevelopment of the fetus and to the formation of a functional fetal

cardiovascular system. Its impairment has been shown to corre-spond to the development of some congenital heart defects [28,29]as well as to cardiovascular dysfunction later in life, as proposed inthe Barker hypothesis [30].

From the onset of the fetal heartbeat at around 22 days ofgestational age, until the cessation of organogenesis at 9 weeks,both the vitelline circulation from the yolk sac and the umbilico-placental circulation are functional. At the end of this period, theyolk sac regresses and the umbilico-placental circulation becomesdominant. At around 10e12 weeks of gestational age, trophoblasticinvasion of the placenta begins. This appears to be influenced byhypoxia, cytokines, growth factors, enzymes, and other angiogenicsubstances [31]. Under the control of these biochemicalmodulators,fetal trophoblastic cells from the surface of the blastocyst invade theuterine spiral arteries, lining them with fetal endothelial cells,increasing their diameter and reducing impedance. This produces alow-resistance, high-flow, low-velocity vascular bed, allowingcopious maternal blood flow and efficient gas and nutrient ex-change. Dysfunction of the placenta and assessment of the placentalcirculation using Doppler ultrasound is discussed in detail by Pruetzet al. in this issue of Seminars. Processes such as poor trophoblastinvasion ormaternal ischemia lead to high placental impedance andhypoxemic vasodilatory adaptation in the fetus to ensure adequateoxygenation of vital organs. This in turn produces a raised rightventricular pre- and afterload, exacerbated by abnormal retrogradediastolic flow in the aortic isthmus secondary to cerebral vasodi-latation, and eventually failing coronary perfusion with potentialmyocardial damage. Many factors may affect placental develop-ment and trophoblast invasion, such as maternal diet, blood pres-sure, and ingestion of alcohol, lithium, or cigarettes [28].

6. Magnetic resonance imaging (MRI) quantification of fetalblood flow

Advances in MRI technology, particularly the development ofmetric optimized gating, have recently allowed quantification ofblood flow in the great vessels of near-term fetuses in utero [27].Metric optimized gating is retrospective processing of non-electrocardiogram-gated MRI data where over-sampled data aresorted with a range of possible heart rates until artifacts areminimized and the true heart rate is identified. Despite the tech-nical challenges of this process, the measurements are relativelyrobust. These studies therefore provide normative ranges for flowvolumes and distributions, and allow verification of true physi-ology, matching closely with values taken from instrumented ani-mal models. Prsa et al. have shown that the mean combinedventricular output in the near-term fetus is 465 mL/kg/min, witharound 55% of this output coming from the right ventricle, and 45%coming from the left. In the near-term fetus, around 25% of the rightventricular output (15% of the combined ventricular output) rea-ches the lungs through the right and left pulmonary arteries; theremaining 75% of the right ventricular output flows through theductus arteriosus to supply the abdomen, lower body and theplacenta. Around 75% of the relatively well-oxygenated left ven-tricular output (30% of the combined ventricular output) perfusesthe head, neck, and arms, with the remaining 25% joining theductus arteriosus flow in the descending aorta. Around 50% of thedescending aortic flow (30% of the combined ventricular output)flows to the placenta [27].

7. Functional differences in the fetal circulation

The fetal heart also has significant functional differences relativeto the adult heart, which profoundly alter its function.

A. Finnemore, A. Groves / Seminars in Fetal & Neonatal Medicine 20 (2015) 210e216214

7.1. Compliance

How the compliance of the fetal heart compares to the moremature circulation is contentious. It had previously been acceptedthat the premature myocardium is a less compliant structure thanthat of the term infant and that this has an impact on preload andcontractility [32]. However, more recent data [33] have suggestedthat fetal compliance may be significantly higher (more thandouble) than in the adult. This increased compliance appears to bepredominantly due to differences in titin subtypes. The fetal (N2BA)isoform represents nearly all cardiac titin throughout gestation,then rapidly disappears after birth [16,33]. Compared to the adult(N2B) isoform the fetal isoform has a high intrinsic compliance andthis provides the fetal myocardium with a low passive tension andreduced myocardial stiffness (Fig. 3).

