Thoracic and Lumbar Spine Construct Design

Published on 27/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2057 times

Chapter 141 Thoracic and Lumbar Spine Construct Design

Gandhi Varma, Darrel S. Brodke, Edward C. Benzel

Construct design is a process that formulates a specific blueprint for an orderly and thoughtful assembly of implantable spinal instrumentation, designed to correct instability or deformity or both of the spinal column.¹ Construct design requires specific understanding of the deformity or instability and the biomechanical forces acting on the pathologic alignment. An understanding of corrective forces and where they must be applied also is required. A keen knowledge of the anatomy and pathoanatomy is required to preserve neuralgic function and avoid adjacent segment injury. Although skillful assembly of the mechanical construct is a definite prerequisite, ultimate success is determined by the orderly thought process for designing the construct, based on personal experience, experience of others, and laboratory data. Creating a preoperative plan, or blueprint, can focus this design process. Spinal instrumentation surgery must not be assumed to be strictly “mechanical” or “routine”; rather, it requires serious and meticulous planning to ensure success.

Nomenclature

The nomenclature of spinal instrumentation is complex and often confusing at the outset. Factors that contribute to this complexity are the numerous components that constitute assembly, the numerous choices for purchase sites in the spine, and the variations in the mode of assembly of the hardware. There are four major categories: (1) anchors, devices that attach the construct to the bony spine (e.g., pedicle screws, cables, hooks); (2) longitudinal members (e.g., rods or plates); (3) connectors, devices that connect anchors to the longitudinal members or connect two longitudinal members (cross-connectors); and (4) accessories (e.g., washers or spacers). Most spinal instrumentation systems have all four of these components. The skill in designing a construct is reflected in the optimal choice of implants that result in biomechanically stable architecture.

Indications for Spinal Instrumentation

Spinal implants may be considered as internal supports that immobilize the spine until bony fusion occurs. In contrast to external orthoses that serve similar functions, spinal implants provide direct control of spinal segments and have a much broader scope.

The goals of spinal instrumentation are threefold. The first goal is the immediate restoration of stability so that the patient may be prepared for early rehabilitation efforts. Immediate stability often decreases pain and may improve early function. It may also increase the success of bone union or fusion. The second goal of instrumentation is indirect decompression of neural structures, often accomplished by controlled distraction. The instrumentation may also be used to restore or maintain physiologic alignment of the spine. The third goal of spinal instrumentation is the correction of deformity to prevent pain or neurologic compromise and the neutralization of pathologic, deforming forces. Surgeons designing spinal instrumentation constructs should clearly delineate which of the aforementioned goals, or which combination of goals, they are attempting to achieve.

Construct Types (Modes of Force Application)

The six fundamental construct types are simple distraction, three-point bending, tension band fixation, fixed-moment arm cantilever-beam fixation, non–fixed-moment arm cantilever-beam fixation, and applied-moment arm cantilever-beam fixation (Fig. 141-1).

FIGURE 141-1 The application of forces to a construct can be broadly categorized into simple distraction (e.g., Knodt rod; A); tension band fixation (e.g., spinous process wiring; B); three-point bending using the middle point as a fulcrum (e.g., Harrington rods; C); non–fixed-moment arm cantilever beam design, in which axial load cannot be resisted because of nonrigid fixation between the screw and the longitudinal member (e.g., lateral mass plates; D); fixed-moment arm cantilever beam design, in which axial loading forces can be resisted because the screws are rigidly fixed to the longitudinal member (e.g., locking cervical plate systems; E); and applied-moment arm cantilever beam design, in which complex forces can be applied (e.g., universal spinal instrumentation; F).

Development of Construct Blueprint

The preoperative development of a blueprint for implant placement, based on the composite information obtained from clinical assessment and imaging studies, ensures a definitive plan and saves time in the operating room. Some flexibility in this plan may be required after surgical exposure of the bony spine because of unexpected findings. For instance, minor fractures at the implant-anchor site may necessitate deviation from the original plan.

