Introduction

The microstructural effects of aging on the spine may have dramatic consequences on both the individual vertebrae and the vertebra as a constituent within an osteoligamentous structure, that is, a functional spinal unit (FSU). Additionally, the cervical, thoracic, and lumbar regions of the spinal column may be adversely affected by the deleterious effects of senescence. The consequences may cover a spectrum of physical quality-of-life factors ranging from the relatively benign to those that dramatically alter the health of a patient. When clinicians are faced with deteriorating conditions severe enough to warrant surgical intervention, additional considerations must be made for the properties of senescent spines. Therefore, the biomechanical capabilities of the spine should be examined with careful consideration for age along with this caveat: biomechanical changes do not necessarily become symptomatic.

Biomechanical measurements can be affected by numerous indicators, and it is important to distinguish which are related to global measures, for example, body mass index, and which may be relevant specifically to the local spinal elements, e.g., friability of a vertebral body. Two distinct but related indicators should be evaluated with spinal pathologies: the advancement of age and degenerative changes resulting in anatomical transmutation that potentially leads to abnormal loading of the spine. Anatomical changes may be attributed to the primary degenerative conditions associated with age. Miller et al reported an approximate 10% occurrence of severely degenerated intervertebral discs in 50-year-old males, with an increase to 60% in 70-year-olds.¹ The degenerative conditions result in several anatomical changes and, of particular importance to an aging population, is the potential for constriction of the spinal canal diameter. The cause of the constriction may be from a single specific etiology or from a combination of factors, including spinal canal stenosis, disc herniation, osteophyte growth into the canal, hypertrophy of the ligamentum flavum, and calcification of the posterior longitudinal ligament and the ligamentum flavum.

A combination of interrelated mechanobiological conditions and associated kinematic response of the spine due to degenerative diseases is also known to occur with age. Changes in proteoglycan concentration within the intervertebral disc along with matrix disorganization result in a cascade of events over time that affect the anatomical structures within an FSU. The range of motion (RoM) and the ability to absorb and transmit load in the spine are biomechanical capabilities that may be compromised by microstructural changes within the anterior and posterior columns. Under the worst conditions, the degenerative pathology within a FSU results in a significantly different kinematic response to physiologic motion, and abnormal loading may occur.

Aging and Degenerative Changes on the Effects of Biomechanical Range of Motion

The relationship between age, degeneration, and RoM has been studied both in human cadaveric FSU testing and in clinical studies. The instability of the lumbar spine was proposed by Kirkaldy-Willis and Farfan to be categorized into three diskrete stages of degenerative change. In order of progression, the clinical assessment of the lumbar spine categorized pathologic changes as temporary dysfunction, the unstable phase, and finally, stabilization.² Well-defined, controlled, biomechanical testing and clinical studies involving well-documented patient profiles have tested various aspects of this initial hypothesis on spinal instability.

Traditional methods of comparing the effects of age, degeneration, or subsequent treatments have been subjected to biomechanical characterization through the flexibility test method. The methodology of flexibility testing has been well described in the literature, originating with Panjabi’s early description of load input utilizing pure moments.³ Subsequent comparisons, particularly relevant in fixation instrumentation via flexibility testing, have described the performance of these devices relative to the intact spine. often with high mean age donor specimen. Additionally, comparisons between fixation treatments, as well as comparison of fixation treatments from laboratory to laboratory, have been possible. The standardization of the pure moment test protocol by Goel et al has contributed to the repeatability despite biologic variability inherent in cadaveric testing.⁴

It is important to understand the rationale of the test methodology when considering clinically relevant biomechanical studies. The basis of the traditional flexibility test, or pure moment testing, is to apply a uniform moment across all FSUs in a given specimen. Figure 8-1 is an example of a mounted lumbar specimen that will be subjected to flexion-extension bending. The ability to extrapolate the biomechanical effects to clinical outcomes is dependent on study design and successful interpretation of the resulting data. Clinically relevant biomechanical testing in the appropriate form is an important parameter for clinicians to consider in the triage of patients with spinal pathologies.

FIGURE 8-1 Biomechanical test setup subjected to flexibility protocol, with a lumbar specimen mounted in flexion-extension test.

