Formation of Blood Clot on Biomaterial Implants Influences Bone Healing

Hemostasis of bone healing

Bone repairs itself through similar phases of normal wound healing and some bone tissue-specific processes. Typically, there is an overlap of four major phases: hemostasis, inflammation, proliferation, and remodeling.^1,2 As a vascularized tissue, an injury in bone is associated with the instant damage of blood vessels in the periosteum, endosteum, and surrounding soft tissue. ³ Hemorrhage and edema at the injured site disrupt oxygen and nutrient supply to osteocytes and eventually cause necrosis of broken bone ends.^1,4,5 To prevent excessive blood loss and maintain blood flow to the rest of the body, hemostasis is triggered. Hemostasis is a tightly regulated process and involves three events that occur in rapid sequence: vasoconstriction, formation of a platelet plug, and coagulation.^6,7

Vasoconstriction

Constriction is the immediate response of a damaged vessel to its own injury. When blood vessels are damaged, extracellular matrix proteins underlying the endothelium are exposed to whole blood. Subendothelium collagen (principally type I and III), von Willebrand factor (vWF), and fibronectin interact with blood platelets through various glycoprotein receptors, thereby supporting platelet adhesion to the damaged site.^8–13 Moreover, these interactions stimulate platelets to release contents from their granules, including serotonin and thromboxane A₂. ¹⁴ These vasoactive factors induce contraction of vascular smooth muscle cells, ¹⁵ and together with cytokine-stimulated endothelial cells, damaged blood vessels are constricted to reduce extravasation of blood constituents (Fig. 1).^6,16 The significance of vasoconstriction is to allow time for platelet plug formation and coagulation to occur.

OPEN IN VIEWER

FIG. 1.

Vasoconstriction and platelet plug formation in response to blood vessel damage. Color images available online at www.liebertpub.com/teb

Coagulation

The main purpose of coagulation is to produce a stable hemostatic clot by forming fibrin mesh on the platelet plugs. Coagulation system can be activated by intrinsic and extrinsic pathways. The intrinsic pathway begins when blood protein prekallikrein, high molecular weight kininogen, and factor XII (FXII) contact with thrombogenic surfaces and undergo cleavage to produce the active forms.^25–27 On the other hand, the initiation of extrinsic pathway is dependent on the presence of tissue factor (TF). TF is highly expressed on or released from various cells (e.g., endothelial cells, activated platelets, and monocytes/macrophages) following vascular damage and inflammatory stimuli such as endotoxin, tumor necrosis factor α, and interleukin-1α.^28–30 For both pathways, a series of proteolytic reactions where an enzyme precursor becomes active will trigger the activation of another precursor in the downstream cascade. It has been revealed that activated platelets provide the phospholipid surfaces for the assembly and function of the enzyme complexes, and hence greatly propagate the coagulation activation.^29–31 Both pathways converge to a common final pathway in which prothrombinase complex (factor Xa–factor Va) cleaves plasma protein prothrombin to produce active enzyme thrombin.^32,33 Thrombin then catalyzes the conversion of fibrinogen into insoluble fibrin, including fibrinogen bound to the platelet plugs, thereby restricting formation of fibrin mesh to the location of damaged vessel (Fig. 2).^34–36 Besides platelets, other blood cells and proteins are also incorporated into the growing clot. This eventually produce a more stable and definitive hemostatic clot between the broken bone ends, namely a hematoma (Fig. 3). ³⁷ The hematoma not only preserves the integrity of the vessel, stabilizes the broken bone ends but also supports subsequent inflammation, granulation tissue formation, and bone remodeling.

OPEN IN VIEWER

FIG. 2.

The intrinsic and extrinsic pathways of coagulation system. The intrinsic pathway starting with surface contact activations of prekallikrein, high molecular weight kininogen (HMWK), and factor XII (FXII) is shown as a linear cascade of zymogen activation steps, leading to the formation of intrinsic tenase complex (FIX–FVIIIa). In parallel, the extrinsic pathway is initiated by tissue factor (TF) generated during trauma. TF activates factor VII (FVII) into FVIIa, and forms extrinsic tenase complex (TF-FVIIa). Both tenase complexes from respective pathways merge at the common pathway in which factor X (FX) is converted to factor Xa (FXa). FXa, which in turn binds to activated factor V (FVa) forming the prothrombinase complex (FXa–FVa) that converts prothrombin to thrombin. Thrombin as the end product of coagulation activation subsequently catalyzes the formation of fibrin from fibrinogen. Phospholipids (PL) membrane of platelets and calcium ion (Ca²⁺) serve as cofactors of the process. Color images available online at www.liebertpub.com/teb

OPEN IN VIEWER

FIG. 3.

