Pediatric pulmonary valve replacements: Clinical challenges and emerging technologies

Abstract Congenital heart diseases (CHDs) frequently impact the right ventricular outflow tract, resulting in a significant incidence of pulmonary valve replacement in the pediatric population. While contemporary pediatric pulmonary valve replacements (PPVRs) allow satisfactory patient survival, their biocompatibility and durability remain suboptimal and repeat operations are commonplace, especially for very young patients. This places enormous physical, financial, and psychological burdens on patients and their parents, highlighting an urgent clinical need for better PPVRs. An important reason for the clinical failure of PPVRs is biofouling, which instigates various adverse biological responses such as thrombosis and infection, promoting research into various antifouling chemistries that may find utility in PPVR materials. Another significant contributor is the inevitability of somatic growth in pediatric patients, causing structural discrepancies between the patient and PPVR, stimulating the development of various growth‐accommodating heart valve prototypes. This review offers an interdisciplinary perspective on these challenges by exploring clinical experiences, physiological understandings, and bioengineering technologies that may contribute to device development. It thus aims to provide an insight into the design requirements of next‐generation PPVRs to advance clinical outcomes and promote patient quality of life.

through the heart 16,17 ; however, the lack of ideal replacement prosthetics generates a requirement for repeat interventions throughout early life. 18,19 Prosthetics based on metals are prone to thrombosis and require lifelong anticoagulation, introducing additional burdens that persist past childhood, 20,21 while those based on biological materials exhibit underwhelming durability due to their immunogenicity and structural degradation. 22,23 Prosthetic failure is generally exacerbated in the pediatric population due to their active lifestyle and robust immune responses accelerating adverse biological responses and structural degradation. 24,25 Furthermore, the somatic growth of pediatric patients is not accommodated by current PV prosthetics, which all operate at a fixed diameter. 12,14,26,27 A growing mismatch between patient and prosthetic sizes gradually impairs hemodynamic performance and cardiovascular health. 24 Taken together, these failure modes result in an unacceptably high requirement for reoperation in the pediatric population, leaving the patient and family under sustained physical, emotional, and financial hardships. [28][29][30] Compared to unaffected children, those with CHD spend twice as long in hospital at 10 times the expense, and their risk of early mortality is 10-fold. 28 Within 10 years of RVOT intervention, pediatric patients can expect to spend more than 80 days in hospital at a cost of over $200,000 (USD, 2017), with a large portion of this time and money dedicated to reintervention. 29 Thus, there is an urgent demand for the development of novel pediatric PV replacements (PPVRs) that may reduce the current reoperation requirements for pediatric patients with CHD.
Furthermore, over 90% of CHD patients now survive into adulthood, resulting in a growing lifetime prevalence of adult congenital heart disease 9,15,31 ; for example, the number of people aged between 50 and 69 years old living with CHD has increased by 117% since 1990. 3 Heart valve reinterventions are the most common procedure in adult CHD, 15   Truncus arteriosus Lack of separation between the left ventricular outflow tract (LVOT) and RVOT, resulting in a single valved conduit that separates into the aorta and pulmonary artery 8 Pulmonary, aortic 1.0 The pulmonary artery is separated from the main trunk, and continuity between the right ventricle and pulmonary artery is typically achieved with a valved conduit 8 Congenital aortic valve diseases For example, bicuspid aortic valve 10 Aortic, pulmonary 4.6 Aortic valve disease be treated by transplantation with the patient's own pulmonary valve in the Ross procedure, thus requiring a valve replacement in the PV position 11 Note: Note that disease frequencies underrepresent the frequency of PPVR implantation as patients inevitably require multiple procedures. discussion of the bioengineering advances that may facilitate the innovation of next-generation PPVRs.

