Biomechanical analysis of trivergent, a new posterior lumbar spinal fixation system

Abstract

Background

Pedicle screw fixation provides stability after posterior lumbar spine surgery. Traditional pedicle screw devices face several complications, including loose and broken screws. A strategy addressing these issues is to eliminate sources of failure and optimize the biomechanical advantage of using a “triangle” in these devices. The purpose of this study was to measure the biomechanical properties of the Trivergent spinal fixation system and compare it to a predicate construct (ie, an FDA approved similar device against which testing is compared), Here, we report the biomechanical results of a new plate-based, triangular fixation device.

Methods

The ASTM (American Society for Testing and Materials) protocol included 3 forms of testing: static compression bending, static torsion, and dynamic compression bending. 3 predicate constructs, 3 Trivergent 60° constructs, and 3 Trivergent 75° constructs were tested in both static compression bending and static torsion. 3 predicate constructs, 4 Trivergent 60° constructs, and 7 Trivergent 75° constructs were tested in dynamic compression bending.

Results

In static compression bending, the Trivergent 60°, Trivergent 75°, and predicate constructs tolerated average ultimate loads of 480 N, 550 N, and 400 N, and demonstrated average ultimate moments of 26.5 N-m, 30.2 N-m, and 18.6 N-m, respectively. In static torsion, the Trivergent 60°, Trivergent 75°, and predicate constructs tolerated average yield torques of 44 N-m, 34 N-m, and 20.5 N-m, and demonstrated an average torsional stiffness of 14 N-m/deg, 13 N-m/deg, and 5.22 N-m/deg, respectively. In dynamic compression bending, the Trivergent 60°, Trivergent 75°, and predicate constructs tolerated maximum loads of 300 N (maximum load tested), 230 N, and 160 N, respectively.

Conclusions

Hardware failure in posterior spinal pedicle screw fixation devices is common, and includes issues such as screw migration and rod breakage. The Trivergent device was designed to address these complications by incorporating characteristics such as screw divergence and cortical bone engagement. This study provides evidence of the biomechanical strength of Trivergent, supporting its potential ability to improve outcomes for posterior lumbar fusion surgery. Future cadaveric and human studies would help to substantiate the utility of Trivergent in treating degenerative disc disease, spinal stenosis, and other pathologies of the lumbar spine.

Introduction

Pedicle screws are commonly used in spinal surgeries to provide stability while bony fusion of the vertebral bodies occurs. These fixation systems effectively promote surgical arthrodesis [1]. However, there have been reported complications, including bent and broken screws, screw pullout, neurovascular injury, and disc degeneration adjacent to the instrumentation site [[2], [3], [4], [5]]. These issues are exacerbated in osteoporotic patients, who have weaker bone structure [3]. Rates of pedicle screw loosening have been reported to occur in 4.1% to as high as 54% of spine surgery patients, with the highest rates occurring in osteoporotic bone [[6], [7], [8], [9]].

In addition to these concerns, the traditional lumbar screw placement procedure involves a cumbersome assortment of instrument trays and variably sized screws and rods. Strategies aimed at addressing these issues have included the use of multiple spinal level fixation points [10], cement augmentation [11], hydroxyapatite screw coating [12], alternative screw trajectories [13], placement of multiple rods, and the development of more effective minimally invasive screw placement techniques [5,14,15].

Trivergent is a newly designed spinal fixation system which is intended to stabilize lumbar L2-S1 segments. Lumbar stabilization with the simple 3-piece design requires fewer instruments than other techniques and addresses complications commonly seen with classic pedicle screw fixation. In this construct, small plates replace the rods. Pedicle screws “diverge” through the plate and enter the pedicles along a cortical bone trajectory (CBT), maximizing screw purchase and holding strength. Clinically, the CBT has demonstrated lower rates of screw loosening failure and better functional outcomes compared to traditional trajectories[13,16]. Additionally, studies have shown that CBT provides superior biomechanical advantage compared to traditional techniques [17,18]. Trivergent was designed in an attempt to minimize these potential causes of rod failure and combines the CBT screws with facet trajectory screws, aiming to optimize outcomes.