A simple scheme should be used that provides (1) information about the level of the lesion or the level of the unstable segment or segments, (2) the types of implants to be used (anchors, longitudinal members, and cross-connectors), (3) the length of stabilization required on either side of the lesion, and (4) the mode of load bearing by the construct. The scheme guides selection of the appropriate implant components in advance, improves the intraoperative communication between surgeons and assistants, and enhances the chances of success.

Although the concept of construct design encompasses similar principles in all anatomic regions of the spine, designing a thoracolumbar construct poses more challenges than most cervical constructs. Various constructs, using a variety of anchors in different bony landmarks, each used in various mechanical modes (i.e., compression, distraction, neutralization, distraction followed by compression, or distraction and compression at different segmental levels), may be used in a successful strategy. In consideration of these complex decision-making dilemmas, this chapter focuses on thoracic and lumbar fixation design strategies.

Line Drawing of Proposed Construct

A simple dorsoventral or lateral line drawing of the spine provides a framework for the clear definition of the operative plan. Often only a dorsoventral drawing is necessary, although clear consideration of any sagittal plane deformity is vital. The line drawing provides the blueprint for surgery (Fig. 141-2).² This drawing can be easily obtained from a CT scan or from radiographs.

FIGURE 141-2 A blueprint of the proposed surgery is helpful in clarifying the treatment plan for both the surgeon and the operating team.

(From Benzel EC: Construct design. In Benzel EC, editor: Spinal instrumentation, Chicago, 1994, AANS Publications, pp 239–256.)

The convention used in this chapter, with regard to a dorsoventral line drawing, dictates that the left side of the drawing portrays the left side of the patient (i.e., the drawing portrays the patient as viewed from behind). This portrayal is in accordance with the most common surgical approaches for complex instrumentation constructs and decreases the chance for confusion.

Level of Lesion and Level of Fusion

The designation of the level of the lesion or location of instability, the levels to be fused, and the type of fusion should be placed next on the line drawing. The level of instability or lesion is designated by an “X,” and the precise extent of proposed bony fusion is designated by a hatched outline. The number of unstable motion segments should be assessed carefully, as should associated deformity. These factors determine the number of levels to be spanned with the construct. The choice of implants also affects this decision.

Hook constructs, often used in the past throughout the spine and still used at the present time in the thoracic region, should incorporate three spinal levels above and two spinal segments below the limits of the lesion (3A-2B rule). For the past decade or so, hooks have mainly been reserved for situations in which the pedicles are very small or for additional support in osteoporotic patients along with pedicle screws. If the patient has a marked angular kyphotic deformity, and if three-point bending is considered in an attempt to reduce the deformity, inclusion of four or more spinal levels above the lesion is common (4A-2B rule) and may provide a more functional lever arm. Such long constructs are suited mostly for lesions in the middle and upper thoracic regions, although thoracic pedicle screws are widely used even in these regions.

In the lower thoracic spine (T8-10), the thoracolumbar junction (T11-L1), and the lumbar region (L2-5), pedicle screw constructs are often preferred. The size of the pedicles and the increased stiffness these screws provide make them an ideal choice. Short-segment fixation, with the inclusion of only one vertebra immediately above and below the lesion, is appropriate if the anterior, load-bearing column is intact and kyphotic deformity is not present. The bone structure must also be of sufficient strength. With increasingly sophisticated fixation choices available, it must be remembered that a rigid, stable construct is the goal, and biology and biomechanics surrounding the implants must be taken into account.

