Mayer
J
Ei
. Tissue Engineering for Cardiac Valve Surgery.
Cohn Lh, ed. Cardiac Surgery in the Adult. New York: McGraw-Hill, 2008:1649-1656.
| ||||||||||||||||||||||||||||||||||||||||||||
CHAPTER 70 |
Tissue engineering is a developing science, which brings together engineering and biology in an attempt to develop replacement tissues beginning with their individual cellular components. The impetus for our work on tissue engineered cardiovascular structures derives from the need to replace cardiovascular tissues that failed to develop normally during embryogenesis or have become dysfunctional as a consequence of disease, as well as from the fact that currently available replacement structures have significant limitations. In the cardiovascular sphere, the structures most often affected, excluding atherosclerotic coronary disease, are the cardiac valves and the great vessels. Diseases of the heart valves and large "conduit" arteries account for approximately 60,000 cardiac surgical procedures each year in the United States, but all of the currently available replacement devices have significant limitations.1,2 Ideally, any valve or artery substitute would function in similar fashion to the normal valve or artery to allow blood to pass through it without narrowing or leakage, but would also have the following characteristics: (1) durability, (2) growth potential (for infants and children), (3) compatibility with blood so that thrombus would not form on its surface and activation of inflammatory cascades would not occur, and (4) resistance to infection. None of the currently available devices constructed from either prosthetic or biologic materials meet these criteria. Prosthetic heart valves are very durable, but they require anticoagulation to reduce the risk of thrombosis and thromboembolism.1,2 Notwithstanding anticoagulation, the incidence of thromboembolic complications is not zero.1,2 Biologic valves, whether they are of allograft or heterograft origin, remain subject to structural deterioration after implantation.2–4 Neither prosthetic nor biologic valves have any growth potential, and this limitation represents a major source of morbidity for pediatric patients who must undergo multiple reoperations to replace valves and/or valved conduits as the patients grow. The tissue engineering approach to overcome these shortcomings offers the possibility of creating replacement cardiovascular structures from cells and scaffolds with the result that the tissue-engineered construct is a living structure with the capacity for growth, repair, and remodeling similar to normal tissues. This chapter summarizes some of the progress that has been made in tissue engineering research as it relates to cardiac valves and conduit arteries, and to then outline the areas where additional efforts must be focused in order to make cardiovascular tissue engineering a clinical reality.
CELL TYPE OR TYPES FOR TISSUE-ENGINEERED HEART VALVES ENGINEERING OF CELLULAR PHENOTYPE STRUCTURAL ORGANIZATION SCAFFOLD CYTOKINE SIGNALS BIOMECHANICAL SIGNALS IN VIVO MATURATION References
Much of the strength and flexibility in normal tissues is due to specialized proteins and polysaccharide-protein complexes (extracellular matrix) that are produced by the cells in the tissue.5,6 For cardiac valves, the biomechanical demands are particularly high, as there are approximately 40 million openings and closures of the valves per year. The normal valve presents minimal resistance to opening and no pressure gradient during systolic forward flow. In diastole this same structure must close promptly and completely to prevent valvular regurgitation, and must resist pressure differences between the diastolic arterial or pulmonary artery pressure and the ventricular diastolic pressure. The extracellular matrix of the normal semilunar heart valve is not homogeneous,5 and the arrangement of the extracellular matrix seems uniquely designed to provide a high degree of flexibility during systole but a high degree of strength to resist the diastolic pressure load.5 In order for a tissue-engineered cardiac valve to function effectively, it seems intuitively obvious that extracellular matrix production, composition, and remodeling processes are critical, and should result in a structure that is similar to the native valve. The determinants of extracellular matrix composition in a given tissue are incompletely understood, but are determined to some extent by the types of cells in the tissue, by the interactions among these cell types, by a variety of cytokine signals that the cells receive, and by the biomechanical signals that the cells receive.5
One potential source of insights for the development of a tissue-engineered heart valve (TEHV) is the process by which the normal valve develops from the embryonic to the mature adult form. A recent publication by Aikawa and colleagues7 demonstrates that the normal semilunar valve undergoes significant in vivo maturational changes during development, which include changes in the extracellular matrix composition, collagen fiber alignment, and cellularity, and the development of a nonhomogeneous layered architecture. The mature semilunar valve structure with a more dense layer of collagen on the valve surface facing the sinus of Valsalva, (fibrosa layer) a middle layer of glycosaminoglycans, (spongiosa) and an elastin-rich layer facing the ventricular cavity and the flow from the ventricle to the great artery (ventricularis) seems uniquely designed to allow optimal semilunar valve function.6 However, fetal semilunar valves and even those obtained from children are more cellular and do not have a fully mature valve structure.7 Thus the semilunar valves undergo significant maturation in vivo under continuous biomechanical and other signaling conditions, and studies of tissue-engineered pulmonary valves from our laboratory have shown a similar evolution in histologic appearance, extracellular matrix composition, and organization of the collagen fibers in the extracellular matrix.8 The signals and genetic changes controlling the early embryonic development of mammalian semilunar valves have recently been reviewed by Armstrong and Bischoff,9 and it has become evident that the genetic mechanisms and signaling pathways are quite complex and tightly controlled, and are not completely understood at this point. The insights gained from an understanding of normal semilunar valve development is of potential application to the development of TEHVs, although it seems that the complexities of normal valve development will be difficult to replicate completely.
