Bone Substitutes

Rationale for Bone Substitutes

The ability of the skeletal system to heal is remarkable. When bone healing is impaired, however, the use of various types of bone grafts including cancellous autografts, vascularized cortical autografts, and cortical allografts, or bone transport techniques has been recommended. Owing to limitations inherent in each technique, the search for bone substitutes remains a medical necessity and has been the subject of extensive research.

Cancellous bone autografts, while considered the gold standard for bone repair, are fraught with limitations, including donor site morbidity (infection, pain, hematoma), limited availability, and high cost owing to increased surgical time. The impact of such limitations is exacerbated if several harvesting sites are required to obtain a sufficient graft volume. Furthermore, cancellous bone autografts are poorly suited for the treatment of segmental defects because of their tendency to undergo central resorption in cortical defects larger than 4 to 5 cm in people and ~3 cm in dogs [1,2]. Vascularized bone transplant, routinely used in specialized human surgical units, requires a major microsurgical operative procedure and a sophisticated infrastructure. Therefore, this technique is rarely used in veterinary medicine. Conversely, cortical allografts are widely used with many successes reported in both the human and veterinary orthopedic literature [3-11]. However, limited supply, high cost of bone banking, risk of rejection, and compromised mechanical properties make the continued reliance on cortical allografts a serious concern [12,13]. The early attraction of distraction osteogenesis techniques has been somewhat mitigated by their shortcomings including local morbidity, limited response of soft tissues to distraction, and cumbersome follow-ups.

Therefore, a critical need exists to develop artificial bone substitutes to act as replacements for biologic bone grafts. The ideal bone substitute should be biocompatible, osteoconductive, osteoinductive, osteogenic, structurally similar to the bone to be replaced, easy to use, and affordable.14 Further, it is advantageous if the bone substitute is also bioresorbable; however, this is not a limiting factor.

Properties of a Bone Substitute

An osteoconductive material promotes bone apposition on its surface and functions as a three-dimensional scaffold supporting vascular and cellular ingrowth, leading to new bone formation. An osteoinductive scaffold material provides a biologic stimulus that induces the recruitment and differentiation of host and transplanted cells into osteoprogenitor cells and then into osteoblasts. An osteogenic material contains living cells directly capable of new bone formation. Bioactivity is defined as the ability of a material to develop a strong, direct bond with the neighboring bone tissue, without the formation of the fibrous interface typically induced by an implanted foreign body [13].

Approaches to Bone Substitutes

The use of bone substitutes, or more generally, bone tissue engineering, is based on three strategies: (i) matrix-based approaches, (ii) cell-based approaches, and (iii) factor-based approaches.

Matrix-Based Approaches

Matrix-based approaches utilize structural scaffolds to replace the missing bone. Consequently, repair of osseous tissue depends on the recruitment of endogenous osteoprogenitor cells. Optimally, these tissue-engineered scaffolds have a porous structure that facilitates bony ingrowth; however, their lack of biologic activity can limit their use.

Biomaterials

Calcium Phosphate Ceramics

Calcium phosphate (CaP) ceramics are inorganic synthetic materials with no inherent osteoinductive or osteogenic properties. Because of their osteoconductive behavior in vivo, they have been shown to induce a biologic response similar to that of bone and are therefore deemed bioactive materials. For example, when implanted into healthy bone, osteoid and, subsequently, new bone are produced directly onto the surface of the CaP ceramic without any soft tissue interface. Calcium phosphate ceramics are extremely strong in compression but are also brittle and possess poor tensile strengths. Calcium phosphate scaffolds should be tightly packed in rigidly stabilized bone in order to protect them from shear and tensile stresses as well as to optimize bone incorporation [15].

