Tissue engineering is an emerging discipline that blends the fields of biomaterials, bioengineering, and cell biology and has the goal of inducing tissue repair or regeneration. What differentiates this from more traditional biomaterials fields is that implants are designed to induce a response from the recipient and provide a biological tissue replacement. Typically, a temporary biomaterial scaffold, cells, and/or cell signals are combined in vitro or in vivo to elicit the desired response. This approach sharply contrasts the use of permanent prostheses, which are intended to remain inert over long periods of time. Tissue engineering is being explored for virtually all fields of medicine and surgery, and it is likely that as this field develops, the need for biological grafts and traditional prostheses will be greatly diminished.
The success of tissue-engineered ACL regeneration depends on three principal factors, each with several variables: the characteristics of the implant (scaffolds, cells, and cell signals), surgical procedure (multivariate), and the postsurgery rehabilitation protocol. One logical approach to ACL regeneration is to mimic "spontaneous regeneration" that occurs in limbs of lower vertebrate species and certain tissues in adult humans. In these cases, regeneration recapitulates embryonic development to a certain extent. Examples include limb regeneration in urodeles via local cell dedifferentiation to form a blastema (22) and bone regeneration in humans via reserves of embryonic stem cells. Taking cues from nature, our group and other investigators are combining resorbable scaffolds, cells, and cell signals to form a "blastema-like" template for tendon (32) or ACL regeneration. To reiterate, the feasibility of this tissue engineering approach has been established (33,34), but true ACL regeneration has not yet been realized.
Similar to a scaffold erected around a building under repair, tissue engineering scaffolds are designed to provide temporary mechanical support and serve as a substratum for the attachment, migration, and activities of the "workers" (cells). The success of a scaffold-based implant depends on the well-coordinated timing of implant resorption and new tissue in-growth. Scaffolds for ligament reconstruction are typically biomimetic, designed to be similar to normal ligaments: complex composite materials with continuous fibers aligned in parallel within a deformable matrix. Various textile configurations, e.g., yarns or braids, also provide constructs with different mechanical and biological properties.
Tissue engineering scaffolds for ACL reconstruction can be of natural or synthetic origin but must satisfy diverse design criteria. The criteria involves mechanical properties, biocompatibility, and a "moderate" resorption rate to gradually transfer mechani-
cal loads to in-growing neoligament tissue. As in all engineering designs, there are trade-offs with optimizing scaffold performance, and it is difficult, if not impossible, to satisfy all the design criteria. For example, currently available synthetic materials are generally deficient relevant to tissue interactions, whereas for natural materials, the mechanical properties and resorption rates are difficult to reproducibly control.
Clearly, there are special mechanical considerations for tissue engineering applications in the musculoskeletal system, because these tissues (and their replacements) must bear significant mechanical loads. In fact, several early attempts to develop tissue-engineered tendons or ligaments failed because the biomaterial scaffold had inadequate strength and stiffness. Examples include the use of relatively weak collagen gels seeded with cells in vitro. Although these constructs appeared to be "ligamentlike" based on histology and/or biochemical markers, they lacked the key ingredient required for a ligament: mechanical strength.
Recently, the concept of functional tissue engineering emerged, emphasizing the need to consider the role of biomechanics in tissue engineered constructs (35). How these principles are important in the design of ACL regeneration scaffolds is briefly discussed. The first four principles involve understanding the mechanical state of the normal tissue: (1) measuring stress/strain histories in vivo; (2) measuring the mechanical properties under subfailure and failure conditions; (3) selecting and prioritizing a subset of these mechanical properties; and (4) setting standards to evaluate surgical outcomes. For ACL regeneration scaffolds, it is critical to consider these mechanical factors.