| PHYSIOLOGY OF THE PULMONARY VALVE
The PV is an important structure in right heart function as it mediates unidirectional blood flow from the right ventricle into the pulmonary artery by preventing backflow during diastole 16,32 (Figure 1a). PV dysfunction is characteristic of many CHDs and their surgical repairs, such as in Tetralogy of Fallot, 1,6 truncus arteriosus, 8 and transposition of the great arteries. 1,34 In health, the PV consists of three crescentshaped leaflets anchored to a crown-shaped annulus, itself adjoined to the myocardium of the RVOT. 35 Anatomically, the sinuses describe the spaces between each leaflet and the RVOT wall, the commissures describe the points along the annulus where two leaflets meet, and the interleaflet triangles describe the regions underneath each commissure 35 (Figure 1b). At the microstructural level, the leaflet extracellular matrix (ECM) exhibits three distinct layers: the fibrosa, spongiosa and ventricularis 16,32,33 (Figure 1c). The fibrosa layer is closest to the arterial surface of the leaflet and is characterized by a dense network of circumferentially aligned Type I collagen fibers, 35,36 which continues into the annulus via the commissures. 32,35 When the valve is open, the collagen occupies a corrugated configuration, which uncrimps as the valve closes. 36 On the ventricular side of the leaflet, networks of radially aligned elastin fibers form the ventricularis layer. 36,37 In between the fibrosa and ventricularis, the spongiosa is predominated by highly hydrated proteoglycans and glycosaminoglycans. 37,38 A suite of other proteins (such as fibronectin, periostin, and vitronectin) are also present throughout the ECM to maintain mechanical integrity and mediate cellular activity. 38 The valvular surface is lined with a circumferentially orientated monolayer of valvular endothelial cells, while its interior is populated by valvular interstitial cells. 37,38 Pathological valvular interstitial cell differentiation into an osteoblast-like phenotype stimulates the upregulation of osteogenic genes and deposition of calcium phosphates, resulting in calcific valvular diseases. 16,32,38 As this review aims to provide a resource for the development of next-generation PPVRs, an understanding of their biomechanical requirements is important. In this respect, the structural and functional similarities between the PV and the aortic valve, collectively termed semilunar valves, 35,37 allow the translation of aortic-specific biomechanical studies to this exploration of the PV. The PV opens and closes with each cardiac cycle (Figure 2a), summing to at least 3 billion cycles over an individual's lifetime. 16,17 During systole, the PV opens to eject blood into the pulmonary artery and closes during diastole to prevent backflow. 16,32 Systolic opening is induced by the F I G U R E 1 Pulmonary valve structure. (a) Unidirectional blood flow through the heart is facilitated by four heart valves (pulmonary, aortic, mitral, and tricuspid). The pulmonary valve mediates flow between the right ventricle and pulmonary artery. contraction of the right ventricle, generating pressures around 25 mmHg to drive blood through the PV. [43][44][45] Once blood has been ejected, closure is facilitated by the rapid formation of fluid vortices in the sinus regions, occurring shortly after the inertial flow along the arterial wall reverses direction 40 (Figure 2b). At peak diastole, the pressure gradient across the PV from the arterial surface is around 6 mmHg. 43,44 Thus, heart valves are inherently passive structures, opening and closing in response to the dynamic pressures of its cardiac environment.
Within each cycle, the PV is subject to tensile, flexural and shear stresses. 37,46 Tensile stress is applied during diastole to guard backflow, wherein elastin components initially carry the load while the collagen is still corrugated (Figure 2c, top). 36 At around 7% strain, 47 these collagen corrugations flatten, absorb the tensile load, and transfer it into the pulmonary root 36 (Figure 2c, bottom); ultimately, the collagen-rich fibrosa bears the majority of the tensile load. 37 This microstructural heterogeneity establishes a nonlinear stress-strain profile 46 and anisotropic tensile properties; leaflets tend to be stiffer and stronger in the circumferential direction but more extensible in the radial direction 41,48 (Figure 2d). Flexural stresses predominate the stress cycle, as flexure is the principal deformation mode in heart valves. 36 Based on the assumptions of Euler-Bernoulli simple beam theory, leaflets bent around the circumferential plane are twice as stiff when bending against their natural curvature as compared to bending with their natural curvature. 42,49 Conversely, no differences in flexural stiffness are observed when bending around the radial plane 42  Figure 2e). Functionally, this facilitates valve opening but prevents inversion during closing. 42,49 This is permitted by the heterogeneous microstructure of the leaflets, with the stiffer fibrosa on the concave surface and the extensible ventricularis on the convex surface. 36,37 Finally, blood flowing through the open valve imparts shear stresses on the leaflets, which are largest during early systole; for example, under normal adult resting conditions (70 bpm and 5 L/min cardiac output), the aortic valve typically experiences maximum stresses of around 7 Pa on the inflow leaflet surface 50,51 and 2 Pa on the outflow surface. 52

