Abstract and Introduction
Abstract

Bone fusion can be achieved by one or more of three methods: in situ, onlay, and interbody fusion. Interbody implants provide the spine with the ability to bear an axial load. They function optimally when placed along the neutral axis and produce little, if any, significant bending moment. Interbody implants may be comprised of bone, non-bone materials such as acrylic, or a combination of both such as in interbody cages.

In this report the authors' goal is to provide some insight into the theoretical, as well as practical, biomechanical factors that influence bone fusion, focusing on interbody implants. They review the concept of stress shielding and its impact on fusion. With the attendant biomechanical nuances of the different regions of the spine, they discuss region-specific strategies involved in successful fusion. Finally, they review intraoperative techniques that will improve the chance of achieving a successful arthrodesis.

Introduction

Bone fusion may be achieved using one or more of three surgical strategies: in situ, onlay, or interbody fusion. In situ fusions are used when native bone is allowed to come in contact with other native bone that was previously prevented from doing so because of intervening soft tissue. A common clinical example is the denuding of facet cartilage performed in conjunction with placing a lateral mass plate system in the cervical spine. Onlay fusion techniques rely on a decorticated graft bed and the subsequent application of cortical and cancellous autograft. Interbody implants provide the spine with the ability to bear an axial load; they function optimally when placed along the neutral axis and thus produce little, if any, significant bending moment. Interbody implants may be comprised of bone, nonbone materials such as acrylic, or a combination of both, such as interbody cages. All three methods may be used alone or in combination with other implants that can be applied through anterior and/or posterior applications.

The application of interbody implants has more complex biomechanical implications than the other two implants and thus is the focus of this discussion. Interbody implants benefit from the compression-related augmentation of the bone healing-enhancing forces, as predicted by the Wolff law.

The theoretical, as well as practical, biomechanical factors that influence bone fusion are presented here. First, we review of the concept of using bone as an implant material. Second, with the attendant biomechanical nuances of the different regions of the spine, we discuss region-specific strategies. Third, we review the intraoperative techniques that will improve the chance of achieving a successful arthrodesis.

Bone as an Implant

Theoretically, there is an optimal distribution of load between a spinal implant and a bone graft. Such ideal balance provides stability, yet allows the graft to "see" the necessary compressive forces for healing, as delineated in the Wolff law. Seventy percent of the load should optimally be applied through the bone graft in both the thoracic and lumbar spine (BC Cheng, CSRS, 1997). When a disproportionate amount of the load is borne by the implant, the chance of achieving a solid arthrodesis diminishes. Stress shielding is defined as an implant induced reduction of bone healing enhancing stresses and loads to such a degree that stress reduction osteoporosis, or non-union, may result. In addition to the most common culprit -- spinal instrumentation -- inactivity, such as bedrest, and spinal bracing may also play a role.

In a head-to-head comparison, one group of authors reviewed their clinical experience with four anterior cervical strategies: 1) strut graft combined with halo brace; 2) anterior cervical plate combined with strut graft; 3) anterior strut graft combined with posterior instrumentation and fusion; and 4) anterior strut graft combined with a kickplate; the pseudarthrosis rates were 20%, 40%, 0%, and 7%, respectively (ES Doh, CSRS, 1998). Several conclusions may be drawn from these findings. Fusion in which no instrumentation was used in this series was relatively efficacious. This has been borne out in other studies as well (A Hildibrand, NASS, 1997).The stress shielding of the rigid plate most likely contributed to the high pseudarthrosis rate. The addition of a posterior fusion appears to prevent settling while promoting fusion.

A kickplate functions as a dynamized implant, permitting subsidence while theoretically preventing graft dislodgment (TJ O'Brien, CSRS, 1996). Unfortunately, a kick-plate resists angular deformation poorly.

Region-Specific Strategies

Lumbar Spine Constructs

In the lumbar spine, interbody fusion techniques have several theoretical and proven advantages over posterior onlay grafting techniques, including: 1) a decreased incidence of pseudarthrosis; 2) an accelerated rate of fusion; and 3) an increased axial load-bearing ability.

The lumbar spine, however, can provide a hostile environment for successful bone fusion as it bears substantial axial, torsional, and translational loads. Translational loads are particularly relevant at the lumbosacral junction where the local anatomy acts to convert axial loads into translational and angular resultant force vectors . Furthermore, it is difficult to resect a disc completely or to perform a complete corpectomy. On one hand, aggressive endplate debridement may result in the loss of integrity of this vital component of the axial load- bearing complex; on the other hand, incomplete debridement may result in pseudarthrosis due to inadequate preparation of the graft bed.

Posterior Lumbar Interbody Fusion. First popularized by Cloward, PLIF provides a technique by which to prepare the disc space and achieve an interbody fusion via a posterior approach. He emphasized careful endplate preparation, meticulous graft preparation, and the attainment of a high surface area of contact between the endplate and the bone graft . He was able to achieve excellent results by using bone as his only graft material. Unfortunately, others have been unable to duplicate his success, resulting in the provision of other adjunctive procedures, such as pedicle screw fixation, to increase the clinical results of PLIF. Nonetheless, much can be gleaned from the techniques of Cloward. Above all, subsidence is governed by the cross-sectional area of contact and the relationship of the strut to the margin of the endplate. In newer methods of PLIF the authors have modified both the mechanical (structural support) and biological functions (bone graft for fusion) of the traditional PLIF concept by making use of rectangular cages that increase the surface area of contact with bone, while providing structural support through the cage.

Anterior Lumbar Interbody Fusion. Difficulties with PLIF and the development of laparoscopic techniques have led to a rise in the popularity of anterior lumbar interbody fusion techniques. Here too, approach-related complications and similar pseudarthrosis rates have led others likewise to add posterior spinal implants (for example, pedicle screw fixation). Allograft, in which either fibula or femoral ring is used, seems to perform particularly well in this application.