Several studies have determined actual stresses and/or strains generated in the ACL during knee loading in animals (36) or humans (37,38). The loads generated during normal usage are well below the failure load of the ACL and vary within the antero-medial and posterolateral bundle as a function of loading conditions. The mechanical properties of isolated femur-ACL-tibia complexes have also been studied under subfailure and failure conditions. The structural properties (breaking load, stiffness, and extensibility) and material properties (ultimate strength, modulus, and ultimate strain) of various tendons and ligaments have been determined from uniaxial constant strain rate failure testing (39,40). The ACL is anisotropic, nonlinear, and viscoelastic (experiences creep and stress relaxation), making it even more difficult to match its properties. To obtain design criteria, these mechanical properties must be prioritized, according to the most important to restore normal knee function and mechanics. The stiffness certainly is critical because it influences load sharing between the implant and neoligament tissue. If the implant is too stiff, the neoligament tissue will not bear significant loads, and may atrophy from stress-shielding. Viscoelastic properties are also significant because the knee is loaded cyclically. Finally, standards should be set to evaluate surgical outcomes; these must be based largely on mechanical performance and, in clinical studies, patient satisfaction. Again, the reader is referred to the excellent seminal article by Butler, Goldstein, and Guilak for a more complete treatment of this topic (35).
Another level of difficulty is that the tissue-engineered ACL is designed to gradually degrade, lose strength, and transfer mechanical loads to newly formed neoligament tissue. The optimal rate of scaffold strength loss is not known, but two extreme cases are unsatisfactory. If the implant remains too stiff for too long, stress shielding will occur, and the neoligament tissue may not mature owing to lack of mechanical loading. At the other extreme, if the implant is initially too weak, or resorbs too rapidly, mechanical failure will occur before neoligament tissue can form and bear loads. Ideally, the mass resorption of the implant would coincide with the strength loss so that a "nonfunctional implant" would not take up space needed for neoligament tissue formation.
It is important to note that biological properties are also critical to the scaffold design. The scaffold must be biocompatible and capable of supporting or inducing tissue in-growth and remodeling. Thus, only scaffolds containing biocompatible polymers were investigated, including collagen and various synthetics.
Type I collagen is a candidate biomaterial for this application because of its unique physical, chemical, and biological properties. Collagen is a triple-helical protein that self-assembles into strong rope-like fibers that provide mechanical stability to all tissues of the body. Collagen is the major structural component of connective tissues (41), especially the tendon and ligament (12). Collagen can be extracted from tissues and processed into high-strength fibers (42) and scaffolds with mechanical behavior similar to that of autografts (43). Collagen is widely used in cell culture studies because fibroblasts attach, proliferate, and secrete matrix on collagen scaffolds. Collagen has numerous biological functions, serving as the natural scaffold supporting cell attachment, proliferation, and matrix synthesis within the body. The resorption rate of collagen can be controlled by the extent of crosslinking (44-46). Collagen is not highly antigenic (47) and is chemotactic for fibroblasts and other cells involved in tissue repair (48). Several groups have explored the use of tissue-derived biomaterials and/or purified collagen for reconstruction of tendon (32), meniscus (49), or ACL (50-52); this section focuses on efforts from the author's laboratory.
In our studies, collagen fibers (50-100 |m diameter) for ACL reconstruction scaffolds are made by extruding a 1% (w/v) acid-insoluble bovine dermal collagen dispersion into saline solution, rinsing, and drying under tension. Fibers are crosslinked to improve the mechanical properties and control the resorption rate after implantation. The wet tensile strength of individual extruded collagen fibers (40-80 MPa) is comparable to that of native ligament tissue. Composite collagen scaffolds for implantation are made by embedding several hundred to several thousand (depending on the animal model) parallel continuous collagen fibers within a collagen or synthetic polymeric matrix.
Feasibility studies demonstrated that collagen fiber-based scaffolds can induce tissue ingrowth and gain strength (similar to biological grafts) following reconstruction of the Achilles tendon (43) or ACL (53). Collagen scaffolds for ACL reconstruction were made by aligning 225 crosslinked type I collagen fibers (60-|m diameter) in an uncrosslinked type I collagen matrix. The ACL was removed in skeletally mature rabbits and reconstructed using an acellular collagen scaffold implanted through the anatomic ACL attachment sites. Polymethylmethacrylate cement was used to secure the scaffolds within the femoral and tibial bone tunnels.
In our initial study (53), about 50% of the collagen scaffold implants ruptured before adequate neoligament tissue was formed, partly because of surgical and reha bilitation factors. Most failures were likely a result of unrestricted knee motion postsurgery (animals were not immobilized), which cut the device at the sharp bone tunnel interface.