A stable hemostatic blood clot is composed of platelet aggregates, erythrocytes, and leukocytes entrapped in the fibrin network.

Blood clot

Network of fibrin fibers

Thrombin (factor IIa), as the end product of coagulation activation, is produced by cleaving prothrombin fragment from prothrombin (factor II). In mediating conversion of fibrinogen to fibrin, thrombin sequentially cleaves fibrinopeptides A and B from fibrinogen, resulting in the generation of fibrin monomers.^38–40 The monomers interact in a half-staggered end-to-end fashion and become double-stranded protofibrils. Following lateral aggregation of the protofibrils, fibrin fibers are formed, branched out, and eventuated in a three-dimensional network on the platelet plugs.^41–43 Coagulation protein factor XIII that is activated by thrombin forms cross-links between neighboring fibers in the network. Hence, the resultant clot is strengthened against flow, mechanical, and proteolytic impacts (Fig. 4).^44–46

OPEN IN VIEWER

FIG. 4.

Conversion of fibrinogen to fibrin is mediated by thrombin. Fibrinogen is a trinodular protein consisting of two sets of three different polypeptide chains: Aα, Bβ, and γ, assembling with their N-termini in a central E domain. Thrombin releases the fibrinopeptides A and B from E domain of fibrinogen sequentially, forming the fibrin monomers. Interacting in a half-staggering and end-to-end manner, fibrin monomers polymerize into protofibrils. By lateral aggregation and branching out of protofibrils, fibrin polymers are formed in three-dimensions. A stable fibrin clot is formed by factor XIIIa-mediated cross-linking between the γ chains in D-domian in the fibrin network. Color images available online at www.liebertpub.com/teb

Cellular and molecular components

During fibrin generation, platelets and other blood cells such as erythrocytes and leukocytes are entrapped in the growing clot. ⁴⁴ Activated platelets degranulate a series of cytokine and growth factors including platelet factor 4 (PF-4), platelet-derived growth factor (PDGF), and transforming growth factor-beta (TGF-β).^16,47–49 These factors are revealed to stimulate chemotaxis of neutrophils, monocytes, fibroblasts, and osteogenic progenitor cells, which originate from the periosteum of broken bone into the damaged site.^3,5,50–55 The inflammatory cells, neutrophils and monocytes, enhance removal of necrotic cellular debris, pathogens, and foreign materials via phagocytosis. In addition, the monocytes present in the tissue differentiate into long-life (up to months) macrophages,^56,57 which release reactive oxygen species and growth factors: fibroblast growth factor (FGF) and epidermal growth factor (EGF), to mediate further growth of fibroblasts, new blood vessels, and epithelial cells. ³ Subsequently, in the clot microenvironment fibroblasts synthesize collagen to form a new matrix while mature endothelial cells organize into new capillaries from pre-existing vessels, replacing the hematoma with granulation tissue.

On the other hand, an array of proteins is also incorporated into the clot during fibrin formation. For instance, vWF, fibronectin, collagen, albumin, tissue-type plasminogen activator (tPA), plasminogen activator inhibitor (PAI), α2-antiplamin, and FGF-2.^42,58 These proteins facilitate attachment and migration of cells on fibrin fiber into the area, control clot dissolution prior to granulation tissue formation, and remodeling of woven bone by dual actions of osteoblasts and osteoclasts.^54,59–61

Despite bone healing being a highly complex process influenced by biological and mechanical factors; it is conceivable that the underlying mechanism is largely driven by the early microenvironment of which a hematoma forms the critical component to influence mesenchymal stem cell (MSC) migration, blood vessel formation, and osteogenesis via the release of growth factors such as bone morphogenic proteins (BMPs). Figure 5 summarizes the sequence of bone healing events.

OPEN IN VIEWER

FIG. 5.

