5.1 Coflex and Coflex-F in non-fusion surgery
The subject study found that (1) the Coflex device can provide stability of the surgical segment in most motions, except in flexion; (2) the rivets of the Coflex-F link bone and implant and can provide stability in all motions, especially in flexion; (3) in flexion, the disc stress distribution of the surgical segment is improved by the use of rivets; (4) in flexion, the Coflex-F influenced the adjacent segments; and (5) in extension, all implants influenced the adjacent segments.
In the subject study, the Coflex device in the defect model was found to provide stability in most motions, except in flexion. The instability of the Coflex device in flexion causes stress concentration at the anterior regions of the disc annulus (close to the superior and inferior sides of the endplate). Wilke et al. [59] suggested that the key for the Coflex device to provide stability in flexion is based on whether the teeth on the wings of the Coflex can provide sufficient anchorage to the spinous process. Two factors can improve this stabilization effect.
First, the surgeon must tighten the teeth on the wings against both edges of the spinous processes. Second, the bone density of the spinous processes should be strong enough to provide sufficient anchorage. However, both conditions are not always guaranteed.
For numerical analysis, the coefficient of friction in the interface between the implant and spinous processes was difficult to obtain. It is hypothesized that the teeth on the wings of the Coflex device will prevent implant slip motion in the spinous processes, and therefore a higher coefficient of friction (0.8) was used in the contact interface. In addition, this study also tested different coefficients of friction (0.4, 0.8, 1.2, and 1.6) to seek its influence on the effect of teeth on the wings of the Coflex device. The results show that the influence of the coefficient of friction is negligible.
The Coflex device was implanted between the interspinous processes located at the
80
posterior structure of the spine to resist instability in extension. By comparison with Tsai’s results in cadaveric experiments [57], our data show discrepancies in lateral bending and axial rotation. It is inferred that these were caused by individual differences among cadaveric specimens and differing experimental conditions. In the subject study, a partial L3-L4 interspinous process was removed to provide sufficient space for the implant, and the spinous process interface was modeled as a perfect contact and was able to transmit both tensile and compression forces. This assumption is different from the results of cadaveric experiments.
Kettler et al. [55] reported implantation of the Coflex-F can provide stability for all motions in lumbar spine stability. In the subject study, we also showed that the rivet connecting the metal wings and bony spinous process provides more security than the conventional Coflex device. Therefore, the rivet can improve the load transmission on the posterior spinal structure to decrease the stress concentration on the disc annulus at the surgical segment in all motions.
There are limited reports about implanting the Coflex device in the long lumbar segment model. The potential side effects in the adjacent segments need to be addressed. In 1-year outcome evaluation, Kong et al. [58] reported that the Coflex device reduced the ROM at the surgical segment but did not affect the ROM at the adjacent segments. The subject study, using a long lumbar spine segment model of an implanted Coflex device, showed that the ROMs are increased at both adjacent segments in extension but are unchanged in other motions. Therefore, the Coflex device increased annulus stress at both adjacent segments in extension. However, the Coflex-F constrained the surgical segment in all motions and it increased ROM at adjacent segments, especially in flexion. Therefore, the Coflex-F increased annulus stress at both adjacent segments in flexion and extension. The Coflex-F and pedicle screw fixation have the same effect on the adjacent segments in both flexion and extension. In addition, the remote adjacent L1-L2 segment and adjacent L2-L3 segment demonstrate the same effect in all forms of implantations.
5.2 Coflex-F in fusion surgery
The subject study found that (1) The Coflex-F device combined with ALIF can provide stability similar to a pedicle screw fixation in combination with TLIF or ALIF. (2) larger stress at the cage-bone interface for the Coflex-F combined with TLIF, thus causing the exclusion of the pedicle screw fixation.