| CURRENT PEDIATRIC PULMONARY VALVE REPLACEMENTS
The first-line treatment for congenitally diseased PVs is almost always repair in order to preserve the native tissue. 12,14 For example, pulmonary stenosis may be treated with balloon pulmonary valvuloplasty, wherein the expansion of a transcatheter balloon forces the stenotic leaflets open. 9,53,54 However, PV replacement is often inevitable, 12,14 such as when valvuloplasty or valvectomy eventuates in significant pulmonary regurgitation. 9 A range of heart valve replacements (HVRs) have been introduced since the 1950s, including mechanical valves, bioprosthetic valves homografts and autografts ( Figure 3). In recent years, an emerging family of polymeric valves have been explored; however, none have yet achieved clinical and commercial translation.
No ideal PPVR is currently available, with all contemporary options exhibiting limitations that preclude extended durability, thus requiring patients to undergo multiple surgeries.

| Mechanical valves
Mechanical valves are constructed from rigid materials such as steel, titanium alloys, and pyrolytic carbon. 56 In the pediatric population, the use of mechanical valves is restricted to implantation in the systemic circulation with adequate outcomes. 61,62 Importantly, mechanical valves are rarely used in the pulmonary circulation due to incompatibilities with right ventricular pressures and an exacerbated requirement for anticoagulation. In general, novel PPVRs would likely be utilized in place of bioprosthetic valves; however, if their performance and durability are suitable then they could also be considered in cases Introduction timeline of HVRs with application in the pediatric population. The first ball-and-cage mechanical valve was introduced in 1952, 55 with the first generation of bileaflet mechanical valves introduced in 1963. 56 The first homograft was implanted in 1962, 57 the first use of an autograft in the Ross procedure was reported in 1967. 58 The first chemically-treated bioprosthesis was described in 1968 59 and the first transcatheter bioprosthesis was successfully deployed in 2000. 60 where mechanical valves would be selected, warranting a brief discussion on the limitations of mechanical valves. Thrombosis is a persistent concern for mechanical valves, attributed to their artificial metallic construction that promotes biofouling 20,24,63 and nonphysiological flow profile that induces cavitation. 64 Prevention of thrombosis is managed by pharmaceutical treatment with vitamin K antagonists 20 and more recently a range of novel oral anticoagulants such as dabigatran, rivaroxaban, apixaban and edoxaban. 65 However, lifelong anticoagulation predisposes bleeding risks, 24 requires patients to adhere to a strict pharmaceutical routine, and necessitates regular blood tests to monitor treatment effect and to adjust dosage. 20 Furthermore, all mechanical valves function at a fixed diameter, thus precluding the accommodation of somatic growth that occurs during early life. While some surgical strategies involve deliberately dilating the native annulus to implant an oversized prosthetic, this process involves additional incisions that can inadvertently damage ventricular function and the conduction pathway. 12