Time sequence of four main phases of bone healing: hemostasis (bleeding and blood clotting), inflammation, proliferation, and remodeling. Color images available online at www.liebertpub.com/teb

Insights from platelet-rich plasma

With advances in the understanding of platelet biology, platelet-rich plasma (PRP) has been developed as a current approach in bone tissue engineering. PRP is a fraction of plasma in which platelets are concentrated in a small volume of plasma.^62,63 The rationale behind the use of PRP is to provide autologous platelets, which secrete their storage pool of growth factors at high concentration to expedite bone regeneration and soft tissue repair.^16,64

Indeed, activated platelets release a range of osteogenic and angiogenic growth factors from their α-granules, including PDGF, TGF-β, platelet-derived epidermal growth factor, platelet-derived angiogenesis factor (PDAF), insulin-like growth factor-1 (IGF-1), and PF-4 (Table 1).^21,48,65 These growth factors are known to have positive effects on bone healing by stimulating the proliferation and differentiation of undifferentiated mesenchymal cells and osteoblasts,^66–69 angiogenesis, and chemotaxis for inflammatory cells.^48,70 Primarily, it is believed that the initiation of bone regeneration begins with the release of PDGF-AB and TGF-β1 after platelet aggregation.^68,71

Previous studies have shown that growth factor concentrations directly correlate with platelet number in PRP.^63,72 After dual gradient density centrifugation of whole blood to obtain PRP, platelets are activated by the addition of thrombin and calcium chloride, leading to platelet degranulation and formation of a PRP gel.^62,73 The application of PRP offers theoretical advantages over the delivery of a single recombinant growth factor since PRP releases high concentrations of multiple native growth factors in their biological ratios. The complex and interdependent nature of growth factors (i.e., TGF-β, PDAF, and IGF-1) suggests that more than one signaling pathway of bone regeneration could be targeted with the use of PRP.^65,71 Also, better handing characteristics and in vivo stability could be achieved when PRP gel is combined with particulate grafts.^74,75

In vitro studies have shown that PRP supernatants support the viability and proliferation of human fetal osteoblast-like cells, ⁷⁶ alveolar bone cells, ⁷⁷ porcine articular chondrocytes, ⁷⁸ and human endothelial cells.^79,80 Extensive animal studies have also investigated the effect of PRP gel alone on bone regeneration,^79,81–83 or in combination with bone grafts and graft substitutes.^74,84–88

However, there is some inconsistency in the literature regarding the benefits of PRP. While some studies reported significant increases in bone formation and maturation rates,^79,83–85 others did not observe any improvement or even inhibition of new bone formation.^74,77,81,82 One of the possible reasons is the differential use of platelet concentrations among studies. A typical PRP is defined to have a five-fold increase platelet concentration (∼1,000,000/μL) over the physiological level. ⁸⁹ Platelet concentration varies greatly due to different baseline values of animal species and the preparing procedures of PRP.^75,90 In fact, the platelet concentration required for a positive effect on bone regeneration appears to span a very limited range. Weibrich et al. ⁸¹ reported that advantageous effects of PRP on peri-implant bone regeneration in rabbits only occurred when a platelet concentration of ∼1,000,000/μL was used. At lower concentration (164,000–373,000/μL), the effect was suboptimal, whereas higher concentration (1,845,000–3,200,000/μL) led to a paradoxically inhibitory effect. This finding was supported by in vitro work of Choi et al. ⁷⁷ and Tomoyasu et al. ⁹¹ who studied the effect of platelet concentration in PRP alone on human alveolar cells, and in combination with BMPs on human osteoblasts respectively.

Different protocols for platelet activation may be another reason for the discrepancy of results.^62,92 Concentrations of thrombin and calcium for platelet activation were shown to affect the release of growth factors, endothelial cell division,^79,93 and the adhesive property of PRP clots on soft tissue. ⁹⁴ However, the mechanism of how these activators vary the properties of the PRP clot is not fully understood. This may be associated with the thrombin concentration, which alters platelet activation and fibrin polymerization, leading to different kinetics of growth factor release and clot structure. ⁹⁵