It is thought that these variations in the protein structure of themyocardium enable the fetal heart to produce an adequate outputeven at the reduced filling pressures (3e4 mmHg [34]) seen inutero [33]. The high compliance may also be important to allowintrinsic stretch of the sarcomeres, which is thought to trigger theirproliferation and hence the growth of cardiac structures [33].Maturational changes in collagen make-up and rate of calciumextrusion are also likely to contribute to altered ventricularcompliance in the fetus [35].

7.2. Passive versus active ventricular filling

Ventricular filling is biphasic, consisting of phases traditionallyreferred to as passive and active. In the adult circulation, ‘passive’filling predominates, though as discussed above this is more com-plex than simple relaxation of the ventricle, and is governed byrecoil of titin, untwisting of ventricular torsion, rotational flowpatterns in the atria, and likely many other factors. Ventricularfilling in late diastole is then augmented by atrial contraction. Therelative contributions of these phases are readily studied byDoppler echocardiography in both adults and fetuses. In the fetalcirculation the maximum velocity of passive filling (E wave) isconsistently lower than the peak of active filling (A wave), pro-ducing an E/A ratio of <1 [36]. As gestation advances, E/A ratioincreases, but is rarely>1, so active filling predominates throughoutgestation.

As with compliance, multiple factors are likely influencing E/Aratio. Low E/A ratio was previously cited as an indicator of lowcompliance ‒ a stiffer ventricle being less likely to fill spontane-ously [35]. However, with the increasing awareness of the actions oftitin, it could also be argued that a low E/A ratio is a marker ofhigher compliance. As shown in Fig. 3, a highly compliant fetalmyocardium will have a shallow slope of the length/force curve,such that, as the sarcomere is stretched, little additional force isgenerated. But similarly, at lengths below the slack length (as seenat end systole), a compliant myocardium will also have low recoil.Less recoil could then manifest as less passive filling, and a lower E/A ratio.

Second, the venous return to the left atrium is low throughoutgestation since pulmonary blood flow is minimal, such that fillingvolumes of the left side of the heart depend almost entirely onforamen ovale shunt. Low preload is known to be strongly associ-ated with poor passive ventricular filling and low E/A ratio. Sincethe pulmonary blood flow ‒ and therefore the left atrial venousreturn ‒ increases through gestation, this could partly explain therise in E/A ratio seen with advancing maturity.

Third, the normal rotational flow patterns, which may maintainthe kinetic energy of inflowing atrial blood, have been shown to bedisrupted in the neonatal heart [37], and may also therefore beinefficient in the fetal heart.

7.3. Heart rate

The normal fetal heart rate is between 110 and 160 beats perminute. This means there is less time available for diastolic filling.In addition, as at higher heart rates, ejection time is relatively wellpreserved, whereas filling time is disproportionately reduced.

7.4. Contractility

The fetal heart has a reduced inherent contractility. Less of thefetal myocardium is made up of contractile elements [32,38]. Thesarcoplasmic reticulum and the T tubules are immature in thefetus and the calcium shifts for contraction and relaxation aremore reliant on the extracellular levels [35]. Perhaps to counteractthis immaturity, the fetal heart displays a higher percentage of afetal troponin I isoform, which raises calcium sensitivity [39].However, troponin C, which is responsible for maximum forcegeneration in response to calcium levels, is low through much ofgestation, only increasing in late gestation. The fetal left ventricleis also likely to have fewer b-adrenoceptors [40], a prime driver ofcontractility in response to circulating catecholamines. Similarly,sympathetic innervation of the fetal heart is significantlyreduced [41].

8. The transitional circulation in the preterm infant

Immediately prior to birth the placenta is a low resistancevascular bed ensuring that low pressure deoxygenated blood canreturn for exchange, while the lungs represent an area of highresistance and the pulmonary flow is therefore diverted. If born atterm, a healthy baby's first few breaths inflate the lungs, rapidlydropping the resistance they present to the circulatory system andincreasing pulmonary blood flow. At around the same time (seeKluckow and Hooper in this issue of Seminars for a detailed dis-cussion of the importance of ventilation around the time of cordclamping), the umbilical cord is cut, blocking the low-pressureplacental bed. Due to a sharp increase in left atrial return fromthe pulmonary veins, left-sided atrial pressure rises rapidly, func-tionally closing the septal layers of the foramen ovale [42]. Theductus arteriosus will then gradually close under the influence ofoxygen and the metabolism of prostaglandins in the fully func-tioning lungs. At the end of this transition the full return of deox-ygenated blood from the body to the right heart is directed to thelungs for oxygenation, and the full return of oxygenated blood fromthe lungs to the left heart is now directed to the body to supply thetissues [42].

