Enhancement of Fracture Healing

Physical Stimulation of Bone Healing

It has long been known that the mechanical environment of a long-bone fracture can influence the biologic process of healing [1]. Classically this has been demonstrated by the development of a large periosteal callous through the pathway of endochondral ossification in fractures stabilized with relatively flexible fixation that allows some motion at the fracture site. This is opposed by healing with minimal periosteal callous formation by direct intramembranous ossification when fractures are rigidly fixed, allowing virtually no motion between fracture ends. Which of these two pathways is preferable has been debated extensively as reviewed in chapter 85 and 86 on “bone healing” in this text. More recently, researchers and clinicians have been manipulating these responses to mechanical environment to optimize or stimulate fracture healing. This section summarizes the current techniques being investigated and used clinically, and the proposed mechanisms of action of these techniques.

Induced Micromotion (IMM)

This concept has been extensively studied by Goodship and Kenwright in the sheep tibia model [2,3] and in at least one human clinical trial [4]. This scheme uses an external fixator with pneumatic driven activator that provides axial tension and compressive forces that can be controlled in displacement, speed, and duration. A positive effect was seen in an ovine 3-mm transverse osteotomy gap model in callous formation and increasing stiffness of healing if stimulated by 1 mm of induced displacement when compared with nonstimulated controls [2]. More recent work suggests that the positive effects of IMM are greatest at relatively moderate or higher rates of induced motion (40 and 400 mm/sec versus 2 mm/sec) [3].

This positive effect was greatest when IMM was initiated 1 week postoperatively when compared with delayed initiation (after 6 weeks) and when studied at 12 weeks. Consequently, these authors feel IMM has the optimal positive effect on the early phases of the repair process, which in their model appears to be endochondral ossification. A subsequent clinical trial on 39 human tibial fractures treated with IMM did show a significant decrease in healing time compared with nonstimulated controls [4]. Another study [5] on externally induced motion was performed on 2-mm femoral osteotomies in rats. A specialized external fixator was used to provide cranial-caudal bending of the osteotomy from day 7 to day 18, 3 times a week. Finite element analysis found the cranial aspect of the healing osteotomy experienced approximately 7% tensile strain, while the caudal aspect of the osteotomy experienced 3.5% compressive strain. Overall, the tensile strains tended to promote endochondral ossification. They hypothesized by increasing the pool of mesenchymal progenitor cells and demonstrated increased bony bridging, whereas the compressive strains suppressed chondrogenesis and encouraged intramembranous ossification.

Dynamization

Dynamization is a concept that was initially described by De Bastiani and incorporated in the design of a human external fixator system (Orthofix™) [6]. De Bastiani proposed providing initially rigid fixation to allow revascularization and initial fracture healing followed by releasing the axial stabilization while continuing bending and torsional stabilization. Whereas many human clinical cases have been treated using this technique, most experimental studies have failed to show an overall significant positive effect on osteotomy healing in dogs [7]. In one study that did report a positive overall effect with dynamization performed 7 days postoperatively, the effect may have reflected increased rigidity resulting from total osteotomy gap collapse [8]. Another study using a veterinary external fixator design suggested dynamization after two weeks tended to cause larger callous size and callous density, but may impede remodeling at the fracture gap [9]. Statistically significant differences were not found with densitometry, absorptiometry, or CT throughout the period of the study, and a lack of torsional stiffness or maximal strength difference was found at the end, 13 weeks postoperatively.