Factors influencing clot structure

Effect of thrombin concentration on fibrin architecture

Thrombin concentration present at the time of blood clotting has been shown to profoundly influence the fibrin architecture compared with fibrinogen concentration, pH, and ionic strength.^43,96,97 Clots formed at low thrombin concentration (<1 nM) are composed of thick fibrin fibers in a loose configuration, while those formed at high thrombin concentration are composed of thin fibers in a tight configuration. ⁹⁸ Using turbidimetric analysis of plasma, it has been shown that the altered thrombin concentration contributes to different clot structure through fibrin polymerization process. An increase in thrombin concentration leads to a shorter time required for protofibrils to grow to a sufficient length before they aggregate. It also causes an increase in maximum rate of turbidity development and a decrease in the maximum final turbidity, indicating a faster fibrin formation and a decrease in fibrin size. ⁵⁸ Indeed, the pattern of in situ thrombin generation is shown to follow the initiation, amplification and propagation phases of coagulation. These phases are in turn profoundly affected by environmental factors.^99–103 As such, a dynamic change in the thrombin concentration (1 nM to greater than 500 nM) ³² may lead to significant differences in kinetics of fibrinopeptide release, protofibril and fiber formation.^97,98,104 This is also tied into the fact that normal plasma clots usually display a bimodal distribution of the fiber diameters.^105,106

In the presence of polymers, clots produced with fibrinogen solutions was also found to contain heterogeneous fibrin structure due to changes in protofibril aggregation rate, the number and size of fibers formed compared to controls with an addition of a single thrombin concentration. ¹⁰⁷ In line with this, it has been demonstrated that the surface functional groups significantly affect the efficacy of adsorbed fibrinogen to convert to fibrin.^108–110 With similar amounts of adsorbed fibrinogen, a denser fibrin network with more branches was found on methyl (−CH₃) surface associating with a larger amount of fibrinopeptides released at a faster rate compared with sparse fibrin observed on carboxyl (−COOH) surfaces.^110–112 The extent of fibrinopeptides release and fibrin proliferation have been shown to be related to surface-dependent fibrinopeptide availability. Approximately 2.7-fold more accessible fibrinopeptide A was found on fibrinogen adsorbed on −CH₃ surfaces for thrombin cleavage than those on −COOH surfaces. Hence, this may explain the higher efficacy of fibrin proliferation observed on −CH₃ surfaces. ¹¹³ Furthermore, different alkyl chain length of poly (alkyl methacrylates) have also displayed a major effect on regulating the rate of thrombin generation and fibrin deposition. ¹¹⁴ In accordance to this, surfaces presenting −COOH/methyl (−CH₃), ethyl (−CH₂CH₃) or butyl [−(CH₂)₃CH₃] functionalities at varied compositions are shown to provide another level of control of rate of thrombin generation, in addition to fibrin thickness and density throughout three-dimensional whole blood clots. The fibrin architecture at the edge and the center of the whole blood clots are also dramatically different (Fig. 6).

OPEN IN VIEWER

FIG. 6.

Scanning electron micrographs (5000×) of whole blood clot structures formed on surface with 33% −COOH/67% −CH₃ (a, b) and glass surface (c, d). The edge (a, c) and the center (b, d) of clots, scale bar represents 20 μm.

Relationship between clot structure and viscoelasticity

Individual fibrin fiber has been illustrated to possess extensibility and elasticity. ¹²³ The effect of fiber thickness on the properties of individual fibers is not known yet. Instead, it has been demonstrated that altered fibrin structure may modulate the clot rigidity as a whole, depending on fiber thickness, length, density, degree of branching, and cross-linking.^58,98 In particular, cross-links that reinforce fibrin contacts within the clot increase the elasticity of individual fibers and the overall clot elasticity.^46,123,124 Specifically, the clots formed on 33% −COOH surfaces with different alkyl groups are revealed to have changes in rigidity (Fig. 7).

OPEN IN VIEWER

FIG. 7.

Compaction studies of clots formed on surfaces with same content of−COOH (33%) but different alkyl groups compared to the uncoated glass surfaces. After 2 h incubation, the clots formed in the incubation vials were transferred to Eppendorff centrifuge tube (2.0 mL; Hamburg, Germany) and centrifuged at 6000 g for 60 s. The volume of fluid expelled from the network by centrifugation was measured and expressed as a percentage of the initial volume of the clot and was termed the compaction coefficient.¹⁶⁷ The lower compaction coefficient, the higher clot rigidity. Data were presented as mean of six replicates of each surface with SD. *p≤0.001. Color images available online at www.liebertpub.com/teb