In our laboratory at the Children’s Hospital, Boston and in the laboratories of other investigators, attempts to create TEHVs have been made over the last decade with varying degrees of success,10–18 constructing these valves from a variety of cell types and scaffolding materials. Most of the in vivo studies have been carried out in the lower-pressure pulmonary circulation, which is more forgiving of an imperfectly functioning valve construct. The longest successful in vivo implant from our laboratory is 8 months, but the optimal process for engineering of a tissue-engineered pulmonary valve remains incompletely defined. If a tissue-engineering approach to the creation of a heart valve substitute is to be successful, several basic questions must be addressed: (1) What cell type or combination of cell types is necessary to allow the production and maintenance of an appropriate extracellular matrix? (2) To what extent can cellular phenotype be altered or "engineered" to replicate cells found in the normal valve? (3) How can these cells be spatially organized during the development of the tissue-engineered structures until sufficient extracellular matrix is produced by the cells in the construct? (4) What biochemical signals are necessary during the development of these structures to ensure proper extracellular matrix production? (5) What mechanical signals are necessary for the optimal tissue development? (6) Should a tissue-engineered valve construct be completely developed and "mature" prior to implantation, or can there be further "maturation" of a tissue-engineered construct in vivo after implantation in a fashion similar to the age-related maturational changes in the normal valve?7 Progress has been made in addressing each of these questions, and this progress will be reviewed.
CELL TYPE OR TYPES FOR TISSUE-ENGINEERED HEART VALVES
A considerable effort has been made to explore the optimal cell source (or sources) for tissue-engineering applications. From a conceptual standpoint, the cells for a TEHV can be derived from fully differentiated cells with the capability to synthesize collagen, elastin, and glycosaminoglycans or from less committed, pluripotent stem cells with the potential to differentiate into multiple cell types. A second-order decision regarding cell source is the choice of autologous versus allograft cells. This choice will have significant implications if the tissue-engineering approach is to become a clinical reality. In order to avoid any confounding effects of immune rejection in our laboratory, we have based all of our research on the use of autologous cell lines, and it is our intuition at this time that if a tissue-engineering approach to development of a cardiac valve is to be successful, then it will likely involve the use of autologous cells, unless a nonantigenic cell source is identified or selective immune tolerance can be established. The ultimate choice of cell type(s) will likely reflect a complex interplay between what is favored in terms of basic cell biology, what tissues are readily available for harvest in the individual patients, and what is clinically acceptable to both patient and physician.