| Bioprosthetic valves
Bioprosthetic valves are constructed from biological tissues, often in combination with metallic or polymeric supports, and demonstrate more physiological flow profiles and reduced susceptibility to thrombosis than mechanical valves. Bioprosthetics are commercially available for surgical or transcatheter delivery, 66,67 and are treated with glutaraldehyde to preserve structural integrity and reduce immunogenicity. 24 Various bioprosthetics have historically been used as PV replacements, including the Hancock II (a stented porcine aortic valve), 66 the Carpentier-Edwards PERIMOUNT (a stented bovine pericardial valve) 66,68 and the Freestyle (a porcine aortic root). 69 In 1999, the bovine jugular vein conduit (BJVC) emerged as a promising graft for RVOT reconstruction due to the valve being naturally integrated within the conduit. 70 A BJVC for surgical delivery was commercialized as the Contegra valve (Medtronic, USA), with utility in RVOT diameters between 12 and 22 mm. 69,70 The advent of transcatheter PV replacement (TPVR) in 2000 60 introduced a non-invasive alternative to surgery, 54,67,71 while achieving the same hemodynamic performance. 72 Today, TPVR is preferred to surgical methods 54 80,81 and Venus P (Venus MedTech, China). 82,83 All these devices utilize an hourglass-shaped self-expandable nitinol stent, with the Harmony and Venus P devices incorporating valves made from porcine pericardial tissues 25,82 and the Alterra Adaptive Prestent providing a landing site for the SAPIEN valve. 81 The longevity of bioprosthetic valves is limited by their susceptibility to structural valve deterioration (SVD), 22 which describes the progressive degradation of the bioprosthetic tissue, resulting in hydrodynamic dysfunction. 23,77 Calcification is typical of SVD 84 and leads to leaflet stiffening, 25,85 while various other hallmarks include immune responses and the foreign body reaction (FBR), 23,83,85 mechanical degradation, 23,82,86 and the deposition of advanced glycation end products. 87 As the nonviable bioprosthetic tissue lacks repair capabilities, these myriad factors accumulate toward significant structural failure. 77,88 The exacerbation of SVD in the younger population is generally attributed to the more active lifestyle of this demographic accelerating the wear on the prosthetic. It may also emerge from the more vigorous immune system of younger patients, [23][24][25] and the lack of growth potential in contemporary fixed-diameter bioprostheses. 14 While the Melody valve can be slightly over-dilated to correct increasing transvalvular pressures, [89][90][91] there is no evidence for any longterm capacity in accommodating growth. Bioprostheses may also fail due to infective endocarditis or thrombosis. 22 Infective endocarditis is an infection of the blood facing surfaces around implanted HVRs, 92,93 with reintervention typically required due to its severity. 71,72,94,95 It is predominantly a concern after transcatheter procedures, with an incidence between 7.5% and 17% after TPVR. 96,97 Endocarditis occurs more in BJVCs than in homografts, and the transcatheter Melody valve exhibits significantly greater susceptibility than the surgical Contegra valve. 98 Finally, thrombosis is relatively rare in bioprosthetics, and it is remarkedly less concerning than in mechanical valves with events often limited to the first 3 months after implantation as the device undergoes endothelialization. 99,100 In fact, the mortality associated with bioprosthetic valve reoperation is significantly less than that associated with mechanical valve bleeding events. 100

| Homografts
Homografts are transplanted from human donors, demonstrating considerably greater durability and better freedom from reoperation than bioprosthetics. 101 However, they suffer SVD at unpredictable rates and valve competence in the medium term is variable. Reoperation to replace the homograft is frequent and younger age at implantation continues to be a risk factor for reoperation, 101 again likely due to somatic outgrowth and immune responses. 13 Some groups have suggested that decellularizing the homograft may improve longevity by removing immunogenic material. 102 For example, decellularized homografts implanted in the pulmonary position of adolescents and young adults have reported slightly better durability than standard homografts and bovine jugular veins. 103 However, there is no clinical consensus on the importance of decellularization, with many studies demonstrating no benefit to homograft longevity. 104,105 Furthermore, immunogenicity is not entirely eliminated by decellularization, and implants may still experience immune responses against residual cellular components or remnants of detergents used in the decellularization process. 106 Nonetheless, homografts are an excellent graft in comparison to bioproshtetics, 19,101 although their application remains imperfect due to their limited availability and inevitable SVD.