There are multiple reasons why this process can be disruptedwhenever infants are born prematurely. The key factors are dis-cussed below.

8.1. Low inherent contractility

The immature myocardium has impaired contractility due to adecreased proportion of contractile elements, altered calciumrelease, altered titin function, decreased b-adrenoceptor numberand decreased sympathetic innervation [39e41]. Animal and hu-man studies conclusively show that the innate contractility of thepretermmyocardium is impaired [32,43]. Preterm infants also havehigher levels of serum troponin T, a marker of myocardial injury.The rise in troponin T is particularly marked in those who havesuffered from hypoxia either in utero or at birth or in infants withcontinuing respiratory distress syndrome [44]. This suggests thatthere may be a level of myocardial damage or even programmedapoptosis occurring in these infants that may affect the contractionand relaxation properties of their hearts.

Practice points

� The fetal and neonatal circulations differ significantly

from the circulation seen at any other point in develop-

ment; interpretation of clinical and imaging findings

should be made in that context.

� Therapeutic interventions during the transitional circula-

tions should target specific pathologies in preload,

contractility, systolic function, diastolic function, after-

load, and shunting through the fetal pathways.

� As inter-individual differences in circulatory physiology

are broad, response to therapeutic interventions should

be regularly reassessed in every patient with evidence of

circulatory failure.

Research directions

� Technological advances, particularly in fetal imaging,

should examine the changes in fetal circulatory function

in the presence of diseases such as placental failure,

congenital heart disease, and twin‒twin transfusion

syndrome.

� Improvements in clinical assessment of circulatory

physiology after birth are urgently required; a range of

invasive and non-invasive methodologies are entering

the clinical arena including but not restricted to echocar-

diography, MRI, impedance electrical velocimetry moni-

toring, near-infrared spectroscopy, and stroke volume

variability.

A. Finnemore, A. Groves / Seminars in Fetal & Neonatal Medicine 20 (2015) 210e216 215

8.2. Poor tolerance of high systemic vascular resistance

The left and right ventricles both face an abrupt rise in vascularresistance at the time of cord clamping. Unless born after cho-rioamnionitis or with sepsis, high resting a-adrenergic tone in theperipheral vasculature in preterm neonates maintains constrictionof the capacitance arterioles in many vascular beds, furtheraccentuating high vascular resistance. Echocardiographic studies inpreterm infants have shown that the immature myocardium notonly has reduced contractility, but that it is also particularly sen-sitive to increases in afterload [45].

8.3. Impaired diastolic filling

As seen in the fetus, the low compliance of fetal isoforms of titinmean that early diastolic recoil is ineffective [33]. Rotational flowpatterns are disrupted, potentially having an effect on preservationof kinetic energy of inflowing blood during atrial filling [37]. Highheart rate means that there is limited time available for ventricularfilling. Most inotropes and vasopressor-inotropes are also chrono-tropic, and will therefore further limit the time available for ven-tricular filling [46]. Although the reliance of ventricular filling onatrial contraction reduces through gestation, it is still significantlymore important for diastolic filling in neonates than in older chil-dren and adults [47], with an E/A ratio consistently below 1.0.Preterm neonates appear to have notably compromised early dia-stolic left ventricular filling, even when compared to term infants[48]. Infants who have suffered intrauterine growth restriction alsoshow a significant persistence of impaired passive filling [49].