Considerable early success in fabricating a TEHV was achieved in our laboratory using myofibroblasts and endothelial cells derived from the systemic arteries of immature animals.10–14 These cells were chosen by virtue of their anatomic location within the arterial limb of the cardiovascular system and their known ability to synthesize structural extracellular matrix components such as collagen and elastin. A comparison of myofibroblasts from the wall of the ascending aorta with those from discarded segments of saphenous vein revealed that the latter cells exhibit superior collagen formation and mechanical strength when cultured on biodegradable polyurethane scaffolds.19 We were able to construct TEHVs based on these cells from the systemic blood vessels that functioned for periods of up to 4 months in vivo.10,12 However, enhanced collagen formation may be a double-edged sword. The rapid formation of new tissue in the early culture period could give rise to uncontrolled proliferation and synthesis of matrix elements, leading to decreased flexibility and potential tissue shrinkage. Early studies with dermal fibroblasts showed that leaflets constructed from these cells did develop tissue contraction, which limited the ability of these leaflets to coapt with other leaflets effectively.21 In addition, mature cells may present a problem of senescence in long-term cell cultures in vitro, which limits the ability to produce sufficient numbers of cells to seed a TEHV construct. Finally, the prospect of harvesting segments of artery from an otherwise normal peripheral circulation in order to obtain cells for a TEHV represents a somewhat unattractive solution clinically, and therefore led to the search for alternative cell sources.
The search for a more clinically palatable source of cells led us to investigate the use of stem cells for tissue-engineering applications. Our initial experience was gained with autologous endothelial progenitor cells (EPCs) that were isolated from the circulating blood of lambs and then seeded onto decellularized arterial segments.20 These seeded arterial grafts were then implanted as an interposition graft in the carotid artery of the donor lamb.20 These grafts remained patent and functional for up to 130 days.20 Cebotari and colleagues have recently reported an initial experience with seeding EPCs onto homograft valves followed by implantation into two children.39 In our subsequent studies, Sutherland and associates used ovine bone marrow mesenchymal cells to seed a biodegradable scaffold formed into a three-leaflet valve within a conduit19 (Fig. 70-1). These valved conduits were implanted to replace the pulmonary valve for periods up to 8 months and functioned well hemodynamically. An echocardiographic image of the valve in open position in vivo is shown in Fig. 70-2. Importantly, these valve leaflets underwent a "maturational" process after implantation18 similar to our earlier findings using myofibroblasts and endothelial cells from systemic arteries,12 and in both sets of experiments a layered histologic appearance developed during the time after implantation. Other investigators have also used bone marrow as a cell source for tissue-
engineered cardiovascular structures. Matsumura and associates have shown that bone marrow cells labeled with green fluorescent protein can be seeded onto a copolymer of lactic acid and 
-caprolactone and found that these seeded cells contribute to the histogenesis of a tissue-engineered vascular graft. Their tissue-engineered vascular grafts remained patent and at explant the constructs contained green fluorescent protein–labeled cells which expressed both endothelial-specific and smooth muscle cell–specific markers.21 The group at Tokyo Women’s Medical College has also carried out implants of tissue-engineered vascular grafts in children undergoing repair of certain types of congenital heart disease using a similar technique with encouraging early results.23
|
|
ENGINEERING OF CELLULAR PHENOTYPE
Among the more interesting observations on mesenchymal stem cells is that the phenotype of these cells seems to be dependent on the local environment in which they come to reside.29 However, the factors in these local environments that control differentiation into various phenotypes remain unknown.29 For EPCs, there is evidence that these cells have the ability to transdifferentiate in response to various signals in a tissue-engineering environment. Dvorin and colleagues have shown that transforming growth factor (TGF)-β1 can induce EPCs to express 
-smooth muscle actin (-SMA) after these cells had been seeded onto a tissue engineering scaffold of polyglycolic acid (PGA)/poly-4-hydroxybutyrate copolymer.30 This -SMA expression is not characteristic of endothelial cells or EPCs, and its expression suggests that the EPCs have undergone a phenotypic change into a cell type resembling cardiac valve interstitial cells. Human aortic valve endothelial cells, but not vascular endothelial cells, respond to TGF-β1 in a similar fashion, suggesting that EPCs may be an appropriate cell type to serve as valve endothelium.30 It is noteworthy that endothelial transdifferentiation to a mesenchymal phenotype is a critical step during the normal embryonic development of cardiac valves, and Dvorin and associates have speculated that the ability to recapitulate some normal steps in embryonic valve development is relevant to the choice of cell types for tissue engineering of cardiac valves.30 More recent studies by Sales and coworkers in our laboratory have shown that exposure of EPCs to TGF-β1 while being grown on tissue-engineering scaffolds not only begin to express -SMA, but also demonstrated increased production of laminin, fibronectin, tropoelastin, and collagen when compared to EPCs seeded on the same scaffold without TGF-β1.38
An additional issue for cell selection for a TEHV is whether better results will be obtained by seeding the scaffold with more than one cell type and whether the different cell types are seeded at the same time or sequentially. In our initial studies, both endothelial cells and vascular smooth muscle cells were seeded sequentially10 onto a PGA scaffold with favorable results. In subsequent in vitro studies, we observed an interaction between human EPCs and vascular smooth muscle cells when these cells were co-seeded onto PGA/poly-L-lactic acid (PLLA) scaffolds such that microvascular tube formation was observed when these two cell types were co-seeded versus when EPCs alone were seeded onto the same scaffold material.31 The potential for interactions among cells of differing origin and phenotype adds another dimension of complexity to the tissue-engineering process. It is quite clear that normal embryologic development of semilunar valves involves important interactions between cells of different embryologic origins and between cells and the extracellular matrix surrounding them.9 It has seemed reasonable to us to conclude that similar interactions could play an important role in the evolution of a TEHV.
