Journal Logo


The Biology of Bone Grafting

Khan, Safdar N. MD; Cammisa, Frank P. Jr MD; Sandhu, Harvinder S. MD; Diwan, Ashish D. MD, PhD; Girardi, Federico P. MD; Lane, Joseph M. MD

Author Information
Journal of the American Academy of Orthopaedic Surgeons: January 2005 - Volume 13 - Issue 1 - p 77-86
  • Free


The transplantation of animal tissues into humans has been attempted since the time of Hippocrates. Evidence suggests that the ancient Hindus and Egyptians also undertook transplantation experiments; however, the first documented xenograft is attributed to Dutch surgeon Job van Mee'kren, who in the 1600s1 attempted to fill a defect in a soldier's cranium with a piece of dog's skull. In 1821, the first experimental autogenous bone grafting procedure was performed successfully in Germany in experimental defects created in animal skulls. Sir William MacEwen introduced allografting in 1879 by successfully replacing the proximal two thirds of a humerus in a 4-year-old boy with bone procured from other patients.2

The science of bone grafting has evolved significantly, particularly in the past two decades, with the fundamental understanding of osseous healing now incorporating principles of cellular and molecular biology. Bone grafts are used in virtually every aspect of reconstructive orthopaedics, from the simple treatment of fractures to extensive limb salvage procedures and complex spinal reconstructions. Several factors dictate the successful incorporation of grafted bone, including the type of bone graft used (Table 1), the site of implantation, the vascularity of the graft and the host-graft interface, the immunogenetics between the donor and the host, preservation techniques, local and systemic factors (Table 2), and the mechanical properties that depend on the size, shape, and type of graft used. Given the increased clinical demand for bone grafting, a review of the biologic events surrounding implantation of the various bone grafts is essential to understand the mechanisms of repair that lead to graft incorporation.

Table 1
Table 1:
Bone Graft Activity by Type
Table 2
Table 2:
Local and Systemic Factors Influencing Graft Incorporation

Functions of Bone Grafts

Bone grafts serve a combined mechanical and biologic function; depending on the desired clinical outcome, one function may be more important than the other. For instance, massive osteochondral grafts in limb salvage procedures for tumors serve a predominantly mechanical support function, whereas autogenous bone graft derived from the iliac crest for posterolateral spine fusions provides a biologic stimulus for new bone formation, with little or no mechanical function. A complex relationship exists at the host-graft interface, and to ensure the desired clinical result, the surgeon must be aware of the properties of both the graft and the recipient site.

Osteogenesis is the synthesis of new bone by cells derived from either the graft or the host. When correctly handled, cells from cortical and cancellous grafts can survive the transfer to the host site and form new bone that is critical in the initial phase of bone repair. The properties of cancellous grafts, which consist of an intimate trabecular structure lined with osteoblasts and a large surface area, make them very attractive at sites where new bone formation is desired. Indeed, this concept of osteogenesis forms the biologic justification of decortication for spinal fusion. Exposing the intramedullary space of the transverse processes, lamina, and pedicles with a burr opens local bone marrow to the fusion site. Marrow elements then provide the fusion bed with osteoinductive proteins, potential osteogenic cells, and a local blood supply.

Osteoinduction is the process by which mesenchymal stem cells (MSCs) at and around the host site are recruited to differentiate into chondroblasts and osteoblasts. Recruitment and differentiation are modulated by graft matrix-derived growth factors whose activity is triggered when bone mineral is removed. These growth factors include bone morphogenetic proteins -2, -4, and -7, which are members of the transforming growth factor-β superfamily. Other factors involved with bone formation include mitogens, such as platelet-derived growth factors, interleukins, fibroblast growth factors, insulin-like growth factors, granulocyte colony-stimulating factors, and granulocyte-macrophage colony-stimulating factors. Angiogenic factors, such as vascular endothelial-derived growth factor, also are released.

Osteoconduction is the process by which an ordered, spatial threedimensional ingrowth of capillaries, perivascular tissue, and MSCs takes place from the host site along the implanted graft. This scaffold permits the formation of new bone along a predictable pattern determined by the biology of the graft and the mechanical environment of the host-graft interface.3 For bone grafting to be successful, osteogenic activity and bone formation alone are insufficient. New bone must be distributed evenly in the grafted volume and must unite with the local host bone. Failure results in discontinuous bone formation without adequate mechanical strength to support function.