Destabilization

The basis of this concept is to provide stabilized fixation initially and decrease that fixation rigidity as the fracture heals and develops strength. It developed as an attempt to combine the early advantages of stable fixation, such as rapid reduction in postoperative pain and increasing speed to limb use while avoiding the potential long-term disadvantages of rigid fixation such as stress protection of the fracture. Destabilization, also known as “staged disassembly,” developed as more rigid fixation was obtainable with external fixators through the use of more complex frames [10], new components [11], and threaded fixation pins [12]. Clinical observation of periodic atrophic delayed and nonunions occurring with these rigid devices stimulated attempts to optimize fracture healing by manipulating the fracture fixation mechanical environment. Early experiments with both transverse [13] and oblique [14] canine tibial osteotomies suggested that destabilization from rigid configurations of external fixator to much less rigid type I configurations (twofold in torsion, threefold in bending, and sevenfold in axial compression) [15] resulted in more callous production but a decreased healing strength when performed at 0, 2, and 4 weeks after osteotomy. Minimal callous production and more appearance of remodeling with greater fracture strength were seen when destabilized six weeks after osteotomy. No effect was seen when the osteotomies were destabilized 12 weeks after osteotomy and compared with contralateral rigidly stabilized controls. These results suggested that too great or early reduction in fixation rigidity can cause the fracture healing pathway to convert from primary intramembranous ossification to secondary endochondral ossification. Destabilization could result in greater fracture strength when performed at six weeks, probably by stimulating the hypertrophy and remodeling phase of primary bone healing. Late destabilization of that model provided no beneficial effect, either because the responsiveness of the healing tissue was inadequate at that time or the study did not go long enough (12 to 15 weeks). It was also noted that problems with fixation, pin loosening, and secondary lameness with infection developed as the fixation pin numbers decreased after destabilization. However, a similar, more recent and well controlled study performed by Dupuis’ laboratory found no significant differences in healing of a 2-mm wide oblique osteotomy when destabilizing from a type II to a type I frame [16]. These authors concluded that the previous positive study results may have reflected excessively rigidly controlled osteotomies, essentially causing stress protection, and that destabilization from a type III may, in fact, be a return to normal bone healing rather than enhanced bone healing [17-19]. However, they also recognize that different clinical situations may require different approaches, and slow-healing fractures may still benefit from a gradual increase in loading by progressive removal of fixation [16].

Increasing Stabilization

A final scenario of fracture fixation manipulation that has been discussed is that of allowing early motion at the fracture site to induce a large periosteal callous through the chondral precursor pathway, and then increasing rigidity to encourage the later stages of ossification. We performed a pilot study of this approach starting with a type I configuration for 4 weeks, followed by conversion to a rigid type III configuration after significant callous production. Overwhelming problems with fixation failure and pin loosening resulting in loss of reduction and poor limb use owing to pain caused the abandonment of this approach; however, with today’s improvement in device design and threaded pin use, the concept deserves further study.

Direct Current Stimulation (DC)

DC uses a constant 20 μA to stimulate the fracture. The cathode is placed at the stimulation site and the anode/battery pack is buried in subcutaneous tissues. Brighton [15,20] showed that at the cathode, PO2 is lower, which appears to favor bone formation; proteoglycan and collagen synthesis are increased. Currently, DC is approved for use in human established nonunions and spinal fusions [21], with up to 83% success reported for tibial nonunions [22]. The implantable battery eliminates the problem of patient compliance, but usually requires removal after six months.

Pulsed Electromagnetic Fields (PEMF)

The PEMF signal was developed to induce electrical fields in bones similar in magnitude and time to the endogenous electrical fields produced in response to strain.21 They may reflect the ability of bone to respond to change in a mechanical environment known as Wolff’s Law. The signal consists of bursts of EM pulses repeated at 15 Hz. Various in vivo studies have shown upregulation of TGF-B and a several fold increase in BMP mRNAs, which increased chondroneogenesis by enhancing differentiation of osteochondral precursor cells [22], suggesting that PEMF will have optimal effect on endochondral bone production. That is consistent with the clinical observation that PEMF is more effective in treating hypertrophic nonunions than atrophic nonunions [22]. Currently PEMF is used in humans as an adjunct to standard fracture and delayed union therapy.