STRUCTURAL ORGANIZATION SCAFFOLD
Although it is has been possible to grow individual types of cells in culture for some time, it is more difficult to induce these cells to assemble or organize into the more complex structural arrangements that are found in normal tissues or to produce normal extracellular matrix components in an organized fashion. For this reason, most attempts to create tissue-engineered structures have started with a scaffold onto which cells are "seeded." Any scaffold for tissue engineering applications must be biocompatible and allow cells to adhere and proliferate. Thus the chemistry of the scaffold itself and the scaffold degradation products must be nontoxic at the outset. If the tissue-engineered construct is to have any growth potential, then ultimately the scaffold must either degrade or be able to be remodeled. A fundamental difference in the various tissue-engineering strategies in different laboratories has centered on this choice of scaffold materials. One option is to use decellularized biologic tissues with the extracellular matrix remaining after the decellularization process serving as the scaffold for cellular attachment and structural organization. The alternative approach is to use synthetic biodegradable polymer matrices to provide these scaffold functions with the anticipation that the cells in the tissue-engineered construct will produce their own extracellular matrix and that the synthetic scaffold will be degraded and eliminated. The disadvantages of decellularized grafts include the relative shortage of available homografts and potential immunogenicity problems that may arise from the use of decellularized xenogeneic tissues. Perhaps more importantly, the density of residual extracellular matrix that is attractive from a structural integrity and early post-implant functional perspective may prevent the penetration of seeded cells into the interstices of the matrix. Furthermore, given the complex and poorly understood interactions between the cellular cytoskeleton and the extracellular matrix in normal tissues, it may be naïve to expect cells seeded onto decellularized tissues to assume those same relationships with matrix that may be altered by the decellularization process or may differ in subtle ways from species to species. There has been some experience gained in cardiac valve tissue engineering using small intestinal submucosa as the scaffold.16 In our laboratory, we have constructed trileaflet valved conduits and seeded them with EPCs. After these conduits were implanted in an ovine model, there was satisfactory short-term function, but there was no penetration of any cells into the substance of the small intestinal submucosal scaffold.