Graft site preparation is important to the success of the grafting procedure, and meticulous adherence to surgical principles is essential. Care must be taken to ensure adequate surface area contact between the graft and recipient site without interposition of soft tissue. Overzealous use of a reamer or burr may cause excessive heat generation, leading to necrosis at the graft site. Efforts must be made to preserve the osteogenicity of corticocancellous autografts, including decreased harvest-to-implant time, storage in covered containers, and attention to hydration.

Autograft Biology

Autogenous bone graft is the transplantation of bone taken from one anatomic site to another site in the same individual. The graft contains osteogenic properties (marrow-derived osteoblastic cells as well as preosteoblastic precursor cells), osteoinductive properties (noncollagenous bone matrix proteins, including growth factors), and osteoconductive properties (bone mineral and collagen). There is complete histocompatibility and no opportunity for disease transmission. As a result, autografts are the most common and favored graft material in musculoskeletal reconstruction. However, there are drawbacks to the use of autogenous bone, including insufficient amounts of graft material, particularly in children and in revision reconstructive procedures; the likelihood of significant postsurgical morbidity at the donor site (ie, rib, fibula, iliac crest), such as infection, pain, hemorrhage, muscle weakness, and nerve injury; increased surgical time and blood loss; and additional cost.4

Autogenous Bone Marrow

The observations of Goujon in 1869 regarding the ability of autologous marrow to induce bone at heterotopic sites led to the initial interest in the potential osteogenic capabilities of marrow.5 Phimester6 applied Goujon's observations clinically for the first time more than 50 years ago. Since then, the use of autogenous cancellous bone marrow graft to stimulate skeletal repair has become standard in many musculoskeletal procedures. The ability of bone marrow to perform its function as a graft depends on the presence of MSCs, ubiquitous undifferentiated cells that divide throughout life. Under specific influences, they may form additional undifferentiated clones or produce transitory cells that amplify and eventually differentiate fully into the various connective tissue types, including bone, cartilage, ligament, tendons, and fat.

Under specific mechanobiologic influences and the presence of boneinducing growth factors, MSCs in bone marrow differentiate into osteoblasts by medullary osteogenesis. The concept of medullary osteogenesis has been applied clinically to fractures of the tibia, where nonunions are frequent. The high incidence of complications incurred in the healing phase of tibial fractures has led many to advocate the early use of bone grafting combined with appropriate internal fixation. In this regard, the use of autogenous bone marrow graft injection is an alternative to open surgical grafting. Bone marrow is aspirated from the iliac crest and injected percutaneously at the fracture location to stimulate osteogenesis. Despite the initial success of treating nonunions with marrow grafts, the injected material tended to diffuse away from the graft site, particularly in large osseous defects. Attempts have been made to counter this drawback by adding demineralized bone matrix (DBM) or other carriers to concentrate the marrow at the target site.7 The limited number of stem cells in the marrow is another drawback of marrow grafting. It has been estimated that marrow contains 1 per 50,000 nucleated stem cells in young adults and as few as 1 per 1 million in the elderly. Efforts have been made to increase stem cell concentration by centrifuging the marrow as well as expanding the mesenchymal population by special culture conditions ex vivo.8

Autogenous Cancellous Bone

By virtue of its large surface area covered with dormant and active osteoblasts, autogenous cancellous bone is very osteogenic, easily revascularized, and rapidly incorporated at the host site (Fig. 1). Although autogenous cancellous grafts lack mechanical strength, their robust biologic activity in inducing and producing new bone provides early stability at the recipient site.

Figure 1
Figure 1:
Histology of morcellized human cancellous bone in the interspace between the transverse processes of L4 and L5, 6 months after implantation in an athymic rat. Note the trabecular continuity between the transverse processes (TP), with evidence of incorporation at the graft-transverse process interface (arrow) (hematoxylin-eosin, original magnification ×10). (Courtesy of Linda E. A. Kanim, Santa Monica, CA.)

Hemorrhage and inflammation immediately after the surgical procedure lead to coagulated blood, with vascular buds infiltrating the recipient site. During the first week after the procedure, the graft is surrounded by a medley of inflammatory cells, including lymphocytes, plasma cells, osteoclasts, mononuclear cells, and polynuclear cells. Small amounts of fibrous tissue also are present. In the second week, fibrous granulation tissue is predominant at the recipient site, and an increase in osteoclastic activity is observed. Invading macrophages remove necrotic tissue within the haversian canals of the graft, which leads to the release of intracellular byproducts that, along with the low oxygen tension and pH of the recipient site, serve as chemoattractants to host undifferentiated stem cells.

