CHAPTER 54 Total Facet Replacement
The many advancements in the understanding of spinal pathology and spinal implants in the last 20 years, coupled with 30 years of extremity total joint arthroplasty, have created a new era in spine surgery. Complete segmental replacement in the spine is now possible. Total disc replacement has been used successfully for almost 15 years.
The degenerative problems of the facet joints and their resultant pathology have remained as a potential pain source, a remaining area of poorly understood and poorly diagnosed axial skeletal disease. Total facet arthroplasty is available for the treatment for central stenosis, lateral stenosis, degenerative spondylolisthesis, pars interarticularis syndrome, and other overlapping diseases caused by damaged or degenerative posterior elements.
The lumbar facet joints have been thoroughly studied. Much is known about their loading, biomechanical function, time-related anatomic changes, tropism and asymmetry, modes of articular wear, and soft tissue attachments.1–35
At its most narrow analysis, the facet joint is designed to function as a shear stop—an anterior shear stop (Fig. 54–1). Facet joints prevent anterior shear forces from destroying the disc, which is primarily designed to absorb compressive loads. The disc is so well designed for its purpose that it is essentially impossible to injure the anulus with purely compressive loads.26 Also, the facets control varying amounts of rotation and limit flexion of the lumbar spine,28–3136 but their most important function is to protect the disc from parallel force vectors.
Viewed in this simplified manner, it becomes clearer why degeneration of these lumbar apophyseal joints produces so much spinal pathology. The loss of even 1 mm of cartilage thickness within the facet joint allows a significant increase in anterior-posterior translational motion to occur. This increased motion causes repeated strain of the multifidus muscle and facet capsule ligaments, increases the shear load on the disc and the posterior and anterior longitudinal ligaments, and allows the neural foramen and lateral recess to collapse with flexion and extension of the spine. At the same time, cartilage wear stimulates spur formation in all directions around the facet joint, as cartilage wear typically does in all degenerative synovial joints. These osteophytes encroach further on the neural elements increasing central and lateral stenosis, and the osteophytes may act as a source of pain when they impinge on each other or the pars interarticularis. Degeneration can continue until facet subluxation occurs, producing further stenosis (Fig. 54–2).
Concomitant with the five degenerative changes in the apophyseal joints—loss of articular cartilage, spur formation, loss of control of anterior shear forces, facet subluxation, and increased anterior-posterior translation—the facets also frequently undergo disadvantageous morphologic changes with aging. These changes undermine the ability of the spine to withstand shear forces. In infant spines, the facets are primarily coronally oriented.25 In adult spines, the lumbar facets generally have a small anteromedial coronal component and a large posterior sagittal component (Fig. 54–3). As the spine ages, the facet becomes more and more sagittally aligned with a smaller and smaller coronal component (Fig. 54–4). The coronal part of the facet joint controls shear forces.28 In a presentation at the International Spinal Arthroplasty Society Meeting in Montpelier, France in 2002, DuPont, using finite element analysis, verified how the shear forces across the disc increase as the facets are directed more and more sagittally.
There is at least one additional force working on the lumbar facet joint. Depending on the position of the spine, the facets absorb 0% (in full flexion) to 33% (in full extension) of the compressive (axial) load at a given level.23,37 An incompetent, degenerative facet joint can no longer absorb its share of the compressive loads, which can narrow the neural foramen further in an up-down direction, especially with the spine in extension.
Although excellent radiologic tools have been developed, including computed tomography (CT), CT myelogram, and magnetic resonance imaging (MRI), they may not be used to their fullest advantage. There is no spinal equivalent to the 30-degree weight-bearing anteroposterior view of the knee (Fig. 54–5) to detect cartilage wear or joint laxity. Similar positional views of other joints of the extremities, such as weight-bearing anteroposterior and lateral views of the feet and the 30-degree angle view of the shoulder, have been developed to aid in diagnosing arthritis and instability but not as well with advanced imaging for the spine.
MRI and CT of the spine are extremely informative diagnostic tests, but most are performed in a supine position during which most patients do not have back pain. Some MRI studies allow imaging in the upright position and in flexion and extension, but these tools have low tesla outputs at the present time. It seems sensible to develop a technique that applies some extension to the lumbar spine and perhaps some axial loading as well to assess more accurately the functional capacity of the facet joints. Flexion and extension films are rarely used in the spine, and when they are used, they require patient cooperation. It may be easier to run a second sequence on MRI with a pillow beneath the lumbar spine to obtain a comparison extension view.
