We conducted a cross-sectional retrospective study of consecutive patients undergoing spinal surgery to evaluate the association between non-traumatic VFs and IVD degeneration adjacent to VFs. VFs were significantly associated with MRI grades of IVD degeneration and the presence of intradiscal VPs of discs adjacent to VFs. Logistic regression analysis showed that VF was one of the independent related factors for adjacent disc degeneration.
Eight clinical studies have evaluated the association between thoracolumbar VFs with spine trauma and adjacent disc degeneration (Table 4). Among these, two studies focused on degenerative IVDs adjacent to a VF in children or young patients who were treated conservatively [6, 20]. Kerttula et al. investigated the occurrence of disc degeneration by MRI in young patients (average-age: 15.5 years) with a history of wedge type VFs [6]. They concluded that wedge type VFs, especially with endplate injury, in young people were significantly associated with the occurrence of disc degeneration [6]. Later, Moller et al. [20] evaluated whether VFs in children (average-age: 12 years) are a risk factor for adjacent IVD degeneration. They used MRI and the Oner classification scheme [21], which mainly classifies morphological changes of IVDs and endplate injuries, and reported no significant association between stable VFs and adjacent disc degeneration [20]. Because of differences in MRI assessments and type of injuries (Table 4), the relationship between VFs and adjacent disc degeneration in young people and children remains controversial. The other six clinical studies had evaluated disc degeneration following thoracolumbar VFs (AO classification [22]: type A1-4) treated with spine surgeries (Table 4). Posterior pedicle screw fixation was performed in four studies [7-9, 23] and instrumented kyphoplasty in two studies [10, 11]. Among six clinical studies, adjacent disc degeneration was evaluated by MRI in five studies [7-10, 23], and by radiograph in one study [11]. These studies have shown that adjacent disc degeneration had significantly progressed at 9 to 32 months after thoracolumbar burst fractures in comparison with those at the time of injury, except for one study reported by Verlaan et al. [23]. Three studies [8, 10, 11] have reported that disc degeneration was predominantly found at the superior adjacent disc. On the other hand, Sander et al. [9] reported that the disc degradation was identified both at superior and inferior adjacent discs after traumatic VFs. Toyone et al. [7] also showed that adjacent disc degeneration had progressed at two years after burst fractures; however, the degeneration of superior and inferior adjacent discs were not separately analyzed. Lastly, Verlaan et al. [23] reported that 10.5% of the superior adjacent discs and 15.8% of inferior adjacent discs showed progression of degeneration at 12 to 18 months after trauma. However, they concluded that no statistically significant progression in adjacent disc degeneration was found for superior or inferior discs.
Table 4. Association between thoracolumbar vertebral fractures with spine trauma and the adjacent disc degeneration
Author
|
Year
|
Subjects
|
Age (averaged)
|
Design
|
Follow-up
|
Type of injury
|
Treatment
|
MRI Assessment
|
Kerttula6
|
2000
|
14
|
15.5 [8.8- 20.8]
|
Retrospective study
|
3.8 Y
|
Wedge-type (14)
|
Conservative
|
Decrease in T2 signal intensity
|
Moller20
|
2007
|
20
|
12 [7 - 16]
|
Observational cohort study
|
40 Y
|
Stable (18), Denis type B (2)
|
Conservative
|
Oner classification Scheme21
|
Verlaan23
|
2013
|
20
|
42 [18 - 74]
|
Prospective trial
|
12 to 18 M
|
AO: A3 (20), A4 (1)
|
PS fixation
|
Pfirrmann classification14
|
Toyone7
|
2013
|
12
|
38 [14 - 59]
|
Prospective consecutive series
|
10 Y
|
AO: A3 (12)
|
PS fixation
|
Borenstein's report (score: 0-3)37
|
Wang8
|
2013
|
26
|
39.6 [21 - 54]
|
Retrospective study
|
9-12 M
|
AO: A3 (26)
|
PS fixation
|
Pfirrmann classification14
|
Sander9
|
2014
|
27
|
37.5 [16 - 59]
|
Retrospective study
|
1 Y
|
AO: A1 (5), A2 (14), A3 (8)
|
PS fixation
|
Original classification (Grade 0-3)
|
Noriega10
|
2016
|
20
|
50.7 [45 - 56]
|
Retrospective study
|
32 M
|
AO: A1 (10), A3 (10)
|
IKP
|
Diffusion-weighted MR imaging
|
Descamps11
|
2019
|
93
|
54 [18 - 83]
|
Retrospective study
|
26.7 M
|
AO: A1 (54), A2 (5), A3 (34)
|
IKP
|
Radiograph
UCLA Grading Scale38
|
Y: years, M: months, PS fixation: Pedicle screw fixation, AO: AO classification19, IKP: instrumented kyphoplasty.
OVF: osteoporotic vertebral fracture
Numbers in the brackets indicate the range of age. Number in the parentheses indicates the number of subjects.
In contrast to these previous studies, the results of the current study showed a significant association between non-traumatic VFs and disc degeneration, at both superior and inferior adjacent levels, in a relatively older population.
