Vertebral Augmentation: The Comprehensive Guide to Vertebroplasty, Kyphoplasty, and Implant Augmentation

31 Biomechanical Changes after Vertebral Compression Fractures

Olivier Clerk-Lamalice

Summary

Spine biomechanics are important to know as the function of the spine plays a direct role in understanding how to treat various types of vertebral compression fractures (VCFs). The degree of stability or instability, the function of the spinal unit, the sagittal balance, and the degree of kyphosis; are all important factors to optimize to ensure adequate stability and function after a vertebral compression fracture. The degree and anatomy of the fracture compression are important elements that can have an effect on the risk of future vertebral fractures. Additionally, the spine can be divided into functional spine units composed of two adjacent vertebrae, the intervertebral disk, the facet joints, and the intervening ligaments. Most of the axial load of the spine is transmitted through the vertebral bodies. The load through the lumbar spine varies dramatically will differences in posture and weight bearing with the greatest amount of spinal local being present when the patient is in the flexed standing position. This concept explains why some people are at risk of vertebral fracture when performing certain common activities of daily life. As the intervertebral disk degenerates, it places more load on the adjacent vertebral bodies and shifts the overall load more posteriorly. As the load shifts posteriorly there is less stress on the vertebral bodies resulting in less bone density and less bone strength. The combination of disk degeneration, less vertebral body strength, and a flexion neutral posture can prominently predispose the patient to a VCF.

Keywords: vertebral compression fractures, biomechanical, sagittal balance, functional spinal unit, compression force, MagerI, fracture mechanical instability

31.1 Introduction

Spine biomechanics has gained significant interest in the past few years. Key concepts such as the functional spinal unit (FSU), mechanical stability, sagittal balance, and correction of the kyphotic angle have been increasingly studied in patients with VCFs. These concepts are of paramount importance to determine the amount and location of polymethyl methacrylate (PMMA) cement or other fill material to be injected within a fractured vertebral body and to understand the importance of correcting of the end-plate deformity/kyphotic angle. By optimizing these parameters, mechanical stability and pain control can be maximized in addition to lessening the risk of adjacent-level fractures. Thus, this chapter contains some of the core knowledge of this book justifying the “How” and the “Why” for every single vertebral augmentation procedure.

An inappropriate application of these principles has previously contributed to studies that failed to demonstrate the superiority of vertebral augmentation versus a sham arm ¹^–³ and caused significant collateral damage in our field by decreasing therapy access while increasing the morbidity and mortality of VCF patients.⁴ On the other hand, when biomechanical factors are carefully considered, optimal results can be obtained, resulting in significant and long-term improvement of the patient’s pain, function, and quality of life.⁵ Spine biomechanics also helps understand why vertebral augmentation decreases the risks of recurrent/adjacent-level fractures and helps determine which patient may benefit from prophylactic vertebral augmentation. We will review here the key biomechanical concepts that spine interventionists should carefully consider prior to vertebral augmentation.

31.2 Current Information based on Recent Literature and State-of-the-Art Practice

31.2.1 Physiologic Loads on the Vertebral Body

When placed under dynamic loading, the spine allows multidirectional movements that can be analyzed with a six-degree-of-freedom biomechanical model. To analyze the spine and its biomechanical characteristics, this structure is often subdivided in smaller physiological motion units called FSUs. The FSU is composed of two adjacent vertebrae, the intervertebral disk and the ligaments providing support and stability. According to a purist’s definition, the FSU does not include muscles or other connecting tissues. Each vertebra of the FSU can move in space according to orthogonal X, Y, Z axes in addition to rotations along each axis. Because of the musculoligamentous apparatus and the orientation of the facet joints, physical limitations are exerted on FSUs. Furthermore, these limitations fluctuate depending on the region of the spine studied (i.e., cervical, thoracic, or lumbar). At the thoracolumbar level (where most of the VCFs happen), the forces applied on the FSU can be simplified in four main categories: compression forces (axial forces), shear forces, bending moment, and axial torque (▶Fig. 31.1).