In vivo testing will further support the use of Trivergent for obtaining better surgical outcomes. In clinical application, Trivergent is inserted through use of a “guide”. The plate attaches to the guide, which then becomes a functional unit. Using either fluoroscopy or image guidance, the guide unit is positioned over the disc space of the adjacent levels for the intended fusion. Using fluoroscopy as an example, Jamshidi needles are then directed through the guide/plate and aligned to traverse through the cephalad or caudal facet into the pedicle. K-wires are passed through the Jamshidi needles into the pedicles. Drilling, tapping and screw placement are then passed over the K-wires. After placing set screws over the K-wires the wires are removed and the set screws are firmly tighted down. Of course, image guidance eliminates the need for K-wires! Fig. 1 is a A/P and lateral schematic image of a correctly positioned bilateral Trivergent construct.

Fig. 1.

Schematic A/P and lateral image of appropriately placed bilateral trivergent stabilization system.

Biomechanical studies are conducted to understand how spinal fixation devices perform under applied loads, and they permit a comparative evaluation of such devices [19]. The American Society for Testing and Materials (ASTM) F1717 protocol has been used to evaluate spinal devices in many biomechanical investigations [[20], [21], [22]]. This in vitro protocol was applied to characterize the biomechanical properties of the novel Trivergent spinal system in comparison to a predicate device.

Methods

Construct assembly

Each bilateral Trivergent Spinal System construct consisted of (6) components: (1) right plate, (1) left plate, and (4) 5 × 45mm screws. Each predicate construct consisted of (14) components: (2) bent rods, (4) low top cannulated screws, (4) Minimally Invasive Spine Surgery (MIS) screws, and (4) hex set-screws. These components were assembled at CorMedical Ventures, Solana Beach, CA (part of Empirical Testing Corp [ETC]) for biomechanical laboratory testing. Biomechanical testing of the Trivergent and predicate constructs was conducted following ASTM F1717-18, a protocol describing various test methods using a vertebrectomy model to assess the biomechanical characteristics of spinal constructs. The assembled construct was placed in ultra-high molecular weight polyethylene (UHMWPE) test blocks for the static torsion and dynamic compression bending testing, or in stainless steel test blocks for the static compression bending testing. Examples of these constructs are shown in Fig. 2. The gap between these blocks simulates a vertebrectomy, considered the worst-case scenario [23]. These test blocks, when compared to cadaver-based models, control for variability in bone density of the vertebral bodies. This serves as a reproducible way to measure biomechanical properties of the fixation systems. The Trivergent constructs were available in 2 versions: 60° and 75°.

Fig. 2.

(A) Untested trivergent spinal system in stainless steel test blocks (for static compression bending test). (B) Untested trivergent spinal system in UHMWPE test blocks (for static torsion and dynamic compression bending tests). (C) Untested predicate system (for static torsion and dynamic compression bending tests). (D) Untested predicate system (for static compression bending test).

Deviations from ASTM F1717-18

For both Trivergent and predicate constructs ASTM F1717-18 states that the tolerance band for building and measuring X (width), Y (height), and L (length) are listed as X.X ± −0.1 mm. These dimensions could not be measured to this level of accuracy and are measured at ETC for reference only. For the Trivergent construct, the ASTM F1717-18 states that the L, X, and Y dimensions for a lumbar device should be 76 mm, 40 mm, and 20 mm, respectively. Due to the novelty of the Trivergent devices, the dimensions tested were 10 mm, 40 mm, and an average of 20 mm, respectively. For predicate construct, test constructs for the predicate implants were modified from the design in ASTM F1717-18 to approximate a single level implantation. To represent a single level of implantation, the longitudinal distance between the central axes of the implant screws was 38 ± 0.5 mm, which is a deviation from ASTM F1717, which recommend a longitudinal distance of 76 ± 0.5 mm. All testing was performed at ambient room temperature. The ASTM F1717-18 protocol outlined the following tests, which were completed with both the predicate and Trivergent system constructs.