Rahmani et al. have recently evaluated whether endplate fracture (injury) and adjacent disc degeneration have a significant association with the occurrence of delayed union following osteoporotic VFs for 139 consecutive patients (average age: 79 years-old) who were treated conservatively [12]. They also evaluated signal changes of adjacent IVDs in MR T2-weighted images at enrollment and at 6 months follow-up based on a modified Pfirrmann grading system and reported that adjacent cranial disc degeneration had significantly progressed at 6 months post-injury. This suggests the possibility that disc degeneration would progress in the relatively short term after osteoporotic VFs.
Next, to evaluate the effect of spinal levels on disc degeneration, the relationship between VFs and disc degeneration at different IVD levels was assessed (Fig. 5). Although there was no significant difference in the percentages of advanced stage of disc degeneration between VF-positive and VF-negative groups in L3/L4 and L4/L5, those percentages in T12/L1, L1/L2 and L2/L3 discs, which would be expected to have less degeneration than lower lumbar levels [24], were significantly higher in the VF-positive group than in the VF-negative group. This suggests that the effect of VFs on adjacent disc degeneration would be more pronounced at upper lumbar levels than those at middle/lower lumbar levels. There is also evidence to support this suggestion from cadaveric studies [25-29]. Dolan et al. [25] performed a mechanical and morphological study to evaluate how spinal level influences disc degeneration arising from endplate fracture. They reported that the effects of vertebral endplate fracture on disc mechanical function, and specifically on disc decompression, were greater at thoracic and upper lumbar levels than at lower lumbar levels.
In addition, we evaluated the association between VFs and the presence of adjacent intradiscal VPs. Intradiscal VPs refer to the radiographic appearance of a lucency caused by the presence of gas, usually found in the lumbar region [30, 31]; this is one of the characteristics of IVD degeneration [15, 32]. Murata and colleagues showed that the presence of intradiscal VPs is associated with the MRI-grade of disc degeneration and radiographic disc height narrowing [15]. The results of the current study showed that the incidence of VPs, especially the island type, in the VF-positive group was significantly higher than that in the VF-negative group; this suggests that VFs have an impact, not only on the extent of MRI-graded disc degeneration, but also on the intradiscal ruptures evaluated by CT imaging as intradiscal VPs.
The results of the analysis of intradiscal VPs by disc level were nearly identical to those of MRI-graded disc degeneration. Lafforgue and colleagues reported that VPs were grouped into collapse-related VPs and degenerative VPs [33]. They reported that collapse-related VPs, which were secondary to vertebral collapse, were located mainly in the thoracolumbar junction. Degenerative VPs, which were the result of disc degeneration, were located in lower lumbar discs. Therefore, in the current study, we speculate that intradiscal VPs in upper lumbar levels would be mainly attributed to VFs (vertebral collapse).
According to the results of logistic regression analyses, age, disc level, and adjacent vertebral fracture were independent related factors for disc degeneration and intradiscal VPs (Table 2). It is well known that being elderly and lower disc level were significant related factors for disc degeneration [24, 34, 35]. The current study showed evidence that VFs are also an independent related factor for adjacent disc degeneration for the population with non-traumatic VFs.
The following three patho-mechanisms are involved in the occurrence of adjacent disc degeneration. First, endplate and IVD injuries directly caused by VFs promote the progression of disc degeneration. Fujiwara and colleagues reported that endplate injuries were observed in 61%, and IVD lesions in 60% of patients with an acute, single osteoporotic VF [36]. Second, the progression of vertebral collapse is thought to cause impaired blood flow in vertebral bodies, to reduce blood flow and nutrient supply to the disc, and to cause disc degeneration [1, 5]. Imanishi and colleagues recently showed using a rabbit lumbar artery ligation model that ischemia of lumbar vertebrae initiated degenerative changes in IVDs [4]. Therefore, an ischemic vertebra is considered to be one of the important factors responsible for IVD degeneration. Third, mechanical stress is also involved in the progression of adjacent disc degeneration. Dolan et al. reported that vertebral endplate fracture reduced nucleus pressure and created abnormal stress distributions in the adjacent IVD, increasing the risk of internal disruption and degeneration [25]. Interestingly, Stefanakis et al. and Zehra et al. performed mechanical and morphological studies to determine whether high gradients of compressive stress within the IVD are associated with progressive disc degeneration [27, 29]. They reported that as the grade of disc degeneration increased, nucleus pressure decreased. However, stress gradients (concentration) in the annulus increased.
A limitation of this study is that most of the subjects were patients who had been given pre-operative radiographs, CT, and MRI for elective spinal surgeries. Therefore, MRI grading of IVDs and percentage of intradiscal VPs would be much higher than those within a general population [24]. Another limitation is that VFs in our study excluded those caused by high energy trauma. Although the evaluation of osteoporosis was not performed in this study, most VFs in the subjects of this study would be osteoporotic VFs. Thirdly, the other risk factors associated with VFs, such as obesity and physical activity, and the clinical outcome including the subject’s low back pain have not been evaluated in this study. Further study would be needed to evaluate the risk factors of adjacent disc degeneration following VFs, and the association of adjacent disc degeneration and clinical outcomes.