Is Traction Spur, Modic Changes And Vacuum Phenomenon Really An Indicator of Segmental Instability In The Lumbar Spine?

A thorough understanding of lumbar segmental motion is valuable to treat patients with degenerative lumbar disease, but kinematics associated with indicators of lumbar intervertebral instability (traction spur, Modic changes (MCs) and vacuum phenomenon (VP)) in the lumbar spine have not been well understood. The purpose of this study is to evaluate the relationships between traction spur, MCs and VP to radiographic measurements in the lumbar spine. A total of 525 lumbar discs from L1-2 to L5-S1 of 105 patients were evaluated. The sagittal translation (ST) and sagittal angulation (SA) were measured from the radiographs taken exion-extension. The anterior disc height (ADH) was measured from the lateral radiographs, and AADH was measured as the difference from supine to sitting position. Logistic regression analyses were used to detect the association between the existence of traction spur, MCs and VP and related factors. Multivariate analysis showed that the traction spur was signicantly related to translational motion (ST>2mm, OR 4.74) and the VP were signicantly related to vertical motion ( ADH >3mm, OR 1.94). These results suggest that the segments with traction spur and VP should be evaluated carefully because these may be a sign of lumbar intervertebral instability. To evaluate the relationships between traction spur, VP, and MCs to radiographic measurements in the lumbar spine, we performed a cross-sectional retrospective radiographic, CT and MRI study to assess discs (intervertebral space) of consecutive patients who underwent spinal surgery. To the best of our knowledge, this study is the rst to evaluate various aspects of indicators of vertebral instability (traction spur, VP and MCs) and radiographic instability in lumbar spine. The results showed that traction spur was signicantly related to translational motion (ST>2mm), VP were signicantly related to vertical motion ( Δ ADH>3mm), however SA was not signicantly related any indicators of vertebral instability (traction spur, VP and MCs). These ndings may be helpful in the indications for lumbar fusion in clinical practice.


Introduction
Spinal instability in the lumbar spine is considered to be a signi cant factor in lower back pain and indicates lumbar fusion surgery. Functional exion-extension radiography is the most widely used method in the imaging diagnosis of lumbar intervertebral instability. 1 Traction spur formation is located 2 or 3 mm from the endplate and has a horizontal orientation. Macnab et al. described traction spur as indicators of intervertebral instability on plane radiographic image and emphasized the characteristics of traction spurs and their associations with unstable lumbar discovertebral junctions and excessive or abnormal spinal mobility. 4 Yadav et al. reported on the importance of traction spurs, stating that traction spurs constituted the useful indicators of vertebral segment instability. 5 The intervertebral disc vacuum phenomenon (VP) refers to the radiographic appearance of uency caused by the presence of gas, usually in lumbar region; this is one of the characteristics of disc degeneration. 6, 7 Liao et al. reported that intradiscal VP should be regarded as a sign of intervertebral instability. 8 Modic changes (MCs) are bone marrow and endplate change visible on MRI of patients with degenerative disc disease. 9 Hayashi et al. suggests that MCs might play a role in the stability of lumbar spine. 10 A thorough understanding of lumbar segmental motion is valuable to treat patients with degenerative lumbar disease, but the relationship between functional radiography and these instability indicators of vertebral instability (traction spur, VP and MCs) is not clear. The current study evaluates the relationship between the segmental instability and degenerative ndings detected by radiography, CT and MRI. The purpose in this study is to determine the relationships between traction spur, VP, and Modic changes to radiographic measurements in the lumbar spine.

