Kinematics of an intervertebral disc with type 1 Modic change: mechanical and non-mechanical causes


 BackgroundType 1 Modic change (MC) is associated with chronic low back pain and attributed to segmental instability. However, the relationship between type 1 MC and segmental instability is unclear. This study sought to clarify the role of mechanical abnormalities in type 1 MC.MethodReview of magnetic resonance images obtained for 727 patients over a 1-year period at our institution revealed 161 cases of type 1 MC. In 86 of these, the following indicators of mechanical abnormality could be evaluated on dynamic radiographs: segmental scoliosis (> 5°), forward slippage (> 1%), and posterior disc opening in flexion. Patients with one or more of these abnormalities were allocated to a mechanical group (n = 62, 70%) and those with no abnormality to a non-mechanical group (n = 26, 30%). The Pfirrmann grade of disc degeneration at the affected level was compared between the groups.ResultsSegmental scoliosis, slippage, and posterior opening was observed in 34, 21 and 37 cases, respectively. Severe disc degeneration (grade IV or V) was present in 43 cases (69%) in the mechanical group and in 12 cases (46%) in the non-mechanical group; the difference was statistically significant (p = 0.04).ConclusionWe propose that there are mechanical and non-mechanical variants of type 1 MC.


Introduction
In 1988, Modic et al. reviewed the magnetic resonance imaging (MRI) ndings in patients with degenerative disc disease and found three types of abnormality: marrow edoema or in ammation (type 1), fatty marrow degeneration (type 2), and sclerotic change (type 3) [1,2]. Modic change (MC) has been identi ed in 35-40% of the population with LBP [3], and type 1 MC in particular is associated with chronic LBP [4,5]. Kuisma et al. found MC on MRI scans for 128 of 228 middle-aged male workers and concluded that type 1 MC was more likely to be associated with back pain [6].
In a radiological review of 70 patients who underwent stabilisation surgery for spinal instability, Ghodsi et al. [7] noted that 52 (74%) had some type of MC and found type 1 in 31 cases. Toyone et al. [8] also found a relationship between MC and spinal hypermobility. Vital et al. performed posterolateral fusion surgery in 17 patients with type 1 MC [9] and found that it converted to normal in 4 cases (24%) and to type 2 in the remaining 13 cases (76%). Based on these data, they suggested that type 1 MC could be related to spinal instability.
Reports in the literature suggest that the pathogenesis of type 1 MC is non-mechanical and may involve the low-virulence microorganism Propionibacterium acnes. A Danish research team rst proposed this pathway in 2008 [10][11][12][13]. Manniche and O'Neill recently reviewed the studies that had demonstrated a correlation between P. acnes and MC type 1 [14] and proposed the following explanation for how P. acnes infection causes this change. Normally, a disc is surrounded by an anatomical barrier that prevents entry of bacteria. For an infection to occur, the disc needs to be leaky. P. acnes can then migrate into the disc, causing disc degeneration and endplate damage, resulting in type 1 MC.
Therefore, based on the literature, there could be two pathways for type 1 MC: mechanical and nonmechanical. The purpose of this study was to investigate the kinematics of intervertebral discs with type 1 MC and whether this type of MC can be classi ed as mechanical or non-mechanical.

Subjects
A retrospective search of our medical records identi ed 727 patients for whom plain radiographs and MRI scans of the lumbar spine were obtained between July 2019 and July 2020. One hundred and sixty-one of these patients were found to have type 1 MC; 88 (44 men, mean age 68. However, depending on the patient's physical constitution, these exposure conditions could be changed by either of two radiological technicians. MRI was performed using a 3.0-T magnetic resonance scanner. The participant was positioned supine and the following scans were acquired: sagittal T1-weighted images from T12 to the sacrum (repetition time 450 ms, echo time 90 ms, slice thickness 4.0 mm); sagittal T2-weighted images from T12 to the sacrum (repetition time 3000 ms, echo time 90 ms, slice thickness 4 mm); and spectral attenuated inversion recovery (SPAIR) T2-weighted images from T12 to the sacrum (repetition time 3000 ms, echo time 90 ms, slice thickness 4 mm).
The 86 patients were divided into a mechanical group and a non-mechanical group. The mechanical group included patients with degenerative spondylolisthesis (vertebral shift > 1% of the anteroposterior diameter of the inferior vertebrae on exion radiographs), posterior opening (kyphotic slip angle > 0° on exion radiographs), and lumbar scoliosis (segmental Cobb angle > 5°). The non-mechanical group included patients without degenerative spondylolisthesis, posterior opening, or lumbar scoliosis. The relationship between MC and P rrmann classi cation was compared between the two groups. All images were reviewed by a general orthopaedic surgeon (Y.Y.) and a certi ed spine surgeon (T.M.), and all diagnoses was made by consensus.

