Pelvic instrumentation may be necessary for the treatment of degenerative disorders of the thoracolumbar spine to obtain and maintain coronal or sagittal balance, or both. Previously, sacropelvic fixation was fraught with complications, most commonly related to failure of instrumentation, failure to achieve a solid fusion, or both. Improved techniques and advances in instrumentation design and materials have improved fusion rates and reduced instrumentation failure. This chapter reviews the indications for fusion to the pelvis, describes the relevant anatomy of the pelvis and the biomechanics influencing fixation and fusion, and reviews the specific techniques.
Sacropelvic fixation must provide rigid stabilization, especially in view of the significant pullout forces encountered at the caudal end of these constructs.
The utilization of proper technique and the applied knowledge of the relevant anatomy reduce the rate of complications.
Bilateral iliac screws coupled with bilateral S1 pedicle screws provide excellent distal fixation for lumbosacral fusions with a high fusion rate.
Additional stabilization can be attained through the use of interbody support. Anterior lumbar interbody fusion and transforaminal lumbar interbody fusion techniques have been shown to increase the rate of fusion at the lumbosacral junction and decrease the risk for failure.
SCOPE OF THE PROBLEM
Early experiences with pelvic instrumentation in the treatment of degenerative disorders of the thoracolumbar spine yielded a high incidence of loss of fixation, failure of instrumentation, and pseudarthrosis. These failures have been attributed to biomechanical factors, sacral anatomy, patient-related factors (i.e., osteoporosis), and instrumentation.
The lumbosacral junction experiences a combination of high shear, torsion, and flexion-extension forces because of its position and orientation, whereas the L5-S1 disc space is the most vertically oriented of any disc in the spine. In addition, the L5-S1 disc has been described as the most mobile region of the lumbosacral spine, although this is debatable. The orientation of the articular facets at L5-S1 also provides less resistance to rotation, which results in increased torsional strain. Nonetheless, fixation devices at the lumbosacral junction can be exposed to approximately 100-N forces during forward bending. These cantilever forces are transmitted to the pelvic instrumentation. This is accentuated when long thoracic and lumbar fusions are extended to the pelvis, thus creating a long lever arm, which concentrates force at the end of the construct. These forces must be overcome to obtain a solid arthrodesis.
In 1992, McCord et al. demonstrated a significant biomechanical advantage with projection of fixation anterior to the instantaneous axis of rotation (IAR) of the lumbosacral junction. ( Fig. 28-1 ). The IAR is marked by the intersection of the middle osteoligamentous column and the L5-S1 disc. Devices that pass anterior to the IAR provide a significant biomechanical advantage by resisting the flexion-extension moment at the lumbosacral junction. The farther the implants extend anterior to the IAR, the greater the maximum moment of failure and stiffness of the construct.
In addition to the unique biomechanical characteristics of the lumbosacral junction, the complex anatomy of the sacrum also influences the ability to achieve solid fixation. The sacrum consists primarily of cancellous bone within a thin cortical shell. The density of the trabecular bone in the sacrum is considerably less than in the lumbar region. Further, pedicle screws placed into the large, cancellous pedicles of the sacrum fail to engage the cortical walls. Therefore, instrumentation placed in the sacrum is at an increased risk for pullout.
In addition, general medical conditions that cause osteoporosis may further compound the poor quality of sacral fixation. Specifically, Zindrick et al. note that the degree of osteoporosis plays a major role in screw pullout strength, especially at L5 and S1.
The advent of segmental fixation and the development of a better understanding of lumbosacral anatomy and biomechanics have yielded improved fusion rates and reduced instrumentation failure. Early instrumentation of the sacropelvis was performed using hooks and sublaminar wires. More recent advancements in techniques and instrumentation have allowed for the use of screw fixation throughout the axial spine, including the sacrum and ilium.
Additional stability can be attained through anterior column support with insertion of interbody spacers. Interbody graft placement at L5-S1 increases compression stiffness at the LS junction, helps to restore normal lumbar lordosis and sagittal alignment, and increases torsional stability. Interbody graft placement can be performed via a posterior approach such as a transforaminal lumbar interbody fusion (TLIF) or a posterior lumbar interbody fusion (PLIF), or via an anterior approach, preferably a paramedian retroperitoneal approach. As a general rule, distal anterior column support is recommended at L5-S1, and also L4-L5 whenever possible in long fusions.