In a cadaveric human lumbar study by Mimura et al, the authors were able to demonstrate a statistically significant difference between RoM in lateral bending, but not in flexion-extension bending, for intervertebral discs with degenerative ratings in whole lumbar specimens under a flexibility protocol.⁵ Biomechanical studies involving age as a variable in the analysis are often shown to be correlated to RoM. Board et al reported on the results of a human cadaveric cervical biomechanical study. Their results suggest that biomechanical flexion-extension in pure moment loading decreases the RoM as a function of the age of the specimen.⁶ These findings agreed with published articles, when extrapolated and compared to equivalent test parameters. In a similar clinical evaluation on bending in the cervical spine involving only males, Sforza et al concluded that young adult males exhibited statistically significant larger flexion-extension RoM compared to their middle-aged counterparts who participated in the study.⁷ Similarly, in a clinical cervical study involving multiple factors including both age and degeneration, Simpson et al determined age to be the most significant factor on RoM.⁸

Confounding these results are clinical considerations in which surgical treatment may be warranted, but subsequent conditions and outcomes related to the specific implant or procedure for the elderly patient may not be clear. For example, symptomatic spine pathology resulting in instability of a FSU and suitable for an instrumented fusion procedure must consider the interaction of the hardware and the patient’s local host tissue. In addition to global metrics of bone quality, the local bone purchase dependent upon the microstructural integrity of bony trabeculation at the index FSU may have undergone severe anatomical changes. These differences affect the load response, exacerbate degenerative pathologies, and require additional considerations for the type of instrumentation suitable for the patient preoperatively. Intraoperatively, additional factors may further alter the structural integrity of the FSU, for example, endplate preparation or pilot hole drilling combined with tapping.

The biomechanical changes inherent to aging are complex in nature. Many steps have been taken toward the understanding the fundamental process of maintaining a healthy spine, including bone healing, the role of the intervertebral disc, and the significance of endplate changes. However, understanding the nature of biomechanical measurement and the clinical relevance of each metric may help further elucidate the suitability of the treatment for the senescent patient and, ultimately, improved treatment options may be developed.

Assessing Anatomical Changes

Accurate measurements of bone strength are essential to the clinical management of a diseased spine. Both the diagnosis of disease, such as osteoporosis, and also its triage, such as the surgical treatment of an unstable spinal motion segment with hardware, would benefit from explicit descriptions of vertebral bone quality. Dual-energy x-ray absorptiometry(DXA)–obtained measures of bone mineral density are widely regarded across many medical diskiplines as the gold standard for assessing fracture risk. The guidelines set by the World Health Organization based on the standard deviation units of bone mineral density (BMD), referred to as T-scores, have limitations that are documented in the literature. Also, BMD has not consistently supported correlations with patient fracture in all risk groups, and additional indicators to further enhance DXA scores would be particularly beneficial to lower-risk patients with higher T-scores.

Two primary reasons for the frequency of DXA measurements are the relatively noninvasive, nondestructive nature of the test and documented correlations associated with DXA measurements. Imaging modalities that assist in the classification of degeneration have been useful in FSU pathophysiology and could be useful in understanding the relationships between aging, degeneration, and biomechanics of the FSU. Therefore, through the use of known techniques in detecting degeneration of the osteoligamentous structures, such as magnetic resonance imaging (MRI) and the Modic classification of vertebral endplate change, stronger correlations may be established between age and degeneration. Ideally, earlier fracture diagnostic capabilities for all risk groups may be added to a clinician’s armamentarium.

Osteoporosis, Aging, and Biomechanical Properties

The use of clinical guidelines based primarily on BMD results has been widely accepted. The ability to identify patients with high risk of fracture via low BMD measurements, defined by T-scores of −2.5 or lower, and to subsequently provide effective pharmacological treatments, has been proved through large double-blinded placebo-controlled trials. Several challenges remain in identifying low-risk population and ultimately a means in cost effectively managing fracture risk. In an examination of 149,524 postmenopausal women 50 years of age and older with fractures, 82% had T-scores above the threshold criterion of −2.5.⁹ Thus, it has been suggested that the value of BMD would be enhanced with additional risk factors for improved diagnostic capabilities.