Both segmental pedicle screw and hook instrumentation techniques are successful in obtaining global balance safely. Balance can be accomplished without neurologic injury, even in adolescent idiopathic scoliosis, where neurologic risk is greatest. Segmental pedicle screw instrumentation offers a significantly better overall major and minor coronal curve correction and maintenance without neurologic problems and slightly improved pulmonary function values for the operative treatment of adolescent idiopathic scoliosis. Pedicle screw constructs may also allow for a slightly shorter fusion length than segmental hook instrumentation.³

Depiction of Implant Components

The type of implant components used in the instrumentation construct should be delineated clearly on the blueprint. The implant component at each implant–bone juncture may be a cable, hook, or screw. The convention used is to designate hooks by a right-angled arrow, with the arrowhead pointing in the direction of the orientation of the hook. The purchase site—and the type of hook—is designated further by “P” (for pedicle), “L” (for laminar or sublaminar), or “T” (for transverse process). Screws are designated by an “X” surrounded by a circle and placed over pedicles. Cable (or wire) is depicted as a loop.

The mode of axial load application (distraction, compression, or neutral) at each implant–bone juncture is depicted as an arrow. The arrow points in the direction of force application for distraction and compression or a horizontal line for neutral. Bending moments are difficult to depict accurately on the line drawing and are described.

The modes of application of each segmental level are depicted with arrows and lines, as previously described. The arrows and lines are drawn lateral to the designations of implant types. If sagittal plane forces are to be applied, they are depicted on the lateral line drawing. Finally, cross-fixator locations can be designated by rectangles with circles.

Mechanical Attributes of Spinal Implants: Construct Type

The mechanism that the construct uses (types of construct) to bear loads is also designated. Six methods of load bearing are associated with the six construct types: distraction, three-point bending, tension band fixation, fixed-moment arm cantilever-beam, non–fixed-moment arm cantilever-beam, and applied-moment arm cantilever-beam. It is difficult to depict this information on the line drawing. This information may be recorded in the space provided at the bottom of the drawing.

Construct Design Strategies

Multiple factors should be taken into account in designing a spinal instrumentation construct.⁴ Consideration should be given specifically to bony integrity, the location of the unstable spinal segment, the implant length with respect to the unstable segment, the need for cross-fixation, the need for dural decompression, the choice of ventral versus dorsal instrumentation, availability of specific instrumentation, metallic composition of instrumentation, and the familiarity of the surgeon with a particular technique. Each factor should be adequately addressed to achieve optimal outcome.

Bony Integrity

Osteoporosis poses a significant problem in spinal instrumentation. Hooks and cables may resist pull-out better than bone screws.^5,⁶ A longer construct is often required in osteoporotic bone. More anchors are needed to distribute the load over more segments, decreasing the load at any single site. Interbody stability, either from fully collapsed discs or interbody implants, may protect the dorsal construct from cantilever pull-out. If pedicle screws are required in a patient with marked osteoporosis, consideration should be given to filling the screw tracks with polymethylmethacrylate (PMMA). Addition of PMMA significantly increases screw pull-out strength and is accomplished by first placing and then removing the pedicle screw. The pedicle must be carefully assessed to ensure no cortical penetration has occurred that would allow the methyl methacrylate to leak. PMMA is placed using a large syringe under minimal pressure, and the pedicle screws are rapidly reinserted. The spinal canal should be inspected to ensure the methyl methacrylate has not inadvertently extruded into the canal with potential compression of neural elements. The instrumentation construct is then completed.

Biomechanical studies have shown that pedicle screw augmentation with PMMA using transpedicular and kyphoplasty techniques increased the pull-out failure load twofold to threefold in osteoporotic vertebrae.⁷ Pedicle screws augmented using the kyphoplasty technique had significantly greater pull-out strength than screws augmented with the transpedicular augmentation technique.⁷ Another option for osteoporotic spine fixation is use of expansive pedicle screws. The expansive pedicle screw can progressively compress bone at the screw-bone interface by the expansion of the ventral two thirds of the screw, which is thought to provide greater screw thread engagement of vertebral bone than a conventional pedicle screw of the same size. Biomechanical tests have shown that expansive pedicle screws can significantly improve the axial pull-out force compared with conventional screws in osteoporotic bones.⁸^–¹⁰