Relationship between clot structure and fibrinolysis

Alterations in clot structure have been demonstrated to affect the clot susceptibility to fibrinolysis. In fibrinolysis, fibrin fibers are digested by enzyme plasmin, which is produced from cleavage of inactive plasminogen by tPA (Fig. 8). ¹²⁵ Fibrin fibers are shown to be transverse cut across, rather than by progressive cleavage uniformly around the fibers. ¹²⁶ A study by Collet et al. ¹²⁷ showed that while individual thick fibers are lysed more slowly than thin fibers, clots with a loose conformation of thick fibers were lysed more rapidly than those with tightly packed thin fibers. These findings, as confirmed by later study of Bhasin et al. ¹²⁸ indicate that the network conformation is a more important determinant for fibrinolysis rate compared with fiber thickness. The changes in network conformation are believed to regulate the lysis rate by influencing the fibrin density, tPA bindings on fibrin, and transport of fibrinolytic components throughout the clot.

OPEN IN VIEWER

FIG. 8.

Plasmin-mediated fibrinolysis. Inactive plasminogen is converted to active plasmin by tissue-type plasminogen activator (tPA). Plasmin digests fibrin and generates fibrin degradation products (FDPs): D-dimers and E fragments. Antifibrinolysis system, which includes proteins such as plasminogen activator inhibitor (PAI) 1 and 2, α2-antiplasmin, and thrombin-activatable fibrinolysis inhibitor (TAFI), inhibits the fibrinolysis at different steps. Color images available online at www.liebertpub.com/teb

Since fibrin architectures at the clot exterior and interior are different, a suspended clot system is a more feasible method to study the impact of fibrin structure modification on fibrinolysis. In this case, the exogenous fibrinolytic enzymes would initially interact with fibrin at the clot exterior and lysis would proceed from the clot exterior to interior. It is found that clots formed on 33% −COOH surfaces with different alkyl groups undergo different initial rate of fibrinolysis, which indeed is in good agreement with differences in fibrin thickness and density observed at the clot exterior (Fig. 9).

OPEN IN VIEWER

FIG. 9.

Release of fibrin degradation product (D-dimer) after 1 h of clot lysis. After 2 h incubation, the clots formed on various surfaces are suspended in 3 mL of phosphate-buffered saline (PBS) containing human plasminogen (Glu-plasminogen, 5.4 μg/mL final concentration; American Diagnostica, Inc., Stamford, CT). ²⁰⁶ Lysis was induced by adding recombinant tPA (0.25 μg/mL final concentration; American Diagnostica, Inc.) at 37°C with gentle agitation. Aliquots of 300 μL were removed at timed intervals and centrifuged at 1000 g for 3 min. The same volume of PBS was supplemented after samplings. The amount of D-dimer released from the clots was measured using IMUCLONE^® D-Dimer ELISA (American Diagnostica, Inc.). Clots in PBS only were used as controls of spontaneous fibrinolysis. Data were presented as mean of triplicates of each surface with SD. *p≤0.001. Color images available online at www.liebertpub.com/teb

Differences between a PRP gel and a hematoma

While PRP gels are mostly formed by adding a fixed and high amount of thrombin to platelet-concentrated plasma, it is conceivable that a PRP gel would be a platelet clump tightly networked by thin fibers. These abnormal clot structures and properties, and their impacts on growth factor release, cell proliferation, and physical stability are likely attributed to the negative effect on bone healing as previously reported. The confusion between the differences in cellular components and structures between a PRP gel and a hematoma, and how these differences relate to their potentials in enhancing bone healing may be the source of inconsistent results.

A hematoma contains mostly erythrocytes, ∼5% of platelets and <1% of leukocytes. ¹²⁹ However, a PRP gel contains theoretically no other blood cells but platelets, possessing nearly a reverse ratio of erythrocytes and platelets compared to a hematoma. In fact, platelets and circulating blood cells, both their number and interactions play an important role in clot features. It has long been proposed by Ulevitch and Johnston ¹³⁰ that erythrocytes participate in the intrinsic pathway of coagulation, which is believed to be associated with the negatively charged phospholipids of the cells.^131,132 Also, it has been demonstrated that erythrocytes interact with platelets by promoting platelet aggregation, and inversely activated platelets also enhance erythrocyte agglomeration.^133,134 In addition, leukocytes have been suggested to influence coagulation by expressing TF, activating platelets, and factor X,^{24,29,135,136} and cleavage of tissue factor pathway inhibitor.^137,138 All these studies are consistent with the findings of Thor et al., ¹³⁹ where whole blood induced higher levels of thrombin generation and platelet activation than PRP on clinically used titanium, and that the response of PRP could be partially restored in the presence of erythrocytes.^140,141