The present study used an FE lumbar model of the L1-L5 segments to compare the effects of the Coflex-F device and traditional bilateral pedicle screw fixation at the surgical segment after TLIF and ALIF implantation. According to the ROM results, the Coflex-F device combined with the TLIF model had lower stability than all the other models, especially in both directions of lateral bending and axial rotation. On the other hand, the pedicle screw fixation combined with the ALIF showed the highest stability among all model.
The primary factor in the Coflex-F results is the fixed position of its implantation. The motion segment, composed of two adjacent vertebrae and the associated soft tissues, is the functional unit of the spine. Each motion segment has three joints. It has a triangular stack of articulations, with symphysis joints between vertebral bodies on the anterior side and two sliding facet joints on the posterior side. The Coflex-F has rivets joining its wings to the spinous processes. The rivets can attach the implant more rigidly to the posterior spinous processes. However, the vertebral bodies of anterior side sustain the majority of the weight.
Therefore, the rivets cannot provide sufficient stiffness in the motion segment for two adjacent vertebrae because the locations of attachment are within the posterior element, which is not as strong as vertebral bodies. However, pedicle screw fixation can fix vertebral bodies, and therefore provide sufficient stiffness in the motion segment for two adjacent vertebrae.
The geometry of the Coflex-F devise supports a different function—its U-shaped structure retains the same design and flexibility of the Coflex, thus making it more flexible and deformable than pedicle screw fixation. Figure 4.22 shows the von-Mises stress of the
82
Coflex-F and the pedicle screw for various loading cases. In all of these cases, the Coflex-F has higher stress than the pedicle screw when combined with TLIF or ALIF. Figure 4.23 showed the von-Mises stress of the Coflex-F and the pedicle screw for various loading cases.
The Coflex-F has higher von-Mises stress than the pedicle screw. Therefore, the fixed position and geometry of implantation have a great influence on stress distribution for Coflex-F
ALIF and TLIF are two common surgeries for achieving interbody arthrodesis. In the present study, a posterior instrumentation in combination with ALIF can provide higher stability than a posterior instrumentation in combination with TLIF. The ALIF procedure with anterior surgical approach allows expansion of disc space; it can use a larger cage to increase the contact area of cage-bone interface. The larger contact area distributes the load over the cage-bone interface area of the vertebra bone. Consequently, an ALIF cage does not create stress concentration on the cage-bone interface at the surgical segment. On the other hand, the TLIF procedure prohibited the use of a large cage, because a cage pathway would create limitations for the surgery. The TLIF procedure can only utilize cages with long and thin contact area on the cage-bone interface. Therefore, the TLIF cage surfers from stress concentration on the cage-bone interface at the surgical segment.
In extension, the stress concentration of all the models diminished between the cage-bone interfaces. In flexion, the stress concentration of all the models increased at the anterior side of the cage-bone interface. A posterior instrumentation combined with ALIF has higher stress concentration than a posterior instrumentation combined with TLIF. This is primarily due to flexion or extension motion. The posterior instrumentation and interbody cage share the same extension motion. Posterior instrumentation sustains most of the load transferred in extension, therefore reducing the stress concentration of all the models. In contrast, the anterior interbody cage sustains most of the load transferred in flexion, therefore resulting in the stress concentration in the ALIF model at the cage-bone interface, especially with Coflex-F implantation. The Coflex-F sustains larger moment than pedicle screw fixation
because the fixed position of the Coflex-F in the posterior interspinous processes causes a longer moment arm.
PEEK material has recently gained popularity for use in implants because of its mechanical properties. One of the PEEK material’s biggest advantages is its modulus of elasticity (E = 3.5 GPa) which is closer to cortical bone (E = 12 GPa) and cancellous bone(E
=0.14 GPa) compared to that of titanium (E = 113 GPa). Vadapallis [106] performed a finite element investigation to study the effect of different spacer material property. The results from that study indicate that PEEK spacers provide initial stability similar to titanium spacers, and therefore might minimize the chances of subsidence. The present study uses two cage materials: titanium for the ALIF cage and PEEK for the TLIF cage for stability. This study’s results are identical to Vadapallis's results, i.e. both cage materials provide similar stability when combined with pedicle screw fixation. However, the materials of these two cages do not provide similar stability when combined with the Coflex-F, the titanium cage (ALIF) provide higher stability then PEEK cage (TLIF).