Static compression bending test

(3) Predicate constructs, (5) Trivergent 60° constructs, and (5) Trivergent 75° constructs were tested in static compression bending. Outcome data for all constructs included: Linear fit offset (in millimeters [mm]), Compressive bending stiffness (Newtons per millimeter [N/mm]), Compressive bending ultimate displacement (mm), Compressive bending ultimate load (Newtons [N]), and Compressive bending ultimate moment (Newton meters [N-m]). The failure mode was also recorded. The static compression bending tests were performed in displacement control at a rate of 25 mm/minute. Load and displacement data were recorded by the test system software. The ramp waveform was performed until test block impingement occurred. Statistical analysis included calculation of the mean and standard deviation of each measured value.

Static torsion test

(3) Predicate constructs, (5) Trivergent 60° constructs, and (5) Trivergent 75° constructs were tested in static torsion. Outcome data for all constructs included, Linear fit offset (degrees), Elastic angular displacement (degrees), Angular displacement at 2% offset yield (degrees), Yield torque (N-m), and Torsional stiffness (N-m/degree). ASTM F1717-14 states to place aluminum blocks between the UHWMPE blocks and the base plates. Here, stainless steel anti-rotation blocks were used to prevent corrosion of these blocks.

The failure mode was also recorded. Stainless steel inferior and superior spacer blocks were used to prevent rotation of the test blocks. The static torsion tests were performed in angular displacement control at a rate of 60°/minute, with a static axial compression preload of 0 N held in load control for the duration of the test. Torque and angular displacement data were recorded by the test system software. The test was performed in the clockwise direction when looking down on the test setup. For the Trivergent constructs, the ramp waveform was performed until the load cell limit of 90 N-m was reached. For the predicate constructs, the ramp waveform was performed until test block interference occurred. Statistical analysis included calculation of the mean and standard deviation of each measured value.

Dynamic compression bending test

(3) Predicate constructs, (4) Trivergent 60° constructs, and (7) Trivergent 75° constructs were tested in dynamic compression bending with applied loads of (160 N, 200 N, and 230 N), (180 N, 200 N, 250 N, and 300 N), and (200 N, 230 N, 250 N, 270 N, 310 N, and 350 N), respectively. A cyclic load at a constant frequency of 5 Hz was applied to each specimen. Data was recorded by the test system software. The loads were maintained with a constant sinusoidal load amplitude control at a constant load ratio (R=minimum load/maximum load) equal to 10.

Testing was terminated when the specimen reached the endurance value of 5,000,000 cycles or failure defined as a fracture that could be seen under 10x magnification or a maximum displacement of 10 mm from the initial displacement at the minimum desired load level. The distance between the hinge pin centerlines was measured on the left and right sides of the constructs pre-test for all constructs and post-test for runout constructs only. The pre and post-test mass was measured for all constructs. The failure mode of each specimen and the corresponding cycle count were recorded.

Results

Static compression bending test

(5) 60° Trivergent Spinal System constructs, (5) 75° Trivergent constructs, and (3) predicate constructs were tested in static compression bending. Composite load-displacement curves for the Trivergent 75° construct and predicate construct are reported in Fig. 3. The Trivergent curve demonstrated consistently tolerated loads across all specimens, while the predicate constructs failed to tolerate loads beyond 300 to 400 N. For all constructs, the test stopped with test block interference. No observed failures occurred in any of the Trivergent 60° or 75° constructs prior to test block contact, as shown in photos comparing example post-test structures in Fig. 4 (only 60° construct shown).

Fig. 3.

Load-displacement curves produced in static compression bending testing. (N, newtons; mm, millimeters). (UPPER) Trivergent 75° composite of 5 constructs (756246-AC1 – 756246-AC5). (LOWER) Predicate composite of 3 constructs (800319-AC1 – 800319-AC3).

Fig. 4.

Static compression post-test comparison of predicate (LEFT) versus Trivergent 60° (RIGHT).

The failure mode for the predicate construct was slipping of the tulip about the pedicle screw, as shown in Fig. 4. Data for comparison of all constructs is reported in Table 1, statistical analysis included calculation of the mean and standard deviation of each measured value. The table data shows that, in measuring compressive bending ultimate load, the predicate construct, Trivergent 60° construct, and Trivergent 75° construct tolerated 400 ± 32 N, 480 ± 29 N, and 550 ± 24 N, respectively. In measuring compressive bending ultimate moment, the predicate construct, Trivergent 60° construct, and Trivergent 75° construct tolerated 18.6 ± 1.5 N-m, 26.5 ± 1.61 N-m, and 30.2 ± 1.33 N-m, respectively.