The other half of the implanted scaffolds induced formation of functional, intact neoligament tissue connecting the femur and tibia. At 4-wk postimplantation, neoligament tissue was composed of newly deposited collagen, fibroblasts, and inflammatory cells within the implanted collagen fiber scaffold. At 20-wk postimplantation, the dehydrothermal-cyanamide crosslinked scaffolds appeared to be completely resorbed (based on light microscopic observations) and were completely replaced by aligned functional neoligament tissue. The strength of the neoligament at 4-wk postimplantation was less than that of an unimplanted collagen scaffold. However, at 20 wk, the neoligament strength had increased to 17 MPa, exceeding the strength of the unimplanted scaffold. Acellular collagen scaffolds initially lost strength, then strength increased from tissue ingrowth and remodeling, similar to the behavior of biological grafts.
The bone tunnel attachment sites contained new bone at the periphery of the surgical drill holes, approaching the fibrous neoligament. In a follow-up study, we further evaluated the bone tissue response to acellular collagen fiber scaffolds. Results suggest that bone and fibrocartilage tissue in-growth or "biological fixation' might effectively secure the scaffolds within the surgical bone tunnels (54).
Collagen fiber-based scaffolds have the potential to be developed into ACL regeneration scaffolds, but further work is needed. The use of proteinase inhibitors may improve control of the collagen degradation rate. For example, when matrix metallo-proteinases are inhibited in dermal wounds, the wound strength increases (55). Our laboratory has explored various crosslinking methods to improve collagen fiber scaffold performance for ACL reconstruction, including chemical methods (carbodiimides and glutaraldehyde; 56,57) and so-called "physical" methods (dehydrothermal treatment and ultraviolet light; 45,46). Typically, chemical crosslinkers compromise biocompatibility to some extent, whereas physical methods can partially denature collagen, leading to a more rapid resorption rate.
Collagen-based materials have many advantages, but they are difficult to reproduc-ibly fabricate, and there are no ideal sterilization methods for collagen (58). Thus, we are also exploring the use of synthetic polymeric fiber-based scaffolds and collagen-synthetic hybrid scaffolds for ACL regeneration.
Synthetic resorbable polymeric fibers are appealing for this application because critical properties, e.g., strength, strength retention, and degradation rate, can be tailored during processing. We are evaluating both the polyarylate and polycarbonate families of tyrosine-derived polymers (59). By making small variations in the backbone structure and pendant chain length, a broad series of physical properties (strength, stiffness, degradation rate, Tg, and surface tension) can be customized within these families of polymers. Polycarbonates and polyarylates are typically amorphous polymers, soluble in a wide selection of organic solvents. Both have glass transition temperatures (<100°C) significantly lower than their decomposition temperature (>300°C), allowing for various processing modes, such as conventional solvent casting, evaporation techniques, extrusion, compression molding, and fabrication into complex shapes by injection molding.
Poly(DTE adipate), one of the polyarylates, was initially selected based on favorable mechanical and resorption properties found previously for rods and films (60). Fibers of these polymers had not been made previously. As a first screening test for these new polymeric fibers, we evaluated the strength retention, mass loss, and molecular weight (MW) loss of poly(DTE adipate) fibers as a function of time in vitro. Poly(DTE adipate) polymer (100 kDa = initial MW) was spun into 150-^m diameter fibers. Fibers had an initial strength of 48.1 ± 7.4 MPa with a MW of 88 ± 2 kDa. Fibers (n = 8 per group) were incubated in phosphate-buffered saline at 37°C for 0, 1, 2, 4, 8, 16, and 32 wk. After 1 wk of incubation, the strength decreased to about half of the initial value. By wk 16, strength decreased to 10 MPa. Mechanical integrity was lost by 32 wk. There was a continuing decrease in MW throughout the experiment but no mass loss. Poly(DTE adipate) fibers had slightly higher initial strength and degradation rate than reported for films. The mechanical properties and degradation rate of poly(DTE adipate) fibers were within the range suitable for use in an ACL reconstruction device.
Based on these in vitro results, our next step was to evaluate in vivo strength retention and the tissue response to poly(DTE adipate) fibers. Because this polymer degrades hydrolytically, rather than enzymatically, similar degradation behavior was expected in vivo. Scaffold implants (4-cm length) were made by aligning 100 poly(DTE adipate) fibers in parallel. Scaffolds were sterilized by cold ethylene oxide. The implants were placed subcutaneously in five mature male rats that were sacrificed at 2, 4, or 8 wk. At all time periods, there was significant fibrous tissue in-growth surrounding individual fibers. No unusual inflammatory response was observed throughout the length of the study. The poly(DTE adipate) scaffolds had 73% MW retention at 4 wk in vivo. By 8 wk, MW retention significantly decreased to 36% of the initial value.
