The healing of fractures and nonunions has significant science background to it; however, the application of the products in the surgeon’s hands should be considered an art in the science of bone healing. The surgeon must choose adequate fixation for stability and to promote healing by not making the construct too stiff. If a bone graft substitute is necessary, the surgeon must choose the type of bone graft substitute depending on patient factors and surgeon factors involving the treatment of the fracture.
Bone grafts and bone graft substitutes are necessary in trauma surgery for a variety of reasons. Acutely, there are various bone grafts and substitutes used in fracture fixation to provide scaffolding for periarticular fracture fixation. Subacute use of bone grafts and substitutes occurs when the fracture does not heal. When fractures fail to progress in healing, each case needs to be evaluated as to why fracture healing did not occur.
Let’s first look at what happens when a fracture occurs. The blood supply in bone comes from three sources: the intramedullary artery, periosteal vessels, and metaphyseal vessels. The injury itself disrupts some of the blood supply. The more severe the soft tissue injury accompanying the fracture, the more severe the disruption of the blood supply. With fracture fixation, the blood supply is disrupted during surgery by stripping the periosteal vessels with plate fixation or reaming with intramedullary nail placement. Recently, fracture fixation has become more biologically friendly attempting to preserve the periosteal vessels as much as can be possible. In addition to the vessels, the surrounding soft tissue envelope is just as important.
With a fracture, there is the release of numerous growth factors to induce the healing process. There are mesenchymal stem cells found in the bone marrow, the periosteum, and soft tissue and they differentiate into chondrogenic and osteogenic cells. Osteoblasts form bone directly to the intramembranous ossification, chondroblasts begin to form soft callous, and bone is formed through endochondral ossification. These processes are under the influence of many growth factors including platelet-derived growth factor (PDGF) and vascular endothelial growth factor (VEGF) transforming growth factor (TGF)-beta. The vascular supplies are very important for the formation of bone. It is necessary throughout the healing process for the healing to occur, including with nonunions.
The fracture hematoma that is present affects the oxygen tension of the tissues and it is the oxygen tension that has a role in the type of healing along with the fracture stability with the fracture treatment. Osteoprogenitor cells proliferate in areas of increased oxygen tension and decreased strength, so that when you have an area of high oxygen tension and no strain, you get formation of woven bone directly. In regions of intermediate strain and low oxygen tension you have a different environment and callous formation occurs. A fibrocartilaginous callous bridges across the fracture site. The cartilage then can reduce the strain and lead to bone formation.
The stability imparted by the fixation chosen also influences healing. Primary bone healing occurs with rigid stability and secondary healing happens with relative stability. Primary healing has apposition of fracture ends when there is bone-on-bone contact. Cutting cones cross the fracture, the osteoclasts create a tunnel, and the osteoblasts fill in the tunnel. New Haversion systems provide a pathway for blood vessels. The stiffness and strength of this construct increase as osteoblasts form new bone. With secondary fracture healing, you get healing with a callus. The vascular response varies according to the oxygen tension on the tissues just discussed and the mechanical stability. If you have an area where there is continued strain, that can be difficult to form bone and you may be more likely to get a fibrous nonunion.
Patient factors that affect bone healing include the use of nicotine. Nicotine affects the oxygen environment, thus giving you a low oxygen environment, and you are more likely to have fibrous tissue form and the healing process can take longer or the fracture may not heal at all. Medications that the patient takes affect the healing process. Nonsteroidal anti-inflammatory agents affect bone healing by disrupting the inflammatory cycle, which occurs early in fracture healing. Patients with diabetes have delayed healing and, in addition, they may have additional nutritional factors that affect their healing. Brinker and colleagues found a high percentage of male patients treated for nonunion had a nutritional abnormality. Most importantly, the patient’s soft tissue injury is key in promoting fracture healing. Gaston and colleagues evaluated tibia fractures and found that the most significantly affected fracture healing was the soft tissue injury. The soft tissue envelope is not to be taken lightly and disruptions in handling the soft tissues can affect the healing process. Likewise, when a nonunion occurs, it is important to create the environment to allow healing to begin to occur.
In order for bone healing to occur, you need an adequate blood supply and adequate mechanical stability. There are different three types of nonunions. Atrophic nonunions do not have adequate biology for healing. Oligotrophic and hypertrophic nonunions are nonunions that have callous formation present and are making attempts to heal ; however, the union is fibrous tissue and the fracture needs stability to promote healing. With any nonunion it is important to rule out infection as a cause. Preoperative laboratory values should be obtained including erythrocyte sedimentation rate (ESR), C reactive protein (CRP), and complete blood count (CBC) with differential. Further radiographic studies may be helpful if infection is suspected.
Once the nonunion is diagnosed, planning begins to promote healing. Further surgery is often necessary and may involve bone grafting. The choice of whether to use autograft or a bone graft substitute must be made. Bone graft substitutes represent a $2.5 billion industry. No matter what type of bone graft is chosen, there must be a conducive environment, biologically and mechanically, to promote healing.