8.4. Persistence of fetal shunt pathways

High prostaglandin levels, relatively low oxygen levels and theinherent immaturity in the muscular wall of the ductus arteriosusare among the leading factors delaying effective ductal closure inpreterm infants. In the vast majority of preterm infants pulmonaryvascular resistance drops to below systemic levels within minutes/hours of birth, such that direction of blood flow within the ductusreverses to become left to right [42]. Particularly in the current eraof antenatal steroids and surfactant administration there is goodevidence that high volume left-to-right ductal shunt can occur evenon the first day of postnatal life [50]. This causes significant pul-monary hyperperfusion, giving a risk of pulmonary hemorrhage,and in a number of preterm infants will also cause systemichypoperfusion. For a comprehensive discussion of the hemody-namic role of the ductus arteriosus, see the article by Evans in thisissue of Seminars. The foramen ovale can also remain patent inmany premature infants. Due to the large amount of bloodreturning to the left atrium in many preterm neonates with pul-monary overcirculation caused by the sustained ductal patency, thedirection of shunting reverses to become left to right, potentiallyworsening their pulmonary hyperperfusion and systemichypoperfusion.

9. Long-term effects of preterm delivery on heartdevelopment

It is becoming clear that the consequences of the immatureheart undergoing the circulatory transition do not end in thenewborn period.When the hearts of ex-preterm infants are studiedin early adulthood, they have significantly reduced chamber length,increased myocardial mass, and impaired systolic and diastolicfunction [51]. Provisional work in the newborn period suggests thatthese changes may be even more marked in the extremely preterminfants who currently survive to discharge from the neonatal unit.

When studied by cardiac MRI, extremely preterm infants have in-creases in left ventricular mass of more than 50% compared withterm controls [52].

That pathologic remodeling occurs following preterm birth is tobe expected given the requirement for cardiac growth in thehealthy fetus. Fetal myocytes are in mononucleated form, andrespond to stress by hyperplasia as opposed to the pattern seenlater in life when the mature binucleated myocytes respond withhypertrophy. Lamb models show that the hearts of animals bornpreterm have a five- to seven-fold increase in extracellular matrix, a20% increase in cardiomyocyte volume, and a significant increase inthe proportion of myocytes remaining in the mononucleated pro-liferative state, even at term-corrected age [53].

The triggers for and functional significance of the remodelingseen following preterm birth are currently largely unknown, but itis feasible that cardiac remodeling puts surviving ex-preterm in-fants at risk of heart failure and ischemic heart disease.

10. Conclusion

The developing cardiovascular system has functionally signifi-cant alterations in preload, contractility, afterload, diastolic filling,and intracardiac flow patterns. Whereas these changes occur underphysiological circumstances in utero, they become rapidly patho-logical in the setting of preterm birth. Abnormal physiology is thenfurther compounded by persistence of the fetal shunt pathways. Anunderstanding of the development of the circulatory physiologycan assist the practising clinician with management of circulatoryfailure in the premature infant.

A. Finnemore, A. Groves / Seminars in Fetal & Neonatal Medicine 20 (2015) 210e216216

Conflict of interest statement

None declared.

Funding sources

None.

References

[1] Ho SY. Anatomy and myoarchitecture of the left ventricular wall in normaland in disease. Eur J Echocardiogr 2009;10:iii3e7.

[2] McCain ML, Parker KK. Mechanotransduction: the role of mechanical stress,myocyte shape, and cytoskeletal architecture on cardiac function. PflügersArch 2011;462:89e104.

[3] Katrukha IA. Human cardiac troponin complex. Structure and functions.Biochemistry 2013;78:1447e65.

[4] Kobayashi T, Solaro RJ. Calcium, thin filaments, and the integrative biology ofcardiac contractility. Annu Rev Physiol 2005;67:39e67.

[5] Fukuda N, Granzier HL, Ishiwata S, Kurihara S. Physiological functions of thegiant elastic protein titin in mammalian striated muscle. J Physiol Sci 2008;58:151e9.

[6] LeWinter MM, Granzier H. Cardiac titin: a multifunctional giant. Circulation2010;121:2137e45.

[7] Nagy A, Cacciafesta P, Grama L, Kengyel A, M�aln�asi-Csizmadia A,Kellermayer MSZ. Differential actin binding along the PEVK domain of skeletalmuscle titin. J Cell Sci 2004;117:5781e9.

[8] Linke WA, Hamdani N. Gigantic business: titin properties and functionthrough thick and thin. Circ Res 2014;114:1052e68.

