Biomechanical testing of a unique built-in expandable anterior spinal internal fixation system

Surgical procedures for the treatment of thoracolumbar spine diseases include anterior, posterior, and combined anterior and posterior fusion, and adequate immobilization of the spine is critical for successful fusion. The use of anterior thoracolumbar implants and anterior instrumentation for stabilization has become more common in the past decade for reasons including direct visualization and ready access to the anterior column, improved distraction, and placement of a large interbody fusion device [1, 2]. At present, there are 3 systems used for anterior thoracolumbar fixation [1]. One is the screw-rod system, which is represented by the Kaneda® SR Spine System (DePuy Synthes), another is the screw-plate system, which is represented by the Z-Plate Atl™ Anterior Spinal Fixation System (Sofamor Danek) system, and the last is a screw system, which is represented by the MACS HMA (Aesculap, Tuttlingen) system and consist of porous hollow titanium screws with an outer diameter of 12 mm for monocortical use [3]. Though all systems provide adequate anterior fixation, they all have unique strengths and weaknesses.

While the screw-rod system is relatively stable biomechanically, the size is relatively large and the implant profile is relatively high, which can damage peripheral neurovascular structures during or after surgery [1, 4]. In addition, the screw-rod can only be used for vertebrae superior to L3 due to the psoas muscle and pelvic ring. The screw-plate system is relatively simple in structure, easy to manipulate during surgery, and is associated with a relatively short operative time. However, its biomechanical stability is poorer than the screw-rod system, and the plate is less likely to completely match the lateral side of the vertebral body and often leaves gaps [1]. In addition, stress is concentrated at the locking site between the plate and the screw, which can result in internal fixation failure and kyphosis during the late stage after surgery [1]. Although the screw system overcomes the shortcomings of the screw-rod system and the screw-plate system to some extent, the screw has a large diameter (12 mm), and it cannot be used for people with small vertebrae, especially children and adolescents [3].

Expandable screws have been used for improved graft fixation in anterior cruciate ligament reconstruction [5]. Richter et al. [6] reported that a monocortical expansion screw provided the same biomechanical stability as a bicortical 3.5-mm AO screw and better biomechanical stability than a cervical spine locking plate when used in anterior internal fixation of the cervical spine. Studies have also shown that expandable pedicle screws provide improved pullout strength as compared to standard pedicle screws, especially in osteoporotic bone [7–9].

We designed a built-in expandable anterior thoracolumbar fixation system to overcome the shortcomings of current anterior thoracolumbar systems. The fixation system is made of a titanium alloy, and composed of an expandable cylinder screw with an internal cone-shaped space. When an insert is placed into the internal space of the screw, the distal end of the screw expands. The purpose of this study was to compare the biomechanical stability provided by the new anterior fixation system with that of 2 commonly used anterior fixation systems, the Z-Plate and the Kaneda, in a porcine partial vertebral corpectomy model.

This study examined the biomechanical properties of a newly designed built-in expandable anterior spinal fixation system. The results showed that the 3D ROM afforded by the new device was similar to that of the Z-Plate and Kaneda systems, though the Z-plate and Kaneda groups had significantly higher left axial and right axial rotation ROM as compared to expandable device. The maximum torque of the expandable device was greater than that of a common screw, as was the maximum pullout force of the device in both the unexpanded or expanded state.

With the built-in expandable anterior thoracolumbar fixation system the vertebral screw and the rod are completely implanted within the vertebra. This avoids disturbing blood vessels, nerves, and soft tissue around the vertebra, thus overcoming a shortcoming of the screw-rod system. The connecting rod is closer to the center of mechanical transmission, and its torque is relatively short compared to traditional internal fixation devices, which prevents easy screw breakage. In addition, implantation is not affected by the shape of the vertebra, which overcomes a shortcoming of the screw-plate system. Because the system is completely built-in, it can be implanted via the lateral, anterolateral, and anterior approaches. Expansion of the distal end of the fixator increases both the holding force and the anti-rotation capacity of the vertebral screw, overcoming the poor anti-rotation capacity of the single screw-rod system. The aperture on the leaflet in the distal end of the vertebral screw, and the gap between leaflets, connect bone fragments within the vertebral screw and bone outside of the screw, thus promoting bone ingrowth to achieve true permanent fixation, i.e., the system is designed such that the implant and vertebral body become fused together, and thus the implant cannot be removed. Lastly, the vertebral screw is designed to have a long tail, which can be used for distraction and compression of the intervertebral disc space, and can be broken off after surgery.

Fresh adult pig spines are similar to the human spine, they are easily obtained, and specimens of the same weight, age, and size are simple to identify. For these reasons the porcine spine is an ideal animal specimen for testing the biomechanical properties of internal fixation devices in the thoracolumbar spine [16]. The characteristics of spine motion indicate that the load of the specimen is the moment applied to the cephalic and caudal ends. Because of the complexity and individual differences of in vivo spine loading, the moment is closely related to the amount of regular exercise an individual receives. In addition, the moments on different segments of a spine specimen are different. The loaded torsional moment can at least guarantee normal ROM of a spine specimen.

Though the ROM of the whole spine is large, the amplitude of the motion of each segment is relatively small. The true load-deformation curve of a functional spinal unit is 2-phase and non-linear [15]. At the initial portion of the curve, the deformation is relatively small and the corresponding gradient is also relatively small. As the load increases, the deformation resistance increases. The first stage of motion, the initial portion of the curve with a small gradient, is termed the NZ. When the load increases and exceeds the limitations of the NZ, the resistance to deformation increases significantly. This is the second stage of the curve, and is termed the EZ. The sum of the NZ and EZ represent the total physiological activity (ROM) of a functional spinal unit [15]. However, evaluation of a multi-segmental spinal unit (MSU) is more valuable when examining the biomechanical effect of different instrumentation systems [13, 15].