The initial mechanical properties of synthetic tyrosine-derived polyarylate and polycarbonate fibers are in the same range as those of collagen fibers; however, the synthetics have improved strength retention in comparison to collagen. The synthetic fiber scaffolds have also shown excellent biocompatibility and tissue in-growth in a subcutaneous implantation model. Based on our preliminary results and previous work by Kohn's group (60), we selected tyrosine-derived polyarylates (intermediate-strength retention) and polycarbonates (prolonged-strength retention) as model polymer systems to test the effects of scaffold strength retention on neoligament formation. The strength retention of these polymers can be tailored by subtle changes in the polymer backbone and pendant chain. These groups of synthetic resorbable polymers are currently being investigated for use as ACL regeneration scaffolds (61). Other groups have used polylactide-glycolide polymer-based fibers (62) and silk fibers (63) as experimental ACL reconstruction scaffolds.
Hybrid (Natural-Synthetic) Scaffolds
Another strategy to improve the properties of fibrous composites is to combine natural and synthetic fibers and/or modify the matrix surrounding the fibers. We fabricated hybrid (natural-synthetic) composites by embedding parallel collagen fibers within a polylactic acid (PLA) matrix (64). PLA was applied by dipping 500 collagen fibers in a 10% solution of l-PLA in chloroform and drying overnight under vacuum.
The mechanical properties, resorption rates, and subcutaneous tissue reactions were determined for collagen-PLA and collagen-collagen composites. The tensile strength and modulus of collagen-PLA composites were twice that of collagen-collagen composites. Subcutaneous fibrous tissue in-growth was improved, and implant resorption was slightly delayed in the collagen-PLA composites.
ACL reconstruction surgeries were also performed in rabbits using collagen-PLA composite implants. After 4 wk, neoligament tissue was observed in seven of eight implants; however, four neoligaments had ruptured either in the midsubstance (n = 2) or at the bone tunnel interface (n = 2). These results and our previous work suggest that resorbable polymeric composite scaffolds are potentially useful for ACL reconstruction. Again, these resorbable implants must be protected from excessive mechanical loading during the formation of host neoligament tissue.
We are currently investigating various types of synthetic polymeric matrices and fibers as scaffolds for ACL reconstruction. Others developed a hybrid tissue-engineered ACL device, consisting of a poly(L-lactic acid) fiber braid filled with collagen (65).
In summary, scaffolds for ACL regeneration must meet rigorous mechanical and biological design criteria. Natural and synthetic fibers are potentially useful as scaffolds, and a combination of the two may be needed to satisfy the design criteria. Of course, scaffolds should interact with cells and cell signals to form new tissue and enable ACL regeneration.
Although scaffolds provide temporary mechanical support, the long-term success of tissue-engineered devices depends on cells to deposit and remodel new matrix; cell signals are critical to this process. Growth factors have tremendous potential to improve tissue repair or induce regeneration, but these soluble factors may need to be incorporated into delivery vehicles (scaffolds) with appropriate release rates. Mechanical loads provide critical signals that influence normal musculoskeletal tissue development, growth, remodeling, and repair. When deprived of mechanical loads, musculoskeletal tissues atrophy; increased loads cause hypertrophy. Yet, until very recently, mechanical loads were not routinely applied to developing tissue analogs in vitro. Our laboratory is investigating cell seeding and various cell signals, including delivery of growth factors, genes, and mechanical loads to enhance ligament analog formation in vitro and ligament regeneration in vivo.
Development of Cell-Seeded "Ligament Analogs"
Toward the long-term goal of providing an embryonic or regenerative environment, a blastema-like structure or supply of locally active cells, we hypothesized that neoligament formation can be influenced by preseeding a scaffold with autogenous fibroblasts in vitro, creating a viable "ligament analog" for ACL reconstruction. Similar "tissue analogs" have been developed by others to repair large defects in skin and cartilage. In theory, preseeding a scaffold with autogenous cells might improve neoligament formation in the short term by controlling extracellular matrix deposition and the recruitment of local cells and may improve in the long term by influencing scaffold resorption and neoligament remodeling.