| REOPERATION REQUIREMENTS
No commercially available PPVRs are ideal and RVOT reconstruction will often require repeat operations later in life to replace dysfunctional prosthetics. 18

| BIOFOULING
The nonspecific adsorption of blood proteins onto foreign materials, herein termed biofouling, occurs within seconds of exposure to blood 121 (Figure 4a). This is thought to be associated with a suite of undesirable biological pathways that can preface HVR failure and reoperation, including thrombosis, immune responses and the FBR, calcification, and infection.

| Thrombosis
Thrombosis is a universal challenge for all blood-contacting medical devices due to its notorious role in device failure, vessel occlusion, and systemic complications. 63  induce the contact activation system, albeit at a much slower rate. 121 The mechanisms underlying thrombosis on these surfaces are not thoroughly understood, however, may be related to their tendency to denature adsorbed proteins. 121 There is also evidence suggesting that the contact activation system may be triggered via the increased generation of neutrophil extracellular traps around hydrophobic surfaces, as these biological scaffolds are known to directly promote both FXII activation and platelet interactions. 125 Thus, thrombosis on various biomaterial surfaces may be inhibited by avoiding the adsorption of blood proteins such as fibrinogen, vWF, FXII and C1q.

| Immune responses and the foreign body response
The FBR consists of immune-mediated inflammatory and fibrotic pathways stimulated by the implantation of foreign biomaterials. 126,127 The first step in the FBR is the adsorption of blood pro- simultaneously recruited and differentiate into proinflammatory M1 macrophages, 131,132 becoming the dominant immune cell within days to mediate chronic inflammation. 126,127 In bioprosthetics, this contributes to SVD as structural proteins are degraded through proteolytic enzymes or reactive oxygen species. 23 Where biomaterials resist degradation, such as those ideal for polymeric valves, persistent inflammation stimulates macrophages to fuse into foreign body giant cells (FBGCs) to augment phagocytic performance while avoiding apoptosis. 126,131 Importantly, FBGC formation is dependent on the extent of early fibrinogen adsorption and fibrin polymerization. 128 Alongside FBGC formation, anti-inflammatory M2 macrophages differentiate and stimulate myofibroblasts to deposit a collagenous matrix around the implant. 126,127 This can surround the material with a fibrotic capsule, 126,131 impairing the mechanical utility of the implanted device and isolating it from the host. 126,127 This fibrous growth may stiffen valve leaflets, and thus degrade the mechanical performance of HVRs by limiting leaflet movement. 23 Thus, the inhibition of complement adsorption and those proteins associated with thrombosis might aid in mitigating immune responses and the FBR.

| Calcification
Calcification is a common failure mode in native heart valve tissue, 32 bioprosthetics and many polymeric valve prototypes 116,[133][134][135] ; however, its mechanism differs depending on the material (Figure 4d). In bioprosthetic valves, calcification occurs due to abnormally high calcium ion (Ca 2+ ) levels in the intracellular space of devitalized tissues, enabled by the glutaraldehyde crosslinking process that damages the cell walls and deactivates cellular Ca 2+ efflux channels. 23,25,82 Interactions between intracellular Ca 2+ and phosphorus-rich cellular compounds, such as the cell membrane or nucleic acids, then form calcium phosphate crystals. 23,25,82 Additionally, residual free aldehydes from glutaraldehyde crosslinking can react with circulating Ca 2+ to induce calcification. 23,25,82 On polymeric valves, the mechanism of calcification is less clear; however, it has been associated with thrombosis, 134 mechanical stress and failure, and regions of low flow. [133][134][135][136] To this end, calcification on polymeric materials may be closely linked to thrombus formation, suggesting that inhibiting thrombus initiation may indirectly mitigate calcification.