Location of the Unstable Spinal Segment

The transition zones in the spine are key locations. These include the junctions of the occiput and cervical spine, the cervical and thoracic spine, the thoracic and lumbar spine, and the lumbar spine and sacrum. Each transition is associated with a change in anatomy, a change in the orientation of facets, and a change in inherent rigidity. Ideally, the construct should not end at an intermediate junction (e.g., cervicothoracic and thoracolumbar junctions). Ending the construct and the fusion at an intermediate junction may lead to higher loads on the implants, higher failure rates, and a greater likelihood of adjacent segment problems.

The closer the lesion or unstable segment is to the occiput or sacrum, the shorter the applied lever arm is to the terminal end of the construct. More rigid fixation is required at the terminal end of the construct.¹¹ Often, multiple points of fixation are used in the sacrum or occiput. Long constructs that end at the sacrum have a high rate of pseudarthrosis at the terminal end (L5-S1).¹² A very high flexion-extension bending moment exists on the S1 screws, which leads to loosening of the screws. The addition of iliac screws, alar screws, or S2 pedicle screws to the construct significantly decreases the bending moment on the S1 pedicle screws and increases the fusion rate.^12,¹³

Implant Length

Generally, the shortest construct that provides adequate stability is preferred. Use of a short construct preserves as much motion as possible in the normal segments of the spine above and below the construct. Also, long constructs may place undue stress on the inferiormost aspect of the construct and the immediately adjacent motion segment. Although long constructs in the relatively immobile thoracic spine would not jeopardize physiologic motion, they should still be avoided if possible. A relative exception is the presence of gross instability.

A long construct may be required to provide greater purchase of adjacent segments and greater stability. In the treatment of gross spinal instability or long curvatures (scoliosis or kyphosis), longer constructs may apply greater force overall and less force at each individual segment. Adding bone anchors, rod diameters, or both can significantly increase overall stiffness.¹⁴

Specific issues arise regarding the length of instrumentation in circumstances in which it is desired to keep the fusion short. Two options are available: (1) use longer instrumentation—use a short fusion and remove the instrumentation later, or (2) keep the fusion and the construct short.

Instrumentation Fusion Mismatch (“Rod Long, Fuse Short”)

Instrumentation fusion mismatch describes the discrepancy between the number of spinal levels incorporated within an instrumentation construct and the number of spinal levels undergoing bony fusion. This technique was particularly popular in the past, when only less rigid anchors were available. Consequently, construct length was increased to gain stability. Generally, only the spinal segments immediately adjacent to the unstable segment are fused.

The rationale for this approach is that a long instrumentation construct is used to achieve reduction and correction of deformity, while fusing only the minimum number of segments necessary. The hardware is later removed (after 1 year), when the fractures have healed, and the fusion has consolidated, releasing the instrumented, but not fused, segments from immobilization. There is some concern that the facet joints included within the instrumented nonfused segments may undergo degenerative arthropathy.¹⁵ An additional disadvantage of the “rod long, fuse short” construct is the potential for the implants to loosen. Hooks and cables allow more movement at this juncture than screws and may not pose as great a problem. This design has generally fallen out of favor. If this mode of construct is desired, screws should not be used as anchors in the nonfused segments.