In the thoracolumbar spine, the majority of the axial load is borne by the vertebral bodies. Each vertebral body is composed of two structural constituents that sustain that load: the cancellous core and the cortical shell. First, the cancellous core is a dense network of inner trabeculae oriented vertically and horizontally. The vertical trabeculae support the vertebral body thin cortex (measuring approximately 0.4 mm) and resist to the axial loading by transmitting forces from the upper to the lower end plate. On the other hand, the horizontal trabeculae reinforce and provide support to the vertical trabeculae by preventing sideway displacement under compressive loading with and without shear forces. Although the thickness of the cortical bone is not substantial (approximately 0.4 mm), the bone mass attributable to the cortical shell is surprisingly large when compared to the rest of the vertebral body, estimated to be up to ~40% of the total bone mass and providing approximately 45% of the structural support.⁶ Also, the cortical shell allows less plastic deformity than the cancellous core before fracture (~2 vs. ~10%). Thus, it is unlikely to have a fracture of the core of the vertebral body without a fracture of the shell.

Fig. 31.1 Simplified forces applied on the spine: axial torque, bending moment, shear force, and compression force.

Fig. 31.2 Functional spinal unit (FSU) load-displacement curve. The relation between the load applied on the FSU and the displacement is not linear. The range of motion (ROM) of the spinal joints includes an initial neutral zone (NZ) with relatively large displacements at low load. Because of the tension provided by the ligaments and capsule, the elastic zone (EZ) has more resistance to movement leading to increased load per unit of displacement.

The physiologic range of motion (ROM) of an FSU consists of a neutral zone (NZ) and an elastic zone (EZ) (▶Fig. 31.2). These characteristics (NZ and EZ) are intrinsic to any two consecutive vertebra⁷ and have very different biomechanical behaviors. In the NZ, the spine undergoes relatively large motions with minimal force loading due to the laxity of ligaments and capsule. In the EZ, there is resistance to movement leading to increase load per unit of displacement⁸ until failure. Several studies have demonstrated that the NZ increases with injury, muscle weakness, and degeneration. This increase in the NZ tends to lead to FSU instability. The NZ may be brought within a physiological limit with osteophyte formation, surgical fixation/fusion, and muscle strengthening.⁸

When a person stands upright, the mass of the trunk, head, and arms presses vertically on the lower lumbar spine with a force of approximately 55% of the body weight, corresponding to 385 N for a 70 kg man.⁹ The measurements of the various components of spinal loads are complex and vary depending on the region of the spine evaluated, the posture (flexion vs. extension), and the angulation of the motion segment. Sato et al¹⁰ measured in vivo spinal loads at the L4–L5 level with intradiskal pressure measurement. Those pressures were of 91 kPa in the prone position, 151 kPa in the lateral position, 539 kPa in the upright standing position, and 623 kPa in the upright sitting position. The maximum intradiskal pressure in the flexed standing position was more than twice the upright standing position with a value of 1,324 kPa. The spinal load calculated for healthy 73 kg subjects were then: 144 N prone, 240 N in lateral decubitus, 800 N upright standing, and 996 N upright sitting.¹⁰ These results demonstrate that the vertebral body loads fluctuate significantly according to the position of the patient. This concept explains well why patients with severe osteoporosis are at risk of developing VCFs when performing simple daily tasks, such as walking, getting down the stairs, or holding groceries.

31.2.2 Impact of Disk Degeneration on the FSU

The role of the disk is to sustain compressive forces in addition to supporting shear and tensile forces.¹¹ The disk is an important component of the FSU as it redistributes loads from the inferior end plate of the superior vertebra to the superior end plate of the inferior vertebra. As disk degeneration increases, the intradiskal pressure measurements decrease¹⁰ and higher loads are transmitted from one vertebral body to the other (creating peak loads) including to a posterior load redistribution that results in a greater load placed on the facet joints and vertebral arch (▶Fig. 31.3). The loss of normal nucleus pulposus microarchitecture can first be appreciated on MRI by a T2-weighted signal intensity drop of the nucleus pulposus of the disk which can be quantified with the modified Pfirrmann grading system (▶Table 31.1).

