21 Sagittal Balance Treatment Strategy in a Posterior Approach

10.1055/b-0039-171417

21 Sagittal Balance Treatment Strategy in a Posterior Approach

Fethi Laouissat and Pierre Roussouly

Abstract

The purpose of this chapter is to show how to manage a spinal surgery for degenerative or deformity issues by a posterior approach to restore an adequate sagittal alignment and balance of the spine.

21.1 Introduction and Controversies

Sagittal balance has become an increasingly important matter for spine surgeons when dealing with spinal degenerative or deformity conditions. All spine surgeons agree on the fact that painstaking preoperative planning must be established before entering the operating theater because “failing to plan is planning to fail.”

However, despite the fact that tactical and technical basic skills are well known throughout the spine surgeon community, strategy and philosophy of sagittal spinal alignment are still controversial. Questions surrounding some crucial points regarding the relationship between spine curvature and pelvis are still discussed.

It is well recognized that pelvic incidence (PI) is a constant parameter that determines the shape of the pelvis. Pelvic rotation around the hip axis permits adaptation of the sacral plateau by retroversion or anteversion. This adaptation is framed by a geometrical relation: PI = SS + PT (SS, sacral slope; PT, pelvic tilt) ¹ (Fig. 21‑1).

**Fig. 21.1** Sacropelvic parameters: pelvic incidence (PI), pelvic tilt (PT), and sacral slope (SS).

To guide surgical planning, a widespread parameter relating PI and lumbar lordosis (LL) (measured between L1 and S1) has been proposed to enable a patient-specific approach to treatment goals by quantifying the mismatch between pelvic morphology and lumbar curve. ² The threshold identified as a spinopelvic sagittal alignment goal was PI-LL <10°. However, the PI-LL mismatch appears to be a “one size fits all” theory, as well as the Schwab-SRS classification cut-offs. ³

On the other hand, Roussouly et al ⁴ investigated sagittal spinal shapes in asymptomatic populations and defined four sagittal types of lumbar curvatures. The Roussouly classification was refined by Laouissat et al ⁵ by integrating a fifth type of sagittal shape, which held the anteverted pelvis concept. The difference between the two concepts is fundamental. The “PI-LL mismatch” approach appears to be mathematical; it gives only an “ideal” global angle of lordosis without an instruction about curvature repartition. The L1-S1 definition of LL is another paradigm that treats all kinds of lordotic shapes in the same manner: “one size fits all.”

Introduction of lordosis shape variation according to PI by Roussouly et al ⁴ emphasizes not only angle relationships but also lordosis curve organization and distribution of LL and thoracic kyphosis. Based on this concept, sagittal balance treatment tends to restore better balance parameters such as PT and C7PL and, therefore, aims to restore the shape most approaching normal shape according to PI. In pathology, the fundamental reference is the pelvic shape regarding the PI value (high- or low-grade PI), which is, as a constant shape parameter, the only signature of the previous spinal shape (before degenerative changes).

We may roughly divide pelvic shapes into two classes: low PI (<50°) and high PI (>50°). For low PI, it is necessary to restore type 1 or 2 shapes, and for high PI, type 3 or 4 shapes (Fig. 21‑2).

**Fig. 21.2** Subdivision of the sagittal spinal curvatures according to the Roussouly classification. SS, Sacral slope.

This said, the authors of this chapter aim to give to the reader some technical notes and points of view based on their experience in a comprehensive manner to help spine surgeons set up their preop planning.

21.2 Reduction Technique by Posterior Approach

21.2.1 Evaluation of the Length of Fusion

Short Fusion (One or Two Levels)

This option is efficient mainly when unbalance may be linked to painful compensation as a result of a local spinal disease (instability, stenosis). However, for a fixed or structural deformity, short fusion is insufficient for balance restoration.

Long Fusion (More than Three Levels)

Proximal instrumented level: this is probably the most controversial issue and the most difficult decision. There are three major levels: L2, T10, and T2-T4. The most problematic is the longer fusion to the upper thoracic spine. It is generally imposed by a severe thoracic kyphosis that needs to be controlled by the extended fusion.