Capacitive Coupling (CC)

This approach uses 2 surface electrodes placed on alternate sides of the fracture; the induced field is driven by an oscillating electrical current [21]. Field strength is calculated to be 0.1 to 20 mV/cm continuously applied. Stimulation of bone cell proliferation has been observed but the physiology of how electrical signals stimulate has been difficult to demonstrate in the laboratory [23]. CC is typically used with casting in treatment of nonunions with a 60% to 77% success rate and improvement from 65% to 85% success rates in obtaining spinal fusion [23].

Combined Magnetic Fields (CMF)

CMF appears to affect calcium ion transport across cell membranes and activate secretion of growth factors (insulin-like-growth factor-II), increasing cell proliferation [23,24]. However, more study is needed to fully explain its stimulation effect. Currently CMF is applied for 30 minutes a day in cases of nonunion and to stimulate spinal fusion [21]. Treatment of osteoarthritis and neuroarthropathy has been suggested but awaits more understanding of its mechanism of action [21].

Low-Intensity Ultrasound (LIUS)

Low-intensity ultrasound with high-frequency short bursts has been shown to accelerate fracture healing, but no specific stage of healing appears to be more sensitive [25]. A minimal heating effect (less than 1 degree C) may increase some enzymatic activity. Ultrasound does affect the exchange rate of potassium ions and increases the release of intracellular calcium and appears to stimulate proteoglycan synthesis in rat chondrocytes [26]. Also a greater degree of blood flow has been shown in canine ulnar fracture models with a generalized stimulatory effect [27]. In humans it is applied for 20 minutes daily. Current clinical indications are aimed at reducing the healing time of fresh fractures to avoid delayed unions and subsequent loss of fracture reduction [28]. It has also been used in treatment of nonunions and increasing strength of the new bone produced by distraction osteogenesis [21].

Shock-Wave Therapy

Shock wave uses a high-energy variable-frequency mechanical impulse applied with a transducer that can focus the impulses to concentrate on deeper structures. A rapid positive pressure wave is followed by a variable negative pressure that may cause cavitation of soft tissues or disruption of large structures. It was originally used as lithotripsy for the noninvasive fragmentation of kidney stones [29]. Several studies have shown positive and variable effects on the healing of nonunions [28], and one clinical study found a positive effect on late healing of acute tibial osteotomies in dogs [301]. Although the mechanism of action has not been clarified, it is proposed that shock-wave application produces microtrauma with hematoma creation. Subsequently, the induced neovascularization stimulates connective tissue proliferation and activity [31]. Generally, 6000 to 12,000 shock impulses are applied at a treatment that may be repeated periodically. Currently, shock-wave therapy is recommended only as an adjunctive to traditional treatments for chronic nonunions, particularly those with compromised local circulation. A temporary weakening of the osseous structures may result from the trabecular microfractures produced. There also appears to be a significant analgesic effect so caution in post-treatment activity must be used to avoid catastrophic fracture until the treated bone has an opportunity to respond and strengthen.

Biologic Stimulation of Bone Healing

Most fractures in companion animals heal in an acceptable time period with adequate stabilization of the fracture fragments. Factors that contribute to delay or prohibition of fracture healing exist. Major categories of these factors are deficiencies of vascular supply, deficiencies in the vigor of the osteochondral response, and deficiencies in stability or physical continuity. Use of cells, bioactive factors, and/or supportive matrices can enhance fracture healing by providing or stimulating the deficient factor. Bioactive materials can be grouped based on their properties that enhance healing: osteogenesis, new bone formation; osteoinduction, recruitment and differentiation of bone-forming cells; and osteoconduction, mechanical support or scaffold for bone-forming cells. These therapeutic options can be used alone or in synergistic combinations. Some of the bioactive factors discussed here are available for clinical use; however others are not, and have only been used in experimental fracture or defect models.

Materials that are Primarily Osteogenic and Osteoconductive

Autogenous cancellous or corticocancellous grafts are frequently used by veterinary surgeons. These grafts are the gold standard for enhancing fracture healing against which bone substitutes are compared. Autograft is readily available from the metaphysis of several long bones or the ilium. The volume of graft is limited and donor site morbidity occurs. When handled properly, autograft provides mesenchymal stem cells and differentiated osteoblasts and osteocytes that can form new bone. The mineralized bone in the autograft provides osteoconductive surfaces on which bone-forming cells can lay down new matrix. Bone grafts are addressed more extensively in Bone Grafting.