It is important to understand the terminology of bone grafting. Osteoconductive bone grafts are those that provide scaffolding for new bone growth. New bone growth occurs by creeping substitution. This is a more passive response. Osteoinductive bone graft substitutes form bone in a more active process. Anything classified as osteoinductive will induce new bone formation where bone would not normally form. An osteogenic bone graft involves generation of bone directly from bone-forming cells.
No matter what product is chosen, the ultimate goal is fracture healing. No one product represents a magic potion to promote healing. The product chosen must be right for the environment and mechanical stability of the fracture.
The importance of mechanical stability on healing is known. Bone healing is an active process and stability plays a key role in this process. Bone formation is a continuous process and initiates early on in embryogenesis and continues throughout life. Bone remodeling after fracture continues in response to the stresses that are placed on it. When a fracture occurs, some heal with callus in which there was some motion or, with primary bone healing, no callous formation occurs.
Le and colleagues looked at which changes in the signaling process cause cartilage or bony callous. The authors wanted to evaluate the mechanism in which mesenchymal stem cells (MSC) sense the disparities between a stable and unstable fracture environment and how the healing occurs. The authors found that Indian hedgehog (IHH), which regulates the aspects of chondrocyte maturation during early fetal and early postnatal bone development, was expressed earlier in a nonstabilized fracture model as compared with those with a stabilized fracture. IHH exerts its effect on chondrocyte maturation through feedback involving bone morphogenetic proteins and transcription factors. Stabilizing the fracture decreases or minimizes the cartilaginous phases of bone repair that correlates with a decrease in IHH signaling. The authors concluded mechanical stabilization influences mesenchymal cell differentiation to bone. Because the mechanical stabilization is under the control of surgeons and it can affect signaling pathways important for bone healing, this aspect of fracture stabilization is key in minimizing nonunions.
Autograft often comes to mind as being the gold standard for bone grafting because of its osteogenic and osteoconductive properties. Autograft does not contain a significant amount of bone morphogenetic protein (BMP), so it is not considered osteoinductive. The most common site for autograft is the iliac crest. Other sites to consider are the greater trochanter, proximal tibia at Gerdy’s tubercle, distal tibia, the distal radius, and olecranon. Autograft contains rich sources of progenitor cells and growth factors, which are essential to help in the healing process. With cancellous bone grafts, about 15% of the osteoblasts and osteocytes survive and are capable of producing early bone. The surface area of the cancellous bone allows ingrowth of blood vessels and influx of osteoblast precursors. There is an active process of bone formation and resorption with eventual remodeling of the new bone.
After the autogenous cancellous cell bone is placed at the nonunion site and there is bleeding bone induced by taking down the nonunion, the cascade to promote healing begins. There are vascular buds that infiltrate the site and the graft is surrounded by inflammatory cells. If nonsteroidal anti-inflammatories are used during this phase, they would reduce the inflammatory response and thus inhibit the bone-forming cells. A small amount of fibrous tissue is formed later in the second week. In the fibrous granulation tissue predominant at the recipient nonunion site, there is an increase in osteoclastic activity observed. The macrophages remove necrotic tissue within the Haversian canals of the graft and there are intracellular by-products, which attract undifferentiated stem cells.
The secondary phase of cancellous incorporation involves osteoblasts lining the edges of the trabeculi and laying down a seam of osteoid, which surrounds the necrotic bone. The necrotic bone is gradually resorbed by osteoclasts. New bone continues to form. Then bone remodeling occurs over several months in response to the stresses placed on the bone.
Autograft can be an ideal choice in an aseptic nonunion in a healthy, young patient. You know the source of graft, and if there are no concerns regarding bone quality or metabolic deficiencies, autograft is a viable choice. The fracture has mechanical stability and needs a good biologic environment. By taking down the nonunion to a healthy tissue bed, autograft can provide the osteoconductive and osteogenic properties to promote healing.
With autograft, there are complications to consider in up to 40% of patients. These complications include limited donor site availability and donor site morbidities, which may include pain, iatrogenic fracture, and the need for increased hospital stay (which affects the economics of fracture care). With advancing technology and choices available for bone graft substitute, autograft may no longer be the gold standard.
Bone marrow aspirate
Because we know the success of autograft, the consideration for obtaining cells from the iliac crest through a bone marrow aspirate became highly attractive in terms of minimizing complications involved in the harvesting of the bone graft. This is not a new concept. In 1869, Goujon demonstrated the osteogenic capacity of bone marrow. The potency of the osteogenic capabilities of the bone marrow was considered. It is known the quality of this is donor dependent. The marrow has been estimated to contain 1/50,000 nucleated stem cells in younger adults and 1/1,000,000 in the elderly. Thus, women older than 35 with osteoporotic bone and/or elderly patients can have fewer stem cells available for formation of bone. In addition, with the increased knowledge of nonunions having a metabolic deficiency, we would be harvesting bone marrow from someone who is metabolically deficient.