Neovascularization occurs within the graft as early as 2 days after implantation; as it continues, there is a repopulation of the marrow spaces with primitive MSCs of both donor and recipient origin. Under osteoinductive influences, these cells begin to differentiate into osteogenic cells (Table 3). These osteoinductive influences include cytokines, growth factors, and prostaglandins. As a result, nonsteroidal anti-inflammatory drugs are contraindicated during the acute healing phase because of their inhibition of prostaglandin synthesis, reduction of immune and inflammatory responses, and inhibition of boneforming cells at endosteal surfaces. The primary phase of hemorrhage, inflammation, revascularization, and osteoin-duction occurs as a continuum, with active bone formation and resorption by 4 weeks of implantation.

Table 3
Table 3:
Influence of Growth Factors on Graft Incorporation and Bone Healing

In the secondary phase of cancellous autograft incorporation, the osteoblasts line the edges of the dead trabeculae and lay down a seam of osteoid, which eventually surrounds a central core of dead bone.3 The radiodensity of the graft site initially increases; however, overall radiodensity gradually decreases as the necrotic bone is subsequently resorbed by osteoclasts. Concurrently, hematopoietic marrow cells form new bone marrow within the transplanted bone. Radiographic evidence of graft incorporation may be judged by the loss of sharp margins between the graft and recipient bone that occurs because of remodeling. Remodeling can take several months to complete.9

Biomechanical variations during cancellous autograft incorporation may be ascribed to the underlying biologic process. Cancellous autografts are incorporated by the formation of new bone on a necrotic bed. As a result, the mechanical properties of the construct are initially strengthened. As the necrotic bone is resorbed and substituted, the mechanical strength of the graft-host junction gradually returns.

Autogenous Cortical Bone

The early stages of cancellous and nonvascularized cortical graft incorporation are similar; however, the most apparent difference is in the rate of revascularization and degree of osteoinduction (Fig. 2). Revascularization is hampered by the dense architecture of the graft and the relatively scarce number of endosteal cells available for the formation of end-toend anastomoses, as is seen in cancellous grafts.9 Incorporation thus is initiated by osteoclasts instead of osteoblasts. Extensive resorption begins as early as 2 weeks after surgery and increases until about 6 months; thereafter, the rate of resorption gradually declines to normal levels at 1 year. Both early revascularization and resorption preferentially follow peripheral haversian canals and interstitial lamellae. Resorption of the inner cortical material progresses, although at a much slower rate. Resorption ceases when a certain cavity size of the central, osteonal canal is reached; thereafter, osteoblasts appear and begin to lay down new bone. Appositional bone growth, well underway by 3 weeks, further limits osteoclastic activity.

Figure 2
Figure 2:
Histology of human cortical bone from a laminectomy site in the space between L4 and L5, 6 months after implantation in an athymic rat. Note discrete chunks of unremodeled cortical bone surrounded by fibrous tissue in the mass. Remodeling and incorporation of cortical bone is evident with the transverse process (TP). New bone is woven into the graft-transverse process interface (arrow) (hematoxylin-eosin, original magnification ×10). (Courtesy of Linda E. A. Kanim, Santa Monica, CA.)

The end result is complete resorption of the graft with concomitant replacement with viable new bone, known as creeping substitution. Cortical autografts demonstrate creeping substitution most prominently at the graft-recipient site. The substitution progresses transversely and parallel to the long axis of the graft. As a result, initial repair is greater at the grafthost junction; repair then proceeds to the mid regions between the graft-host interfaces.10 This is radiographically evident by increasing radiolucencies in the initial 6- to 12-month period by irregular peripheral bony margins. As osteogenesis continues, radiodensity increases initially at the graft-host interface before extending onto the mid regions of the graft.

Unlike cancellous autografts, cortical grafts remain a combination of necrotic and new bone for a prolonged period. Remodeling of nonvascularized cortical grafts initiated by osteoclastic activity may lead to bone loss, resulting in up to 75% reduction in mechanical strength.11 This weakness persists for months to years after surgery depending on the size of the bone graft used. However, vascularized cortical autografts heal quickly at the graft-recipient junction because the resorption and remodeling process closely resembles that of normal bone. Residual weakness of the construct is minimal.11