Furthermore, cell-associated fibrin is revealed to be more resistant to fibrinolysis than distally located fibrin.^142,143 This might be due to soluble proteins released from the cells, which can regulate the equilibrium between clot formation and dissolution. FXIII and PAI-1, released from platelets, are known to increase the resistance of the clot to fibrinolysis.^144–146 On the contrary, hemoglobin and neutrophil elastase and cathepsin G, are shown to enhance fibrinolysis.^147,148 These studies agree that during in vivo myocardic infarction, normal erythrocytes-rich clots are readily dissolved by enzymatic lysis, whereas platelet-rich clots are more resistant to be degraded.^149–151

So far, the innovative uses of PRP in bone tissue engineering solely focus on the biological value of platelet growth factors. There are conflicting results of PRP with bone healing and its utility remains unsolved. Recently, a second-generation of platelet concentrates has also been introduced based on leukocyte content and fibrin architecture including pure platelet-rich plasma, leukocyte- and platelet-rich plasma, pure platelet-rich fibrin, leukocyte- and platelet-rich fibrin.^152–155 Little is known about how these two parameters influences the intrinsic biology of these products,^156,157 and not to mention other cellular components in the clots are too often neglected.

Taken together, a PRP gel is different from a natural hematoma in both cellular components and structure (Fig. 10). It is likely that these differences contribute to different molecular and cellular activities and mechanical stability at the injured bone, ultimately dictating the outcome of peri-implant bone healing.

OPEN IN VIEWER

FIG. 10.

Schematic pictures illustrating differences in cellular components and fibrin scaffold between (a) platelet-rich plasma gel and (b) a normal hematoma. Color images available online at www.liebertpub.com/teb

The strategy of bone tissue engineering to treat critical-sized defects (CSDs) is to provide three key elements for bone healing^158,159: the scaffolding for osteoconduction, growth factors for osteoinduction, and progenitor cells for osteogenesis.^160–162 Although current approaches have shown some potential in regenerating bone, to date, no engineered material outperforms autograft in bone-forming ability. ¹⁶³ A blood clot displays several characteristics: (i) being biocompatible, (ii) possessing three-dimensional and highly interconnected porous network, (iii) presenting suitable surface chemistry for cell adhesion, proliferation, and differentiation, (iv) being biodegradable at an appropriate rate, and (v) displaying certain extent of mechanical properties. Hence, modification of a blood clot formation on the implant surface appears as a promising alternative to expedite natural healing mechanism enhancing bone healing in severe defects.

Effect of PRP and blood clot on bone healing: importance of appropriate clot structural properties in bone healing

The abnormal clot structure and susceptibility to fibrinolysis are principally causative mechanisms of many thrombotic diseases and bleeding disorders.^{106,128,164–167} Patients with acute ischemic stroke are found to produce in vitro plasma clots that are denser with thicker fibers and more resistant to fibrinolysis compared with controls. ¹⁶⁸ In contrast, hemophilia patients, who are deficient of factor VIII, are shown to produce clots that are much more porous with thicker fibers, and overly more susceptible to fibrinolysis. ⁴³ A dental implant inserted in the jaw bone is a typical example of endosseous (in bone) implants in which its clinical success is greatly influenced by a blood clot structure. A blood clot is formed at the gap between host bone and implant, as a result of coagulation activation. The blood clot not only detains blood flow, anchors the implant to the endosseous wound site, but most importantly supports two types of peri-implant endosseous healing: distance and contact osteogenesis. ¹⁶⁹ Distance osteogenesis occurs when new bone is initially formed on the surfaces of surrounding old bone at a distance from the implant. Conversely, contact osteogenesis takes place when new bone is first formed on the implant surface (Fig. 11). Hence, it is clinically considered as a superior mode of healing in case there is insufficient cortex to provide early stability.

OPEN IN VIEWER

FIG. 11.