Hydroxyapatite (Ca10(PO4)6(OH)2, HA) is structurally similar to the apatite found in normal human bone and, along with beta tricalcium phosphate (Ca3(PO4)2, β-TCP), is the most highly biocompatible and widely used calcium phosphate bioceramic. However, the behavior of HA and β-TCP after implantation is slightly different. The rate of biodegradation of β-TCP is much greater than that of HA. As β-TCP is resorbed, new bone formation fills the area once occupied by the β-TCP scaffold [16]. In contrast, biodegradation of HA is slow and although it may be osteointegrated, fragments may still be present in the bone several years after implantation. When β-TCP was implanted in a canine metaphysis, no inflammatory response occurred and direct bone apposition was noted at 4 to 6 weeks [17]. Because of its biocompatibility and bioresorbability, β-TCP is used as a filling material in defects where autografts are indicated or as a bone expander supplementing autologous cancellous bone grafts. The biocompatibility of HA has also been demonstrated in vivo, where incorporation occurs without inducing an inflammatory or foreign body response [18]. By three weeks, the HA surface is lined with fibroblasts and osteoblasts with an interposition layer of mineralized osteoid laid down directly by osteoblasts [16]. The precise chemical composition of macroporous CaP ceramics influences their osteointegration capability. Significant differences in the amount of new bone were observed depending on the ratio of HA to β-TCP in biphasic CaP (BCP) mixtures, where BCPs containing a higher volume of β-TCP induced better osteointegration when compared with pure HA ceramics [19].

Calcium phosphate materials are osteoconductive but usually not osteoinductive. Under some conditions, however, CaP scaffolds may acquire osteoinductive properties owing to the binding of an optimum amount of endogenous bone morphogenetic protein (BMP) to the material[13,20]. This indirect osteoinductive property depends on both the type of CaP ceramic with different phasic composition and porous structure and the species of animal in which it is implanted [20]. When a BCP scaffold was implanted intramuscularly and subcutaneously, obvious bone formation was detected in dogs and pigs, but no bone formation was observed in rabbits, goats, and rats [20].

CaP Cement

Calcium phosphate materials may also be used as bone cement. In an injectable bone cement, including particles of a BCP (60/40 HA/β-TCP) incorporated in a methyl cellulose carrier gel, bone ingrowth proceeded at a greater rate than in BCP blocks alone [21]. However, the major drawback of this system is the lack of significant initial mechanical properties, which imposes the requirement of an additional method of bone stabilization and leads to difficulty in maintaining the composite within the defect during surgery. To address these concerns, ceramic pastes may be used. This type of injectable CaP solidifies in situ after several minutes via a nonexothermic reaction to form a high compressive strength material. In the metaphysis of long bone in dogs, a cement composed of a monocalcium phosphate monohydrate, β-TCP and CaCO3, was both osteoconductive and biocompatible. At 2 weeks, extensive apposition of woven bone and unmineralized osteoid had occurred on the cement, without interposition of fibrous tissue between bone and cement [22]. However, only a small proportion of the volume of this cement was resorbed and replaced by bone during the first 16 weeks. At 78 weeks, cement was not totally resorbed [22].

The strength of macroporous CaP ceramics having interconnected porosity may be improved by filling the pores with a CaP cement. Using a highly soluble, self-setting calcium phosphate cement made of β-TCP and dicalcium phosphate dihydrate, it has been demonstrated that all the cement is replaced by bone after 4 months. Filling the macroporous ceramic pores with the CaP cement significantly improved the mechanical strength of these ceramics without modifying their integration in the healing bone or its biocompatibility [23].

Coralline Hydroxyapatite

Coralline hydroxyapatite is a natural material derived from sea coral. Certain sea coral species produce a CaP porous structure that is close to normal human bone structure. Coralline substitutes may be natural, e.g., directly harvested from the sea or manufactured by conversion from natural coral. It is processed by a hydrothermal chemical exchange method that converts the coral calcium carbonate to a crystalline material. Goniopora and Porites species of sea coral may be used as biomaterials. Coralline HA-Goniopora has larger macropores, measuring 600 μm in diameter as well as 260-μm intercommunicating pores, a structure close to typical cancellous bone. In contrast, with smaller 230-μm macropores and 190-μm intercommunicating pores, the structure of coralline HA-Porites is similar to that of cortical bone [24].