Allograft Biology

Although autogenous bone derived from the patient's own iliac crest generally is considered to be the optimal source for grafting, factors such as limited quantities of available autograft, donor site morbidity, and, in certain instances, unsatisfactory biologic activity have dictated the search for alternatives. Allografting is the process by which bone is transferred between two genetically dissimilar individuals of the same species. More than 200,000 bone allografts are used in musculoskeletal procedures annually in the United States.12 Allografts are used primarily to support mechanical loads and resist failure at sites where structural support is desired. The development of modern tissue banks has resulted in a supply of high-quality allogeneic tissue for reconstructive orthopaedic surgery. More than 50 tissue banks in the United States are accredited by the American Association of Tissue Banks (AATB). The success of allografting in recent years may be directly related to stringent measures to ensure safety of the transplanted tissue and to improvement in processing technology.13 Combined efforts by the US Food and Drug Administration (FDA) and the AATB have helped ensure that tissue donors are adequately screened and that allograft material is properly processed, labeled, and distributed.

Screening of cadaveric donors begins with a detailed medical, social, and sexual history questionnaire completed by the life partner or next of kin. Any of the following automatically disqualifies the individual as a donor: positive history of exposure to specific communicable diseases, unprotected sexual contact, drug use, neurologic diseases, autoimmune diseases, collagen disorders, or metabolic diseases. The greatest concern with using allograft materials is the possibility of viral disease transmission, including hepatitis C, hepatitis B, and human immunodeficiency virus (HIV).12 Since 1980, two cases of HIV transmission as a result of musculoskeletal allografting have been reported.14 However, the incidence of disease transmission seems to have been halted primarily because of strict donor screening programs adopted by the AATB and mandatory screening blood tests for all donors. The blood tests, mandated by the FDA for all human tissues intended for transplantation, include titers for HIV-1 and HIV-2 antibody, hepatitis B surface antigen, hepatitis B core antibody, hepatitis C antibody, syphilis, human T-lymphotropic virus I antibody, and HIV p24 antigen.15,16 Other tests, such as the polymerase chain reaction test used to amplify minute quantities of viral genomes in transplantable tissues and genome amplification tests used to detect RNA in blood donors, have been introduced in the screening algorithm.

Several methods may be used to process allograft bone, including lowdose (<20 kGy) irradiation, physical débridement, ultrasonic or pulsatile water washes, ethanol treatment, and antibiotic soaking (4°C for at least 1 hour).3 The goal of processing is to remove the antigenic components of the graft to avoid induction of a host immune response, to ensure sterility, and to retain certain biologic and biomechanical functions. In the event of contamination, terminal sterilization by gamma irradiation, electron beam irradiation, or ethylene oxide treatment may be used. However, these treatments have a dose-dependent effect on the mechanical strength of the graft.17

Recently, a fatal bacterial infection with Clostridium species was reported after the use of a femoral condyle allograft for reconstructive knee surgery.18 Hematogenous seeding from bowel flora before harvesting likely contaminated the donor tissue. After a thorough investigation, the Centers for Disease Control identified a total of 26 patients with allograft-associated infections: 13 with Clostridium species infections and 14 with infections associated with a particular tissue processing agency. Factors that may contribute to contamination with bowel flora include the time interval between death of the donor and tissue retrieval, delays in refrigeration, and mechanism of death (eg, trauma). The Clostridium case highlights the problem of bacteriostasis; cultures obtained before and after processing in antibiotic/antifungal solution by the tissue processor were negative.

Sterilization must include methods that can kill bacterial spores (eg, gamma irradiation). Tissue not processed with a sporicidal method should be considered unsterile, and health care providers should be informed of the possible risk of bacterial infection. When a sporicidal method is unavailable, tissue should be cultured before soaking in antimicrobial solutions. If bowel flora are isolated, all tissue from that particular donor that cannot be sterilized should be discarded. Simultaneously, culture methods must be validated to ensure that residual antibiotics do not lead to a false-negative culture result. Both destructive and swab cultures must be considered, and recommended time limits for tissue retrieval must be followed.18

To facilitate storage, allografts may be cooled and transplanted within 24 hours or freeze-dried and stored at -80°C. Freezing alone minimally changes the properties of the graft; however, freeze-drying (lyophilizing) can cause changes in the graft mechanical properties, resulting in microfractures along the collagen fibers in the matrix. As a result, tissues processed by freeze-drying usually require some form of rehydration to regain some of the diminished mechanical properties.19