5.3 Coflex-F in minimally invasive fusion surgery
The subject study found that (1) the TLIF combine with Coflex-F cannot provide sufficient stability of the surgical segment in lateral bending and axial rotation; (2) the TLIF combined with unilateral pedicle screw fixation and translaminar facet screw fixation can provide sufficient stability of the surgical segment all motion as TLIF combined with bilateral pedicle screw fixation.
Several lumbar interbody fusion methods have been used for degenerative disc diseases and instabilities via various approaches, such as ALIF, PLIF, and TLIF. The TLIF was involves removal of one facet joint and a lateral approach to the disc space, thus reducing the potential for nerve injury. The TLIF surgical construct requires less bone and soft tissue dissection, respects neural elements, laminar bone, facet and pars interarticularis on the
84
contralateral side for additional posterolateral fusion, and avoids the morbidity of ALIF and PLIF approaches.
Another minimally invasive fusion device, an interspinous process device Coflex-F may be used instead. This device requires only a minimal incision and disruption of the interspinous ligaments to insert one part of the device with a post which goes through the interspinous space.
Results showed that TLIF combined with Coflex-F provide lower stability than TLIF combined with unilateral pedicle screw and translaminar facet screw fixation. The primary factor is fixed position of Coflex-F implantation and Coflex-F structure as described previously 5.2 sections. Therefore, the TLIF combine with Coflex-F cannot provide sufficient stability of the surgical segment in lateral bending and axial rotation.
In this subject study, ROM results showed the fixation following TLIF of unilateral pedicle screws with a supplemental translaminar facet screw fixation showed no difference in stiffness to that of the standard bilateral pedicle screw fixation. The advantages of surgical procedure for unilateral pedicle screw and supplemental translaminar facet screw fixation were significantly reduced iatrogenic trauma and reduced surgical risks than bilateral pedicle screws fixation.
5.4 limitations
Several limitations in these studies are related to the simplified and idealized material properties during simulation, such as the linearized behavior of the spinal ligaments and pure elastic intact discs without degeneration [93][94]. A degenerative disc is common in many patients before surgery. The various grades of degeneration in the disc, such as delamination, dehydration or reduced disc height, do not allow for exact replication of the unique material properties of a degenerated disc. Therefore, normal material properties were used in this simulation.
In a real spine, the size of vertebrae and the orientation of the facet joint are different depending on each segment. The influence of geometry was not considered here, which might affect the absolute values of the vertebral stresses and facet joint loads.
The degree of gripping force applied between the wings of the Coflex-F device and the spinous process is determined by the clamping force that is applied by the surgeon, which is difficult to measure, and there have been different results presented in previous studies [55][57][59]. In addition, determination of gripping force must also consider bone strength and geometry of the spinous process. In this study, the degree of the gripping force was simplified and only considered the friction conditions between the teeth on the wings of the Coflex-F device and the spinous process. The coefficient of friction used here was based on the results of a previous study into friction parameters between the cage and the bone [93]. In addition, our simplified simulation of gripping force ignored the pre-force between the teeth of the wings and the spinous processes, as well as the inward and outward deformation of both side flanks of the Coflex-F device. Also, the constrained behavior used in the bone-screw interface, the thread of the pedicle screw, and the bone ingrowth into the cage were simplified.
Pretension should occur after inserting the device, which might distract the remaining annulus, reducing the ROM and facet loading at the implant level. This mechanism was not modeled here.
The loading conditions in the present FE simulations were similar to those of the traditional in vitro tests. Thus, muscle contraction and pelvic movement were not included in the present study. Furthermore, FE models should be interpreted only as a trend because of the variability among different human tissues.
86