11.1.1 Ligamentous Disruption

A significant portion of the contribution to ligamentous stability by ventral ligamentous structures is via the anterior and posterior longitudinal ligaments and the annulus fibrosus. Disruption of the anterior or posterior longitudinal ligament or the annulus fibrosus, either by the offending pathologic process or by the surgical approach, can substantially reduce the intrinsic stability of the spine.

Magnetic resonance (MR) imaging techniques have provided a diagnostic tool for assessment of the integrity of ligamentous structures (see Chapter 3). ¹ This assessment, however, is static; it informs the clinician only of the extent of the anatomical continuity of the ligament and the presence of acute soft tissue injury, revealing nothing about the ligament’s strength. Dynamic radiographs (flexion and extension views of the spine) can demonstrate a lack of integrity if excessive movement occurs. However, if subluxation or excessive movement does not occur during dynamic radiographic studies, the presence of spinal stability is not established. Spinal guarding and splinting, or inadequate imaging techniques or suboptimal patient cooperation, can lead to erroneous interpretations in this regard (see Chapter 3). These factors notwithstanding, the ligamentous contribution to stability can usually be reasonably assessed preoperatively.

The extent of the disruption of ventral ligamentous structures by an operative exposure is difficult to assess. Several facts about the anatomy and strength characteristics of the anterior and posterior longitudinal ligaments should suffice for most clinical decision-making scenarios, particularly when combined with the information gained from intraoperative observations.

The anterior longitudinal ligament is a strong ligament. It is also relatively wide (see Chapter 1). If it is not disrupted before surgery, the surgical exposure (even a wide ventral exposure) does not usually disrupt the entire ligament. Therefore, in most cases, the contribution of the anterior longitudinal ligament to postoperative spinal stability is significant. Thus, the tension-band nature of the anterior longitudinal ligament in extension is partly preserved, which limits extension. Therefore, it is a limiting factor in ligamentotaxis (see Chapter 8).

The posterior longitudinal ligament, on the other hand, is weaker than the anterior longitudinal ligament in all regions of the spine. Furthermore, it is waisted (narrower) in the mid-vertebral-body region at each segmental level. The posterior longitudinal ligament in the mid-vertebral-body region is narrower at each spinal level than the dural sac. Therefore, at any level of the spine, a vertebrectomy that adequately decomposes the dural sac is almost certain to disrupt the posterior longitudinal ligament totally at any level of the spine. Thus, the tension-band nature of the posterior longitudinal ligament is disrupted, and its contribution to the limitation of flexion (and distraction) is impaired. This also limits the efficacy of the ligamentotaxis and of ligamentotaxis procedures.

A surgeon may acquire a “feeling” for the extent of ligamentous stability at the time of surgery, following dural sac decompression (vertebrectomy). The application of traction, spinal distraction with instruments such as vertebral body spreaders, or other intraoperative spinal manipulations can provide the surgeon with vital information regarding spinal laxity. This may help to determine whether a spinal implant is necessary as an adjunct to interbody fusion. For example, excessive laxity, as determined by intraoperative distraction maneuvers, may suggest that an interbody bone graft alone will not suffice.

For an interbody strut graft to be immediately effective as a stabilization device, it must be securely positioned in the mortises of the vertebral bodies (i.e., the vertebral bodies above and below the strut). This allows a semirigid fixation of the vertebral bodies abutting the strut (Fig. 11.1a). If ligamentous integrity is not adequate, as demonstrated by excessive laxity during intraoperative stress maneuvers, the strut graft will not be securely affixed in the mortises of the vertebral bodies above and below the strut (Fig. 11.1b). The resistance to distraction provided by intact ligaments allows the vertebral bodies to “clamp down” on the strut graft. This “clamping down” effect is an integral part of most interbody fusion techniques. Spinal distraction, followed by the placement of a well-fashioned strut graft into well-fashioned mortises and then by relaxation of the distraction, allows the “clamping down” properties of the ligaments to become manifest and leads to a strong construct (Fig. 11.1c–e). Thus, a spinal implant, prolonged bed rest, or a bracing adjunct to the decompression–fusion procedure is usually necessary when this ligamentous resistance to distraction is lost. Many spinal implants placed in a distraction mode, including Harrington distraction rods and interbody strut grafts, rely on intrinsic spinal resistance to distraction to obtain optimal security of fixation.

Disc interspace disruption is a cause of spinal instability, although rare. ² This effect is cumulative. ^{3 ,​ 4} It can be readily assessed by MR imaging (Fig. 11.2). However, as an isolated entity, it does not substantially affect the decision-making process, except by necessitating a period of external spinal bracing. ¹ The contribution of the annulus fibrosus to spinal stability, although significant, parallels that of the immediately adjacent anterior and posterior longitudinal ligaments. Its contribution cannot be separated from that of these ligaments. Therefore, no separate biomechanical consideration is warranted. It is worthy of emphasis, however, that the annulus fibrosus–anterior longitudinal ligament–posterior longitudinal ligament complex provides substantial stability to the spine.