Host response to bone grafts is still poorly understood. Several investigators have demonstrated the presence of humoral- and cell-mediated responses associated with fresh osseous allografts. Experimental work has shown immunogenicity reduction when grafts were deep frozen and remarkable reduction when freezedried. An intimate relationship seems to exist between the cells and cytokines of the immune system and the cells of the bone remodeling system during fundamental processes of bone homeostasis, including remodeling and repair. A variety of potential antigens present in allograft bone may elicit a host immune response. This reaction is primarily a cellmediated response to major histocompatibility complex class I and II antigens on specialized antigenpresenting cells. Horowitz and Friedlaender20 reported that allogeneic bone cells activate host T-cells of the suppressor phenotype and induce their proliferation. It has been theorized that thymus-derived T-cells, along with macrophages from the host, recognize alloantigens from the implanted bone graft and respond to them.21 That response may lead to the release of cytokines, some of which share a role in the bone remodeling cycle. The effect of human leukocyte antigen matching and the host response to allograft bone has been studied, but currently there are no clinically significant data. Clearly, the immune system plays an important role in bone graft incorporation, but the exact nature of this relationship is unknown.

Allograft Cancellous Bone

Compared with autograft cancellous bone, allograft cancellous bone is a poor promoter of bone healing, although the autograft may have been procured from sites such as the femoral head, ilium, proximal tibia, or distal femur. In the first 2 weeks after implantation, fresh allografts invoke an extensive host response mediated by lymphocytes and macrophages. This aggressive immune response leads to the destruction and eventual inhibition of the essential osteoinductive growth factor-mediated response requisite for bone graft incorporation. Neoangiogenesis also is delayed, with new vessels surrounded by inflammatory cells leading to occlusion and necrosis. By 8 weeks, fibrous tissue begins to encapsulate the allograft. Depending on the disparity in histocompatibility between donor graft and the host, this response may persist for 8 months or longer.

The host response to bone and cartilage allografting is similar to that observed after transplantation of other organs or soft-tissue allografts. The aggressive immune response and resulting delay in incorporation has led to clinically disappointing results in the use of fresh allograft cancellous bone. Consequently, allografts are frozen or freeze-dried in an attempt to reduce immunogenicity and encourage incorporation.22 Although remodeling and revascularization are inferior compared with fresh autograft cancellous bone, resorption, osteoconduction, and osteoinduction proceed more rapidly with preserved allografts than with fresh allografts. Allograft cancellous chips are incorporated more completely and significantly faster than are allogeneic cortical bone grafts because they are more easily revascularized than their cortical counterparts. The cancellous grafts act as a scaffold onto which the host lays down new bone. The allografts are never completely resorbed by the host osteoclasts and remain entrapped within the host bone many years after transplantation.

Cancellous chips are the most common type of allogeneic cancellous graft. They are predominantly cuboid blocks of cancellous bone measuring 1 cm3. Cancellous chips are commercially available as 40 cm3 of freezedried chips or approximately 90 cm3 of frozen chips. Clinically, allograft cancellous chips have been used in numerous scenarios, such as spinal fusion augmentation and filling of bone defects, particularly in revision joint reconstructions.

Allograft Cortical Bone

Fresh cortical allografts undergo an inflammatory response similar to that of autografts. As revascularization proceeds, the host mounts an inflammatory response against the graft-derived cellular antigens. After the initial nonspecific inflammation, specific lymphocytic infiltration leads to occlusion of the invading host vessels, which then leads to graft necrosis and invasion by granulation tissue. Cortical allograft incorporation occurs by sporadic formation of new appositional bone. The lack of vascularization leads to significant weakness of the graft (compared with cortical autografts) for up to 1 year after surgery. Poor vascularization of large cortical allografts has been attributed to the density of cortical bone, lack of stability (in larger allograft constructs), and immunologic reaction to the grafts. However, smaller segments of allogeneic cortical grafts, such as fibular strut grafts used in cervical spine surgery (Fig. 3), are more rapidly incorporated because of potentially easier revascularization. These grafts are processed to remove the periosteal sleeve and other immunogenic soft tissues as well as donor marrow within the graft.23

Figure 3
Figure 3:
Lateral radiograph 3 months after an anterior cervical fusion, showing nonvascularized fibular allograft used as an inlay strut graft spanning three disk spaces. (Reproduced with permission from Muschler G, Lane J: Orthopaedic surgery, in Habal MB, Reddi AH [eds]: Bone Grafts and Bone Substitutes. Philadelphia, PA: WB Saunders, 1992, pp 384-395.)

