Facet joint, as a crucial element of the three-joint complex in the posterior aspect of the vertebrae column, provides a far-reaching influence on the spine, and plays significant roles in guiding spinal motion and transmitting axial loading [27]. Biomechanical models of isolated spinal segments have confirmed that the load carried by facets can more than 33% of total load borne by the spinal segment, and will be more or less transmitted through facets joints during spinal extending or flexing[28]. Meanwhile, these paired diarthrodial joints transmit shear forces and help the intervertebral discs carry out 16% of the vertical load [29, 30]. Some previous studies have demonstrated that any abnormality of one facet joint could affect others and cause asymmetric stress transmission to both disc and facet, ultimately resulting in abnormal stress concentrations in regions of disc and facet joints [31, 32]. Owing to complicated geometry, researchers considered FO and FT as anatomical parameters to represent the geometric characteristics of facet joints. Some prior studies have found out that there was a relationship between FO and FT and lumbar degenerative diseases, such as lumbar disc herniation, lumbar stenosis and low back pain [22, 33, 34]. Distinct from the lumbar spine, facet tropism in cervical region might be more common in sub-axial cervical spine [35]. Additionally, the lumbar facet articular surface area is relatively smaller compared with vertebral end plate, but in the cervical region the facet surface area is about two-thirds the size of the area of the cervical disc, which means that the biomechanical and pathological role of facet joints in cervical degenerative diseases might be more dominant than that in lumbar.
Although FT has been considered common phenomenon in the sub-axial cervical spine [35], it is still unclear whether there is an association between facet parameters and cervical degenerative diseases, even specific effects. As far as we know, this is the first investigation to analyze the relationship between FO and FT and CSR based on more accurate and practical measuring methods on 3-dimentional views. In this work, middle-aged participants (aged from 30–65 years) were included because various conflicting factors might interfere final results in elderly individuals [26, 36, 37].
In this study, we considered FO, FT and incidence of FT as potentially risk factors to analyze the relationship between these facet parameters and CSR. First of all, differences in several characteristics between CSR and control group were found after inter-group comparison and these statistically different parameters were regarded as confounding factors in the following logistic analysis. Then, bilateral FO, FT and FT (+) on axial, sagittal and coronal planes in CSR group were compared with those in control group at C3/4, C4/5, C5/6 levels, respectively. In terms of FO, there was no statistical significance in the difference of left or right FO on 3-dimentional views between groups except C5/6 axial view (left: P = 0.000, right: P = 0.003), this result was in harmony with the conclusion of a biomechanical visualization study in lumbar spine [23]. In aspect of FT, axial FT of CSR group was larger and shown significant difference between groups at each spinal level (C3/4:P = 0.000, C4/5:P = 0.000, C5/6:P = 0.000), whilst sagittal and coronal FT shown not. However, sagittal FT at C5/6 also revealed significant difference between groups (P = 0.004). To identify which is more influential, further comparison between axial and sagittal FT in C5/6 CSR patients were performed. The result suggested that the degree of axial FT was substantially larger and significantly different from sagittal FT (P = 0.000). In addition, we also compared FT between axial or sagittal view in control group and obtained identical outcome (P = 0.000), meaning that the degree of axial FT was larger in both CSR patients and normal individuals. Thus, as a more common pre-existing anatomical feature, axial FT might play a more important role in pathogenesis of CSR between axial and sagittal FT. With regard to FT (+), the result was similar to that of FT. Axial and sagittal FT (+) were both statistically different between CSR and control group, whereas the number of axial FT (+) was more in CSR patients or normal humans, pointing out that as with axial FT, axial FT (+) might exert a more dominant function on the development of CSR.
Next, the results of univariate and multivariate logistic analysis demonstrated that axial FT (+) might be the risk factor of CSR at C3/4; axial FT (+) and sagittal FT (+) might be the risk factors of CSR at C4/5; axial FT (+), sagittal FT (+) and axial FT might be the risk factors of CSR at C5/6, adjusting for other potentially confounding factors such as sex, hypertension and smoking. Notwithstanding, sagittal FT (+) might be considered as a predisposing factor but its incidence was not equal to that of axial FT (+). Thus, axial FT (+) should be pay more attention to prognostication of CSR in comparison to sagittal FT (+). Of note, FT was also an influential factor in CSR at C5/6 but not at C3/4 and C4/5. FT represented asymmetry between left and right, while FT (+) delegated that asymmetry was equal to or more than 7. As a result, if only FT is larger to a certain extent, FT might have an effect on occurrence of CSR. Therefore, axial FT (+) might show stronger correlation on CSR.