We fabricated ligament analogs in vitro by seeding high-strength resorbable collagen fiber scaffolds with intra-articular (ACL) or extra-articular (patellar tendon [PT])
fibroblasts (66). Fibroblasts explanted from rabbit or human tissues were cultured and seeded onto collagen scaffolds. Fibroblasts attached and proliferated on collagen fiber bundles and deposited new collagen within the ligament analogs in vitro. The cells adhered rapidly to the collagen scaffolds, spread along the long axis of the collagen fibers, proliferated, and remained viable for weeks in vitro. Fibroblast function was a mechanism of the culture substratum (ligament analog vs tissue culture plate) and the origin of the fibroblasts (ACL vs PT). PT fibroblasts proliferated more rapidly than ACL fibroblasts when cultured on ligament analogs. Collagen synthesis by ACL and PT fibroblasts was approximately tenfold greater on ligament analogs than tissue culture plates. The composition, structure, and geometry of the collagen fiber scaffolds may promote collagen synthesis within ligament analogs in vitro. These findings suggest that collagen fiber scaffolds influence cell behavior and collagen deposition, which is potentially beneficial following implantation.
However, possible benefits depend on sustained viability of the seeded cells following implantation into the knee joint. In ACL reconstruction surgery, implants are placed intra-articularly, surrounded by synovial fluid. Because the joint environment is relatively avascular, and nutrition may be compromised, the viability and function of a fibroblast-seeded ligament analog implanted in this environment is of particular interest.
The next question we addressed was whether these living ligament analogs would remain viable after implantation (67). We examined the fate of autogenous ACL and skin fibroblasts that were seeded on ligament analogs and implanted into the knee joint space. ACL and skin fibroblasts that were harvested, cultured, labeled, seeded on collagen fiber scaffolds in vitro, and implanted into the autogenous knee joint remained viable for at least 4-6 wk postimplantation. Thus, fibroblast-seeded ligament analogs may impact neoligament formation and remodeling. Studies are in progress to compare the efficacy of acellular scaffolds vs fibroblast-seeded ligament analogs in a load-bearing ACL reconstruction surgery model.
These results indicate that synovial fluid provides sufficient nutrition to sustain ACL or skin cells transplanted into the joint space. We selected skin as a possible source of fibroblasts because it is easily accessible in a clinical setting, and it has a robust healing potential. Although skin fibroblasts normally reside in an extrasynovial environment, they survived transplantation into the relatively harsh environment of the synovial joint. ACL fibroblasts also survived in the knee joint space, but their potential to improve neoligament formation may be limited by a poor intrinsic healing capacity. For example, relative to fibroblasts of the medial collateral ligament (MCL), ACL fibroblasts have decreased rates of migration, adhesion, proliferation, and collagen synthesis. Thus, it may be necessary to seed with a different type of fibroblast or to use cells genetically modified to optimize neoligament formation.
This preliminary work established the feasibility of harvesting, culturing, seeding, implanting, and tracking fibroblasts on viable ligament analogs grown in the laboratory. This approach would not specifically result in ACL regeneration, but it might enhance repair or formation of neoligament tissue. However, these studies establish the foundation for future studies, such as seeding a high-strength scaffold with pluri-potential mesenchymal cells to create an ACL regeneration blastema. Autogenous mesenchymal stem cell-seeded collagen gels appear to enhance healing of Achilles tendon defects (68). Prockop recently reviewed the potential uses of marrow stem cells for nonhematopoietic tissues, including tissue engineering applications for musculoskeletal tissues (69).
Growth factors are soluble peptides that influence normal tissue morphogenesis, repair, and regeneration. Platelet-derived growth factor (PDGF) and transforming growth factor P (TGF-P) are present in the normal healing response of the MCL and ACL. PDGF or a combination of TGF-P and epidermal growth factor administered locally in a soluble form can improve MCL healing (70,71). In vitro research implies that growth factors may enhance repair by increasing ligament fibroblast proliferation (72-75) and matrix synthesis (76,77).
A potential problem with these soluble growth factors is rapid diffusion away from the site where activity is required in vivo. The duration of growth factor activity may be prolonged by incorporating these soluble factors within a scaffold delivery vehicle, e.g., bone morphogenetic protein delivered via a collagen sponge enhanced tendon healing in a bone tunnel (78).