Fresh allografts incorporate unsatisfactorily; therefore, processed and preserved allografts are favored in clinical practice (Fig. 4). Processing involves the removal of antigenic cells and proteins; preservation techniques include deep-freezing at -70°C or freeze-drying. Deep frozen grafts retain their mechanical properties and may be implanted after thawing; however, as noted, freeze-drying may alter the material properties of the graft.24 Rehydration (reconstitution) of the graft is required before implantation. However, even after rehydration, these grafts remain weak and are vulnerable in torsion and bending.

Figure 4 A,
Figure 4 A,:
Anteroposterior radiograph demonstrating the use of large cortical struts for reconstruction and augmentation of the proximal femur in a revision setting after fracture below a primary stem. B, Anteroposterior radiograph demonstrating the use of a fibular allograft as a mechanical strut to resist compression. The defect resulted from the excision of a low-grade osteosarcoma from the medial calcar of the femur. Autogenous cancellous graft from the contralateral iliac crest also was used. C, Anteroposterior radiograph demonstrating the use of proximal humerus allograft for reconstruction after wide excision of a radioresistant metastatic adenocarcinoma. The rotator cuff was preserved and reattached to the allograft to provide stability and function to the glenohumeral joint. (Reproduced with permission from Muschler G, Lane J: Orthopaedic surgery, in Habal MB, Reddi AH [eds]: Bone Grafts and Bone Substitutes. Philadelphia, PA: WB Saunders, 1992, pp 384-395.)

Massive Osteochondral Allografts

Massive osteochondral allografts, comprising diaphyseal cortical bone, metaphyseal cancellous bone, and articular cartilage, are used primarily in joint reconstruction after limb salvage procedures for tumor resection (Fig. 5). Osteochondral allografts are washed in an antibiotic-free tissue culture medium and cleaned of excessive soft-tissue attachment and bone marrow. Apart from the joint capsules and the ligaments and tendons that may be required during reconstruction, the periosteum and soft tissue covering the graft are meticulously removed. The bone marrow is then washed out of the shaft of the graft. However, epiphyseal cancellous bone and marrow are retained. Radiographs are taken to allow the graft to be matched anatomically with the needs of the recipient. Osteochondral grafts also are deep-frozen to ensure reduction in graft antigenicity. The grafts are stored at -80°C after the articular cartilage has been treated with glycerol or dimethyl sulfoxide for cryopreservation.

Figure 5
Figure 5:
Posteroanterior radiograph demonstrating fresh-frozen cryopreserved osteochondral allograft replacement of the left distal femur after excision of osteosarcoma. Cruciate ligaments were reconstructed using Gore-Tex (W. L. Gore, Newark, DE) synthetic replacements to augment knee joint stability. The posterior capsule and medial and lateral collateral ligament complexes were repaired using allograft and host soft-tissue attachments. (Reproduced with permission from Muschler G, Lane J: Orthopaedic surgery, in Habal MB, Reddi AH [eds]: Bone Grafts and Bone Substitutes. Philadelphia, PA: WB Saunders, 1992, pp 384-395.)

The allograft soft-tissue attachments are connected to the host soft tissues via imbrication. The host tissue allows creeping substitution, leading to slow incorporation of these grafts. Subchondral trabeculae usually are thickened because of new bone formation with relatively poor remodeling. Intertrabecular spaces are filled with fibrous connective tissue. After cryopreservation, the viability of the native articular cartilage remains poor. Satisfactory surgical results are correlated with good anatomic fit of the graft and joint stability. The host-graft interface is the site of potential nonunion in these massive grafts. Other complications include allograft fracture, articular surface degeneration necessitating subsequent arthroplasty, joint instability, and infection.

Clinical and radiographic results of massive osteochondral allografts reflect variable success, ranging from 60% to 90% in several series.25,26 Friedlaender et al25 reviewed the results of 29 patients who had massive frozen osseous or osteochondral allografts for segmental bone loss secondary to either resection of bone tumor or traumatic bone loss. At a mean follow-up of 10.6 years, 23 patients (79%) had achieved an excellent or good clinical result; 6 patients had a fair or poor result. Allograft-related complications included allograft fractures, nonunions, deep infections, and osteoarthritis.

Allogeneic Demineralized Bone Matrix

Mild acid extraction of bone leaves behind growth factors, noncollagenous proteins, and collagen while removing the mineral phase of bone. This demineralized, partially defatted homologous bone matrix provides a suitable framework for cells to populate and produce new bone and also may stimulate the healing response by encouraging MSCs to differentiate into bone-forming osteoblasts. This differentiation occurs via the extracellular matrix, which contains noncollagenous osteoinductive proteins. Hence, at the site of implantation, these factors trigger the endochondral ossification cascade, culminating in new bone formation at the site of implantation (Fig. 6).