Table 1.

Static compression bending.

	Linear Fit offset (degrees)	Compressive bending stiffness (N/mm)	Compressive bending ultimate displacement (mm)	Compressive bending ultimate load (N)	Compressive bending ultimate moment (N-m)
Predicate construct
Mean	0.1	50	18	400	18.6
Std Dev	0.3	2	6	32	1.5
Trivergent (60° construct)
Mean	−0.3	42.6	11.3	480	26.5
Std Dev	0.67	1.01	0.48	29	1.61
Trivergent (75° construct)
Mean	−0.23	42	13	550	30.2
Std Dev	0.165	1	0.62	24	1.33

Static torsion test

(5) 60° Trivergent Spinal System constructs, (5) 75° Trivergent constructs, and (3) predicate constructs were tested in static torsion. Composite torque-angular displacement curves for the Trivergent 75° construct and predicate construct are reported in Fig. 5. The Trivergent curve demonstrates consistently tolerated loads across all specimens, while the predicate constructs failed to tolerate loads beyond 20.5 N-m. For all constructs, the test stopped at the machine capacity of 95 N-m. No observed failures occurred in any of the 60° or 75° constructs prior to machine capacity being reached, as shown in photos comparing example post-test structures in Fig. 6 (only 75° construct shown).

Fig. 5.

Torque-displacement curves produced in static torsion testing. (N-m, Newton-meters; Deg, degrees). (UPPER) Trivergent 75° composite of 5 constructs (756245-TR6 – 756245-TR10). (LOWER) Predicate composite of 3 constructs (800319-TR1 – 800319-TR3).

Fig. 6.

Static torsion post-test comparison of predicate (LEFT) versus Trivergent 75° (RIGHT).

The failure mode for the predicate construct was rotation of all screws in the test blocks, as shown in Fig. 6. Data for comparison of all constructs is reported in Table 2, statistical analysis included calculation of the mean and standard deviation of each measured value. The table data shows that, in measuring yield torque, the predicate construct, Trivergent 60° construct, and Trivergent 75° construct tolerated 20.5 ± 0.8 N-m, 44 ± 5.7 N-m, and 34 ± 9.2 N-m, respectively. In measuring torsional stiffness, the predicate construct, Trivergent 60° construct, and Trivergent 75° construct demonstrated 5.22 ± 0.15 N-m/deg, 14 ± 1.18 N-m/deg, and 13 ± 1.54 N-m/deg, respectively.

Table 2.

Static torsion.

	Linear fit offset (degrees)	Elastic angular displacement (degrees)	Angular displacement at 2% offset yield (degrees)	Yield torque (N-m)	Torsional stiffness (N-m/deg)
Predicate construct
Mean	−0.31	3.9	4.9	20.5	5.22
Std Dev	0.1	0.26	0.26	0.8	0.15
Trivergent (60° construct)
Mean	−0.17	3.2	3.5	44	14
Std Dev	0.091	0.71	0.71	5.7	1.18
Trivergent (75° construct)
Mean	−0.09	2.7	3	34	13
Std Dev	0.08	1.04	1.04	9.2	1.54

Dynamic compression bending test

(4) 60° Trivergent Spinal System constructs, (7) 75° Trivergent constructs, and (3) predicate constructs were tested in dynamic compression bending. Post-test images comparing an example predicate construct to an example Trivergent 60° construct are shown in Fig. 7. The predicate construct could not tolerate loads beyond 160 N. The example predicate construct, tested at an applied load of 230 N, was unable to tolerate the load and showed a partial fracture of the left rod at the inferior contact point of the superior set screw and rod. The Trivergent 60° and 75° constructs tolerated loads up to 300 N (maximum load tested) and 230 N, respectively. The example Trivergent construct, tested at 180 N, tolerated the load until runout without fractures.

Fig. 7.

Dynamic compression bending post-test comparison of predicate (LEFT) versus Trivergent 60° (RIGHT).