Mesenchymal stem cells (MSC) or multipotent adult progenitor cells (MAPC) are pluripotent cells that have a high replicative capacity and the potential to differentiate into many tissues including the osteogenic lineage [30]. Approximately 1 in 25,000 nucleated cells from an adult canine bone marrow aspirate is a MSC [32]. These cells can be directed to develop into an osteogenic lineage during culture expansion by exposing the cells to a number of substances, such as transforming growth factor-β (TGF-β), insulin-like growth factor (IGF), vascular endothelial growth factor (VEGF), dexamethasone, β-glycerophosphate, or ascorbic acid phosphate [33,34]. The MSCs in the marrow of an autogenous graft are exposed to these factors during the inflammatory response from the fracture and surgical intervention. Bone morphogenetic proteins or osteogenic proteins, BMP-2 or BMP-7/OP-1, are potent factors that induce MSC to differentiate into osteoblastic cells [35-37]. Local concentrations of bone morphogenetic protein (BMP) at a fracture site may increase as osteoclasts resorb bone matrix [38]. The multipotent cells that appear early in the fracture hematoma also appear to synthesize BMPs, which creates a cascade of events that results in greater numbers of multipotent cells and increased levels of BMPs [39].

Two services currently purify and expand MSC from either bone marrow (VetCell Bioscience) or lipoaspirates (Vet-Stem). These concentrates have been used to enhance healing of equine tendons and ligaments. MSCs expanded in culture and exposed to an osteoinductive factor have been loaded onto an osteoconductive carrier. Although these osteogenic and osteoconductive combinations have been shown to enhance fracture and bone defect healing in many experimental animal models, they are not currently used in human or veterinary clinical fractures. The problems associated with clinical use of this treatment modality are that 1) 2 to 3 weeks are required to purify and expand the cells following the initial collection, causing a delay before sufficient amounts of patient-derived MSCs are available for re-introduction; 2) patient-derived MSCs may not be suitable for regeneration of damaged tissues if the patient is aged or if has bone marrow neoplasia; and 3) the cost of isolation, expansion, and safety and quality control must be covered for each patient and is cost-prohibitive.

Many surgeons are mixing unconcentrated bone marrow aspirates with osteoconductive material, such as bone allograft or hydroxyapatite to enhance bone healing. The osteogenic potential of concentrated MSCs has been shown to be significantly greater than fresh bone-marrow aspirate. Systems that concentrate MSCs from fresh bone-marrow aspirates instantly in the operating room are commercially available. CELLECT® (Johnson & Johnson) filters fresh marrow through a central chamber that holds osteoconductive material (i.e., cancellous bone or demineralized bone matrix) and concentrates the MSC three- to fourfold.

Periosteum is tissue that covers the external surface of most bones. It is composed of two microscopically distinct layers. An outer, fibrous layer is composed of fibroblasts, collagen, and elastin fibers with a neural and microvascular network [40-42]. The inner, cambium layer in contact with the cortical surface is highly cellular. It contains MAPC, osteogenic progenitor cells and osteoblasts, fibroblasts, tiny blood vessels, and sympathetic nerves [41-43]. Trauma activates the progenitor cells of the periosteum, although most studies have explored the use of periosteum to generate cartilage. Experimental models have shown that the osteogenic potential of this tissue can be stimulated by basic fibroblastic growth factor (bFGF), transforming growth factor-beta (TGF-β) and BMPs[44-46]. One study showed the powerful ability of periosteum to heal 5-cm tibial cortical defects in sheep with new bone production [47].