Materials And Methods
The inclusion criteria were de ned as patients who had undergone lumbar spine surgery at a single institution from August 2014 and March 2016. The exclusion criteria were trauma (n = 3), infection (n = 19), spinal tumors (n = 8), history of lumbar surgery (n = 16), osteoporotic vertebral fracture (n = 18), and unsuitable radiographs to measure (n = 15). Of 184 patients registered during the study entry period, 105 patients of them were completed in this study (60 men and 45 women, average age of 68.0 ± 12.8 years (range 29-89)). The diseases responsible for enrollment of patients in the study included the following: lumber spinal stenosis in 72 cases, herniated lumbar disc in 15 cases, degenerative spondylolisthesis in 9 cases, and others in 9 cases. A total of 525 lumbar discs from L1-2 to L5-S1 were evaluated for all patients. The Institutional Review Board of the University of Saga at Saga city approved this study and informed consent was obtained from all participants (2020-04-R-10). This study also adhered to the principles of the Declaration of Helsinki. Written informed consent was obtained from all patients.

Radiographic studies
Plane CT was evaluated with regard to the presence of osteophyte. An anterior lumbar vertebral osteophyte should be > 2 mm or more in length according to the classi cation of Macnab et al. 4 Kasai et al. distributed anterior lumbar vertebral osteophytes into six types based on the direction of extension of each pair of osteophytes across the intervertebral disc space as follows: group A, no osteophytes; group B, the pair of osteophytes extended in the direction of the adjacent disc; group C, there was almost complete bone bridge formation by a pair of osteophytes across the intervertebral disc space; group D, the pair of osteophytes extended in a direction away from the adjacent disc; group E, the osteophytes extended nearly horizontally to the vertebral body border without closing the intervertebral disc space; and group F, ungroupable. 11 In this study, anterior lumbar vertebral osteophytes were distributed into three types based Kasai's classi cation: no osteophytes, include group A; claw spur, include group B and C; traction spur, include group D and E. For each level, a diagnosis was made of no osteophyte, claw spur, and traction spur. When two different osteophytes were present in one segment, only one diagnosis was applied ( rst priority: traction spur; and second priority: claw spur). MCs were classi ed into none or types 1, 2, and 3, according to their signal patterns on T1-and T2-weighted sagittal MR images. 9 VP were evaluated by presence of areas of gaseous radiolucency using CT imaging. The presence of VP was judged as present or not present.
Radiographic parameters, including anterior disc height (ADH), intervertebral slip angle, and distance of slippage, were collected. We measured these parameters using Virtual Place RAIJIN Ver3.8(AZE Ltd., Tokyo, Japan). ADH was measured as the distance between the most anterior point of the upper and lower endplates. Δ ADH was measured as the difference from supine to sitting position. (Figure 1) The distance of slippage was de ned by drawing two lines perpendicularly to the line superior endplate of lower vertebral body. The rst line was sited at the posterosuperior corner of the caudal vertebra, and the second line was dropped from the posteroinferior corner of cranial vertebra. The distance between these two parallel lines was the distance of slippage. The amount of sagittal translation (ST) was obtained as the difference of the distance of slippage between exion and extension. (Figure 2) The intervertebral slip angle was found simply by constructing two lines at the endplates of the affected disc and measuring the angle between them. The sagittal angulation (SA) was also measured as the difference of intervertebral slip angle from extension to exion. (Figure 2) Detailed measurement methods were described previously elsewhere. 12 In this study, excessive motion (instability) was de ned as following; ΔADH > 3mm, ST > 2mm and SA > 10 degree. 12 We measured the following characteristics twice; radiographic parameter (ΔADH, ST and SA), the type of spurs, MCs and VP. When there was a difference, the rst measurement was used.