Two distinct patterns of peri-implant endosseous healing: (a) distance osteogenesis and (b) contact osteogenesis where osteoblasts line initially on the host bone or implant surface respectively and lay down matrix. In distance osteogenesis, the bone surfaces provides a population of osteogenic cells, which differentiate into osteoblasts and lay down a new matrix that grows slowly toward the implant and results in the approximation of the implant surface shape by the newly formed bone. An intervening layer of nonbone cells is often formed between the bone and implant. Contact osteogenesis occurs with osteoconduction and de novo bone formation, which predominantly require recruitment and migration of osteogenic cells toward the implant surfaces. Modified from Davies.¹⁶⁹ Color images available online at www.liebertpub.com/teb

The prerequisite of contact osteogenesis is the continuous recruitment and migration of osteogenic cells to the implant surface through the clot structure. The three-dimensional network of fibrin and structural proteins of the clot serve as physical scaffolds to support cell adhesion and migration. ¹⁷⁰ Addition of a fibrin network has been shown to increase the biocompatibility of collagen scaffold for bone repair due to improved osteoblast adhesion, proliferation, and differentiation. ¹⁷¹ Studies of fibrin gels composed of varied concentrations of fibrinogen, thrombin, and calcium have clearly demonstrated that the fibrin structure influences cell behaviors, including sprouting of endothelial cells, in addition to viability, proliferation, and osteogenic differentiation of MSCs, which is known to play a major role in bone healing.^172–176 Increasing the fibrinogen dilution in forming fibrin gels increased MSC migration out of the gels, suggesting a diluted fibrin gels is more efficient for rapid cell delivery to damage tissue. ¹⁷⁷

Changes in thrombin concentration may also indirectly modulate clot architecture by activating factor XIII to cross-link adhesive proteins to fibrin. Bindings of fibronectin and collagen to fibrin are essential for cell adhesion and migration into the clots and supporting the formation of extracellular matrix at the injury sites, overall aiding the healing process.^178,179 The binding of fibronectin to integrin α₅β₁ is known to promote osteogenic differentiation of MSCs and mineralization in primary osteoblasts. ¹⁸⁰ Moreover, fibrin-bound actin, myosin, and vinculin together with platelet cytoskeleton have been shown to mediate clot retraction (to ∼1/10 of its original volume), and thus wound narrowing,^{42,116,181,182} An altered fibrin structure are reported to affect clot retraction. ¹⁸³

When adhered on fibrin, the preosteoblasts are known to impose contractile forces on the fibrin fibers that attached to the implant surface, where they differentiate into osteoblasts with the stimulation of osteogenic factors such as BMPs, IGF, TGF, and PDGF released from the peri-implant blood clot. The osteoblasts then directly lay down bone matrix. Following mineralization, a collagen-free cement line appears and results in de novo bone formation. ¹⁶⁹ Therefore, the fibrin architecture of the clot is important for effective fibrin retention of implant surface and critically determines the process of contact osteogenesis.^123,170 Furthermore, fibrin gels were demonstrated to be able to bind growth factors in biologically relevant concentration. ¹⁸⁴ Hence, it possibly explains the changes in growth factor release, cell division, and adhesive properties of PRP clots as previously reported by studies using thrombin at different concentrations.

Alternatively, specific functional groups on biomaterial surfaces that modify fibrin structure and susceptibility to lysis of resultant clots are also found to influence the releases of PDGF-AB and TGF-β1 during clot lysis (Fig. 12). ¹⁸⁵ Given significant understanding of the function of both growth factors in support bone healing, it is conceivable that a clot with appropriate structural properties has the pivotal role on affecting the bone healing by influencing macromolecule transport, cell behavior, and new tissue ingrowth. A number of studies have combined the PRP with bone substitute materials such as polycaprolactone-tricalcium phosphate, ¹⁸⁶ biphasic calcium phosphate, ¹⁸⁷ biphasic hydroxyl apatite/β-tricalcium phosphate, ¹⁸⁸ and calcium phosphate cement ¹⁸⁹ to enhance new bone formation. It has been found that PRP can effectively augment the bone formation activity of osteoconductive scaffolds.

OPEN IN VIEWER

FIG. 12.