Fig. 31.3 Load redistribution in a functional spinal unit (FSU) with a normal disk and with a degenerated disk. Degenerative disk disease alters the compression loads on the vertebrae resulting in increased load toward the posterior portion of the vertebral body and neural arch. Adapted from Pollintine et al.¹²

As compressive load within a normal disk increases, the hydrostatic pressure within the nucleus pulposus also increases, and some of the inner pressure (horizontal pressure) is transferred to the annulus fibrosus by circumferential stress (hoop stress). The end-plate forces are normally evenly distributed on the anterior column with less forces transmitted to the posterior elements and facet joints. With increased disk degeneration, there is less load bearing on the ventral end plates (decreasing from 44 to 19%) and more stress on the neural arch, specifically on the facet joints (increasing from 8 to 40%) (▶Fig. 31.3).¹² Load redistribution can be further exaggerated with extension maneuvers (in that context, up to 90% of the compressive forces can occur on the facets).

This anterior-to-posterior redistribution due to disk degeneration gives rise to Stress Shielding. As the load switches posteriorly, as per Wolff’s law, there is less trabecular turnover and remodeling in the anterior vertebral body, resulting in decrease of up to 20% in the trabecular volume and increase of up to 28% in intertrabecular spacing.¹⁴ As discussed earlier, upright flexion results in higher stress on the disk. In patients with disk degeneration, the load delta on the anterior portion of the FSU is even further increased. For example, in flexion versus neutral position, the loads applied to the FSU can increase by up to 300%,¹² further predisposing osteoporotic patients to VCFs.

31.2.3 Biomechanical Changes after Vertebral Compression Fractures

The vertebral end plate itself or its trabecular support is the most common region of the vertebral body to fracture under stress (AO-type A1 or MagerI-type A1.1).¹⁵^–¹⁸ Because of numerous vascular channels transferring metabolites to the nucleus pulposus, the end plates represent areas of weakness within the vertebral body that are even more pronounced in osteoporotic patients.¹⁹ Once an end-plate fracture happens, the disk loses its inner pressure.²⁰ This loss of internal pressure results in increased compressive loading of the anterior wall during active flexion predisposing to adjacent-level fracture²⁰ but also, when the patient is upright, there is an increase in the compressive stress along the posterior vertebral body, posterior annulus fibrosus, and neural arc.²¹^,²² These changes further contribute to the downward spiral of Stress Shielding seen with disk degeneration, thereby predisposing the patient to adjacent-level VCFs.

In wedge-type compression fracture (AO-type A1 or MagerI-type A1.2), segmental kyphosis contributes to stress shift toward the posterior arc of the affected and adjacent levels. This kyphosis amplifies the Stress Shielding of the adjacent level and also contributes to the adjacent-level fracture risk ranging from 12 to 20%.²³^–²⁶ In summary, untreated fractures result in altered FSU biomechanics that predispose to adjacent-level fracture both by creating altered peak loads amplified with flexion of the spine and with increased Stress Shielding, resulting in a VCF “domino” effect.

31.2.4 Technical Considerations during Vertebral Augmentation: Correcting Fracture and FSU-Related Instability

As just mentioned, after a VCF the FSU loads are displaced posteriorly (similar to degeneration of the intervertebral disk). For this reason, whenever possible, PMMA cement should be instilled within the posterior third of the vertebral body to provide further structural support and avoid collapse of the middle column (▶Fig. 31.4). To avoid this collapse, experts in interventional pain management pay great attention to correct the kyphotic angle, realign end plates, and instill cement uniformly within the vertebral body from pedicle to pedicle, superior end plate to the inferior end plate, and the anterior cortex of the vertebral body to the posterior cortex. This process is also referred in our practice as “Civil Engineering of the Spine.”