Internal fixation of the spine mainly aims to provide sufficient stability before rigid fusion of the spine occurs [1, 2]. High immediate stability after internal fixation increases the fusion rate of interbody bone grafting [17], and reduces the rate of internal fixation failure. All spinal fixation systems provide biomechanical stability; however, stiffness of the spine after instrumentation has been shown to be related to the design of instrumentation system, rather than if it is a plate or rod style system [18, 19].

We found that the maximum torque and pullout force of the expandable device were greater than that of a common fixation screw. Both anterior and posterior internal fixation screws are affected by axial pullout force, flexion force, and rotation force. Screw loosening and pullout are the result of the combined effects of these 3 forces. Most biomechanical experiments use axial pullout force as an indicator of screw holding strength [20]. Screw pullout force is related to screw shape, diameter, and depth of insertion, and bone density, and the most important factors are screw diameter, length, and bone density [20]. All screws used in the current study were 30 mm in length, which eliminated the effect of screw length on the results. The maximum axial pullout force (F-max) of the screw depends on the shear stress between the screw and bone; namely, when the contact area between the screw and the bone is large, the shear stress is large. Liu et al. [21] consider that axial pullout of a screw may induce a torque ‘spinning out’ the screw because there is a certain inclination in the screw thread, and this may accelerate screw loosening. Therefore, when only the axial pullout force is examined there is a limitation in the evaluation of anterior screw stabilityRotational torque depends on the friction at the interface between the screw and the bone, and is represented by f = μ × N, where f is the force of friction, μ is the friction coefficient of the screw-bone interface, and N is the positive pressure of the screw-bone interface. The rotational torque of a common screw depends on the force at the screw-bone interface, and its size is relatively limited.

In general, pullout force increases with enlargement of the screw diameter. In a cadaveric study, Willett et al. [22] reported that the mean pull-out force for a 6-mm screw was 597 N, significantly greater than the mean force of 405 N for a 5 mm screw. Large diameter screws compress the cancellous and trabecular bone during screw insertion to form a dense bone layer, and thus the screw thread is embedded in dense bone thereby increasing the pullout force of the screw [23]. Study has shown no significant difference in the maximum axial pullout force of screws less than 1 mm in diameter, but there is a significant difference when the screw diameter is more than 1 mm [24]. Krag et al. [25] reported that screwing-in torque increased significantly when the diameter of a screw increased from 6 mm to 8 mm.

The diameter of the unexpanded screw in the current study was 6.5 mm, which is larger than that of common screws, and the diameter of the screw tip becomes larger after expansion. The F-max of the common screw was 1,826.67 ± 260.25 N, that of the unexpanded anterior fixation screw was 2,333.49 ± 310.14 N, and that of the expanded anterior fixation screw was 3,035.48 ± 252 N, which was 30.09% more than the pullout force of the unexpanded screw and 65.14% more than that of the common anterior fixation screw. By inserting an inner core into the expandable anterior thoracolumbar fixation screw, the screw tip expands. This increases the screw diameter to increase the holding force, and pushes the leaflet in the distal end of the screw to create a “claw” shape that increases the contact area with the bone, also increasing the holding force.

A greater pullout force is especially important in patients with osteoporosis. Study has shown a positive correlation between bone mineral density (BMD) and F-max; namely, the greater BMD, the greater the fixation strength [26]. Li et al. [27] defined a BMD below 0.9 g/cm² as osteoporosis, and found that the F-max and bending moment of screws were 1,062.8 ± 72.2 N and 2.6 N^.m, respectively, in bone with a BMD above the cutoff, and 232 ± 92.4 N and 0.49 N^.m in osteoporotic bone. Okuyama et al. [26] reported that the F-max of a screw decreases 60 N when the BMD decreases 0.01 g/cm² in osteoporotic vertebrae. Similarly, Yamagata et al. [28] reported that the F-max of a screw decreases l0 kPa when the BMD decreases 100 mg/cm². Cook et al. [29] applied expandable screws to human spine specimens and found that there was positive correlation between BMD and F-max. In the current study, we used a randomized design to help exclude the impact of bone density on the results.

There are limitations of this study that should be considered. The number of specimens tested was small, and while fresh adult pig specimens closely resemble the human spine, the number of thoracic vertebra is not the same as in the human spine (the pigspine has 15 thoracic vertebra). Because this study used pig spine specimens, and the absolute values for torque and pullout force will be different than if testing were done in vivo. Because the diameter of the screws used in the MAC system is large and can damage the vertebral body, we did not include MAC system in the study. In an adult thoracolumbar spine screws 6 mm or 6.5 mm are commonly used. In China, the diameter of most screws used with the thoracolumbar anterior single screw-rod system are 5.5 mm, which is why 5.5 mm screws were used in this study. The difference in screw diameter will have affected the pullout strength results. In this study the pull- out strength of a single screw was compared among the 3 systems. The results may be different clinically as in the Kaneda and Z-plate 2 screws are used in each vertebra which may increase the overall pullout force. We did not perform fatigue testing or 3-point bending testing of the new system. Lastly, a finite element analysis should be carried out to study the load-carrying capacity of the screw, and provide a theoretical basis for further improvement of the design.

The built-in anterior fixation system tested provides better axial rotational stability as compared to other 2 systems tested, and greater maximum torque and pullout strength than a common fixation screw. The system can be installed with less manipulation of surrounding tissue than rod or plate systems.