From the biomechanical viewpoint, four screws and two longitudinal rods form a three-dimensional construct to stabilize the instrumented segment. The two trajectories of screw insertion result in two biomechanical responses: the bony contact of posterior element, cancellous core, and cortical shell along the screw length and the stabilizing base that is spanned by the paired screws. The two types of TT and CBT screws showed distinct modes of bone contact. Significant variations in screw pull-out strength and the higher purchasing ability of the 5.5 mm diameter CBT than TT screws were demonstrated previously 5,7-9, but this was not the case with the 3.5 mm diameter CBT screw stress.
The dynamization of screw fixation aims to provide flexibility to the adjacent segments and suppress the postoperative ASD problem. There was less ROM constraint (−50% flexion, −28% extension in CBT dynamic and −58%, −30% in TT dynamic) and lower stress sharing (−34%, −28% in CBT dynamic and −38%, −23% in TT dynamic) in dynamic CBT in flexion and extension. Similar to previous findings in bending, there was weaker fixation strength in bending in both dynamic and static CBT compared with TT (Figs. 3C and 5C). 5,7 Interestingly, in rotation, both static TT and CBT showed a 33% reduction in normalized disc stress. In dynamic TT and CBT, however, the disc stress adversely increased by 18% (Fig. 5D), which may be attributed to pretension in the cord.
The entire lumbosacral model provides more detailed information about biomechanical behaviors of fixed and adjacent segments. Interesting, the adjacent segment compensation in rotation is different from the other motion (Fig. 3D). The rotation ROM did not show compensation in L3-L4 after static TT and CBT fixation. This research accounted for this behavior by comparing facet force and disc ROM in intact and static TT and CBT models in the time curve of the finite-element analysis (Fig. 8). For static TT and CBT, the faster increase in L3-L4 facet force shows earlier contact of the paired facet after fixation and thus deteriorates the kinetic and kinematic compensation for rotation.
Among four fixation methods, the TT static behaved as a more constrained stabilizer to the fixed segments. These surgery-related findings indicate that the choice of the static or dynamic fixators is potentially dependent on the stability demand of the fixed segment and degeneration degree of the adjacent segments. If the structural integrity of the fixed segment is not the first requirement, the CBT dynamic might be a recommended option in a situation of mild or moderate degeneration at the adjacent segments. However, a static CBT and even a static TT fixation might be adopted if the adjacent segments are still healthy and the fixed segment requires stabilization.
The stress distribution for all fixations showed that stress was concentrated near the screw hub, at the junction of the threaded and unthreaded regions, corroborating reports showing sites with most failure on the TT screws. 24 Interestingly, the static and dynamic CBT fixations consistently showed higher von Mises stress distribution than their counterparts, which indicated that the use of the CBT screws was more prone to screw-bone interface failure. However, the results of loosening the CBT screws were in contrast to previous studies, which revealed compatible or even better pull-out strength and toggle strength of the CBT than TT. 7-9 The simulation in this study showed that the different geometry and mode of bone contact in CBT may subject it to tremendous regional stress, resulting in loosening (bone contact failure) or breakage (screw fatigue). This warrants the need for additional investigation on interface failure using data on screw threads and bone destruction.
There are numerous reports in the literature that attest to the superiority of the CBT screws over TTs for fixation of osteoporotic bone.3,5,6,7,8,9 For our study, we divided the vertebral body into three zones, the cancellous core, cortical shell, and posterior element, which correspond to the Young’s modulus of three distinct vertebral components. The TT screw was within the core and the CBT screw served to anchor the cortical shell (Fig. 2). Due to the extreme nonlinearity of the entire lumbosacral column (L1-S1), this study did not use micro-computerized tomography7 to evaluate the trabecular bone, as our intent was to avoid too complex a model, which inevitably leads to divergence when performing finite-element analysis. Consequently, osteoporosis within the cancellous core was not simulated because the capacity of TT fixation for immobilization (e.g., stability and holding power) was not reduced. Consequently, the predicted results that were based on the assumption of healthy bone might overestimate the capacity of the TT screw to a greater extent than the CBT screw. The biomechanical impact of the osteoporotic vertebral cores will be investigated in future studies.
In clinical practice, adjacent disks are prone to mild or even moderate degeneration; this is observed even when the condition does not require instrumentation. The morphological and structural changes in the adjacent disks in this case make them stiffer than the healthy ones. This had been simulated and described in our previous report 22, 25. In this study, we assumed that the adjacent disks were healthy so as to evaluate the dynamic effects and trajectories of the TT and CBT screws with respect to the adjacent disks. A stiffer disk can suppress transferred kinematic and kinetic changes from the instrumented segment. Consequently, the assumption of the healthy L3/L4 and L5/S1 disks will maximize the implant-induced effects on the non-stiffened adjacent disks.
