Construct Length

Choosing an appropriate length of construct is critical to optimizing patient outcome and minimizing the likelihood of revision surgery. A construct that is too long takes away more normal motion than necessary through excessive instrumented segments and requires a larger operation. This places the patient at increased risk of perioperative complications and likely higher pseudarthrosis rates. On the other hand, an undersized construct can transfer abnormal loads onto adjacent noninstrumented segments (which in the setting of complex spinal pathology are often abnormal), leading to adjacent segment failure.

Although choosing a construct length should always be a patient-dependent decision, there are a few guidelines that can be used. The construct should include the entire Cobb angle of a coronal curve and should not stop at the apex of a kyphosis. In other words, instrumentation of any spinal curve in any plane should include the entire curve. This avoids placing abnormal stress on a segment that is not normally acting in transition, and reduces the risk of adjacent segment failure and iatrogenic deformity. Similarly, great care and thought should be taken when stopping a construct between a mobile and nonmobile segment. Again, the stresses placed on the adjacent segment in this setting can increase the risk of proximal junctional kyphosis.

Including the L5-S1 disc space should always be considered in any long construct ending in the caudal lumbar spine. Indications for fusion to the sacrum include L5-S1 spondylolisthesis, previous L5-S1 dorsal decompression, stenosis (central or foraminal) at L5-S1, oblique takeoff at L5, and severe L5-S1 disc degeneration. When a construct is stopped at L5, there is a significant risk of degeneration of the L5-S1 disc, especially in the adult degenerative/deformity populations. This risk can be as high as 69% in patients receiving thoracolumbar fusions in the 5- to 15-year follow-up and furthermore leads to significant postoperative sagittal imbalance. Therefore, we maintain that although fusion to the sacrum (or pelvis) does decrease a significant amount of motion, this motion loss should be carefully weighed against the risk of subsequent adjacent degeneration at L5-S1 in a long construct ending at L5.

Fusion to the Sacrum and Pelvis

Long construct fusions to the sacrum can be problematic from a fixation standpoint. The sacrum is composed primarily of cancellous bone, which results in decreased pullout strength compared to other pedicle screws. Many surgeons, therefore, have come to place so-called tricortical S1 pedicle screws that capture both anterior and posterior cortical bone for increased screw strength. Biomechanical studies have shown that the highest bone mineral density, and therefore the greatest insertional torque, was in the anterior sacral promontory. When placed in this fashion, there is a nearly 99% increase in insertional torque.

Because of the pseudarthrosis risk at L5-S1, many surgeons advocate the use of interbody support at this level. L5-S1 interbody fusion increases biomechanical stability, restores junctional lordosis, improves L5-S1 fusion rates, increases disc and foraminal height, and decreases foraminal stenosis. Angulation of the L5-S1 disc space results in a long moment arm, at the end of which is the C7 vertebral body. Restoration of junctional lordosis at the L5-S1 disc space via placement of an interbody graft is therefore a powerful technique to correct sagittal imbalance.

The decision to use anterior lumbar interbody fusion (ALIF) versus transforaminal lumbar interbody fusion (TLIF) should be made on a case-by-case basis. Although the TLIF can be performed from the usual dorsal approach to the spine, ALIF requires an anterior approach that is typically performed by an access surgeon and can be fraught with approach-related morbidity. This can result in longer operative times or the need for a staged procedure. The extensive anterior exposure of the disc afforded by ALIF, however, allows for a much larger graft with a larger fusion footprint and better restoration of lordosis. TLIF affords a shorter operative time and has been shown to lend greater correction of anterior-posterior scoliotic curves in a study of matched cohorts (likely because of the extensive dorsal soft tissue exposure and complete facetectomy needed for the TLIF approach).

In addition to placement of an L5-S1 interbody graft and use of tricortical S1 pedicle screws, iliac fixation has been advocated as another method of allowing solid L5-S1 arthrodesis. Indications for iliac screw placement include constructs greater than three levels ending at the sacrum, revision surgery for L5-S1 pseudarthrosis, high-grade spondylolisthesis, and pathology that does not allow for adequate sacral fixation. Tsuchiya and colleagues evaluated 67 patients with both high-grade spondylolisthesis and long constructs to the sacrum with iliac fixation and found an overall fusion rate of 95.1% compared to rates of 22% to 89% reported in other studies without such augmentation.

Osteotomies and Sagittal Deformity Correction

Sagittal balance has been shown to be the single most important factor affecting outcomes in adult deformity surgery. One measures sagittal balance by drawing a plumb line from the C7 vertebral body and measuring the horizontal linear distance from this line to the posterior sacral promontory. A sagittal balance greater than 5 cm correlates with poor long-term outcomes, and therefore the major goal of a deformity correcting operation is to reduce any positive sagittal balance toward within the normal range.