Statistical analysis
All statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). More precisely, it is a modi ed version of R commander designed to add statistical functions frequently used in biostatistics. 13 Statistically signi cant differences in radiographic parameter (ΔADH, ST and SA) between the type of spurs, MCs were assessed using one-way analysis of variance (ANOVA) followed by post hoc multiple comparisons using the Bonferroni method. T-test was used to evaluate the VP and radiographic parameters. After the variables were categorized, multivariate logistic regression analyses were used to compute odds ratios (ORs) and 95% con dence intervals (CIs) and detect the association between the existence of radiographic parameter (ΔADH, ST and SA) and instability related factors (spur type, MCs and VP). The variables in the multivariate model were that were sex, age, spur type, MCs and VP. The level of signi cance (p value) was set at 0.05.
Inter-observer agreement between the two board certi cated spine surgeons of JSSR (The Japanese Society for Spine Surgery and Related Research Reserved), and intra-observer agreement by one reader were analyzed in 30 discs using kappa statistics. The intra-and inter-observer agreement for the type of spur, MCs and VP were analyzed.
Prevalence of traction spur, odic changes and vacuum phenomenon The traction spurs were found 219 discs (41.7%), MCs were found in 106 discs (20.6%) and VP was found in 179/525 discs (35.4%) of the 525 discs analyzed ( Table 1). The results of the radiographic measurements are shown in Table 2.
Relationship between traction spur, odic changes, vacuum phenomenon and vertical motion (ADH) There was signi cant difference between ADH and spur type. ADH was signi cantly higher in traction spur than the others (p<0.001). There was signi cant difference between ADH and MCs. ADH was signi cantly higher in Modic type3 than Modic type0 and type2 (p<0.001, p=0.036). ADH was signi cantly higher in VP. (Table 3) Factors related to ADH >3mm were evaluated using logistic regression analysis to control confounding factors. Using crude analysis between ADH >3mm and <3mm, signi cant differences were found in traction spur (p=0.013) and VP (p<0.001). Next, a multiple logistic regression model was used to adjust for age, sex, spur type, MCs and VP. After adjustment for potential confounding factors, signi cantly elevated adjusted odds ratio (AOR) was observed in segments with VP (AOR 1.94, 95% CI 1.13-3.33, p=0.0164) ( Table 4).

Relationship between traction spur, odic changes, vacuum phenomenon and sagittal translation (ST)
There was a signi cant difference between ST and spur type. The ST was signi cantly higher in traction spur than the others (p<0.001). There was no signi cant difference between ST and MCs or VP. (Table 3) Using crude analysis between ST>2mm and <2mm, signi cant differences were found in spur type (p=0.002) ( Table 5). Next, a multiple logistic regression model was used, signi cantly elevated ORs were observed in segments with traction spur (AOR 4.74, 95% CI 1.79-12.6, p=0.00174) ( Table 5).
Relationship between traction spur, Modic changes, vacuum phenomenon and sagittal angulation (SA) There was no signi cant difference between SA and spur type, and MCs. The disc with VP was signi cantly smaller than disc without VP (Table 3).

Discussion
To evaluate the relationships between traction spur, VP, and MCs to radiographic measurements in the lumbar spine, we performed a cross-sectional retrospective radiographic, CT and MRI study to assess discs (intervertebral space) of consecutive patients who underwent spinal surgery. To the best of our knowledge, this study is the rst to evaluate various aspects of indicators of vertebral instability (traction spur, VP and MCs) and radiographic instability in lumbar spine. The results showed that traction spur was signi cantly related to translational motion (ST>2mm), VP were signi cantly related to vertical motion (ΔADH>3mm), however SA was not signi cantly related any indicators of vertebral instability (traction spur, VP and MCs). These ndings may be helpful in the indications for lumbar fusion in clinical practice.

The association between traction spurs and radiographic measurements
Our multivariate analysis showed traction spur were signi cantly related to ST>2mm. However, we did not nd a signi cant association between traction spur to ΔADH>3mm and SA. Disc degeneration with disc height narrowing is considered related to spinal instability. 14 On the other hand, when disc height is collapsed, there is a natural tendency to restabilize the motion segment and as a result, spondylolisthesis becomes less likely to progress. 15 Murata et al. found that severe disc degeneration was less signi cantly related to angular displacement and had a tendency to stabilize the motion segment. 16 The values of 10°f or sagittal angulation and 3 or 4 mm for sagittal translation are typically used to infer instability. 12,17 In this study, because traction spur associate with >2mm ST, traction spur indicates intervertebral instability. We speculate that the signi cant association at ST>2mm, but not >3mm or >4mm, is because traction spur occurs relatively early in disc degeneration.