[9] Granzier HL, Labeit S. The giant protein titin: a major player in myocardialmechanics, signaling, and disease. Circ Res 2004;94:284e95.

[10] Berne R, Levy M. Physiology. 3rd ed. St Louis: Mosby, Inc.; 1993.[11] Nichols CG, Hanck DA, Jewell BR. The Anrep effect: an intrinsic myocardial

mechanism. Can J Physiol Pharmacol 1988;66:924e9.[12] Pladys P, Wodey E, Beuch�ee A, Branger B, B�etr�emieux P. Left ventricle output

and mean arterial blood pressure in preterm infants during the 1st day of life.Eur J Pediatr 1999;158:817e24.

[13] Kluckow M, Evans N. Relationship between blood pressure and cardiac outputin preterm infants requiring mechanical ventilation. J Pediatr 1996;129:506e12.

[14] Rothfeld JM, LeWinter MM, Tischler MD. Left ventricular systolic torsion andearly diastolic filling by echocardiography in normal humans. Am J Cardiol1998;81:1465e9.

[15] Helmes M, Trombit�as K, Granzier H. Titin develops restoring force in ratcardiac myocytes. Circ Res 1996;79:619e26.

[16] Lahmers S, Wu Y, Call DR, Labeit S, Granzier H. Developmental control of titinisoform expression and passive stiffness in fetal and neonatal myocardium.Circ Res 2004;94:505e13.

[17] Al-Naami GH. Torsion of young hearts: a speckle tracking study of normalinfants, children, and adolescents. Eur J Echocardiogr 2010;11:853e62.

[18] Kilner PJ, Yang GZ, Wilkes AJ, Mohiaddin RH, Firmin DN, Yacoub MH. Asym-metric redirection of flow through the heart. Nature 2000;404:759e61.

[19] Watanabe H, Sugiura S, Hisada T. The looped heart does not save energy bymaintaining the momentum of blood flowing in the ventricle. Am J PhysiolHeart Circ Physiol 2008;294:H2191e6.

[20] Markl M, Kilner PJ, Ebbers T. Comprehensive 4D velocity mapping of the heartand great vessels by cardiovascular magnetic resonance. J Cardiovasc MagnReson 2011;13:7.

[21] Dawes BYGS, Mott JC, Widdicombe JG. The foetal circulation in the lamb.J Physiol 1954;126:563e87.

[22] Di Tommaso M, Seravalli V, Martini I, La Torre P, Dani C. Blood gas values inclamped and unclamped umbilical cord at birth. Early Hum Dev 2014;90:523e5.

[23] Kotaska K, Urinovska R, Klapkova E, Prusa R, Rob L, Binder T. Re-evaluation ofcord blood arterial and venous reference ranges for pH, pO2, pCO2, accordingto spontaneous or cesarean delivery. J Clin Lab Anal 2010;24:300e4.

[24] Edelstone DI, Rudolph AM, Heymann MA. Liver and ductus venosus bloodflows in fetal lambs in utero. Circ Res 1978;42:426e33.

[25] Murphy PJ. The fetal circulation. Contin Educ Anaesth Crit Care Pain 2005;5:107e12.

[26] Schneider DJ, Moore JW. Patent ductus arteriosus. Circulation 2006;114:1873e82.

[27] Prsa M, Sun L, van Amerom J, et al. Reference ranges of blood flow in themajor vessels of the normal human fetal circulation at term by phase contrastmagnetic resonance imaging. Circ Cardiovasc Imaging 2014:663e71.

[28] Linask KK, Han M, Bravo-Valenzuela NJM. Changes in vitelline and utero-placental hemodynamics: implications for cardiovascular development.Front Physiol 2014;5:390.

[29] Midgett M, Rugonyi S. Congenital heart malformations induced by hemody-namic altering surgical interventions. Front Physiol 2014;5:287.

[30] De Boo HA, Harding JE. The developmental origins of adult disease (Barker)hypothesis. Aust NZ J Obstet Gynaecol 2006;46:4e14.

[31] Kn€ofler M, Europe PMC Funders Group. Critical growth factors and signallingpathways controlling human trophoblast invasion. Int J Dev Biol 2010;54:269e80.