Coralline HA has a completely continuous three-dimensional pore structure with a high degree of uniformity in both macro- and micropore size. Coralline HA exhibits many of the criteria for a functional bone graft substitute: it is readily available, is easily contoured, has an adequate compressive strength, and allows ingrowth of new bone. However, no evidence exists of rapid degradation [16].

Bioactive Glasses

Bioactive glasses are materials consisting of calcium, phosphorus, and silicone dioxide. They possess osteointegrative and osteoconductive properties. Their mechanical strengths are significantly greater than porous CaP ceramics. However, when drilling or shaping, they are susceptible to fracture and consequently are difficult to fix to the bone. Therefore, they are primarily used for filling bone cavities instead of as segmental bone replacements. Bioglasses have also found use as bone-graft expanders [25].

Demineralized Bone Matrix

Demineralized bone matrix (DBM) is produced by acid extraction of the bone mineral phase. The preservation of the organic phase of the bone matrix confers DBM with some osteoconductive and osteoinductive properties. The osteoinductive capacity of DBM has been demonstrated by Urist [26]. Demineralized bone matrix has poor biomechanical strength and can have variable osteoinductive capability depending on the processing steps, the sterilization method, and the final formulation [27]. To date, the true osteoinductive potential of DBM, in humans, has been challenged because ectopically implanted DBM is not capable of reliably inducing bone formation.

Bone formation that takes place directly on DBM may start during the inflammatory phase of healing, indicating early differentiation and osteoblast activity. During this early phase of healing, osteoprogenitor cells are recruited from the bone marrow, the endosteum, the periosteum, and/or the vasculature [28]. This osteoinductive ability of DBM may be attributed to the presence of BMP 2 and BMP 7 [29]. Growth factors such as insulin-like growth factors-1 (IGF-1) and transforming growth factor-beta 1 (TGF-β 1) have also been found in DBM [30]. Early new bone formation is also related to the osteoconductive capacity of DBM that allows osteoblast precursors to adhere to a collagen matrix similar to endogenous cortical bone matrix. Indeed, the surface of DBM looks like the surface of the injured bone, which is the site of bone deposition during healing [28]. A final advantage of DBM is its rapid resorption rate.

Polymer Composites

Polymer composites with CaP ceramics (HA and β-TCP), particularly polylactic (PLA) and/or polyglycolic acid (PGA), have been investigated. Generally, these composites are used as delivery vehicles or carriers for growth factors.

Porosity

Scaffolds for osteogenesis should mimic bone morphology, structure, and function in order to optimize integration into surrounding tissue. Pores are necessary for bone tissue formation because they allow migration and proliferation of osteoblasts and other mesenchymal cells, as well as vascularization via micropore channels. Some controversy exists in the literature over the optimum pore size for bone tissue engineering scaffolds. For coralline HA, scaffolds containing larger pores (~500 μm) have been associated with greater bone ingrowth than implants with smaller pore size (200 μm) [24]. Similarly, when using macroporous biphasic CaP ceramics, it was shown that implants having larger, 565-μm sized pores, provided more abundant newly formed bone than those with smaller, 300-μm sized, pores [30]. Although pore sizes of approximately 100 μm are sufficient to allow cell migration into the implant, capillary invasion of the scaffold, which is essential to enhance new bone formation, requires pore sizes of 300 μm or greater. Indeed, while smaller pores (90-120 μm) may induce hypoxic conditions leading to osteochondral formation before osteogenesis, larger pores (~350 μm) will promote a more effective neoangiogenesis, leading to direct osteogenesis [31,32]. Interestingly, in some studies, the morphology of bone-formation porous HA blocks with 300 to 400 μm pores resembled that of the Haversian system in bone remodeling. This pore size coincides with the average diameter in the Haversian system (approximately 300 μm), indicating that the optimal pore size for bioceramic scaffolds should closely mimic the pore size of normal bone [31].