Lateral Intertransverse Fusion. With a lateral intertransverse fusion, the fusion mass is situated at a significant distance from the instantaneous axis of rotation and the neutral axis. Even when a solid fusion is achieved, the resulting flexibility that persists can often result in clinical failure. Furthermore, the obligatory lateral soft-tissue retraction and associated soft-tissue injury may be a causative factor in the loss of lumbar lordosis and chronic back pain.

Flat-Faced Fusion Cages. Although their shapes in other planes may be varied, all flat-faced cages share the common attribute of presenting a flat surface to the accepting fusion bed (for example, the endplate region). Metal alloy cages have a much higher modulus of elasticity compared with that of bone, with carbon fiber cages and femoral ring allografts exhibiting a more intermediate value. The latter, by more closely matching the modulus of the endplate, reduces the chance of subsidence.

Although the surface area of contact varies among the different cages, it is almost always greater than that of the round-faced cages (JW Brantigan, NASS, 1997); however, the greater the area of contact, even with allograft cages, the less area there is for autologous bone contact. The bone mineral density of the endplate region also has significant biomechanical implications (HS An, NASS, 1997). Compared with an interbody strut or a round-faced cage, a flat-faced cage provides a significant advantage regarding angular deformation .

Round-Faced Cages. Originally developed by Bagby for application in equines (GW Bagby, NASS, 1997), TIFCs provided a method by which lumbar interbody fusion could be readily achieved, either with allograft dowels or metallic cages (GW Bagby and SD Kuslich, NASS, 1997). Threaded interbody fusion cages became popular, in part due to their ability to stiffen a motion segment acutely and their relative ease of application. They may even be inserted using minimally invasive techniques.

Although there have been studies in which persistent good results have been demonstrated,TIFCs have had their critics. Some have questioned the use of very young- age study populations (mean 41.5 [HA Yuan, unpublished data] and 42.1 24 years of age), the definition of a solid fusion (less than 5° of motion in the sagittal plane), and the absence of lucency surrounding the cage. Biomechanical concerns include the possible sequelae of disrupting the anterior and posterior ligaments during insertion (NM Grosland, NASS, 1997) and the cage's ability to withstand loading in shear manifest in the L5-S1 disc space once a patient assumes an upright posture .

With these cages a round face is presented to the fusion bed. As indicated by Cloward, maximizing the surface area of contact is critical. The degree of bone-on-bone contact varies among the clinically available TIFCs. For a given cage, placement in a lordotic spine may vary the quality of bone-on-bone contact throughout the length of the cage. Finally, the observation that explanted human cages have contained predominantly collagen and woven bone raises doubts about the integrity and quality of intracage bone (B Cunningham, NASS, 1999).

Subsidence-Related Biomechanics

Bone Shaping and Fitting

The ultimate outcome of a surgical procedure hinges on several factors. Although an implant may fail of its own accord, more commonly failure may be directly or indirectly attributed to the surgeon. A surgeon may choose the wrong operation or the wrong implant, or fail to use proper techniques. We have attempted to provide a basic theoretical knowledge of the relevant biomechanics, as well as clinical spine region-specific applications. This should help with choosing the right operation and the best implant/ construct. We now turn our attention to those intraoperative techniques that can literally make or break a fusion.

Good "carpentry" of bone components is critical for optimizing bone fusion outcome. The creation and shaping of a mortise in the VB is equally as important as the precise fitting and shaping of an interbody bone graft in minimizing the chance of dislodgment and other forms of failure. Three factors directly affect the incidence and extent of subsidence: 1) the closeness of fit of the bone graft in the VB mortise; 2) the surface area of contact between the bone graft and VB; and 3) the character or quality of the contact surfaces.

Closeness of Fit

Just as we learn in childhood, that square pegs do not set firmly in round holes, squared-off bone grafts do not fit well in a round mortise and vice versa. A poor fit increases the likelihood of two types of adverse outcomes: 1) nonunion due to an inadequate surface area of contact ; and 2) excessive subsidence due to the concentration of stresses and loads at the points of contact between the VB and the bone graft . On the other hand, maximizing the surface area of contact and optimizing the closeness of fit between the bone graft and the VB minimize stress concentration and, hence, minimize the chance of nonunion or excessive subsidence.

Surface Area of Contact

The extent of subsidence is inversely proportional to the surface area of contact between the bone graft and the VB. The larger the surface area of contact, the less the subsidence occurs, and vice versa. A basic example illustrates this principle: the force required to penetrate a Styrofoam block with the eraser end of a pencil is much greater than that required for the sharpened end.

Quality of the Contact Surfaces

Two qualities determine the efficacy of the contact surfaces: 1) the extent of endplate preservation; and 2) the proximity of the point of contact to the edge of the VB. Emery, et al.,have observed that simply burring the end-plate results in higher fusion rates without increasing the degree of clinically significant settling. Anyone who has stood on an aluminum soda can has realized the power of the "boundary effect." The VB cortex is superior to the softer inner cancellous bone at bearing axial loads. The size of the load that may be borne when all or a portion of the load is placed over the VB margin is dramatically increased by the buttressing effect of the cortical boundary. Thus, the greatest biomechanical advantage is achieved with a bone graft that makes contact with the entire VB surface and is the same size, making contact with all of the cortical margin.

Conclusions

Bone fusion involves the successful integration of several factors. First, one must understand the relevant biomechanical theory including the concept of bone as an implant. Second, one must be familiar with the variety of available constructs and the appropriate techniques required to apply them. Here, we have tried to describe these in a region-specific manner. Third, one must be proficient at the intraoperative skills that may make the difference between successful fusion and pseudarthrosis.