[32] McPherson RA, Kramer MF, Covell JW, Friedman WF. A comparison of theactive stiffness of fetal and adult cardiac muscle. Pediatr Res 1976;10:660e4.

[33] Walker JS, de Tombe PP. Titin and the developing heart. Circ Res 2004;94:860e2.

[34] Johnson P, Maxwell DJ, Tynan MJ, Allan LD, Hospital C, Road G. Intracardiacpressures in the human fetus. Heart 2000;128:59e63.

[35] Rychik J. Fetal cardiovascular physiology. Pediatr Cardiol 2004;25:201e9.[36] Pacileo G, Paladini D, Russo MG, Pisacane C, Santoro G, Calabr�o R. Echocar-

diographic assessment of ventricular filling pressure during the second andthird trimesters of gestation. Ultrasound Obstet Gynecol 2000;16:128e32.

[37] Groves AM, Durighel G, Finnemore A, et al. Disruption of intracardiac flowpatterns in the newborn infant. Pediatr Res 2012;71:380e5.

[38] FriedmanW. The intrinsic physiologic properties of the developing heart. ProgCardiovasc Dis 1972;15:87e111.

[39] Reiser PJ, Westfall MV, Schiaffino S, Solaro RJ. Tension production and thin-filament protein isoforms in developing rat myocardium. Am J Physiol1994;267:H1589e96.

[40] Kim MY, Finch AM, Lumbers ER, et al. Expression of adrenoceptor subtypes inpreterm piglet heart is different to term heart. PLoS One 2014;9:e92167.

[41] Friedman WF, Pool PE, Jacobowitx D, Seogren SC, Braunwald E. Sympatheticinnervation of the developing rabbit heart. Circ Res 1968;23:25e32.

[42] Hines MH. Neonatal cardiovascular physiology. Semin Pediatr Surg 2013;22:174e8.

[43] Toyono M, Harada K, Takahashi Y, Takada G. Maturational changes in leftventricular contractile state. Int J Cardiol 1998;64:247e52.

[44] Trevisanuto D, Zaninotto M, Altinier S, Plebani M, Zanardo V. High serumcardiac troponin T concentrations in preterm infants with respiratory distresssyndrome. Acta Paediatr 2000;89:1134e6.

[45] Takahashi Y, Harada K, Kishkurno S, Arai H, Ishida A, Takada G. Postnatal leftventricular contractility in very low birth weight infants. Pediatr Cardiol1997;18:112e7.

[46] Cox DJ, Groves AM. Inotropes in preterm infants ‒ evidence for and against.Acta Paediatr Suppl 2012;101:17e23.

[47] Mori K. Pulsed wave Doppler tissue echocardiography assessment of the longaxis function of the right and left ventricles during the early neonatal period.Heart 2004;90:175e80.

[48] Koz�ak-B�ar�any A, Jokinen E, Saraste M, Tuominen J, V€alim€aki I. Development ofleft ventricular systolic and diastolic function in preterm infants during thefirst month of life: a prospective follow-up study. J Pediatr 2001;139:539e45.

[49] Crispi F, Hernandez-Andrade E, Pelsers M, et al. Cardiac dysfunction and celldamage across clinical stages of severity in growth-restricted fetuses. Am JObstet Gynecol 2008;199. 254.e1e8.

[50] Groves AM, Kuschel CA, Knight DB, Skinner JR. Does retrograde diastolic flowin the descending aorta signify impaired systemic perfusion in preterm in-fants? Pediatr Res 2008;63:89e94.

[51] Lewandowski AJ, Bradlow WM, Augustine D, et al. Right ventricular systolicdysfunction in young adults born preterm. Circulation 2013;128:713e20.

[52] Cox D, Bai W, Price AN, Edwards AD, Groves AM, Rueckert D. PS-011 quan-titative assessment of preterm left ventricular anatomical development andremodelling using neonatal cardiac MRI and atlasing techniques. Arch DisChild 2014;99:A115e6.

[53] Bensley JG, Stacy VK, De Matteo R, Harding R, Black MJ. Cardiac remodelling asa result of pre-term birth: implications for future cardiovascular disease. EurHeart J 2010;31:2058e66.