Implant-Host Interactions

Biointegration of a bioceramic scaffold implant depends on its mechanical stabilization. Micromotion favors the formation of fibrous tissue and precludes biointegration even if the implant is bioactive, like CaP ceramics [33]. For CaP bioceramics, a strong chemical similarity exists between HA and the bone mineral crystals. These materials have the capability to initiate nucleation and crystal growth of CaP at their surface from biologic fluids by a dissolution/precipitation process. Thus, a layer of nanocrystalline carbonated apatite associated with specific bone proteins is deposited onto the CaP ceramic and enables osteoblast adhesion and activity [33]. This process is similar to bone remodeling. The neoformed layer would present a composition analogous to the cement line at the edge of an osteon [33]. Thus, the bond between implant and bone is strong and mechanical tests demonstrate that 4 weeks after implantation, fractures occurred within HA ceramic and not at the implant-bone interface [34].

When a bioactive material is implanted, inside the connective tissue that has replaced the initial hematoma osteoblasts differentiate in close proximity to the biomaterial. Such osteoblasts are immobilized at the material surface as they would be at the surface of bone trabeculae. Owing to the polar activity of these cells, the extracellular matrix is synthesized by the cell pole in contact with the material, leading to the deposition of an osteoid matrix at the material surface. Both the immature bone formed in contact with the material and the scaffold itself are subject to remodeling and are subsequently degraded by osteoclasts and giant cells. New bone gradually replaces the de novo bone material, leading to osteointegration.

Incorporation

At 24 weeks following implantation of small cylinders of several types of ceramics into rabbit cortical bone, there was no degradation of coralline HA-Goniopora, 27% degradation of coralline HA-Porites, and 46% degradation of β-TCP. New bone occupied 56% of the coralline HA-Goniopora implant, 53% of coralline HA-Porites, and 45% of the β-TCP implant [24]. In large segmental diaphyseal defects, HA and β-TCP alone are not suitable as bone substitutes. In a 2.5-cm long segmental canine radial defect, use of these ceramics resulted in nonunions in many of the limbs 24 weeks postoperatively. In contrast, all the defects filled with autologous cancellous bone grafts achieved union [35]. This study clearly showed the limits of the inherent osteoconductive or osteoinductive capability of CaP ceramics. Osteointegration leading to healing of the bone defect requires active participation of host cells and a favorable recipient site environment that provides osteoprogenitor cells and neovascularization of the implant. In epiphyseal or metaphyseal cavities of relative small volume, CaP ceramics alone may be integrated because (i) packing of the ceramic may induce mechanical stabilization of the material, and (ii) the neighboring cancellous bone is well vascularized and contains mesenchymal stem cells capable of osteoblastic differentiation. In large segmental diaphyseal defects, the behavior may be different because (i) mechanical stabilization is difficult to obtain, and (ii) the recipient site is not capable of providing osteoprogenitor cells in a sufficient quantity to completely colonize the implant. With porous scaffold materials, osteogenic cells do not differentiate homogeneously within the biomaterial. The thickness of the scaffold may be critical as it may impair revascularization at the center of the biomaterial and, therefore, limit cell survival. Moreover, the number of cells supplied by the granulation tissue may be inadequate in the central area of extensive defects because of the remoteness of source tissues [36].

Cell-Based Approaches

Cell-based therapies entail transferring cells that have osteogenic potential directly to the repair site to facilitate synthesis of new bone, thereby reducing the reliance on local osteoprogenitors. These cell-based strategies rely on the implantation of (i) fresh autologous bone marrow, (ii) purified, culture-expanded mesenchymal stem cells (MSCs), or (iii) differentiated osteoblasts. Mesenchymal stem cells and progenitor cells are present in almost all normal tissues. Stem cells are resting cells that can be activated by biochemical signals to divide and differentiate. This cell division provides two daughter cells that are not identical: one daughter cell returns to the original resting state of the mother cell while the second daughter cell proliferates, producing an abundance of progenitor cells. These progenitors are subsequently triggered to differentiate to form a mature tissue [37]. A heterogeneous population of cells, including actual stem cells and progenitors derived from these cells, is present in bone marrow, periosteum, bone trabeculae (within Haversian canals of cortical bone), adipose tissue, and muscle.