In preoperative planning for deformity correction, measurements of kyphosis, lordosis, and sacropelvic parameters are performed using degrees, whereas global sagittal balance is measured in centimeters. The amount of correction needed, therefore, must be converted from the linear amount of sagittal plane imbalance to degrees of angular correction needed. This is most reliably calculated using a trigonometric method described by Ondra and associates. First, the C7 plumb line is drawn, and then a line is drawn from the posterior sacral promontory to the center of the C7 vertebral body. The angle between these two lines represents the overall degree of sagittal plane correction needed to be in proper sagittal balance ( Fig. 86-1 ).

A preoperative planning method for converting a required linear amount of sagittal plane imbalance into degrees of angular correction needed. A, A vertical plumb line (PL) is drawn from the C7 vertebral body and a posterior sacral perpendicular line (PSPL) is drawn vertically from the posterior-superior sacrum. The correction distance (CD) is then drawn between these two lines. An oblique line is subsequently carried from C7 to the PSPL where it intersects with the pedicle to be subtracted. Angle alpha is then calculated using the simple trigonometric equation shown. This is the desired angle of correction for the patient’s spinal imbalance. B, For actual osteotomy planning, line L is first drawn from the PSPL to the anterior vertebral body at the level of the inferior pedicle wall. This will be the base of the osteotomy. Using angle alpha, the cephalad border of the osteotomy is extrapolated and the starting height of the osteotomy along the lamina ( D ) can be determined. The osteotomy height along the posterior vertebral body wall ( E ) is similarly determined.

Osteotomies are cuts made through the spine that allow its mobilization and realignment to achieve deformity correction—typically kyphosis correction and lordosis enhancement. Several spinal osteotomy techniques have been used and commonly discussed in the scientific literature, and a comprehensive anatomic classification system for spinal osteotomies has been developed ( Fig. 86-2 ). This system divides spinal osteotomies into six grades. Grades 1 and 2 include partial and complete facet joint resections, respectively (Smith-Petersen, Ponte, Chevron, or posterior column osteotomies). These osteotomies are performed by removing varying degrees of bone, and ligament from the posterior column at one or multiple levels while leaving the anterior and middle columns intact. With dorsal compression maneuvers, the anterior column can open through the disc space and provide 5 to 10 degrees of lordosis per level. Important caveats for performing grade 1 or 2 osteotomies include ensuring complete removal of any loose superior facet fragments so that compression does not entrap the exiting nerve root and the fact that correction via posterior column osteotomies requires a mobile disc.

Anatomic classification of spinal osteotomies.

Grade 3 osteotomies involve complete resection of the posterior elements and pedicles as well as wedge resection of the vertebral body (i.e., standard pedicle subtraction osteotomies, or PSOs), whereas grade 4 osteotomies also include complete resection of the proximal disc (i.e., “extended” PSO). The wedged resection through the vertebral body in both of these osteotomy types causes a hinge effect with the fulcrum at the anterior aspect of the vertebral body. Dorsal compression thus results in bone-on-bone closure that can provide 25 to 35 degrees of lordosis. Grade 4 extended PSOs can provide even more correction.

Grades 5 and 6 osteotomies include single and multilevel vertebral column resections (VCRs), respectively. The VCR is the most aggressive osteotomy, with complete removal of a vertebral level and its adjacent discs (as well as the associated rib heads when performed in the thoracic spine). A grade 6 osteotomy extends beyond the scope of a single-level VCR and can incorporate a partial or complete resection of additional vertebral bodies. Closure of this type of osteotomy is typically not bone on bone, relying on an anterior pivot such as a cage or bone graft. This technique can achieve 50 to 60 degrees of correction and is ideal for angular deformities in the thoracic spine.

Use of a temporary rod during the performance of any three-column osteotomy is mandatory to prevent intraoperative spinal column translation and potential neurologic deficit. After completion of the osteotomy, compression is required to close the resultant defect and gain the desired correction. This is often performed by compressing against the adjacent pedicle screws, which can place extra force against these screws and result in pedicle fracture with screw failure. This can be prevented by using an alternative method in which two temporary rods are placed on each side of the construct with their free ends meeting at the osteotomy site. Domino connectors are then placed across the rods and tightened on both sides of the construct, and the rod is then secured to the pedicle screw heads. The screw caps on either the cephalad or caudal rod are then loosened and compression subsequently occurs across the connector as opposed to the pedicle screw heads. It is important that bilateral compression of the same rod (either cephalad or caudal) occurs simultaneously so that symmetrical correction can occur. Others have described a “claw construct” technique using laminar hooks rostral and caudal to the osteotomy site with closure along a temporary rod that spans the hooks instead of the pedicle screws. This takes all the force off the screws entirely until the permanent rod is placed.

Closure of any of the three-column osteotomies (grades 3 to 6) should be performed with great care and vigilance. Although rare (2.6%), intraoperative spinal cord deficit has been reported as the most common intraoperative complication and motor deficit has been described as the second most common postoperative complication (12.1%) in a large multicenter series of patients undergoing three-column osteotomy for deformity correction. Prior to closure, attention must be paid to the extent of dorsal decompression to prevent compression of the spinal cord from dural and ligamentous buckling. Similarly, the exiting nerve roots above and below the resected pedicle in a PSO should also be closely inspected prior to and after closure.