It is commonly taught in academic centers that PMMA cement should be instilled only in the anterior two-thirds of the vertebral body. The latest biomechanical data argue against this teaching since more posteriorly located structural support is required to avoid posterior wall collapse. However, additional experience and care is required to reliably perform this technique and avoid extravasation in the ventral epidural space. The degree of comminution, the MagerI fracture type and the experience of the interventionist should all be taken into consideration before attempting to more aggressively fill the vertebral body. This adequate PMMA filling will help correct the motion segment instability.

Indeed, after a VCF there are two types of instability that can be improved with vertebral augmentation:

• Fracture mechanical instability.

• FSU instability.

Fracture mechanical instability is when the FSU loading results in microscopic or macroscopic trabecular/bony displacements within the vertebral body. These displacements result in painful nociceptors stimuli transmitted via the basivertebral nerve. Mechanical stability of the fracture can only be achieved when the fracture/cleft is completely filled with PMMA cement, avoiding then all painful stimuli generated from abnormal movement. New bone will be able to grow within the cement porosity and around the cement to heal the fracture.

FSU instability is when a compression fracture alters the mechanical characteristics of the building blocks of the FSU. For instance, a compression fracture will result in decompression of the intradiskal pressure and impairs the ability of the disk to distribute load evenly to the adjacent vertebral body. Also there is an increase in the NZ,⁸ resulting in increased ROM that further facilitates peak loads (for instance with flexion maneuvers). Previous studies suggest that in addition to restoring spinal sagittal alignment, the ability to restore height and kyphosis of the fractured end plate should be the ideal end point to normalized load transmission across the fractured level.²⁰^,²⁷

Fig. 31.4 Follow-up after polymethyl methacrylate (PMMA) vertebral augmentation. (a) Sagittal CT-scan reformat demonstrates PMMA vertebral augmentation of L1, L2, L3. (b) Sagittal CT-scan reformat shows fracture of the posterior portion of the vertebral body with retropulsion of the bone fragments (white arrow) resulting in symptomatic moderate spinal canal stenosis.

Furthermore, other mechanical changes are noticed after VCF, such as decreased strength and stiffness. Restoring biomechanical characteristics of the FSU is highly dependent on the volume of PMMA injected. At least 16% of the vertebral body volume needs to be filled with cement to restore strength and 29% of the vertebral body volume needs to be filled to restore stiffness.²⁸ Interestingly, even after injection of 8 mL of PMMA at the lumbar level, the vertebral body still had not reached the prefracture stiffness level. Stiffness of the vertebral body is important to obtain after vertebral augmentation to restore intradiskal pressure and normalize the FSU load distribution. However, if the disk nucleus pulposus is severely degenerated (i.e., modified Pfirrmann grades 7 and 8), axial loading will be transferred directly to the next vertebral body end plate without being dampened by hoop stress propagating the force from the nucleus pulposus to the annulus fibrosus. In those cases, prophylactic vertebroplasty can selectively be performed to reduce the risk of adjacent-level fracture.²⁹

31.2.5 Technical Considerations during Vertebral Augmentation: Quantity of Cement to Inject

This topic is discussed in greater detail in Chapter 20 of this book but the amount of cement injected within a vertebral body is not only the most important predictor of pain relief in balloon kyphoplasty³⁰^,³¹ but it is also of paramount importance in restoring the FSU stability. The recommended amount of cement to be injected varies significantly in the literature and in clinical studies. The vast majority of vertebral augmentation publications use insufficient cement amount according to the biomechanical data standards. For instance, in the 2009 Buchbinder NEJM study, an average of 2.8 ± 1.2 mL of PMMA cement was injected to treat patients. Considering that the L4 and L5 vertebral bodies have a volume of approximately 45 mL in men and 30 mL in women,³² to restore the vertebral body stiffness after a fracture 29% of the vertebral body volume should be filled, corresponding to a minimum cement volume of 13 mL in men and 8.7 mL in women.