Modic change
Signal changes in the marrow of the vertebral body adjacent to the end plates visualised by MRI were described by de Ross et al. [15] in 1987. According to Modic et al. [1,2], these changes are visible as three different types. Type 1 change is seen on T2-weighted images as areas of increased signal intensity and on T1-weighted images as low signal intensity extending from the vertebral endplates into the vertebral body. Type 2 change is observed as increased signal intensity on both T1-weighted and T2-weighted images. Type 3 change is presumed to represent bone sclerosis and is visualised as decreased signal intensity on both T1-weighted and T2-weighted images. When we could not diagnose MC as type 1 or type 2, patients with a high signal intensity on SPAIR were diagnosed to have MC type 1. In a previous report, only type I MC showed a signi cant association with LBP [16]. Therefore, in this study, we evaluated type I MC only (Fig. 1).

P rrmann classi cation
Intervertebral disc degeneration was assessed on T2-weighted sagittal images using the P rrmann method [17] ( Table 1). The P rrmann grading system assesses degenerated intervertebral discs by MRI based on asymmetry in disc structure, distinction of the nucleus and annulus, the signal intensity of the intervertebral discs, and intervertebral disc height, and assigns a grade of I-V for disc degeneration (Fig. 2). In this study, grade IV or higher was considered to indicate degeneration

Non-mechanical and mechanical groups
Segmental scoliosis, slippage, and posterior opening were observed in 34, 21, and 37 cases, respectively.
At least one of these abnormalities was present in all cases in the mechanical group (n = 62, 70%) but in none of those in the non-mechanical group (n = 26, 30%).

P rrmann classi cation
The P rrmann grade was II in 4 patients  The case was an 83-year-old woman with a diagnosis of degenerative scoliosis of the lumbar spine based on ndings in plain radiographs (Fig. 3a, 3b). A SPAIR MRI scans showed in ammation of the endplate (i.e., type 1 MC, Fig. 3c). Decreased signal intensity was seen on T1-weighted images and increased signal intensity was seen on T2-weighted images. (Fig. 3d, e).
Case 2: Non-mechanical type 1 MC at L4/5, P rrmann grade III The patient was a 31-year-old woman in whom plain radiographs indicated an almost normal lumbar spine with no scoliosis or disc collapse. A SPAIR MRI scan showed in ammation of the endplate (i.e., type 1 MC). Decreased signal intensity was seen on T1-weighted images and increased signal intensity was seen on T2-weighted images. (Fig. 4).

Discussion
This study had two important ndings. First, 62 (70.5%) of the 88 cases of type 1 MC had abnormal kinematics parameters (the mechanical group) and the remaining 26 cases (29.5%) did not have any such abnormalities (the non-mechanical group). Second, severe disc degeneration was found in 43 cases (69.4%) in the mechanical group but in only 12 cases (46.2%) in non-mechanical group. Crockett et al. [18] reviewed the pathology of patients with type 1 MC and divided them into three types according to whether the injury was due to mechanical stress, in ammation, or infection. The initial report by Modic et al. [1] also stated that type 1 MC could be associated with mechanical stress, and their histological analyses revealed disruption and ssuring of the endplate in patients with type 1 MC. Furthermore, there are reports in the literature indicating a close relationship between type 1 MC and spinal instability [7][8][9].
In this study, the kinematics parameters analysed were segmental scoliosis, slippage, and posterior opening, which were found in 34, 21, and 37 cases, respectively. Overall, 70% of our 88 patients had at least one abnormal kinematics parameter. Therefore, our data also suggest that type 1 MC could be caused by spinal instability. Akazawa et al. [19] reviewed 25 patients with adolescent idiopathic scoliosis who had reached middle age without undergoing corrective surgery and found that these patients had a similar prevalence of disc degeneration but were more likely to have MC. Other studies have similarly reported a higher rate of MC at the unfused scoliotic segment following surgery for adolescent idiopathic scoliosis [20] and at the level of the degenerative scoliosis [21].
Aunoble et al. [22] performed fusion surgery in 20 patients with L5 spondylolisthesis and found MC in 50% of cases while Li et al. [23] analysed the MRI scans of 204 patients with degenerative spondylolisthesis and found endplate defects that were closed related to MC in 47% of cases. Meanwhile Ghodsi et al. [7] detected MC in 74% of 70 patients with spinal instability, which they de ned as slippage and hypermobility. In our study, we assessed hypermobility by measuring posterior opening on exion radiographs. According to our ndings, slippage and hypermobility should be related to occurrence of type 1 MC.