Finally, the side of disc herniation was associated with the side of larger axial FO, meaning that the direction of herniation preferred the side of greater FO. It has been demonstrated in a previous cervical study that the greater facet angle at the left or right side did not affect the side of herniation in patients with cervical disc herniation [25]. This could be explained by the different experimental design, study population and methodology. In our study, we selected cervical spondylotic radiculopathy, in which not only disc herniation but other secondary changes of cervical discs such as osteophyte stimulated the nerve roots. And secondary changes of cervical discs might carry out considerable or even more abnormal load in the development of CSR. A prior study [23] has revealed biomechanical mechanism between FT and side of lumbar disc herniation, thus our finding should be further investigated and confirmed by biomechanical study of cervical spine. Also, it has been reported that there was an association between lumbar FT and disc degeneration [9, 32, 38]. Still, more longitudinal studies were needed to verify the relationship among FT, cervical disc degeneration and CSR, and their intrinsic and specific mechanism.
Biomechanically, the interplay between facet joints and cervical disc may be similar to lumbar but not the same. A rat model study reported that participating factors to disc herniation were not only axial compression but flexion and torsion[39]. It is commonly held that cervical facet joints are 3-dimentional, including that axial cervical spine is responsible for rotation movement, sagittal cervical spine is responsible for flexion-extension, and coronal cervical spine is responsible for lateral bending [24, 25]. Previous biomechanical findings proposed that the biomechanical effects of FT rather than FO contributed to the increases in disc stress and facet joint contact at the corresponding segment [32, 40]. Consequently, FT should be the first topic of interest in clinical practice of predicting cervical degenerative diseases. Ke et al [23] suggested that under flexion, extension and torsion movement, disc stress and facet joint contact force increased with increasing axial FT in the lumbar biomechanical modelling. Increased axial FT caused more severe stress asymmetry and held a stronger effect on disc stress, eventually aggravating the disc degeneration and causing clinical degenerative diseases. In particular, the increasing trend was most obvious for the torsion moment [23]. Rong et al [41] indicated that under flexion, extension, lateral bending and axial rotation moments, sagittal FT increased disc pressure and facet contact force. The increased loading on the intervertebral disc could cause micro-injury to the disc. The accumulation effect of these repetitive micro-injury could give rise to initiating or accelerating the degenerative process accordingly. However, the incidence of sagittal FT is not common as axial FT in clinical practice. As far as we known, there was no study exploring the biomechanical role of coronal FT in cervical spine, which is imperative to implement. Ke et al [23] found that when FT existed, stress was concentrated on the disc region ipsilateral to the facet joint with smaller axial FO, which means the side of disc herniation were supposed to prefer the side of smaller axial FO. This result was not in line with our finding. This may be the fact that cervical loading transmitting is different from lumbar. More specifically, axial FO of cervical spine would be near or more than 90°, while axial FO of lumbar spine might be much smaller, generally being an acute angle. This might result in distinct biomechanical mechanism in the prognosis of cervical diseases. Moreover, movement pattern of cervical spine could be more complicated than that of lumbar. Further both numerous clinical and biomechanical studies of cervical spine should be performed. The finite element analyses in both cervical and lumbar spine concluded that FT might be anatomical risk factor for inducing disc degeneration [23, 41]. Although the specific mechanism of facet parameters inducing CSR was not the scope of this study, we hypothesized that axial FT (+) might deteriorate cervical disc degeneration and then induce CSR. There exist two explanations: 1) In cervical region, axial rotation was performed more frequently than flexion-extension, thus producing more load on axial facet surfaces. 2) As is reported in the prior study, the increasing trend of disc stress was most obvious for the torsion moment, contributing to more far-reaching cumulative effects on disc degeneration with repetitive rotation. Equally, sufficient evidences are imperative to confirm these hypotheses.
Several limitations existed in the present study. Firstly, the number of cases was relatively small. Further multicenter studies with larger sample sizes are necessary to carry out. Secondly, the findings of this studies were based on the CSR patients undergone clinical intervention, which might represent greater FT or more FT (+) in CSR population. Thirdly, given its nature as a retrospective study, CT and MRI images were not completely controlled by protocols, which may lead to bias. Besides, although the measuring methods in our studies were more comprehensive and is not influenced by cervical curvature or head position, there still existed measurement bias. Finally, this study only revealed superficially positive correlation between FT (+) and cervical disc degeneration, further studies should focus on inherent biomechanical effects and more specific association among FT (+), cervical disc degeneration and CSR.