Figure 6
Figure 6:
Histology of demineralized bone matrix after implantation at 6 months in an athymic rat. Note large mass of remodeled bone with discrete cortex and excellent incorporation with the transverse process. (Courtesy of Linda E. A. Kanim, Santa Monica, CA.)

The osteoinductive properties of DBM are influenced by several factors. Processing methods greatly influence osteoinductivity. Choosing the appropriate demineralizing agent is important. For example, 0.5 to 0.6 M hydrochloric acid is the most commonly used agent. Hydrochloric acid used in combination with alcohol yields noninductive DBM. Acetic acid, lactic acid, and nitric acid also have been shown to produce poor quality DBM. Sonication during processing adversely affects the osteoinductivity of DBM. Demineralizing time is an essential variable that requires strict control. Remaining mineral in the tissue affects the host cell response to the osteoinductive matrix proteins. As a result, the mineral content in the implant must be reduced to at least 40% of the normal level.22 The size of the DBM particle also is a critical factor that affects the osteoinductive response.24

Processed DBM composites became generally available for clinical use in 1991 and since then have been used widely in orthopaedic bone grafting. Grafton DBM Gel (Osteotech, Eatontown, NJ), the first widely available preparation, consists of DBM combined with a glycerol carrier and is easily applied from a standard syringe onto a surgical site. Variations of the Grafton composite also are available, including putty and flexible forms (Grafton DBM Putty and Grafton DBM Flex), but clinical data assessing their performance are limited. Various formulations now being prepared contain demineralized corticocancellous chips, presumably for greater osteoconduction and osteoinduction. Other bone processing facilities have made available preparations of DBM with alternative carriers. Dynagraft (Citagenix, Montreal, Quebec, Canada) combines DBM with a Pluronic reverse-phase copolymer carrier, which becomes firmer as it warms to body temperature. Osteofil (Regeneration Technologies, Alachua, FL) combines thermoplastic, collagen-based, hydrogel carrier matrix. This non-water soluble composite is easily extruded through a syringe when warmed to 46°C to 50°C but becomes firm when cooled to body temperature. Allomatrix (Wright Medical Technologies, Arlington, TN) contains DBM with calcium sulfate pellets and a carbomethylcellulose carrier. DBX (Synthes, Paoli, PA) is another widely used DBM preparation that uses a hyaluronic acid-based carrier. Clinical data from randomized, multicenter clinical trials are needed before safely claiming biologic efficacy for any of these DBM allografts.

Impairment of Bone Graft Healing

A variety of factors has been associated with impairment of bone graft incorporation. Smoking inhibits cellular proliferation and causes vasoconstriction. Systemic steroid use leads to inhibition of the differentiation of progenitor cells down the osteoblastic pathway. The effects of nonsteroidal anti-inflammatory drugs are well known; they inhibit prostaglandin formation, leading to diminished local blood flow, thereby delaying graft resorption. Malnutrition, especially calcium and phosphorus deficiencies, have been associated with delayed mineralization of new bone.27


Bone graft is commonly used in reconstructive surgery. Although autogenous grafts typically are utilized, limited amounts of autograft, donor site morbidity, and occasional unsatisfactory biologic activity have led to increased use of allografts. Meticulous donor screening and allograft bone processing are necessary to ensure patient safety. Incorporation of a graft within the host depends not only on the type of graft but also on the site of the transplant, quality of the transplanted bone and of the host bone, host bed preparation, graft preservation techniques, systemic and local disease, and mechanical properties of the graft. Understanding of the key biologic events surrounding the incorporation of the various types of bone grafts continues to evolve. The most interesting developments in the next decade will be the potentially widespread clinical use of a variety of biologic devices to augment bone grafts. Recombinant human growth factors, both rhBMP-2 and rhBMP-7 (OP-1), already are being studied experimentally in spine and traumatic clinical applications. FDA approval of rhBMP-2 in anterior lumbar spinal fusion has made the concept of biologic arthrodesis a clinical reality. Developments in gene therapy research may in the near future lead to the use of minimally invasive interventions for bone graft healing.


