Introduction

Spinal arthrodesis is a method of treating traumatic and nontraumatic injuries of the spine and is performed on all segments of the spine. The goal of the procedure is to fuse 2 adjacent vertebral bodies to strengthen and stabilize the segment. Fusion is a complex process because of the staggered sequence of the 2 important components that contribute to a successful result: resorption of the graft material and formation of new bone within the graft.

Currently, autogenous iliac crest cancellous bone graft is the gold standard in grafting material for fusion mass formation in the spine, especially in the lumbar region. However, there are disadvantages to using this material. First, there is a limited supply of the autograft, and multilevel fusions require substantial amounts of bone graft, which the iliac crest harvest sites may not be able to accommodate. Second, patients who have had previous graft harvest procedures or who have compromised bone quality are poor candidates for graft harvest. Another issue of concern is the substantial donor site morbidity associated with all graft harvesting procedures. Persistent (chronic) pain at the iliac crest graft site occurs in approximately one third of patients. Although autograft is considered the gold standard in bone grafting, spinal fusion using autogenous bone graft is unsuccessful in 5% to 35% of patients.

Because of the inherent limitations of autogenous bone graft, alternatives such as allografts, bone graft substitutes, and recombinant human bone morphogenetic proteins (BMPs) were developed. The differences among these materials affect their clinical application. The most basic difference is determined by whether these materials are bone graft extenders or bone graft replacements, and they are classified as either extenders or replacements according to their inherent properties.

Bone Graft Properties

There are 3 distinct properties of graft materials, and each property has a specific role in bone healing. The first is osteogenicity, which signals the presence of osteoblasts, or bone-forming cells, that directly deposit bone. Only freshly harvested autogenous bone grafts or bone marrow aspirates contain these cells. The second property is osteoconductivity, which is the ability of the material to act as passive scaffolding that supports new bone formation and bone growth. A number of inert materials, such as ceramic grafts, have this property while others, such as allografts, are capable of undergoing creeping substitution. The third property is osteoinductivity. Osteoinductivity, first described by Urist in the mid-1960s, is defined as the presence of differentiation factors that facilitate the recruitment and differentiation of mesenchymal stem cells and specifically induce them to form osteoblasts that deposit de novo bone. These differentiation factors are termed BMPs. Figures 1 and 2 illustrate the differences between osteoconductive and osteoinductive bone graft.

Autologous iliac crest bone graft possesses all 3 characteristics (osteogenicity, osteoinductivity, and osteoconductivity). To eliminate the need for harvesting iliac crest autograft in spine fusion surgery, a bone graft replacement must possess the properties of both osteoinductivity and osteoconductivity.

Bone Grafting Materials

Available bone graft technologies consist of either bone graft extenders or bone graft replacements.

Graft extenders. The graft extenders act as a scaffold support for new bone formation. These materials cannot induce new bone formation; they function solely in a passive role by providing a matrix into which bone formation can occur. These products cannot function as bone graft replacements because they lack cellular elements and, because they do not contain precursor cells, they also lack the ability to initiate new bone formation (osteoinduction). These products can only function as extenders of autogenous bone graft; usually, they are osteoconductive materials. Commonly used graft extenders include calcium phosphates, such as ProOsteon (Interpore Cross International; Irvine, California), Master Graft, (Medtronic Sofamor Danek; Memphis, Tennessee), and VITOSS (Orthovita; Malvern, Pennsylvania); calcium phosphate and collagen composites, such as Osteoset (Wright Medical Technology; Arlington, Tennessee); and calcium sulfate granules.

Platelet concentrates, known as platelet rich plasma (PRP), also contain growth factors, but are not osteoinductive because they do not include BMP. Types of PRP include a variety of cytokines, such as platelet-derived growth factor (PDGF), transforming growth factor (TGF-beta1), insulin-like growth factor-1 (IGF-1), and vascular endothelial growth factor (VEGF). These proteins play a role in bone formation, but none of them, either individually or in combination, are capable of inducing new bone formation. PRP is believed to be osteopromotive when used with autograft; however, data from oral surgery studies indicate that neither vital bone production nor interfacial bone contact was achieved when PRP was used as an osteopromotive agent. Commercial examples of this technology include AGF (Interpore Cross International) and Symphony (Johnson & Johnson/Depuy; Warsaw, Indiana).

The other type of bone graft extender is demineralized bone matrix (DBM), which contains trace amounts of naturally occurring BMPs. These mildly osteoinductive materials are allograft-derived formulations. Commercial examples of these products include Grafton (Osteotech; Eatontown, New Jersey) and Osteofil (Medtronic Sofamor Danek). DBM is capable of extending or slightly enhancing the activity of the autograft. As an extender, DBM is added to the autograft to allow the surgeon to increase the graft volume and to improve the healing success of autogenous grafts.

Autograft replacements. Bone graft replacements are the second type of bone graft material available for clinical use. Because graft replacements are osteoinductive, they circumvent the need for harvesting autograft bone. Presently, BMPs are the only clinically available differentiation factors that are osteoinductive.

BMPs are capable of inducing any portion of the bone formation cascade. It is this unique property that allows these proteins, when combined with a suitable carrier, to be used as a bone graft replacement. Although allograft DBM contains trace amounts of naturally occurring BMP, the amount is 1 million times less than that found in recombinant BMP technologies, such as rhBMP-2, and is present only if the DBM is properly demineralized and sterilized. Autograft replacement applications require a relatively high concentration of BMP. An adequate concentration and volume of BMP is found only in INFUSE Bone Graft. (Medtronic Sofamor Danek). Note the difference in BMP concentrations in Figure 3.

Evaluating bone graft replacements. Criteria have been established for the evaluation of bone-grafting technologies (Figure 4). With bone replacements, spine surgeons should consider the following 3 guidelines before making a decision on any new grafting material for clinical use: (1) evidence of osteoinduction in a lower-order animal; (2) bone formation in higher-order animals using clinically relevant models; and (3) completion of prospective randomized human clinical trials.

An example of evidence of osteoinduction in a lower-order animal is ectopic bone formation in a rat. However, this model does not demonstrate or imply clinical efficacy. The best attempt to demonstrate efficacy is the intramuscular model, in which the product is placed in a soft tissue pouch in a rat's thigh muscle. If the material is capable of forming a nodule of bone, it is considered to have osteoinductive capabilities. BMPs and DBMs are capable of forming bone in this environment. Osteoconductive materials such as pure ceramics, combination ceramic and collagen carriers, and platelet rich plasmas fail at this level. The pure osteoconductive carriers do not induce new bone formation, nor do they enhance bone formation (Figure 1).

Although a variety of materials are capable of stimulating ectopic bone formation in rats, very few of them can do so in more challenging animal models. A bone graft replacement must stimulate successful spine fusion in higher-order animals to be considered a clinical option. The rhesus monkey model is the most predictive of a comparable outcome expected in human trials. Several studies have reported positive results with rhBMP-2 in spine fusion applications in the rhesus model.

Prospective, randomized trials are necessary to document the clinical outcomes in patients in whom the osteoinductive substitute is used in the same manner that it will be used clinically. While preclinical studies do have predictive value, they do not replace clinical outcomes documented in large clinical trials. As Table 1 shows, the BMP dose and time required to achieve comparable bone formation increases substantially when research progresses from rabbits, to dogs, to monkeys, and ultimately, to humans.

Conclusions