Discussion

For decades, pedicle screw failure has been a frustrating complication after placement of spinal hardware in patients with pathologies ranging from degenerative diseases to tumors and fractures [[2], [3], [4], [5],24]. These instrument problems are especially prevalent in osteoporotic patients who have reduced bone densities. Regardless of these potential issues, instrumentation has been shown to improve rates of vertebral arthrodesis [2,25]. Therefore, a goal in advancing spine surgery is to design novel devices which have improved longevity. This biomechanical study demonstrated that the Trivergent spinal fixation system has improved performance on static compressive bending, static torsion, and dynamic compression bending tests when compared to a predicate rod construct.

Static compressive bending

In vivo, axial forces acting on spinal fixation devices have been shown to be mainly compressive forces [26], highlighting the clinical importance of static compressive bending tests. These tests simulate the ultimate load and ultimate moment when a constant axial load is applied to the spine. The ultimate load is the maximum force the construct can withstand before deformities occur, while the ultimate moment is the moment which occurs at this maximum force. In this study, both the 60° and 75° Trivergent systems tolerated greater static compressive bending measures than the predicate construct.

Structural differences likely played a role in Trivergent’s increased compressive bending ultimate load. The Trivergent device uses fixed screw heads, in contrast to the predicate construct’s polyaxial screws with tulip-shaped heads. Polyaxial screw heads were designed to allow for more mobility and accommodation of rod placement when compared to fixed heads. A 2004 study by Stanford et al. showed that the more complicated design of polyaxial screws, although useful for surgeons, may have reduced strength and ability to tolerate higher biomechanical loads [27]. Authors used a contemporaneous version of the ASTM F1717 vertebrectomy model to show that polyaxial screws have lower static compression bending yield strength, consistent with our results [27].

Another study demonstrated 76.62% greater average yield strength and lower elastic displacement of fixed screw heads [28], also consistent with the Trivergent results. In the current study, the failure mode of the polyaxial screws in the predicate construct was defined as slipping of the tulip about the pedicle screw shaft, highlighting the propensity for the polyaxial tulip-head component to fail prior to the entire construct undergoing plastic bending. This is consistent with other biomechanical studies[28] and further supports the idea that the predicate’s polyaxial screw design significantly contributed to its weaker static compressive bending measures.

Static torsion

Static torsion testing assesses the yield torque, defined as the maximum rotational force applied to the construct before deformation occurs, and the torsional stiffness, or the ability of the construct to resist twisting. Here, the Trivergent system showed higher yield torque and torsional stiffness. Major contributors to this difference may have been Trivergent’s divergent screw trajectory, which was designed to mimic the cortical bone trajectory (CBT). The CBT is a newer surgical technique that aims to maximize screw contact with high density cortical bone. It has been shown that the most significant factor predicting the ability of a pedicle screw to resist loosening is the bone mineral density [29]. Pedicle screws derive 60% of their fixation strength from the cortical bone and 40% from the weaker cancellous bone and anterior cortex bone [30]; thus it is reasonable to expect that maximizing cortical bone purchase would improve construct stability.

The results here are consistent with an in vivo study which demonstrated that CBT pedicle screws had increased insertional torque compared to traditional screws [31]. Another consideration for static torsion is the biomaterial used in the implant. Two research studies found that titanium screws have higher maximum torque and angular stiffness than stainless steel screws, as measured by torsional testing [32,33]. Additionally, when compared to stainless steel, titanium-alloy has been shown to have higher bioactivity, flexibility, and bone ingrowth [34]. Considering that both the Trivergent and predicate screws were made of titanium-alloy, the screw CBT certainly played a larger role in the improved torque response of the Trivergent device.

Dynamic compression bending

The strength of a spinal construct’s bone-screw interface decreases over time as repetitive daily stress loading occurs. The dynamic compression bending test identifies a construct’s capacity to tolerate this stress. Ideally, a spinal construct will provide sufficient resistance to fatigue failure for at least the time it takes for bony arthrodesis of the vertebral bodies to occur. This is estimated to be approximately 4 months, translating to one million cycles under physiologic loads in the dynamic compression bending test [35]. Here, both the 60° and 75° Trivergent constructs successfully tolerated higher applied loads without fracture when compared to the predicate construct. This may be understood by comparing screw diameters. The Trivergent screw outer diameter was 5 mm, while the predicate construct used a 5.5 mm screw.