Of all potential sources of progenitor cells, the most readily available is the bone marrow, which can be harvested via percutaneous aspiration or as a small core of cancellous bone. Pluripotent progenitor cells may evolve according to several pathways such as cells from liver tissue, central nervous system tissue, or osteoblasts. The differentiation of these cells down one pathway depends on their biologic environment and is influenced by factors such as oxygen tension, nutrient concentration, neighboring cells, mechanical stimuli, and the chemical composition of the surrounding extracellular matrix [37]. In addition, the osteoblastic differentiation pathway is modulated by a broad range of inducible factors, such as IGF and members of the TGF-β superfamily, including BMP-2 and BMP-7 [37].

When a cell-matrix composite graft is used, it is essential to ensure that the environment created by the matrix, i.e., surface texture, pore size and geometry, three-dimensional architecture, and degradation properties, is compatible with the survival of the cells [37]. It is also important to consider the biologic environment of the implanted bone substitute. The survival of implanted cells depends on the capacity for oxygen and other nutrients to diffuse into and out of the site through the implant. Bone marrow harvested by aspiration contains on average from 350 to 1000 progenitors cells per milliliter [38-39]. Mesenchymal stem cells appear to be present at a frequency of approximately 1 in 105 to 106 nucleated marrow cells [39,40]. An increase in the volume of bone marrow aspirate from 1 to 4 ml decreases the concentration of progenitor cells by 50%. Thus, it has been suggested that, in humans, the maximum number of progenitor cells may be obtained in four 1-ml aliquots rather than in one 4-ml aliquot [38]. Aspirates of bone marrow may be mixed with carriers, such as CaP ceramics or DBM, and this composite graft packed into bone defects. Use of the composite bone marrow-DBM in canine nonunions has been shown to induce healing at a rate at least equal, if not superior, to that obtained using standard autologous bone grafting techniques [41]. The addition of bone marrow aspirate to HA or β-TCP dramatically improved the outcome of these scaffold implants in the treatment of experimentally induced segmental radial defect in dogs [35]. The biomechanical and radiographic parameters of β-TCP with bone marrow were roughly comparable to those of cancellous bone autografts at 12 and 24 weeks. At 12 weeks, 94% of defects filled with CaP augmented with bone marrow achieved union, whereas nonunion was observed in 90% of defects filled with CaP alone [35]. It has been estimated that the number of osteoprogenitors provided by bone marrow aspirates is only 20% of what would be needed to permit colonization of the entire implant by new bone [37]; however, it is possible to improve the efficacy of aspirated bone marrow by concentrating the marrow-derived cells.

Centrifugal density separation yields an approximately 4-fold concentration of marrow-derived cells, which has been shown to significantly increase bone formation [42]. Mesenchymal cells can be isolated from bone marrow and expanded ex vivo without any apparent modification in phenotype or loss of function [43]. Using culture systems, ex vivo expansion produces a significant increase in the number of MSCs that can be delivered back to the surgical site [43]. With a ceramic carrier, cultured MSCs promote faster and more extensive new bone formation than fresh bone marrow because of the 300-fold increase in the number of MSCs yielded by the culture expansion process [44]. In a critical-sized canine ulnar defect stabilized by internal fixation, porous HA beads loaded with cultured cells from cancellous bone cores were implanted. Invasion by bone tissue was found in the stromal cell-bearing implants but not in control implants containing only HA granules [36,45]. Similarly, when BCP cylinders augmented with cultured mesenchymal stem cells from bone marrow aspirates were implanted in a critical-sized (21 mm) canine femoral defect stabilized by internal fixation, bone healing was achieved in 12 to 16 weeks. In contrast, no substantial bone regeneration occurred over the 16 week period when ceramic cylinders alone were implanted [44]. Under appropriate culture conditions, these MSCs may differentiate into osteoprogenitor cells and grow as a monolayer culture. Next, the matrix may be loaded with these cells and implanted, constituting a hybrid bone substitute. Marrow cells grown in primary culture for 10 days and transferred to porous coralline HA for a 2-week culture have shown an enhanced rate and extent of bone formation when compared with samples implanted with undifferentiated marrow cells [46].