Vertebral fracture is the most common result of osteoporosis in postmenopausal women older than 60 years of age. Surgical management through vertebral body augmentation involving the injection of polymethylmethacrylate (PMMA) has been diskussed as a method of fracture treatment in the literature. Understandably, the preferred course should be prevention, as opposed to surgical intervention. In addition, iatrogenic effects from vertebral body augmentation, including adjacent level implications, have not been assessed in well-controlled studies.

Analysis of available data regarding fracture in moderate-risk patient populations shows that the increase in fracture risk with decreasing age-adjusted BMD and other factors, including a prior history of fractures, are also important considerations. In short, not all patients diagnosed with current threshold values for osteoporosis will go on to fracture. Moreover, not all patients above the osteopenic level will be free of fracture related to bone structure and density.

BMD and Implications on Instrumented Procedures

Another use for BMD as measured by DXA is to determine the quality of bone for screw purchase. BMD has been shown to be correlated to pull-out strength, and for many fixation devices, screw purchase plays an important role in providing immediate stability and longer-term fixation. The screw-bone interface is integral to many constructs, such as anterior cervical plating and lumbar pedicle screw fixation, and adequate screw purchase is necessary for treatment of any spinal pathology depending on such instrumentation for stabilization and fixation. In patients showing an insufficient BMD, purchase becomes cause for concern. For the osteoporotic spine, the screw-bone interface may be augmented through various techniques in order to provide additional purchase strength. However, methods such as augmentation through PMMA should be exercised with caution, as complications may arise from the use of bone cement.

Biomechanical measures used to test screw-bone interfaces have been evaluated in a number of different ways. Axial pull-out strength has been frequently reported in the literature, including in human cadaveric spines that would be considered osteoporotic. Figure 8-2 illustrates a common test method for determining axial screw-bone interface strength. However, cyclical loading has been suggested to mimic more realistic modes of failure for implanted constructs. Studies have examined bending failure as an appropriate method of loading.¹⁰

FIGURE 8-2 Method of testing the screw-bone interface strength in axial pull-out.

The limitation of any test protocol is the ability to directly compare against native human conditions. Several of the published studies have considered various test materials including both cadaveric and synthetic test specimens. The utility of such tests should still be recognized but it must be tempered with an appropriate understanding of the clinical ramifications. Testing on cadaveric animal models is a consideration that should be taken into account when evaluating screw-bone interface results. Bending modes of failures are considered more realistic complications, but test protocols are more difficult to execute. This is often due to the difficulty in defining the appropriate test methodology.

The bending moment and the associated load levels are one set of test parameters. A depiction of testing the effects of the screw-bone interface through bending moments in vertebrae is shown in Figure 8-3. The construct configuration is another study design consideration with implications for unilateral versus bilateral constructs with and without crosslinks. Fatigue is also another major factor difficult to mimic in a cadaveric test environment during biomechanical testing. Screw pull-out tests can be performed along the bone screw axis, but the flexion-extension type of bending should be executed under a cyclical protocol that eventually fails the screw-bone interface through off-bone screw axis loading. This results in a markedly different biomechanical response at the FSU and, in turn, may have different complications, for example, screw loosening. Gau et al reported modes of radiological failure in a clinical radiographic study that examined implanted constructs that exhibited “windshield-wipering,” which may be an indication of bending fatigue at the screw-bone interface, and classified them accordingly.¹¹ Interestingly, these were not symptomatic complications.

FIGURE 8-3 Application of cyclical bending moments necessary for creating “windshield-wiper” failures.

The ability to derive a specific BMD measurement has been published in a study by Wittenberg et al¹² The authors hypothesized an equivalent mineral density of 90 mg/ml from quantitative computed tomography (qCT) as a threshold level to expect complications associated with screw loosening and 120 mg/ml as a threshold for fewer problems. This has not been validated in a clinical outcomes trial. Often, it is surgeon perception on the adequacy of bony purchase that governs the decision to instrument a patient with hardware. Additional data to provide a validated standardized DXA metric with positively correlated clinical outcomes for specific threshold levels would provide a higher confidence in BMD measurements as a preoperative indicator for instrumented procedures.

Dual Energy x-ray Absorptiometry and Mechanical Strength

The mechanical properties of both a FSU and its components may be analyzed by a number of different measurements and techniques. For ultimate strength and stiffness property studies, both localized indentation studies as well as compressive failure tests of vertebral bodies en bloc and complete FSUs have been reported in the literature. Due to differences used in the test protocols to determine strength, the correlation between bone mineral content (BMC) and BMD as reflected by DXA measurements have varied with failure loads.