In vitro releases of (a) platelet-derived growth factor (PDGF)-AB and (b) transforming growth factor-beta 1 (TGF-β1) during lysis of clots formed on 33% −COOH/67% −CH₃, −CH₂CH₃, or −(CH₂)₃CH₃ surfaces. Data were presented as mean of triplicates of each surface with SD. *p≤0.001 **p≤0.05. Color images available online at www.liebertpub.com/teb

The efficacy of a blood clot in subsequent response toward bone regeneration and its performance compared to that of PRP has been scarcely investigated in literature. Balaguer et al. who utilized blood clots for better handling of calcium phosphate MP found that the composite had osteoinductive properties in critical femoral defects and subcutaneous sites. ¹⁹⁰ In noncritical defects, Oliveira Filho et al. surprisingly found that the blood clot alone led to advanced stage of calvarial bone repair compared to highly concentrated PRP, whereas the addition of PRP to autografts even reduced the graft capacity of bone healing. ¹⁹¹ This suggests that the presence of a blood clot alone or in combination with biomaterials has the potency to induce bone formation. In an investigation by Giovanini et al., the diminishing effect of PRP on calvarial bone repair was shown to be associated with alteration of collagen matrix composition and α-smooth muscle actin expression, overall intensive formation of fibrous tissue.^192,193 In contrast, the presence of a blood clot was observed to improve challenging repair of cartilage. It induced three significant modifications at early sequences, including increased recruitment of inflammatory and marrow-derived stromal cells, vascularization of provisional tissue, and intramembranous bone formation. ¹⁹⁴ These findings provide some evidence of different underlying mechanisms of osteoconduction and osteoinduction mediated by blood clots and PRPs.

Clearly, the prerequisite of contact osteogenesis is the continuous recruitment and migration of osteogenic cells to the implant surface through the three-dimensional blood clot. Osteoinduction is largely stimulated by activated platelets and leukocytes entrapped in the compartment. They release a range of cytokines and growth factors, creating chemoattractant gradients to recruit undifferentiated or osteogenic cells to the implant site.^70,154,195

Recently, PRP supernatant or platelet derivatives from human have been suggested to be an alternative of cell culture medium. They demonstrated stimulating effects on expansion (colony number and average size) and differentiation (osteogenic, chondrogenic, and adipogenic) of MSCs.^196–198 The incorporation of MSCs in PRP was also found to be applicable in whole skeletal repair regardless of the origin of MSCs. Bone marrow MSCs (BMSCs) have a superior osteogenic potential than that of adipose tissue-derived MSCs (ADSCs). ADSCs were shown to inhibit cartilage regeneration, which is believed to be related to vascular endothelial growth factor-A mediated chondrocyte apoptosis and reduced proteoglycan synthesis. ¹⁹⁹ However, the addition of PRP not only promoted ADSC proliferation and chondrogenic differentiation, ²⁰⁰ but also resulted in functional chondrocytes secreting cartilaginous matrix in osteochondral defect. ²⁰¹ Similarly, the inferior osteogenic potential of ADSCs seeded on mineralized collagen than that of BMSCs in CSD was also shown to be partially compensated with the addition of PRP, ²⁰² Furthermore, dental pulp stem cells grafted with PRP was demonstrated to be osteoinductive, associated with a higher level of bone-implant contact than PRP alone. ²⁰³ These results are promising as they demonstrate a feasible strategy of skeletal repair based on both autologous agents and cells from different origins, which requires relatively few manipulations. It may also serve as a homing system of local osteoprogenitor cells via endogenous growth factors and stem cell interaction.

To achieve a good manufacturing practice of such approach, the precise composition of PRP requires further investigation as its cellular and molecular components are highly interlinked. However, only few studies with whole blood have provided insights on possible difference in molecular mechanisms of PRP and whole blood clots toward bone regeneration. Whole blood has a superior response than PRP on implants in term of thrombin generation and platelet activation due to the presence of erythrocytes.^139–141 In addition, serum has been shown to modulate the effect of platelet preparation on bone marrow culture. Supernatant of activated platelets supported proliferation and activity of tartrate-resistant acid phosphatase (TRAP⁺) multinucleated cells, whereas the addition of serum components decreased osteoclastogenesis but increased osteoblastogenesis. ²⁰⁴ Sphinigosine-1 phosphate, a strong endothelial cell chemoattractant, which is released from platelets during clotting, was also found to exert a higher activity in serum than in plasma, suggesting the link between blood coagulation and angiogenic response of endothelial cells. ²⁰⁵ Hence, the molecular components of PRP and blood clot have varied effects on recruitment, proliferation, and differentiation of cells. This could partially illustrate mechanism of PRP and blood clots in mediating difference sequences of skeletal repair. Overall, the presence of a blood clot with appropriate clot structural properties ensures the bone-implant interface environment to support the bone healing.