| Infection
Infective endocarditis is a common concern after TPVR, 71 97 Early stages involve the association of microbes from transient bacteremia with thrombus at the valve site, eventually forming a microbial-thrombotic vegetation 137 (Figure 4e). Moreover, current models implicate fibrinogen, vWF and platelets as key intermediaries for bacterial adhesion to the endocardial surface. [138][139][140] For example, S. aureus and Staphylococcus lugdunensis directly bind to vWF to overcome the high shear stresses of the valve environment. 139,140 Similarly, S. aureus also depends on platelets and fibrinogen 138,140 ; in fact, S. aureus adhesion has been reported to reduce threefold after medication with aspirin and ticagrelor. 138 Infection commonly results in biofilm formation, which increases microbial tolerance to host immunity and antimicrobial treatment. 141 Thus, infective endocarditis is facilitated by the biofouling of thrombosis-inducing proteins such as fibrinogen and vWF, suggesting that infection may be mitigated by resisting biofouling.

| Surface modifications for antifouling biomaterials
Thrombosis, the FBR, and infection directly depend on biofouling, while calcification is indirectly dependent via its association with thrombosis. This suggests that developing biomaterials with resistance to biofouling might mitigate these downstream consequences, allowing for the fabrication of PPVRs with long-term resistance to biodegradation and dysfunction. Many polymeric biomaterials used in biomedical settings are chemically inert to reduce biological responses. 142,143 Out of the various polymers explored for polymeric valves, polyurethanes are most promising due to their biomimetic mechanical properties that combine strength and elasticity, 115,144 with those based on siloxane groups exhibiting the greatest resistance to biofouling, compared to those based on ester, ether or carbonate groups. [144][145][146][147] While polyurethanes are more resistant to biofouling than materials such as poly(ethylene terephthalate) 148 and polytetrafluoroethylene, 149 they are inferior to poly(styrene-b-isobutylene-b-styrene) 150,151 and ePTFE. 148 Nonetheless, all synthetic biomaterials remain prone to biofouling, 63,121 necessitating additional strategies to diminish this susceptibility. A predominant strategy involves modifying a polymer's surface chemistry to improve resistance to biofouling at the biomaterial-blood interface without compromising the mechanical utility of the bulk polymer. 152 Strategies for surface modifications (Table 2) are classified as either bioactive, which actively counteract biological processes or support material repair, or passive, which aim to eliminate any interactions between the blood and bulk material. 152

| Bioactive surface modifications
The only commercially available bioactive modification is heparin, 152 commonly grafted using the CARMEDA method. 153 Heparin augments the function of antithrombin, which inhibits various clotting factors in the contact activation system, such as thrombin and activated FXII. 153 CARMEDA heparin modifications have been used to extend the lifespan of various blood-contacting devices, including extracorporeal membrane oxygenation circuits, ventricular assist devices, coronary stents and vascular grafts. 153 However, heparin relies on a particular five-sugar sequence that is only present in onethird of molecules in a commercial heparin sample, and inactive molecules may induce thrombosis due to its strong negative charge. 153 Heparin may also be enzymatically degraded or shielded by adsorbed proteins, limiting heparin coatings to short-term applications. 152 Furthermore, a rare but serious side effect is heparin-induced thrombocytopenia, an immune response to heparin complexes that causes extensive thrombosis with significant morbidity and mortality. 154 Nonetheless, bioactive strategies do not prevent biofouling but instead counteract their downstream consequences. While heparin mitigates thrombosis, other specific bioactive modifications would be simultaneously required to neutralize immune responses, calcification, and infection.

| Passive surface modifications
In comparison, passive strategies simply provide a barrier that blocks interactions between the bulk polymer and the blood, 152  Antifouling surface modifications with potential application in next-generation PPVRs.