Studies have shown the failure strength of vertebral bodies as measured by indentation testing differs between superior and inferior endplates; and also between locations on the same vertebral body endplate; for example, posterolateral regions tend to have the highest relative strength. With exceptions, the authors concluded from their study that a decrease in BMC correlated to a decrease in strength. In addition, the same research group¹³ later reported removal of the endplate resulted in a significant decrease in compressive failure strength. However, it was not clear if removal of the endplate affected DXA measurements.

DXA is a measurement reflective of the underlying bone mineralization. In order to determine the effects of surgical site preparation, for example., removal of the cartilaginous endplate for intervertebral spacer implants, the effects of surgical approaches on the structural integrity should be understood. DXA and vertebral strength have been shown to correlate closely in the native state. Vertebral body endplates have been shown to affect failure strength. When overly manipulated, the endplates can potentially result in the collapse of a vertebral body, but the relationship between iatrogenic complications due to surgical preparation and implant stiffness coupled with low BMD patients has not been studied.

The consistency of DXA measurements, particularly as it relates to strength, is dependent upon a number of factors, including artifacts from soft tissue. The correlations are especially problematic with higher BMD content. In a study utilizing DXA and cadaveric spine positioning, Myers et al suggested clinical studies to confirm supine lateral patient positioning would be more effective in determining BMD measurements.¹⁴ The aging phenomenon that occurs within every human body may potentially cause global osteoarthritic changes, including BMC and BMD within the spine, that subsequently affect local DXA measurements. Utilizing animal models to control the homogeneity of specimens has not resulted in more significant correlations between BMD and strength. Contrarily, in a study involving porcine cervical spines,¹⁵ the investigators reported no significant correlation between BMC or BMD with compressive failure strength. Furthermore, large animal models rarely exhibit vertebral body fractures even with reduced BMD levels, and thus would not be characterized into high risk for low-trauma fracture categories.

In conclusion, DXA has been a widely used indicator for osteoporotic patients and for assessing the risk of fracture. Potentially, it has validity as a gauge for the screw-bone interface in axial pull-out, but the more complex modes of loading often found in bone-anchoring devices require a better understanding of the failure modes. In addition, with the current DXA standard as an indicator of bone strength, the implications of implant failures and resistance to fracture are not well defined for T-scores above −2.5. However, other modalities exist that may augment the current metrics in quantifying the usefulness of current BMD measurements.

Modic Classification of Vertebral Endplate Change

Degenerative changes of the lumbar spine have been observed with MRI techniques. Specific signal changes from vertebral body endplates and marrow have been differentiated through imaging techniques that increased tissue contrast. A classification system of MRI scans using two different pulse sequences was published by Modic et al¹⁶ Optimizing T1 and T2 relaxation times in pulse sequences during MRI studies helped define and characterize the imaged tissues. Three different types of change were recognized from T1-weighted and T2-weighted MRI scans of the same spine segment. The following is the accepted classification used for Modic changes:

The interobserver and intraobserver error in a clinical study has been documented and the consistency of this imaging classification system was confirmed.¹⁷ The study involved five independent observers of various clinical spine experience who graded 50 sagittal T1-weighted and T2-weighted MRI scans. The evaluation of the same scans was repeated by each participant following a 3-week interval with no reference to the first assessment. The intraobserver agreement, or consistency between the first and second evaluations by the same observer, was assessed based on Landis and Koch’s use of the kappa statistic,¹⁸ which was equal to 0.71. Additionally, interobserver agreement or consistency among all the observers was calculated to be 0.85 for the study. This study demonstrated the intraobserver agreement was substantial while interobserver agreement was excellent for the Modic classifications.

Although the original imaging studies were designed to investigate degenerative disc disease, the impact of these changes is not well understood nor is the clinical implication. One of the early findings of Modic type 1 change was fissures in the endplates, which were confirmed by histological findings. The intensity changes from MRI scans have been deduced to reflect osteocartilaginous fracture signs. Disc herniations that include components of the endplate, namely hyaline cartilage, are then suggestive of avulsion-type disc herniations. Reportedly, this form of intervertebral disc herniation is predominant in the elderly and may warrant investigations into failure strength.