Activated platelets in the fracture hematoma release several growth factors including platelet-derived growth factor (PDGF) and bFGF. These factors stimulate the proliferation of periosteum-derived cells and contribute to the mitogenic response of the periosteum during callus formation [48-50]. After proliferation these cells differentiate into osteoblasts or chondroblasts to form bone or cartilage to allow bridging of the fracture gap.

Materials that are Primarily Osteoconductive

Processed allogenic bone graft has no osteogenetic capacity and little osteoinductive activity. Its primary purpose is to provide structural integrity and act as a scaffold for new bone formation [51]. In order to minimize its immunogenicity, the graft must be processed in some way, which in turn compromises its mechanical properties. The advantages of allograft are that it is readily available in virtually unlimited amounts and engenders no donor site morbidity. Large cortical allografts are pieces of dead bone, and subsequently, are not completely replaced by new host bone and are more prone to infection. Veterinary Transplant Services Incorporated is a bone bank for canine and feline allograft bone in a variety of structures.

Synthetic bone graft substitutes consist of calcium sulfate or plaster of Paris; hydroxyapatite, tricalcium phosphate, and combinations of these minerals, known as ceramics; and synthetic glasses composed of SiO2, Na2O, CaO, and P2O5. The materials vary in fabrication technique, crystallinity, pore dimensions, mechanical properties, and resorption rate. All synthetic porous substitutes share numerous advantages over autograft and allograft including their unlimited supply, easy sterilization, and storage. However, the degree to which the substitute provides an osteoconductive structural framework for new bone ingrowth differs depending on material composition. Calcium sulfate as cement and pellets is commercially available and its affordability, resorbability, and biocompatibility make it an option for veterinary surgeons. Cerasorb Vet (New Generation Devices) is a veterinary product of pure tricalcium phosphate that is available in granular and block forms.

Hydroxyapatite (HA) porous implants are produced when coral carbonate in the exoskeleton of a marine invertebrate of the genus Porite is converted by a hydrothermal exchange. The new material is mechanically superior while maintaining the internal structure of the coral. It has parallel channels 230 μm in diameter with interconnecting fenestrations of 190 μm diameter that are similar to the structure of cortical bone. Fibrovascular tissue that initially invades HA implants is later replaced by mature lamellar bone [52]. Although some surface resorption of the bulk implant does occur by osteoclast-like cells, major remodeling is minimal owing to the inert, insoluble structure of the HA.

Many ceramic preparations are commercially available and are widely used in human orthopedic surgery. Disadvantages of ceramic implants include brittle handling properties, variable rates of resorption, poor performance in diaphyseal defects, and potentially adverse effects on normal bone remodeling. These shortcomings have restricted their primary use to bone graft extenders and carriers for pharmaceuticals. A particulate bioglass preparation marketed for veterinary use (Consil, Nutramax Laboratories Inc.) is indicated for use in infrabony pockets associated with dental and periodontal disease and defects. Bioactive glass may induce more osteoproduction than other ceramics, but it is slowly resorbed.

Materials that are Osteoconductive and Osteoinductive

Demineralized bone matrix (DBM) is allograft bone that has undergone acid extraction of the mineralized portion of the matrix that increases the availability of inductive proteins. Several osteoinductive proteins, including bone morphogenetic protein, may be present in physiologic and biologic concentrations and proportions. One of the advantages of DBM is that, with time and continued remodeling, it will be completely resorbed and replaced by new host bone. Veterinary Transplant Services Inc. markets canine and feline DBM powder alone or combined with cancellous allograft chips to provide an osteoconductive matrix.