Finally, genetic alteration of seeded cells, including stem cells (79) and other gene therapies, are rapidly emerging technologies that will complement and expand tissue engineering methods to tissue repair. Gene delivery may be advantageous when compared to protein delivery. Tendon-bone integration was improved by bone morphogenetic protein gene delivery (80); therapeutic genes can also be delivered directly to ligaments using various vectors (81,82). Like growth factors, genes should be delivered locally over some period of time to maximize efficacy. The concept of sustained local gene delivery from biomaterials is in its infancy. A "gene activated matrix," a polylactic-glycolic acid or collagen scaffold soaked with plasmid DNA prior to implantation (83) is potentially useful to encourage repair of various tissues. However, it is difficult to control the rate and duration of gene delivery, and delivery vehicles generally have poor mechanical properties (84).
Mechanical loads provide critical signals that influence cells (85,86) and help control musculoskeletal tissue development, growth, remodeling, and repair (87,88). Mechanical loads can increase collagen production by ligament fibroblasts and interact synergis-tically with growth factors to influence fibroblast behavior (89). Yet, until recently, mechanical loads were not routinely used to influence tissue analog development in vitro.
Note the concept of functional tissue engineering (35), and its first four principles, as discussed in the Scaffold Design Criteria section. The last two principles of functional tissue engineering involve interactions between mechanical signals and cells: physical regulation of cells on matrices in vivo and physical stimulation of cell-scaffold constructs in vitro (35). It is now well accepted that mechanical loads have a major role in regulating cell behavior in vitro and in vivo. For example, mechanical loads cause morphological changes in fibroblast-seeded collagen lattices (90) and influence collagen gene expression by ACLs in vitro (91). Mechanical loading also affects graft remodeling and biological fixation of ACL grafts in vivo (92). Therefore, patients undergoing
ACL reconstruction who receive grafts are typically exposed to protected loading or continuous passive motion shortly after surgery (93). Early motion prevents excessive scar formation in the joint and stimulates healing and remodeling of tissues. However, postoperative rehabilitation protocols will likely need to be redesigned for tissue-engineered ligament regeneration devices, because their initial mechanical properties and changes in properties postoperatively are not necessarily similar to biological grafts. The process of tissue regeneration may be prolonged in comparison to rapid scar formation; thus, prolonged rehabilitation may be a prerequisite for ACL regeneration.
Technical, Regulatory, and Social Issues
There are important issues that need to be resolved as the field of tissue engineering transforms from the bench to the bedside (94). The first generation of tissue-engineered products (primarily for skin repair) have successfully overcome many of these challenges (95). Many concerns remain regarding the development, clinical use, and safety of tissue-engineered devices (96).
From a technical standpoint, there are many unknowns about interactions between scaffolds, cells, and cell signals. Can a scaffold meet mechanical and biological requirements? Does cell seeding improve implant performance? If so, by what mechanisms? Are the effects of cells short term, or are there long-term benefits? Implantation studies comparing acellular vs cell-seeded scaffolds are required. Should autogenic (self), allogenic (human donor), or xenogenic (animal) tissues (97) or cells be used? Can stem cells be led to differentiate into various tissue types? How can gene therapy be incorporated into designs? Is gene therapy safe? What are the optimal conditions to harvest, grow, sterilize, store, and ship cells and cell-seeded implants? It is difficult enough to make a few functional devices for ACL reconstruction in the laboratory; imagine the complexity associated with scale-up and manufacture of a tissue-engineered ligament device.
In addition to the technical and manufacturing obstacles, there are challenges with federal regulation of tissue-engineered implants. Regulation of tissue-engineered devices by the FDA in the United States is still in development (98), and there are serious regulatory concerns in Europe (99,100). Are these new implants categorized as tissues, drugs, devices, or combination of these designations? The regulatory pathways for these categories are quite different. Can the same safety and efficacy tests used to evaluate traditional inert biomaterials be used for tissue-engineered devices?
Finally, there are political and socioeconomic issues, including government funding of research, corporate liability, medical insurance reimbursement, and ethical concerns. Currently, the use of embryonic stem cells is controversial (101), and research funding is limited by the US government. Companies may be reluctant to develop novel tissue-engineered implants because of liability issues. Furthermore, a new implant must have significant advantages over the existing standard of care to be considered for reimbursement by medical insurers.
Was this article helpful?