One study showed that larger diameter pedicle screws have increased fatigue load [36]. This is supported by the idea that larger diameter pedicle screws have increased purchase of the stronger cortical bone layers [37]. However, as demonstrated here, the smaller diameter Trivergent screw was able to tolerate higher fatigue loads. As discussed previously, the Trivergent device utilizes divergence and the CBT, increasing engagement with cortical bone. This trajectory, in combination with other structural device factors, likely explains the apparent ability of smaller diameter screws to tolerate higher loads. Additionally, other studies demonstrate that, although larger diameter screws may have greater initial pullout strength than smaller diameter screws, both screw sizes have similar stability after fatigue loading [38,39]. There are certainly multiple structural determinants of fatigue load, and further investigation would be required to determine the extent of contribution of each of these factors to this measure. It is also important to note that smaller diameter screws have a lower likelihood of complications such as breaching of the pedicle [40], illustrating that smaller screw size offers a clinical advantage beyond the fatigue testing results presented here.

Another design difference was the predicate’s rod-based system versus Trivergent’s plate-based system. The plate system involves fewer interfacing components than the screw-rod hardware [39], which may present less opportunity for failure mechanisms [41]. Stress on rod components is thought to be a common source of fatigue failure in spinal constructs; a previous design aimed to address this by increasing the contact area between the rod and set screw [42]. By eliminating the rod component, the Trivergent system was able to tolerate higher dynamic compressive loads without fatigue failure when compared to the predicate construct.

This comparison of rod versus plate structures has also been explored in cadaveric studies; Crawford et al. [39] compared the fatigue failures in hyperextension of a lumbar low-profile screw-plate construct and a cantilevered screw-rod construct, and found no differences after 10,000 cycles. However, this study’s methods differed from the present one as they applied 10,000 total cycles of alternating extension and flexion, followed by hyperextension failure testing. In comparison, fatigue testing in the current study applied up to 5,000,000 cycles of dynamic compression bending forces. As noted earlier in this discussion, it is estimated that the time for bony arthrodesis to occur is approximately 4 months, or 1,000,000 cycles under physiologic loads [35]. Therefore, the fatigue results in Crawford et al. may only estimate several weeks of postoperative wear, while the Trivergent results provide better insight into long term effects.

The biomechanical results in this study highlight the potential clinical utility of a new spinal fixation device. Trivergent performed well on the ASTM battery of tests, demonstrating that it may withstand the stresses encountered by the post-operative lumbar spine. This device’s optimized strength can be attributed to its plate-based design and screw trajectory, which could help it to minimize post-surgical construct complications.

Limitations

This study was limited in that the vertebrectomy block model only assesses the mechanical properties of spinal fixation systems and does not include factors related to the in-vivo bone-screw interface. This is a clinically relevant consideration, as implant loosening can occur as a result of microfractures at the bone-screw interface [43]. Additionally, the dynamic compression bending testing only focused on one mode of failure. Further fatigue testing focusing on other failure modes could provide more evidence of clinical viability. Biomechanical testing is a single component of spinal device evaluation; improved parameters do not always translate to clinical success. Thus, these results would be best complemented by cadaveric studies, finite element studies, and eventually, human clinical trials. The biomechanical success seen in this study establishes a precedent for further evaluation of a promising new spinal construct.

Conclusions

The Trivergent spinal fixation device is intended to address the issues of pedicle screw breakage and loosening, while maintaining a simpler design. The findings in this study support the idea that this new rigid and simple pedicle screw construct has biomechanical advantages over a more complicated multipart predicate construct, as measured by static compressive bending, static torsion, and dynamic compressive bending tests.

Funding

Funding for development of the Trivergent device was received from In Queue Innovations, LLC. Authors with In Queue Innovations contributed to the acquisition of data, analysis and interpretation of data, and administrative/technical/material support.

Declaration of competing interest

One or more of the authors declare financial or professional relationships on ICMJE-NASSJ disclosure forms.