Surface modification Class Mechanism Effect on biofouling Effect on downstream events Limitations
Heparin Bioactive Actively inhibits contact activation system proteins by augmenting antithrombin function 153 None. Biofouling may actually hinder heparin mechanism 152 Reduces thrombus formation 153 Lack of long-term stability. 152 Inactive heparin may be thrombotic. 153 May trigger heparin-induced thrombocytopenia 154 Tethered liquid perfluorocarbons Passive Prevents biomolecule adhesion with a lowfriction liquid surface, consisting of a liquid perfluorocarbon layer associated with a network of molecular tethered perfluorocarbons 155,156 Reduces fibrinogen adsorption 155 Reduces platelet adhesion 155 and thrombus formation. 155 Prevents microbial adhesion and biofilm formation 155,157 Currently no evidence for long-term stability in a cardiac environment Poly(ethylene glycol) Passive Generates a hydration layer via hydrogen bonding and an energetic layer due to water displacement and PEG compression being thermodynamically unfavorable 158,159 Reduces fibrinogen adsorption 160 Extends circulation time of modified nanoparticles in the bloodstream. 158 Mitigates platelet adhesion and microbial adhesion 161 Hydration layer remains susceptible to biofouling 162 PEG oxidatively degraded in vivo 158 Induces anti-PEG antibodies 158 Zwitterions Passive Generates a physical hydration layer via ionic bonding and an energetic barrier due to water displacement being thermodynamically unfavorable 162,163 Reduces fibrinogen adsorption [164][165][166][167][168][169][170][171][172] Reduces platelet adhesion 164,168,[173][174][175][176][177] and thrombus formation. 168,170,171,173,176,178 Reduces antibody responses, 179,180 immune cell recruitment 164,169,175,176 and inflammation. 180 Reduces fibroblast adhesion, 165,166,171,174 tissue formation 164,169 and fibrotic capsule formation. 164 Reduces calcification 168,175,181 and SVD 168,181 in bioprosthetics. Prevents adhesion of bacteria, 170,171,174,176,177,182,183 viruses 182 and fungi 171 and limits biofilm formation 177,182,183 Acidic pHs decrease surface hydration by carboxybetaine-based zwitterions. 184,185 High ionic concentrations in solution hinder hydration by screening the zwitterionic charges. 184 The opposite ions of sulfobetainebased zwitterions may self-associate. 186 Currently no evidence for long-term stability in a cardiac environment However, long-term stability studies are required before such modifications can be applied on indwelling devices such as HVRs.
Alternatively, hydrophilic materials associate a layer of water around the material to generate this passive barrier. 152  An emerging family of hydrophilic materials are zwitterions, which bind water molecules via ionic interactions 152,190 and generate significantly superior hydration compared to PEG. 191 Zwitterions are molecules that contain oppositely charged ionic groups but are electrically neutral overall. 192 Zwitterions bind water molecules in a similar structural state as they would be in bulk solution, thus making displacement by blood proteins thermodynamically unfavorable, generating an energetic barrier in addition to a physical barrier. 162,163 Three common zwitterions explored are those containing sulfobetaine, 164,165,169,171,175,176,179,180,183 carboxybetaine, 164,169,174,178,179,[193][194][195] and phosphocholine 168,170,173,179 moieties. Zwitterionic modifications significantly diminish biofouling, as illustrated by fractional rates of fibrinogen adsorption, 164 grafts has yet to be investigated, although some shorter-term stability studies have been reported. 165,170 For example, a phosphocholine-based layer grafted on to polyvinyl chloride via amine-functionalization demonstrated no significant changes after 30 days of agitation in phosphatebuffered saline. 170 Interestingly, a sulfobetaine-based layer grafter on to polyurethane via the Fenton reaction only demonstrated 7 days of stable attachment in distilled water, 165 suggesting the importance of the grafting strategy in achieving stable surface modifications. Thus, the long-term stability of zwitterionic grafts in the cardiac environment must be further investigated before they may be used in devices such as PPVRs.