Our primary laboratory efforts to develop heart valves and large arteries have utilized the alternative approach of seeding cells onto biodegradable polymer scaffolds that temporarily provide the macroscopic structure and the mechanical stability that are necessary for "tissues" to develop from their individual cellular components. Ideally, these polymer scaffolds then degrade during the time that the cells in the developing "tissue" are producing normal structural proteins and are becoming organized and oriented to replicate normal tissue structure. Initially, the biodegradable scaffolds that we used were based on materials already in clinical use, including PGA and PLLA. Each of these materials has differing characteristics, including the length of time to degrade the polymer and the rate at which polymer degradation and loss of strength occur. PGA was introduced into clinical practice as a biodegradable suture material and marketed under the trade name Dexon (U.S. Surgical, Norwalk, Conn). The ability to extrude PGA into fibers permits it to be fabricated into nonwoven sheets with an open porous structure. Pore size has been shown to be important in the choice of scaffold material for tissue engineering liver for a variety of cell types,32 and it is likely that pore size will also be important for cardiovascular structures. An open pore structure seems to facilitate both cell delivery and subsequent cell proliferation by allowing free access to suspended cells, free diffusion of nutrients and dissolved gases, and removal of waste products of metabolism. These spatial properties combined with a consistent and relatively rapid loss of polymer mass relative to other biodegradable materials have made PGA an attractive choice for tissue engineering. One disadvantage that we have observed is that there is a more rapid loss of strength than polymer mass as the polymer degrades by hydrolysis, and this property limits the amount of time that PGA-based tissue-engineered constructs can be kept in aqueous culture before implantation. This shortcoming of PGA was offset to some degree by impregnating PGA with the thermoplastic polymer poly-4-hydroxybutyrate (P4HB), which allowed flat sheets of scaffolding to be assembled into a trileaflet structure by a series of two or more wraps around a cylindrical mandrel and using heat welding to create a tubular conduit containing a trileaflet valve.12
Despite promising early results using the PGA/P4HB composite, in subsequent studies we experienced some problems with the use of this material, particularly continued problems with loss of structural integrity with longer periods of time in an aqueous tissue culture environment. As the rate of loss of strength increases, there is an increased requirement for the cells in the tissue-engineered construct to produce extracellular matrix at an earlier time point prior to implantation. There were problems with suture retention and actual tearing of the wall of the conduit in some of our experiments with valved conduits developed from PGA/P4HB scaffolds. For these reasons, Sutherland and associates in our laboratory adopted a scaffold composed of a mixture of PGA and PLLA.18 PLLA is the methyl substituted form of PGA, and it was initially developed for the biodegradable suture market. This polymer hydrolyzes at a much slower rate than PGA, and the initial tensile strength of PLLA is less than that of PGA. This composite material was able to be fabricated into a valved conduit with a three-leaflet valve that functioned quite well at the time of implantation, and then over periods of up to 8 months after implantation, and had significantly improved surgical strength and handling characteristics.18 The importance of scaffold degradation time was demonstrated in experiments reported by Stock and colleagues from our laboratory,34 in which the use of polyhydroxy-octanoate as scaffolding material was associated with excessive tissue buildup and leaflet retraction.34
These PGA and PLLA scaffolds for cardiac valves have one additional disadvantage, and that is they are significantly stiffer than a normal valve leaflet and only gradually become less stiff with time.12,18,33 As a consequence of this finding but also because of evidence that mechanical signals influence cellular behavior, a new more elastomeric polymer was designed for tissue-engineered heart valve application by Wang and colleagues at the Massachusetts Institute of Technology. This polymer is a rapidly degrading elastic polymer based on sebacic acid.35 Preliminary experiments have shown that this polymer will support cell attachment and proliferation, but much additional experimental work will be necessary to determine if this material will be suitable as a scaffold for a tissue-engineered heart valve. However, the availability of this more elastic polymer will allow further studies of the effects of mechanical scaffold properties on the development of a TEHV.
In addition to the influence of the chemical, degradation, and mechanical properties of the scaffold material, there is preliminary evidence that the fabrication techniques can affect the mechanical behavior of the scaffold and may be of importance to the development of tissue-engineered constructs. Engelmayr and associates have shown that dimensions of the pores within a scaffold can guide engineered tissue cellular and collagen orientation in two-dimensional systems,36 and it seems reasonable to anticipate that similar phenomena will occur in three-dimensional systems. To date, it has been difficult to explore the impact of scaffold fiber orientation on the development of tissue-engineered cardiac valves since the methods needed to fabricate these scaffold materials into more organized structures than films or nonwoven meshes have been limited. A promising fabrication technique that has recently been evaluated for tissue-engineering applications is electrospinning, which allows the macromechanical characteristics of scaffold materials to be controlled so that they mimic those of a native semilunar valve leaflet.37 As this and other fabrication methods evolve, it seems likely that further exploration of the effects of scaffold mechanical behavior at the macroscopic level on developing tissue-engineered structures will be necessary.