Factor-Based Approaches

Factor-based therapies represent an attempt to overcome the limitation of ceramic scaffold implants used alone by directly providing osteoinductive stimuli. Growth and differentiation factors are carried by the implant that is inserted into the bone defect.

Growth factors and Bone Morphogenetic Proteins (BMPs)

Fibroblast growth factors (FGFs) play a role in angiogenesis and mesenchymal cell mitogenesis. Activity of both FGF-1 and FGF-2 has been identified during the early stages of fracture healing. Using basic FGF in tibia osteotomies in dogs accelerated all the stages of bone repair and stimulation of callus remodeling [47]. It is unclear what the therapeutic role of other growth factors such as insulin-like growth factor and platelet-derived growth factor play in fracture healing [48]. Among the growth factors that have been investigated, BMPs, which are members of the TGF-β superfamily, appear to have the most osteoinductive potential. Bone morphogenetic proteins initiate the bone healing cascade through the recruitment of mesenchymal cells from local bone and soft tissues and also guide the differentiation of mesenchymal cells into osteoblasts. To date, more than 15 BMPs have been described, but only a few seem active in the bone healing process [49] (Fig. 89-1). Two BMPs are produced by recombinant gene technology and are commercially available: BMP-2 and BMP-7 (also known as OP-1). In veterinary medicine, experimental studies in dogs and clinical trials in dogs and cats have been performed to assess the efficacy of BMPs in the treatment of bone defects [50,51], fracture nonunions, and arthrodesis [52].

The efficacy of BMPs was evaluated in the healing of a critical-sized radial defect (2.5 cm) in dogs. In one study, the defect was filled with a cylinder of natural coral (calcium carbonate) alone or enhanced with bovine-derived bone protein and compared to autologous cancellous bone graft [50]. Coral alone did not allow bone union whereas coral implants enhanced with bone protein achieved rates of union, bone formation, and biomechanical strength that were superior during the first 12 weeks, and comparable at 24 weeks with those of autogenous cancellous bone graft. In subsequent study, the effectiveness of recombinant human BMP-2 (rhBMP-2), delivered in a collagen sponge, was evaluated in the same segmental defect model in dogs [51]. None of the defects that were treated with a collagen carrier alone healed. From the day of surgery to 12 weeks postoperatively, the rhBMP-2 implants produced bone at a rate equivalent to cancellous bone autograft and were biomechanically comparable. However, defects treated with rhBMP-2 showed evidence of cyst-like bone voids, whose incidence appeared to be dose-related. The specific mechanism by which these voids developed was not determined, but it clearly appears that further research for the optimal dose of rhBMP-2 protein is needed prior to clinical applications. A nonglycosylated rhBM2/fibrin composite was applied in 41 sites in 38 dogs and cats requiring an arthrodesis or revision surgery for fracture nonunions [52]. Bone healing was obtained in 90% of the treated cases with rhBMP-2. Although cancellous bone autografts are the gold standard for the treatment of these clinical conditions, the number of animals showing radiographic bridging in pancarpal arthrodesis was higher in the group treated with rhBMP-2 than in the group treated with autografts at 17 weeks (100% vs. 59%) [53,54]. Similar results have been obtained in the experimental treatment of segmental defects in animal models.

Figure 89-1. Main BMPs involved in the process of differentiation of Mesenchymal Stem Cells (MSC).

Gene Therapy

Gene therapy involves the transfer of genetic information to cells. When a gene is properly transferred to a target cell, the cell synthesizes the protein encoded by the gene. In general, the duration of protein synthesis depends on the technique used to transfer the gene to the cell. Gene therapy can be applied either systemically or regionally. For bone healing, in normal patients, gene therapy can be used regionally. The gene can be introduced directly to an anatomic site with use of an in vivo technique, or it can be introduced via an ex vivo approach. In ex vivo gene delivery, cells are harvested from the patient, the cDNA is transferred to the cells in tissue culture and the genetically modified cells are then administered back to the patient.