Chen et al provided excellent insight into the ligamentous contribution to cervical spine stability, particularly as it relates to structures affected by anterior cervical decompression operations. ⁵ They demonstrated the substantial contribution of intervertebral disc, unilateral uncovertebral joint, bilateral uncovertebral joint, and posterior longitudinal ligament dysfunction to spinal instability. They concluded that anterior cervical decompression significantly decreases stability. All of the aforementioned structures contribute substantially to such stability. Flexion and extension were, of note, substantially affected by disruptions of the aforementioned structures. ⁵

11.1.2 Bony Disruption

Like instability from the loss of ligamentous integrity, diminished integrity of the vertebral body—whether caused by the spinal pathologic process ⁶ or by surgical bone removal—reduces spinal stability. MR imaging is useful in determining the bony contribution to stability. Plain radiography and computed tomography (CT) are better in this regard ² ; however, the use of sagittal CT reconstructions or sagittal MR images to depict the sagittal plane anatomy cannot be overvalued.

The extent of ventral spinal decompression obviously affects spine stability. A spine that has undergone a complete vertebrectomy obviously is less intrinsically stable than one that has undergone an incomplete vertebral body resection. This is true for both ventral and lateral approaches to the vertebrectomy. Rarely, however, is the entire vertebral body resected. The fraction of the vertebral body, as well as the anatomical position (in the anteroposterior plane) of the portion of the vertebral body resected, significantly affects spinal stability. For example, a standard cervical corpectomy resects the vertebral body incompletely over the entire rostral–caudal dimension of the vertebral body (Fig. 11.3a). Similarly, ventrolateral (Fig. 11.3b) and lateral extracavitary (Fig. 11.3c) decompressions incompletely resect the vertebral body over the entire rostral–caudal dimension of the vertebral body (see Chapter 10). The fraction of bone remaining in the ventral portion (vs the dorsal portion) of the vertebral body partly determines the extent of ventral spinal stability.

The location of the segment resected also affects the extent of iatrogenic spinal destabilization. To illustrate this point, consider the vertebral body to be a cube composed of 27 smaller cubes of equal size (Fig. 11.4). Also assume that posterior column stability is present. Surgical removal of the middle third (i.e., the middle layer of nine cubes) of the vertebral body, as viewed in the sagittal plane, grossly destabilizes the spine (Fig. 11.5a), whereas surgical removal of the middle third, as viewed in the coronal–sagittal plane, does not (Fig. 11.5b). In the former case, the anterior and middle columns of Denis ⁷ are disrupted in the entire cross section of the vertebral body, resulting in loss of stability. In the latter case, only one-third of the integrity of the anterior and middle columns of Denis has been disrupted.

Partial vertebrectomies, as viewed in the sagittal plane, also vary in their destabilizing effect by virtue of the portion of the vertebral body removed. For example, removal of the ventral section (the ventral nine cubes) will most likely have a significant effect on stability, whereas removal of both the middle and the dorsal sections of cubes may not result in a significantly unstable situation if the following components remain intact: (1) the ventral section of cubes, (2) the anterior longitudinal ligament, (3) dorsal column ligamentous integrity, and (4) dorsal column bony integrity (Fig. 11.6). Minimizing the extent of vertebral body resection minimizes iatrogenic destabilization by the surgical procedure. In the case of true ventral surgical approaches, a narrow trough of vertebral body resection results in less vertebral body resection and a lesser width of anterior longitudinal ligament disruption. On the other hand, a narrow vertebral body resection often results in inadequate spinal canal exposure and dural sac decompression (Fig. 11.7). In a similar vein, a natural tendency is for surgeons to decompress the spinal canal more than adequately on the side opposite where they are standing, and to decompress the dural sac inadequately on the same side where they are standing (Fig. 11.8). An “Erlenmeyer flask–like” decompression therefore warrants consideration. This type of decompression compensates for several of the problems outlined here. It involves a narrow decompression ventrally and a wider decompression dorsally (Fig. 11.9a); hence, it allows a wide decompression of the dural sac and neuroforamina. This is accomplished by the surgeon’s compensation for the known natural tendency to inadequately decompress the dural sac on the near side of the patient by decompressing the dural sac from both sides of the table. This provides a good view of each side of the exposed spinal canal (wide decompression) while allowing minimal ventral vertebral body resection to suffice (minimizing iatrogenic destabilization of the spine; Fig. 11.9b). The minimization of ventral vertebral body resection also provides greater lateral support for the strut graft (see Chapter 12 and Fig. 11.9b).

Lateral approaches to ventral dural sac decompression (e.g., via lateral extracavitary decompression of the spine) may also unnecessarily destabilize the spine if excessive vertebral body resection is accomplished. As mentioned above, if the ventral aspect of the vertebral body is surgically undisturbed and the dorsal elements have not been violated, substantial stability may be present. Therefore, the minimization of bone removal should aid in the acquisition of postoperative stability. Preservation of the integrity of the ventral and lateral aspects of the vertebral body is particularly important.

Depicting the vertebral body by dividing it into thirds in each plane (for a total of 27 cubic segments) is also useful for conceptualizing the destabilizing nature of a surgical procedure via the lateral extracavitary approach (Fig. 11.10a). Dural sac decompression should involve the most dorsal plane, only on the side of the exposure (Fig. 11.10b). The middle and ventral planes may be considered for the bone graft. If ventral iatrogenic destabilization is to be minimized, the ventral plane (the ventral nine cubes) should not be surgically disrupted. Therefore, in this hypothetical case, the ventral plane should be left intact and the middle plane used as a site for interbody fusion placement (Fig. 11.10c). This makes additional sense if the surgeon also considers that the middle plane is most likely in line with the instantaneous axis of rotation (IAR) and, therefore, is an optimal position for axial load bearing by the surgically placed strut graft (see Chapter 2).

The effect of the uncovertebral joint on stability is significant. This is particularly true for extension, lateral bending, and torsion. ⁸