Non-mechanical pathway in type 1 MC
In this study, none of the above-mentioned abnormal kinematics parameters were found in 26 (29.5%) of 88 cases, suggesting that there may be another pathway for occurrence of type 1 MC that is not mechanical. The relationship between MC and P. acnes infection was rst reported by Albert et al. in 2008 [10]. Albert et al. subsequently used antibiotics to treat type 1 MC and found signi cant improvement in all outcome measures [11,14]. These ndings supported the notion that bacterial infection plays a role in type 1 MC. In an experimental study, Dudli et al. injected P. acnes into rat tail discs and clearly demonstrated type 1 MC on MRI on day 20 [24]. Chen et al. reported similar ndings in a rabbit model [25]. Therefore, low-virulence disc infection is a potential non-mechanical pathway for occurrence of type 1 MC.

Disc degeneration and type 1 Modic change
The initial report by Modic et al. [1] stated that any type of MC could be related to degenerative disc disease. Luoma et al.
[26] et al. reviewed 28 disc spaces with type 1 MC and initially found that 71% had decreased disc height and that the disc was dark in 57% of cases. During follow-up, they noted a further decrease in disc height in 29% of cases and a further decrease in the disc signal in 46% of cases. Therefore, they concluded that type 1 MC can accelerate the process of disc degeneration.
In the present study, we de ned severe disc degeneration as P rrmann grade IV or V. Severe disc degeneration was found in 69.4% of patients in the mechanical group but in only 46% of those in the nonmechanical group. Our data strongly suggest that there is a close association between type 1 MC that develops by a mechanical pathway and disc degeneration. However, there may be another form of type 1 MC that is not related to the mechanical pathway nor disc degeneration, namely, non-mechanical type 1 MC.

Conclusion
In this study, we attempted to clarify the role of mechanical abnormalities in the development of type 1 MC. Seventy percent of our study population had at least one abnormal mechanical parameter and 30% did not. Severe disc degeneration was more common in patients with mechanical MC than in those with non-mechanical MC. Our data strongly suggest that mechanical-related type 1 MC is closely associated with disc degeneration, as has often been proposed in the literature. However, there may be another form of type 1 MC, which is not related to mechanical pathway nor disc degeneration, namely non-mechanical type 1 MC.

ACKNOWLEDGMENTS
This research received no speci c grant from any funding agency in the public, commercial, or not-forpro t sectors. The authors acknowledge ThinkSCIENCE, Inc. (Tokyo, Japan) for English language editing.
AUTHOR CONTRIBUTIONS Yuji Yamada.: acquisition, analysis, and interpretation of data, and drafting of the manuscript.
Masatoshi Morimoto.: acquisition, analysis, and interpretation of data, and drafting of the manuscript.
Toru Maeda, analysis and interpretation of data.
Syogo Tomiyama, analysis and interpretation of data.
Hirofumi Takami, analysis and interpretation of data.
Naoyuki Yoshida, acquisition of data and critical revision of the manuscript.
Masahiro Kashima, acquisition of data and critical revision of the manuscript.
Koichi Sairyo conception and design of the study; acquisition, analysis, and interpretation of data; and drafting of the manuscript.

CONFLICTS OF INTEREST
The authors declare no con icts of interest.

DATA AVAILABILITY
The datasets generated during and/or analysed during the current study are not publicly available to ensure the privacy of the study participants but are available from the corresponding author on reasonable request. Figure 1 Three types of Modic change.

Figure 2
Page 12/12 P rrmann classi cation of severity of disc degeneration.