Short-Segment Fixation

Generally, hook and cable constructs require long constructs, three above and two below (3A-2B) the lesion. This design provides a longer and more efficient moment arm and a stronger construct.¹⁶ The 3A-2B configuration is a logical compromise between the problems associated with longer constructs and the shorter moment arm achieved with the shorter constructs and has been shown to provide good stability. Pedicle screws may allow for shorter constructs and fewer fusion segments, a single level above and below the lesion (1A-1B). Short-segment fixation provides an increasingly popular alternative, particularly when applied in a compression mode. Reinforcement with fracture-level screw combinations in patients who underwent short-segment fixation can help to provide better kyphosis correction and offers improved biomechanical spinal stability in patients with thoracolumbar burst fracture.^17,¹⁸ Because of the limited fixation, there may be a higher rate of failure¹⁹; this is particularly true if the screw purchase is poor (osteoporosis) or if the loads are too great, secondary to a loss of anterior column stability. Long constructs have been shown to be reliable and effective in treating thoracic injuries, with or without ventral reconstruction. Short-segment pedicle instrumentation constructs have proved to be effective in stabilizing thoracolumbar and lumbar fractures, while limiting the disruption of lower lumbar motion segments. Loss of anterior column integrity leads to fixation failure when short constructs are not supplemented with further fixation or a ventral reconstruction.²⁰ In a very unstable spine, without significant anterior column loss and rotational and shear fractures, pedicle screw fixation with two above and two below should provide adequate stability in most circumstances.

Cross-Fixation

Cross-fixation between longitudinal members generally improves the torsional stability of longer constructs and the lateral bending stiffness (Fig. 141-3).²¹^–²³ It is not as critical to use cross-fixation in short-segment constructs. For a short-segment rod construct with a skipped level (nonsegmental fixation), a cross-fixing device improves torsional stability to that of segmental fixation.²² There is little effect of cross-links on flexion-extension stiffness.²¹ When using long rod constructs with hook anchors, cross-fixation is essential to improve hook stability and torsional and lateral bending stiffness. Interconnecting the hooks on the two longitudinal members with cross-links significantly increases the fixation rigidity and decreases the failure rate. When the distal end is anchored by pedicle screws, the torsional and bending stiffness is much greater and not significantly different with cross-links.²⁴

FIGURE 141-3 Cross-fixation can effectively resist torsional forces on the rods. A and B, Rods without cross-fixation are free to rotate with respect to one another. C, Cross-fixation prevents the rotation.

With long constructs, when cross-fixation is deemed necessary, two cross-fixators are optimal, creating a box construct. Further cross-links do not offer an advantage. Ideally, the links should be placed at the junction of the middle and terminal thirds of the construct.

Axial Load-Bearing Capacity of the Instrumented Spine

A major goal in the reconstruction of an injured spine is to restore its ability to carry axial loads. The construct should be designed so that the axial load is transmitted through the reconstituted bony spine, along the axis of the vertebral body rather than through the metallic implants inserted through the pedicles. The latter design may lead to stress risers within the implant, resulting in its ultimate failure.

Two clinical examples illustrate this point. In a case of grade 1 degenerative spondylolisthesis, there is a long-standing translational deformity (glacial instability) but no problem with axial load bearing. The role of instrumentation is to reduce the deformity, align the spine, and stabilize the motion segment rigidly to facilitate bony fusion. In a patient with a severe burst fracture of the L1 vertebral body with some loss of bone substance, axial load bearing is ineffective. This problem may be corrected by resection of the vertebral body, replacement with a strut graft, and short-segment stabilization with a locking plate or dual-rod device. The goal of spinal reconstruction in the latter case is to restore load-bearing ability and to decompress and stabilize the spine.²⁵ If a short-segment pedicle screw construct is used in such a case without vertebral body reconstruction, there may be significant axial loading of the screws, and they may fail.^19,²⁶

Regardless of the approach used, if a spinal implant is placed in a compression mode, the dural sac should be decompressed first. Potentially, doing otherwise would result in additional compression of neural elements.

Nature of Instability

A basic principle used for the correction of a deformity with instrumentation is to apply forces opposite of the forces that created the deformity. Specific maneuvers for the correction of some specific deformities are described in the following paragraphs.

Kyphotic Deformity

Kyphotic deformities of the spine may be corrected in one of two ways: (1) with a dorsal approach, using a three-point bending maneuver through a long rod construct, or (2) via a ventral approach, with the resection of a vertebral body, interbody grafting, and application of a plate or dual-rod device (Fig. 141-4). If the deficit is in the anterior column, such as with compression or burst fracture, the ventral approach may be more appropriate. If the dorsal tension band is lost, which leads to kyphosis, the dorsal approach may work best. Both approaches may be used if the bone quality is poor or if a fixed deformity requires both approaches.