Based on extensive experience with culturing a variety of cell lines, TEHVs have been grown in media containing fetal bovine serum, which is known to contain multiple factors that promote cell proliferation in vitro. To the extent that cytokines can alter cell phenotype, these signals may provide additional opportunities for engineering the environment to guide the development of a tissue-engineered heart valve. The findings that TGF-β1 could induce endothelial progenitor cells to produce -SMA31 and alter the types and amounts of extracellular matrix being produced,38 are examples of the potential ability to use cytokine signals to affect cell phenotype and protein synthesis. Few studies have been undertaken to explore this approach to engineering the environment in which a tissue-engineered construct takes place. In future engineering of valve tissues, one might envision localizing growth factors within the developing tissue-engineering construct to guide behavior at the microstructural level, particularly in regard to control of the types and amounts of extracellular matrix.
Several lines of in vitro and in vivo evidence indicate that both tissue-engineered and normal valve leaflets respond to mechanical signals with changes at the cellular and extracellular matrix levels. In the natural "experiment" provided by the Ross procedure, in which a pulmonary valve from the low-pressure pulmonary circulation is transplanted into the aortic position and subjected to systemic pressure, significant changes in the phenotype of valvular interstitial cells have been observed, including an increase in the activity of matrix remodeling proteins.9 In experiments from our laboratory reported by Hoerstrup and colleagues,12 the ability of mechanical flow and pressure signals in vitro were shown to increase the production of collagen in tissue-engineered semilunar valves. Lee and associates have shown that in cultures of vascular smooth muscle cells exposed to tightly controlled mechanical strains, mRNA and protein levels for versican, biglycan, and perlecan increased, while those for decorin decreased. Also hyaluronan-versican aggregation was enhanced following deformation.40 Engelmayr and coworkers have also shown that vascular smooth muscle cells seeded onto biodegradable scaffolds and then subjected to flexure increased the production of collagen and vimentin, and had increased tissue stiffness.33 In subsequent studies on bone marrow stem cells, Engelmayr and associates have found that combined flow and flexure signals result in increased collagen concentration and tissue stiffness.41 These natural and laboratory experiments demonstrate that biomechanical signals clearly alter cellular behavior and therefore engineering of the in vitro mechanical environment, and offer an additional method to guide the development of tissue-engineered cardiac valves.
The experiments from our laboratory reported by Hoerstrup12 and subsequently by Rabkin8 indicate that there is an ongoing evolution of tissue-engineered heart valves after implantation into the circulation. This evolution occurs at the macroscopic level with thinning of the tissue-engineered leaflets, as well as at the histologic level with development of distinct layers in the valve leaflet.12 Picrosirius red staining indicated that there was a progressive evolution of the orientation of the collagen in the tissue-engineered valve with increasing time in the in vivo environment. There was an evolution of cellular phenotype with decreased expression of SMemb, SMA, and MMP-13, and increased expression of vimentin, a process which is consistent with an evolution from an "activated" state of these valve cells to a more inactive state. Importantly, the microscopic structure of the valve leaflets evolved from a relatively homogeneous appearance to a layered structure with increased collagen on the sinus side of the leaflet and elastin formation on the ventricular side of the leaflet. The mechanisms by which this in vivo evolution of structure and cellular activity occurs remain completely unexplored, but these observations suggest that a tissue-engineered valve may not have to be a "finished product" at the time of implantation into the circulation.
In summary, the tissue-engineering approach to the development of a living heart valve structure represents an exciting new direction for heart valve research. Numerous refinements of the current approaches will likely be necessary, including the optimization of the scaffold materials, cell types, in vitro biochemical and mechanical conditions, and the development of reproducible valve development processes. The observation of in vivo maturation and/or evolution of these tissue-engineered constructs introduces an additional variable that will complicate regulatory efforts to ensure patient safety. It seems likely that initial clinical experience will be gained in the low-pressure pulmonary circulation since valve failure in this position is likely to be well tolerated based on years of clinical experience in patients with tetralogy of Fallot treated with transannular patches. The challenges for application in the systemic circulation will be much greater, as the consequences of valve failure will be much more likely to be catastrophic.
| ||||||||||||||||||||||||||||||||||||||||||||