| Patient growth and prosthetic-patient mismatch
In general, clinical outcomes of RVOT reconstruction identify a younger age at implantation as a significant risk factor in early PPVR failure and reoperation. 19 In addition to the amplified consequences of biofouling in this age group, this observation may be explained by patient outgrowth of the device. 12,14,18,26,27,196 Clinically, this is described as prosthetic-patient mismatch (PPM), 197 24 and requiring reoperation to upsize the valve. 27 The PPM may also exacerbate biological responses to the device, as a progressively non-physiological flow profile through the HVR may promote thrombosis. 200,201 Additionally, PPM has been correlated to greater rates of SVD in bioprosthetics. 198,199 Younger age at implantation has universally been associated with poorer prognoses, 61,62,74,120,202 which may be partly due to the increased incidence of PPM associated with younger patients. 203 Indeed, cardiovascular structures grow substantially in the first 20 years of life; importantly to PPVRs, the diameter of the PV annulus typically enlarges from 7 to 24 mm over this period. [204][205][206] It should be noted that somatic outgrowth has been proposed as a minor concern compared to valve degradation 207 ; however, this finding more accentuates the challenge of adverse biological reactions than it does diminish the challenge of somatic outgrowth.
Some reports suggest that valve insufficiency may be delayed by oversizing the PPVR implantation, 13,208,209 which typically involves implanting a valve that is two sizes larger than the patient anatomy (e.g., implanting a valve with a 14 mm diameter for a 10 mm diameter anatomy); however, others refute the long-term usefulness of this strategy. 210,211

| Growth-accommodating PPVRs
A growth accommodating PPVR would facilitate remarkable advances in the treatment and management of CHDs by undergoing structural expansion to match its growing environment, while maintaining its hydrodynamic function. Various prototypes have been proposed for this purpose, including biological venous valves, tissue engineered heart valves (TEHVs), and purely synthetic growth-accommodating heart valves (SGHVs) ( and structural growth over time. 225,226 A valve constructed from this material was implanted in the PV position of 4-month-old lambs, where its diameter and EOA gradually enlarged over 1 year as the lambs grew into adulthood. 213 In fact, leaflet height was maintained over the year and the leaflet free-edge actually increased in length, 213 which is a significant observation considering that leaflet contraction has been historically associated with TEHVs. 225,[227][228][229] However, incomplete recellularization was observed and the preprogramed mechanical properties in the recellularized regions were altered in vivo, 213 with both observations reinforcing earlier findings, [223][224][225]230 and a gradual increase in valvular insufficiency was observed. 213 Nonetheless, the valve was able to accommodate the complete growth of the lamb (correlating to a diameter increase of 32%) without progressing past moderate regurgitation, 213  Critically, the clinical translation of TEHVs is largely limited by a lack of cellular control within the implanted scaffold. A common observation is the contraction of neo-tissue in the recellularized leaflets, resulting in leaflet shortening and valve insufficiency. 225,[227][228][229] This is a particular trepidation for pediatric applications as valve insufficiency would likely be exacerbated by vascular growth. Another concern is the significantly heterogeneous remodeling rates both within and between valves, 232,234 which bars certainty regarding the composition and quality of the remodeled valve. Consequently, recent work has explored strategies to better predict and control the quality of the remodeled valve, whether through optimizing the design and fabrication process, 39,235 or better understanding the underlying cellular processes. 232 39,110,216,239,247 In this review alone, simulations have accurately predicted structural changes due to cellular contraction, 39 compared the expansile capability of various biological valves, 216 and analyzed the mechanical performance of a bileaflet SGHV. 110 In silico studies could readily analyze the functional performance of an expanding prototype without committing excess time to manufacture, thus streamlining the iterative development of growth-accommodating PPVRs. As this technology evolves, simulation components may allow the application of artificial intelligence and digital twins, 248,249 wherein the child's growth could be predicted and PPVR expansion modeled with unprecedented patient-specificity. In addition, an emerging range of so-called "meta-biomaterials" offer unique and non-intuitive mechanical properties 250 that might permit PPVR expansion. For example, auxetic materials can expand simultaneously along two perpendicular axes due to carefully engineered micro-or nanoscale architectures, whereas normal materials tend to shrink in the axis perpendicular to that being stretched. 251 Funding acquisition (equal); resources (lead); writingreview and editing (supporting). Sina Naficy: Conceptualization (supporting); funding acquisition (equal); project administration (lead); resources (supporting); supervision (lead); writingreview and editing (equal).

DATA AVAILABILITY STATEMENT
Data available on request from the authors.