By injecting more cement, the risk of PMMA leakage also increases. Cement leakage within the ventral epidural space is the most feared type of extravasation as it may result in paralysis, radiculopathy, or cauda equine syndrome. Cement leakage into the disk should also be avoided whenever possible. This leakage is not dangerous per se, but if large enough, it will create an axial load vector from the treated vertebral body to the adjacent end plate. Cement extravasation within the nucleus pulposus is a predictor of adjacent vertebral body fracture with an odds ratio of 4.633.³³

Patients with severe disk degeneration (Pfirrmann grades 7 and 8) are particularly at risk to have nucleus pulposus PMMA cement leakage, especially when the vertebral fracture is highly comminuted. Fluid hyper-pressurization technique of the disk with saline or anesthetic diskogram (mixture of dexamethasone and lidocaine) is the one technique used to reduce adjacent-level fracture³³ and can be performed whenever necessary to avoid cement leakage.

31.2.6 Biomechanics after Vertebral Augmentation: Adjacent-Level Fracture Risk

The correlation between vertebral augmentation and adjacent level fracture has been debated for many years. Computer-based finite element models initially suggested that the axial loading necessary to cause an adjacent-level fracture was lower in patient with vertebral augmentation.³⁴^–³⁶ However, these computer studies did not put much emphasis on the intervertebral disk. The intervertebral disk is an extremely important building block of the FSU and has a critical role in the risk of adjacent-level fracture after a VCF. Indeed, the fractured end plate alters the pressure profile of the damaged disk (decompresses the disk) resulting in increased compressive loading of the anterior vertebral body wall during active flexion predisposing to adjacent-level fracture.²⁰

Luo et al³⁷ demonstrated a threefold increase in creep deformation within the anterior column of the vertebra adjacent to a VCF. In their study, vertebroplasty was a protective factor of adjacent levels by reversing adjacent-level creep deformations by 52%. Also, the intradiskal pressure was normalized after vertebroplasty, measuring 76% of the baseline intradiskal pressure.³⁷ Other recent biomechanical cadaver studies supported the protective role of vertebroplasty.³⁸^-⁴⁰ In fact many studies have demonstrated that the instillation of cement within a vertebral body increases the nucleus pulposus pressure but never above baseline.²¹^,³⁷^,³⁹^–⁴¹ Furthermore, these studies supported that injecting a larger volume of cement (at least 7 mL) was shown to restore the compressive stiffness of the spine and significantly decrease the stress on the neural arch.³⁹

Clinical trials further demonstrate that vertebroplasty did not increase the rate of adjacent-level fracture and is even a protective factor. Trout et al⁴² evaluated 432 patients and of all the new fractures that occurred, only 40% of new fractures were adjacent to the level augmented; in other words, most new fractures happened in nonadjacent levels and are related to the underlying osteoporosis. Finally, systematic reviews of the published literature to date done by two meta-analyses demonstrated no adjacent-level fracture risk increase in one study,⁴³ while a decrease of the risk in adjacent-level fracture was demonstrated in other study.⁴⁴

31.3 Key Points

• The FSU is used to analyze spine biomechanics. This unit is composed of two adjacent vertebrae, the intervertebral disk, and the ligaments providing additional stability while excluding muscles or other connecting tissues.

• The maximum loading on the FSU happens in the flexed standing position corresponding to more than twice the upright standing position.

• With increased disk degeneration and nucleus pulposus microarchitectural disturbances, there is less load bearing on the ventral end plates (decreasing from 44 to 19%) and more stress on the neural arch, and facet joints (increasing from 8 to 40%).

• After a VCF, the fractured end plate alters the pressure profile of the damaged disk (decompresses the disk) resulting in increased compressive loading of the anterior vertebral body wall during active flexion, thereby predisposing to adjacent-level fracture(s).

• Vertebral augmentation improves both the fracture mechanical instability and the FSU instability.

• Adequate vertebral augmentation includes correction of the kyphotic angle, restoration of the height of the depressed end plate, and complete filling of the vertebral body (pedicle to pedicle, end plate to end plate, and from the anterior cortex to the posterior cortex).

• To restore the vertebral body stiffness after a VCF, 29% of the vertebral body volume should be filled.

• Biomechanical data and clinical studies demonstrate that vertebral augmentation decreases the stress on the adjacent level and can even reduce the incidence of adjacent-level fractures.