Previous studies on CBT screws have focused on non-inferior pull-out and toggle strength in direct comparison to TT screws.5, 8, 9, 21 Osteoporosis may have smaller impact on CBT screws than on TT screws due to the relative cortical trajectories. No research has been published that addresses segmental stability and relative impact on adjacent spinal levels when comparing these two fixation methods. The results in this study showed that static CBT screws provide inferior segmental stability compared with that promoted by TT screws. However, considering its minimal invasiveness, static CBT screws may still be quite useful in short segment fusion after limited spinal decompression or after discectomy. CBT screws may also be used as an alternative fixation method in osteoporotic patients, as the use of larger cage may compensate for the inferior stability.26 Pars fractures compromise the cortical trajectory of CBT; as such, this condition should be considered a relative contraindication for their use.27
Dynamic TT simulates the use of Dynesys. Dynesys is the most widely used dynamic fixation method and is based on use of PSs; however, the clinical results generated by this procedure are not fully clear.20 In our simulation, the dynamic TT provides sound biomechanical profile, decreases disk stress at the instrumented level, and reduces stress compensation in adjacent disc and facet, similar to previous findings.28-31 However, no significant clinical benefits of Dynesys were reported in both short term and long-term studies.32 The gap between the biomechanical studies and clinical results may relate to the destruction of muscle during the surgical approach, disruption of facet joint as well as the impact of instrumentation itself that results in deviation of the motion of specific segments away from what is physiologically within normal limits.
Dynamic CBT is a novel design modification from Dynesys. The simulation shows that dynamic CBT results in only slightly inferior segmental stability compared with dynamic TT. The property makes it a minimally invasive alternative to Dynesys for use in short segment stabilization after discectomy and for low-grade spondylolisthesis. This design may improve the clinical results obtained with Dynesys and reduce the incidence of adjacent segment disease (ASD) by the inherently lower chance of facet joint disruption and/or destruction of the posterior musculature. This method may be valuable for osteoporotic patient who needs dynamic fixation. Future biomechanical studies will focus on its effect on physiological motion of the spine.
As with any finite-element analysis, certain assumption-related limitations were inherent in this study. Some of these limitations, including the morphology and material properties of the tissues involved have been discussed previously.22 However, the pedicle size has substantial impact on the diameter of the inserted screw; this point was not extensively investigated in this study. Only 3.5 mm diameter CBT screws showed the same effect of the slimmest diameter screws currently in clinical use. The biomechanical effects of screw diameter have been considered extensively in the literature. This study focused the use of dynamic fixation methods and comparing different trajectories. CBT, limited by the anatomy of the posterior element, cannot tolerate screws with same diameter as those used in the PS trajectory, which are usually >6.0 mm in diameter. Thus, this research did not standardize the sizes of the CBT and TT screws. Future research will focus on the effect of different screw diameters on both stability and stress.
This study simulated the TT and CBT screws as monoaxial elements, and as such, the curvatures of the longitudinal rods differed from those used in the clinic (Figs. 2C and 2D). For the CBT screw, this renders the two-ended surfaces of the Dynesys spacer as non-orthogonal to the spacer axis. Consequently, the simulation of the dynamic CBT fixation might reflect the actual conditions at the screw-spacer interfaces. In practice, a poly-axial screw can avoid excessive rod curvature and simplifying spacer ends, a point that was not considered in the current study. Due to the high nonlinearity of the entire lumbosacral column (L1-S1), this study is designed to simulate the impact of fixation on segmental stability and adjacent stress rather than on bone-screw failure (e.g., loosening and breakage). Consequently, the simulation omitted screw threads and assumed bone-screw interfaces as bone to raise computational efficiency and convergence. Additionally, the fusion cages were not instrumented into the fixed segment. Given these limitations, the findings obtained in this study that were designed to predict results of transpedicular fixation, might overestimate the screw stresses and underestimate ASD.
Similar to any finite-element method, certain assumption-related limitations were inherent in this study. Some, such as the morphology and material property of the tissues have been previously discussed. 22 However, the pedicle size substantially affects the diameter of the inserted screw and is not extensively investigated in this study. Only 3.5 mm diameter CBT screws were considered to show the effect of the slimmest diameter screws that may be used in clinical use. Future research will work on the effect of different screw diameters on the construct stability and screw stress.
This study simulated the TT and CBT screws as monoaxial types, such that the curvatures of the longitudinal rods differed from that used in the clinic (Figs. 2C and 2D). For the CBT screw, this renders the two end surfaces of the Dynesys spacer non-orthogonal to the spacer axis. Consequently, the simulation of the dynamic CBT fixation might reflect the actual condition at the screw-spacer interfaces. In practice, the poly-axial screw can avoid excessive rod contouring or simplifying spacer ends, which was not considered in the current study. Additionally, the fusion cages were not instrumented into the fixed segment. The predicted results of only transpedicular fixation, which was obtained in this study, might overestimate the screw stresses and underestimate ASD.
This study is, after all, a computational model study. The validation may not reflect the real behavior of human tissue. Factors such as osteoporosis and disc degeneration were not considered in this study. Although, the computational model may reflect some biomechanical behaviors of the spine, to apply the results in the complex clinical practice, further biomechanical studies are needed to confirm the results.