1. Meek'ren JJV: Observations Medico-Chirurgicae. Amsterdam, The Netherlands: H & T Boon, 1632, p 6.
2. Macewen W: Observation concerning transplantation of bone: Illustrated by a case of interhuman osseous transplantation whereby over two thirds of the shaft of a humerus was restored. Proc R Soc London 1881;32:232-247.
3. Stevenson S: Biology of bone grafts. Orthop Clin North Am 1999;30:543-552.
4. Younger EM, Chapman MW: Morbidity at bone graft donor sites. J Orthop Trauma 1989;3:192-195.
5. Chase SW, Herndon CH: The fate of autogenous and homogenous bone grafts. J Bone Joint Surg Am 1955;37:809-841.
6. Phimester D: Treatment of ununited fractures by onlay bone grafts without screw or tie fixation and without breaking down of the fibrous union. J Bone Joint Surg 1947;29:946-960.
7. Bruder SP, Caplan AI, Gotoh Y, Gerstenfeld LC, Glimcher MJ: Immunohistochemical localization of an approximately 66 kD glycosylated phosphoprotein during development of the embryonic chick tibia. Calcif Tissue Int 1991;48:429-437.
8. Lindholm TS, Nilsson OS, Lindholm TC: Extraskeletal and intraskeletal new bone formation induced by demineralized bone matrix combined with bone marrow cells. Clin Orthop 1982;171:251-255.
9. Burchardt H: Biology of bone transplantation. Orthop Clin North Am 1987;18:187-196.
10. Stevenson S: Enhancement of fracture healing with autogenous and allogeneic bone grafts. Clin Orthop 1998;355 (suppl):S239-S246.
11. Goldberg VM, Stevenson S, Shaffer JW, et al: Biological and physical properties of autogenous vascularized fibular grafts in dogs. J Bone Joint Surg Am 1990;72:801-810.
12. Tomford WW, Mankin HJ: Bone banking: Update on methods and materials. Orthop Clin North Am 1999;30:565-570.
13. Voggenreiter G, Ascherl R, Blumel G, Schmit-Neuerburg KP: Effects of preservation and sterilization on cortical bone grafts: A scanning electron microscopic study. Arch Orthop Trauma Surg 1994;113:294-296.
14. Transmission of HIV through bone transplantation: Case report and public health recommendations. MM WR Morb Mortal Wkly Rep 1988;37:597-599.
15. Tomford WW: Transmission of disease through transplantation of musculoskeletal allografts. J Bone Joint Surg Am 1995;77:1742-1754.
16. Department of Health and Human Services (FDA): Human tissue intended for transplantation. Fed Reg 1997;62:40429-40447.
17. Jinno T, Miric A, Feighan J, Kirk SK, Davy DT, Stevenson S: The effects of processing and low dose irradiation on cortical bone grafts. Clin Orthop 2000;375:275-285.
18. Update: Allograft-associated bacterial infections— United States, 2002. MMWR Morb Mortal Wkly Rep 2002;51:207-210.
19. Gazdag AR, Lane JM, Glaser D, Forster RA: Alternatives to autogenous bone graft: Efficacy and indications. J Am Acad Orthop Surg 1995;3:1-8.
20. Horowitz MC, Friedlaender GE: Induction of specific T-cell responsiveness to allogeneic bone. J Bone Joint Surg Am 1991;73:1157-1168.
21. Horowitz MC, Friedlaender GE: Immunologic aspects of bone transplantation: A rationale for future studies. Orthop Clin North Am 1987;18:227-234.
22. Boyce T, Edwards J, Scarborough N: Allograft bone: The influence of processing on safety and performance. Orthop Clin North Am 1999;30:571-581.
23. Guo MZ, Xia ZS, Lin LB: The mechanical and biological properties of demineralised cortical bone allografts in animals. J Bone Joint Surg Br 1991;73:791-794.
24. Thorén K, Aspenberg P: Ethylene oxide sterilization impairs allograft incorporation in a conduction chamber. Clin Orthop 1995;318:259-264.
25. Friedlaender GE, Strong DM, Tomford WW, Mankin HJ: Long-term follow-up of patients with osteochondral allografts: A correlation between immunologic responses and clinical outcome. Orthop Clin North Am 1999;30:583-588.
26. Mankin HJ, Gebhardt MC, Jennings LC, Springfield DS, Tomford WW: Long-term results of allograft replacement in the management of bone tumors. Clin Orthop 1996;324:86-97.
27. Boden SD, Sumner DR: Biologic factors affecting spinal fusion and bone regeneration. Spine 1995;20(suppl 24):102S-112S.
© 2005 by American Academy of Orthopaedic Surgeons