As mentioned earlier, fresh bone marrow, that may be centrifuged or filtered to concentrate MSCs, can be mixed with DBM/cancellous chips to provide a graft material that has osteogenic as well as osteoinductive and osteoconductive properties. Another strategy that provides a synergistic effect among bioactive materials is to mix DBM or DBM and bone marrow aspirate with platelet concentrate. A platelet concentrating system, Symphony PCS (DePuy), uses an automated device to rapidly concentrate platelets from a relatively small volume of a patient’s blood. The concentrate has three to six times the levels of PDGF, IGF, TGF-β, and VEGF. These factors are chemotactic for stem cells, osteoblasts, and chondrocytes and are mitogenic for MSCs and osteoblasts. There is also a graft delivery system, Symphony GDS (DePuy) or GPS (Biomet) that has a specially designed graft chamber and manifold for the delivery of graft materials to an orthopedic surgical site. The system facilitates premixing of graft materials with platelet concentrate, bone marrow, or blood that can then be placed into a fracture site or bone defect.

Materials that are Primarily Osteoinductive

Osteoinductive molecules promote bone formation mediated by cell signaling that prompts multipotent cells to differentiate into osteoblasts. The most widely studied of these cytokines are the BMPs. This family of proteins has several members, but BMPs 2, 4, 6, and 7 (also known as osteogenic protein-1, OP-1) have the most potent osteoinductive efficacy.53 Experimental preclinical studies in a variety of species including dogs have demonstrated healing of bone defects, fractures, and spinal fusions with the application of different BMP-carrier systems. Early studies used DBM or highly purified BMP, and more recent studies use recombinant BMP, usually human recombinant BMP (rhBMP). These water-soluble proteins must be combined with a carrier to maintain sustained delivery of adequate protein concentrations in the desired site.

Much of the current research effort is to find the optimal carrier and optimal dose for different species and specific sites in every species. Bone morphogenetic protein 2 (Medtronic Sofamor Danek) on an absorbable collagen sponge, and the OP-1 implant (Stryker Biotech), OP-1 in a purified bovine type 1 collagen particulate matrix, are the most heavily studied osteoinductive biomaterials. These two products are available for limited applications in human orthopedic surgery, including periodontal lesions, spinal fusions, and tibial nonunion fractures. Although these implants are not approved for veterinary use and their cost would most likely be prohibitive, they have had extra label use in a limited number of clinical cases of delayed or nonunion fractures in dogs. The latest method of delivery of BMPs and other bioactive cytokines to desired sites is by gene therapy. Delivery of the relevant DNA to host cells can be accomplished using nonviral, e.g, plasmid, or viral, e.g., adenovirus, vectors. The vectors can be delivered directly into the host tissues by in vivo transfer, or indirectly by in vitro transfer into harvested host cells that are then returned to the desired site. Bone morphogenetic proteins are addressed more extensively in Secundary (indirect) Bone Healing.

The exact roles in bone formation of other growth factors that have been identified are less clear. Conflicting results in experimental studies using these cytokines for fracture healing and clinical applications of these bioactive factors have been reported. Unlike the BMPs, these growth factors are incapable of de novo bone formation in an ectopic site. Transforming growth factor-beta is a protein found in platelets, bone, and cartilage that may act synergistically with other cytokines to promote differentiation, proliferation, and matrix synthesis by osteoblasts. Osteoclasts degrade bone during fracture repair by acid hydrolysis of the matrix, creating an acidic environment that can activate latent TGF-β [54]. The TGF-β released from the bone matrix can activate osteoblasts to produce new bone matrix.

Basic fibroblast growth factor (bFGF) acts primarily as a mitogen, stimulating increased DNA synthesis and cell division. When injected into a fracture site on a fibrin or hyaluronan carrier, it increased callus size, mineral content, and mechanical strength of the healing bone [55,56]. Insulin-like growth factors I and II stimulate bone formation by promoting proliferation of osteoblastic precursors and matrix production by osteoblasts. Platelet-derived growth factor is released by platelets during fracture healing and is chemotactic for fibroblasts, monocytes, and osteoblasts and stimulates proliferation of mesenchymal cells. Growth and differentiation factor-5 (GDF-5) is a divergent member of the TGF-β superfamily required for normal skeletal development. When loaded on a collagen matrix, GDF-5 induced ectopic cartilage and bone formation, new bone formation in long-bone defects in non-human primates, and fusion of vertebral transverse processes in rabbits [57,58].