Distal instrumented level “fuse or not to the sacrum?”: There is a lot of controversy surrounding fusion to L5 or to the sacrum/pelvis. Fusion to the sacrum is technically demanding. It needs a multifocal fixation for long constructs with several screws in various anatomic positions: sacral plate, sacral alar, iliac, sacroiliac. ⁶ ^, ⁷ ^, ⁸ ^, ⁹ ^, ¹⁰ ^, ¹¹ These extensions of instrumentation are at a high risk of pseudarthrosis. ¹² To enhance fusion, circumferential or “360°” fusions and instrumentations have proved to yield good to excellent results both clinically and biomechanically. ¹³ ^, ¹⁴ An associated L5-S1 cage with anterior fusion is needed if disk height exceeds 6 mm. Fusion to L5 is technically easier but less efficient on balance restoration. In the case of suboptimal alignment of the fused spine above L5 as a distal instrumented vertebra, compensation may occur on the free L5-S1 level, mainly with painful local extension. The initial compensation mechanism starts at the L5-S1 level by bending backward the fused spine to avoid forward C7 tilt. Therefore, local stress forces are focused on L5-S1 facets and posterior elements. On long-term follow-up, diskal degenerative evolution may induce a loss of disk height and an anterior imbalance by L5-S1 flexion (kyphosis) or L5-S1 degenerative spondylolisthesis. Revision with disk height restitution is difficult and, generally, a pedicle subtraction osteotomy (PSO) must be used.

21.2.2 Evaluation of Spine Flexibility

Release strategy depends on the flexibility status of the deformed spine. Preoperative radiological means allow this evaluation. Stress films plus computed tomography (CT) scans are generally sufficient. A CT scan combines a stress film effect, because of the supine position, and intervertebral fusion evaluation. Different stages of fusion may be as follows:

Normal intervertebral status: the level is free of degenerative effect.
Degenerative rigidity: but neither the disk nor the facets are fused and a deep posterior release (total articular process resection plus interlaminar release) is sufficient for mobilization.
Total posterior fusion but disk space is free: indication for Smith-Petersen osteotomy (SPO).
Total anterior fusion by disk ossification associated to posterior fusion: indication for PSO. These techniques are well described in Chapter 23. We want to insist on asymmetrical osteotomies, which are very useful in rigid kyphoscoliosis cases, allowing the correction of both deformities.

Remark

The level of PSO is of great importance for the reduction strategy. Some authors ¹⁵ emphasized the L3 level because of better accessibility and reduction of neurological troubles. But recent studies have demonstrated the relation between height of the apex of lordosis and proximal junctional kyphosis (PJK) occurrence. ¹⁶ We have seen previously that it is necessary to restitute the inferior arch of lordosis and to concentrate the correction of lordosis on L4 or L5. If an L4 PSO procedure is already well known, L5 PSO ¹⁷ is less common and has a bad reputation mainly because of the higher risk of neurological complications. To avoid these, we propose a modified L5 PSO, which performs only a proximal hemipediculectomy that preserves the foramens of L5-S1 and avoids an L5 root trapping when closing the osteotomy. The best indication for such an L5 PSO is a fixed sagittal imbalance in a very high PI that imposes a restoration of a very high SS with a very curved lordosis.

21.2.3 Screws Placement and Mechanics

Following principles of modern instrumentation, pedicle screws are widely used anchors because of their strong fixation in the bone. They are connected to the rods either directly with tulips or by means of a connector. There is an advantage to superpose the screw direction with the vertebra anatomy. Positioning the screw with regard to the rod contour, the corresponding vertebra, follows the same orientation. On the contrary, if two screws inserted in both pedicles of one vertebra are in divergent directions, other than the vertebral plane and axis, it will be impossible to correlate the screw direction with the vertebra orientation. We recommend a perfect position of each screw in each pedicle in symmetry regarding the vertebral sagittal plane and parallel to the upper plateau (Fig. 21‑3).