A vest with pantaloons has been developed that can individually manipulate vertebral bodies to obtain much more information on a segmental level, but this is under development, and weight-bearing extension views of the lumbar spine probably will not be ready for initial usage before 2012. Other investigators have developed CT scan techniques to show positional lateral recess stenosis (Fig. 54–6).
One radiologic sign seen on plain films that draws attention to facet instability is a break in the normally round shape of the intervertebral foramina. This radiologic sign is analogous to Shenton line, which is a continuation of the arc of the femoral neck with the arc of the superior pubic ramus. When the line is unbroken, the hip is located, and the biomechanics are considered optimal.
In the same vein, the inferior arc of the pedicle should follow the posterior aspect of the superior and inferior vertebral bodies associated in its foraminal opening, along with the inferior articular process and the arc of the superior articular process. If this circle is unbroken (i.e., the line is smooth) (Fig. 54–7), the foraminal space, at least in a static film, is probably adequate from a bony standpoint. If the proximal end of the superior articular facet or the superior spurs of the inferior articular surface intrude into the hemicircle formed by the pedicle and the inferior facet, the broken foraminal line would indicate facet joint degeneration and possibly suggest instability. This line can be assessed on flexion and extension films and is a more subtle sign than measuring the number of millimeters of spondylolisthesis. An additional sign within the foramen to aid in diagnosis could be the intrusion sign of the superior endplate into the intervertebral foramen. Figure 54–7 illustrates both abnormalities, the former at L3-4 and the latter at L4-5.
FIGURE 54–7 A, Lateral x-ray of lumbar spine with degenerative disease. Intrusion by inferior articular spurs into foraminal circle is seen at L3-4 (arrow). In addition, L4-5 exhibits foraminal circle distortion from superior endplate of L5. B, Same distortions are shown diagrammatically. Both are radiologic signs of facet or motion segment degeneration (“incompetence”). The other foramina at L1-2 and L2-3 show no interruption in circle subtended by two pedicles, inferior process, and superior process.
Even with improvements in imaging of the spine, the amount, type, and location of pain derived from the history and the physical examination are the most important part of a patient’s evaluation. Imaging studies are for the most part confirmatory only.
There are three basic types of operative approaches to the degenerative lumbar spine. The surgeon can perform some sort of débridement of the soft tissue and bony elements (including the ligamentum flavum, part or all of the lamina and the spinous process, a portion of the pars interarticularis, and up to 50% of the facet joint) that are involved in the stenotic process. The surgeon can perform a débridement-type procedure for the stenotic pathology combined with a fusion type of procedure, be it anterior, posterior, combined, posterolateral, instrumented, or noninstrumented. Finally, the surgeon can perform one of the newer extension-limiting procedures using a “stiff bumper/rope combination” or an interspinous process block.
Disc replacement is primarily for isolated disc pathology and is not indicated for patients with posterior degenerative disease. That being said, some manufacturers, either inadvertently or purposefully, have designed total discs with a single point of rotation. This type of total disc system forces the anterior column of the spine to control shear forces, which is not the intended function of the disc. It is the purview of the facets to control forces acting parallel to the disc.
All of these operations have their successes and their failures with quite a bit of variability as reported in the literature. Wide decompressive laminectomy has been shown to have 57% to 85% good results at 4 years.38–41 Postoperative problems and complications include segmental instability, recurrent spinal stenosis, continued back pain, infection, neural injury, and dural tears.35,42
Decompression and arthrodesis are the mainstays of treatment for degenerative facets. Numerous fusion techniques and associated implantable instrumentation are available, and the operation has improved to the point where a successful radiographic fusion is the rule. As Vaccaro and Ball42 have written, “Though the majority of studies have shown that the radiographic fusion success rate is improved with the addition of internal fixation, the benefits in the majority of degenerative spinal disorders are unclear in terms of patient function.” Some of the complications from combined decompression and fusion include 3% to 6% infection rate,43 continued pain, juxtasegmental instability or fracture (Fig. 54–8), failed fusion, and failed hardware.