FIGURE 141-4 The best construct to use depends on various factors, and the choice must be individualized. Even a single fracture may require different constructs, depending on the severity of the injury. Mild burst fractures (A) can be treated with a dorsal construct (B), using a distraction combined with a three-point bending maneuver (C). A more severe burst fracture should be treated with a ventral construct, using a cantilever design.

Translational Deformity

Translational deformities (e.g., fracture-dislocation resulting from trauma) are generally corrected from a dorsal orientation (Fig. 141-5). Distraction and three-point bending with long hook-rod constructs often work well, as do shorter constructs (two above and two below) with pedicle screws. Translational deformity of a glacial type from degenerative spondylolisthesis is generally easily reducible by mere positioning or with a pull-back technique, using pedicle screws, although adding an interbody implant significantly improves the overall stability. Translational deformity resulting from isthmic spondylolisthesis is much more difficult to correct. It requires complete discectomy and intervertebral distraction, followed by pull-back with screws. Great care must be taken with any reduction maneuver, and monitoring of the exiting nerve roots with electromyography is recommended.

FIGURE 141-5 A and B, Radiographs show the pedicle screw construct for fracture of T5 and T6 with two cross-links. This is an example of a 3A-3B type of construct. C, Intraoperative picture shows a long pedicle screw construct reinforced with a sublaminar hook in an osteoporotic spine reconstruction.

Axial Loading Deformity (Burst Fracture)

Saving motion segments by limiting the number of the fusion segments is a fundamental principle of spinal surgery. In milder forms of fracture, distraction and extension via pedicle screws restores the alignment. Some authors have used monosegmental pedicle instrumentation with placement of pedicle screws directly into the fractured vertebral body to treat thoracolumbar fractures. Controversy exists over whether monosegmental pedicle instrumentation is a more suitable method for thoracolumbar burst fractures than short-segment fixation. Data from the literature confirm that both monosegmental transpedicular fixation and short-segment transpedicular fixation are effective and reliable operative techniques for selected thoracolumbar burst fractures. The former technique shortened the operative time and decreased the amount of blood loss significantly.²⁷ In severe burst fractures with considerable bone loss, the preferred treatment is ventral excision of the vertebral body, strut grafting, and plating to reconstruct the anterior column.

Flexion-Distraction (Seat Belt) Deformity

Flexion-distraction injuries, accounting for 5% to 15% of all major vertebral trauma, are considered mechanically unstable because of the disruption and the separation of the dorsal elements, either ligaments or bones. The presence of severe dislocation indicates rupture of the anterior longitudinal ligament. This deformity is best corrected by compression followed by stabilization. This procedure may be performed via a dorsal approach and usually requires only short-segment fixation.

Scoliotic Deformity

Surgical goals when treating scoliosis include achieving a well-balanced spine in all planes, while working to preserve segments and maintain mobility. Multiple types of instrumentation are available (hooks, sublaminar wires, and screws), although pedicle screws have gained in popularity because they offer superior multiplanar correction, the potential for shorter fusions, and a lower rate of implant failure or pull-out.²⁸ Generally, cantilever bending and rod rotation techniques are employed to reduce the curve magnitude, along with distractive force applied on the concave side of the deformity and compression applied on the convex side of the deformity. In addition, interbody fusions may help in correcting the scoliotic deformity. Spinal cord monitoring is essential because distraction may place the cord at risk. Attention should also be paid to the rotational malalignment, which can often be improved with segmental fixation, using screws, hooks, or wires. Constructs with rostral and caudal pedicle screws were statistically more rigid in torsion than constructs with hooks as distal anchors in scoliosis correction.²⁴