Appropriate vectors must be used to enhance the entry and expression of DNA into target cells. These vectors may be of viral or nonviral origin. Viruses are efficient vectors because DNA delivery is a critical aspect of their life cycle. However, several major concerns are related to the use of viral vectors. First, the possibility exists of recombination with other viruses in the host cell as well as replication and multiplication in the patient. Second, the duration of transgene product (protein) expression by the transduced cells may be limited owing to immune response to viral proteins [48]. Finally, the risk of malignant transformation following viral integration into the host genome is a serious concern [55].

Gene therapy is considered as a tool in the treatment of bone defects because of the possibility of upregulating more biologic expression of individual proteins in specific tissues and cells. Although proteins used in bone tissue engineering can be manufactured, their life span is relatively short after being injected or surgically placed at the target site. Gene therapy provides the gene for the desired protein and then the transfected cells produce a biologically active protein in situ. This technique results in a higher and more constant level of protein production when compared with manufactured proteins. This is particularly important for proteins such as the BMPs [56]. One review reports the results of experimental and clinical studies evaluating use of growth factors using gene therapy for enhancing fracture healing [57]. Although the production of the protein has a relatively short duration (e.g., up to 6 weeks) because of associated immunologic response to the vector used for transduction, this duration is satisfactory and useful to boost healing in the treatment of fractures or bone defects. In small animal models, growth factors delivered by gene therapy resulted in better healing than those delivered as recombinant proteins [57]. Experimental studies in large animals have yet to demonstrate the benefits, without any deleterious side effects, of gene therapy for fracture healing and filling of bone defects.

Carriers for Delivering Growth Factors

Growth factors tend to rapidly diffuse away from the target site when administered alone. To limit this phenomenon and, therefore, to optimize their local osteoinductive potential, various carriers have been evaluated. The choice of the carrier or delivery system is essential to the success of growth-factor therapy. The ideal carrier should have several characteristics, including (i) the ability to deliver the growth factor at the appropriate time and at the proper dose, (ii) the ability to enhance cell recruitment and attachment, (iii) the presence of structural voids to allow for cell migration and angiogenesis, and (iv) be biodegradable without inducing an immune or inflammatory response and without producing toxic byproducts [48]. Matrix requirements for BMP carriers vary from simple delivery systems, such as minipellets, beads, or microspheres, to complex three-dimensional structures with a macro- and microporosity that attempt to mimic the matrix of the repair tissue. Optimally, carriers should be resorbable. The four major categories of BMP carriers are natural polymers, inorganic materials, synthetic polymers, and composite materials [58]. Table 89-1 illustrates their main properties.

The typical pharmacokinetic profile of BMP-2 release differs according to the delivery system. The initial burst delivery of BMP has a half-life of less than 10 minutes and appears to be carrier-independent. The secondary release of BMP is characterized by a half-life of 3 to 5 days with collagen carriers, whereas mineral-based delivery systems have a longer half-life, with BMP detected for up to 5 weeks [58,59]. In vitro and in vivo studies have shown the temporal sequence of BMP expression during membranous bone formation. Bone morphogenetic protein-6 appears first, followed by BMP-4 and later by BMP-2. Considerable overlap exists in the time of expression, suggesting codependency of these factors. This sequential expression shows that the BMPs are not necessarily interchangeable and that it is possible that different BMPs can influence the expression of other BMPs as part of a cascade [60]. Moreover, co-administration of BMP-2/BMP-7 or BMP-2/BMP-6 have been shown to be 5 to 10 times more potent in inducing bone formation than BMP-2 alone [27].

To date, the best strategy for bone repair using growth factors has not been determined. The optimal dose of proteins required to achieve bone repair in clinical studies is not clearly established. The large variability in doses and experimental protocols among studies makes the determination of an optimal dose difficult. Moreover, the ideal delivery matrix for the appropriate factor has yet to be identified. There are three major obstacles to routinely using growth factors for bone repair: (i) the high costs of these growth factors, particularly BMPs, (ii) the selection of the optimal doses, and (iii) the optimization of the carrier matrix or delivery vehicle in order to allow vascular ingrowth, osteoprogenitor cell population, and invasion within the scaffold.