INDICATIONS AND CONTRAINDICATIONS
Ending the construct at L5 should be considered whenever possible. The advantages of stopping at L5 include decreased operative time and blood loss, lower rate of medical complications, maintenance of the L5-S1 motion segment, and decreased risk for pseudarthrosis. Ending a construct at L5 may be considered possible when the L5-S1 disc space is normal in a short fusion and no deformity is present at the lumbosacral junction. However, fusions should not be stopped adjacent to a segment with rotatory subluxation, degenerative or isthmic spondylolisthesis, or posterior column deficiency. When a long fusion is stopped at L5, subsequent advanced degeneration of the L5-S1 disc can occur resulting in pain, as well as deterioration in coronal or sagittal alignment, and possible stenosis of the L5 or sacral nerve roots. Furthermore, loss of implant fixation at L5 may occur with resultant loss of segmental alignment and global balance, and possible need for complex revision surgery. The risk for subsequent advanced disc disease is greatest for patients with preoperative positive sagittal imbalance even with complete correction of sagittal imbalance after surgery.
When performing a spinal fusion to the sacrum, S1 pedicle screws directed into the sacral promontory take advantage of the dense bone of the superior sacral end plate. If the bone quality is good and bicortical fixation is achieved, instrumentation to S1 may be sufficient for short constructs. However, extension of the construct to the pelvis should be considered if there is a long fusion to the sacrum (L3 or cephalad), if bone quality and pedicle fixation is compromised, or if there is significant lumbosacral kyphosis, such as with high-grade spondylolisthesis ( Table 28-1 ). In addition, extension to the pelvis should be included in the management of patients with sagittal imbalance and revision procedures. Some authors have expressed concern regarding the effect of sacropelvic fixation on the sacroiliac (SI) joint; to date, intermediate-term follow-up studies have not demonstrated adverse changes within the SI joint.
|Long posterior spinal fusions (L3 or higher)|
|Lumbosacral fractional curve, oblique takeoff of L5|
|Degenerative disc disease at L4-5, L5-S1|
|Spondylolisthesis (degenerative or lytic)|
|Degenerative disc disease below a fusion|
|Coronal or sagittal imbalance, or both|
O’Brien has characterized three distinct anatomic zones of the sacropelvic region. Each anatomic zone has unique biomechanical characteristics that influence fixation options and the ability to attain stable fixation ( Fig. 28-2 ).
Zone I fixation consists of the S1 vertebral body and the cephalad margins of the sacral ala, in addition to the lower lumbar spine (L4, L5). S1 pedicle screws are the mainstay of fixation in this region. The anteromedial trajectory for S1 pedicle screw fixation allows for a larger safe zone and better pullout strength than the anterolateral trajectory because the superior sacral end plate has the highest bone density within the S1 body. Various techniques for the insertion of S1 pedicle screws have been described including unicortical, bicortical, and tricortical techniques.
Bicortical S1 pedicle screws are placed using one of two techniques, parallel to the end plate exiting the anterior sacral cortex or through the superior sacral end plate into the L5-S1 disc space. Smith compared bicortical S1 pedicle screws placed parallel to the sacral end plate with unicortical S1 pedicle screws and found only a 4.8% increase in pullout strength with bicortical screws. He attributed this to the thin cortical bone (approximately 1 mm) of the sacrum. Luk et al. compared the pullout strength of bicortical S1 screws placed parallel to the end plate with S1 screws places through the superior sacral end plate and found that screws placed through the superior sacral end plate were significantly stronger. This technique has been advocated by various authorities. In comparison, tricortical screws are directed into the sacral promontory, and attain fixation in the dorsal cortex, anterior cortex, and superior end plate. Compared with bicortical screws placed parallel with the end plate, tricortical screws yield a 99% increase in maximal insertional torque. ( Fig. 28-3 ). Regardless, sacral screws should all be placed converging toward the midline.
Zone II fixation includes fixation from the sacral ala to the distal sacrum below S1. Fixation to Zone II is generally considered the weakest distal anchor area. Commonly used options for fixation include alar screws and S2 pedicle screws. Because of the risk for injury to the internal iliac vessels, the lumbosacral trunk, SI joints, and L5 nerve roots, several authors have advocated that alar screws should be placed only unicortically. Nonetheless, this technique can improve sacral fixation by 20% or more. Similarly, S2 pedicle screws must be placed unicortically because of the risk for colonic injury. These are not recommended because they provide little advantage. In addition, foraminal hooks, usually in a claw configuration, and sublaminar wires have also been utilized in this zone but provide little benefit. Problems include loss of fixation and instrumentation crowding. An alternative is the S2-alar screw ( Fig. 28-4 ). This screw uses the dorsal cortex of S2 as a start point but is directed cephalad and lateral into the sacral ala. It is generally placed bicortically in the safe zone lateral to the iliac vessels. A typical screw length is approximately 50 to 55 mm. The advantage of this technique is the bilateral coronal divergence in combination with bilateral S1 screw convergence. In the sagittal plane, the S1 screws and S2-alar screws converge, thus providing additional resistance to pullout.