If this material is injected, concerns include localization of the aspirate in the desired area. In addition, the studies published describe injecting the bone marrow aspirate directly into the nonunion without taking down the nonunion and preparing a bed to accept a bone graft substitute. Connolly and colleagues reported on the use of autologous bone marrow injection in 20 ununited tibial fractures over a 5-year period. This technique was used in conjunction with either a cast or an intramedullary (IM) nail fixation. The authors felt that the marrow stimulated callous formation significant enough to unite 8 of 10 nonunions mobilized with cast and 10 of 10 fractures treated with IM nails. Thus the authors concluded that the bone marrow injection was just as effective as autogenous bone grafting. There were confounding variables in this study. Because there was some mechanical stability provided at the same time, it makes it difficult to know which intervention, the aspirate or treatment chosen, was responsible for the results. In addition, in closely evaluating the results, the percutaneous bone marrow injection did not promote bone healing any more rapidly than with standard grafting, as it took an average time to healing of 6 months after injection.
Once the marrow is obtained, it needs to be processed to maximize its osteoinductivity. Different centrifugation techniques may affect the amounts of bone formation. Hernigou and colleagues evaluated percutaneous autologous bone marrow grafting for tibial nonunions specifically regarding the number of cells, number in concentration of progenitor cells transplanted, and callous volume and healing. They aspirated the bone marrow for 60 noninfected atrophic nonunions. The bone marrow was directly injected into the site without taking down any fibrous tissue and the overall results found that the aspirate contained an average of 612 ± 134 progenitor cells before concentration and an average of 2579 ± 1120 progenitors per cubic cm after concentration. In addition, the authors evaluated the fibroblast colony forming units (CFU) injected into each nonunion. There was bone union achieved in 53 patients and the bone marrow that was injected into the nonunion of the patient group that healed contained greater than 1500 progenitors per cubic cm. They found that the concentration and number injected into the nonunion sites of the patients without healing was significantly lower. The authors acknowledge the variability in osteogenic potential from patient to patient. Thus, the amount of progenitor cells is related to the effectiveness of percutaneous autologous bone marrow grafting. The authors cannot explain a mechanism to account for transformation of fibrous tissue into callous; normally the other fibrous nonunion needs to be taken out. It is like putting a band-aid on a laceration requiring sutures and getting lucky when it heals.
Watson and colleagues evaluated iliac crest aspirate of delayed union and nonunion of long bones in 52 patients over an 8-year period. The patients were aspirated under general anesthesia and the aspirate was centrifuged, the cellular concentrate was obtained, and large-bore needles were directed at the site of bone deficit under fluoroscopy. There were 46 patients available for long-term follow-up over 2 years. The results were quite poor with only 17 (37%) achieving bony union after the first procedure with an average time to union of 4 months after intervention. Two patients healed their nonunion after a second procedure and 6 of 7 patients who received multiple injections had a persistent nonunion despite these procedures. Of a total of 56 procedures performed, 37 (66%) failed. There was no difference between the group that healed versus the group that did not heal with regard to gender, fracture treatment, smoking, or type of nonunion. The use of iliac crest aspiration injection alone was found to be ineffective in the treatment of nonunion.
The studies discussed provide variable results. The effectiveness of the bone marrow aspirate is dependent on the donor, and if the donor already demonstrated an inability to heal, one must wonder about the benefits of this procedure. The use of percutaneous bone marrow aspirate as a bone graft substitute at this time is not promising or supported by high level of evidence studies.
As we discussed, fracture repair involves a process in which you get formation of a blood clot and an environment rich in growth factors to promote healing. So thus the natural thought is that with fracture, the platelets activated and granules released contain many osteopromotive factors. So the thought extended to using platelet-derived growth factor (PDGF) and platelet-rich plasma (PRP) to help promote healing. It was thought if one could deliver the platelets in a concentrated amount this would theoretically contribute to earlier bone repair and fracture healing. Platelets are isolated from autologous blood and this eliminates the concern for any immune response. The PRP therapy has been used since the 1970s. So the question became if this therapy is used by itself or in addition to bone graft material, could this promote healing?
PDGF/PRP are osteopromotive products, not osteoinductive products, and the concentration of the growth factors varies. Proteases used in the preparation of the product may degrade the growth factors. The machines used can destroy the platelets. In addition, PRP is often used in combination with allograft and/or autograft composites. Recent studies found that although components of PRP stimulate migration and proliferation of osteogenic progenitor cells, they also inhibit the osteogenic action of BMPs.
Because PDGF and PRP are often used in conjunction with other products in bone grafting, they were evaluated combined with demineralized bone matrix (DBM) in immunocompromised mice. The authors found that PDGF inhibited bone formation in a dose-dependent manner. PRP also reduced the osteoinductivity of the DBM.
The role of PRP in osteoinductive promotion of bone healing is not promising at this time. The product may have osteopromotive factors and has been found to be effective in certain types of surgery; specifically in wound healing and also in oral maxillofacial surgery. We continue to emphasize the importance of the surrounding tissue in healing of nonunions. The environment in oral maxillofacial surgery is different. The bone is different and subjected to different stresses, and it may be that PRP/PDGF may be enough to promote bone healing. However, the environment in which bone healing is required in humans after fracture repair is different. Most often it is in the diaphysial region and not like the rich cancellous bone in the oral maxillofacial region. In addition, PDGF/PRP may be helpful in wound healing, whether or not this effect helps with fracture healing is not proven at this time.