2.The association between vacuum phenomenon and radiographic measurements
Our multivariate analysis showed VP were signi cantly related to ΔADH>3mm and SA>10 degree. Because instability may create excessive intervertebral distraction and subsequent negative intradiscal pressure, allowing interstitial nitrogen in the surrounding tissues to become gaseous and to accumulate within cleft of the degenerated disc, it is assumed that the vacuum phenomenon is often associated with vertebral instability. 1 The changes in disc height between supine and sitting positions probably represent two factors: disc elasticity and instability due to disc degeneration. Because of the few younger patients in this study, intervertebral discs with ΔADH>3mm mm may indicate instability due to disc degeneration rather than elasticity. The VP was one of the indicators of early to moderate disc degeneration and intervertebral instability, since VP was associated with ΔADH>3mm and SA>10 degree. Clinically, Liao et al. reported that the vacuum sign at the spondylolisthesis segment should be regarded as another sign of instability and suggested that instrumented posterolateral fusion simultaneous with intervertebral fusion with a cage can overcome this situation. 8 Therefore, the traction spur is a sign of instability as well as VP and is a useful indicator in deciding on the adaptation for spinal instrumentation.

3.The association between Modic changes and radiographic measurements
Kirkaldy-Willis and Farfan postulated 3 stages with different conditions of stability and motion in the degenerative lumbar spine: dysfunction, instability, and stabilization. 1 Because instability may create excessive intervertebral distraction and subsequent negative intradiscal pressure, allowing interstitial nitrogen in the surrounding tissues to become gaseous and to accumulate within cleft of the degenerated disc, it is assumed that the vacuum phenomenon is often associated with vertebral instability. Moderate disc degeneration with mild disc space narrowing and osteosclerosis of endplate also have been associated with vertebral instability. Modic changes (MCs) are bone marrow and endplate change visible on MRI of patients with degenerative disc disease. 9 Some previous studies demonstrated the relationship between fusion surgery and MCs and have reported on the instability related to MCs. 18,19,20,21 However our multivariate analysis showed MCs were not signi cantly related to any radiographic parameters. The results suggested that the changes in the endplate are not directly related to intervertebral instability.

4.Limitation
The present study was associated with several limitations. Firstly, the subjects were patients who had undergone preoperative radiographic imaging before elective spinal surgery, and an age-matched general population was not used as a control. Thus, the incidence of traction spurs in our subjects would be much higher in comparison to age-matched individuals in the general population. Secondly, this was a cross-sectional study, which cannot clearly explain the prognosis of developing disc degeneration or spinal instability in the presence of traction spurs at each lumbar spinal segment. We are also planning to investigate the direction of spur formation in a longitudinal study. Another limitation of this study is the relatively small number of lumbar segments with >2 mm sagittal translation and >10 degree sagittal angulation, which might have in uenced the statistical power. Further studies are needed to increase the number of these cases.

Conclusion
Our multivariate analysis showed traction spurs and VP were signi cantly related to vertical and translational motion. In addition, clinically, our study suggests that the segments with traction spur and VP should be evaluated carefully because these may be a sign of a disc degeneration and instability. The presence of these should be taken into consideration when evaluating stability in the lumbar spine. Characterizing the type of spur observed may be one of the important factors to take into account when making a decision for or against spinal fusion.

Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Funding
No funds were received in support of this work. No bene ts in any form have been or will be received from a commercial party related directly or indirectly to the subject of this manuscript. The manuscript submitted does not contain information about medical device(s)/drug(s).  Radiographs of lumbar spine with spondylolisthesis to demonstrate the measurement technique Sagittal translation (ST): rst, the perpendicular distance between parallel lines c and d was measured on radiographs, the distance between these two parallel lines was the distance of slippage (b). The difference of the distance of slippage between exion and extension was ST.
Sagittal angulation (SA): the intervertebral slip angle(α) was the angle between two lines at the endplates of the affected disc. The difference of intervertebral slip angles between exion and extension radiographs was SA.