An Assessment of Morphological and Pathological Changes in Paravertebral Muscle Degeneration using Imaging and Histological Analysis

Ding-Chao Zhu Yuying Children's Hospital of Wenzhou Medical College: Wenzhou Medical University Second A liated Hospital Jia-Hao Lin Yuying Children's Hospital of Wenzhou Medical College: Wenzhou Medical University Second A liated Hospital Jia-Jing Xu The second a liated hospital and Yuying children's hospital of Wenzhou medical university Qiang Guo Yuying Children's Hospital of Wenzhou Medical College: Wenzhou Medical University Second A liated Hospital Yi-Han Wang The second a liate hoipital and Yuying Children's hospital of Wenzhou medical university Chao Jiang The second a liated hospital and Yuying children's hospital of Wenzhou medical university Hui-Gen Lu The second hospital of Jiaxing Yao-Sen Wu (  wuyaosen@wmu.edu.cn ) Yuying Children's Hospital of Wenzhou Medical College: Wenzhou Medical University Second A liated Hospital https://orcid.org/0000-0002-3844-2950


Introduction
Low back pain (LBP) is a complex and multifactorial disorder commonly found in middle-aged and elderly individuals [1]. Previous studies have demonstrated that a number of lesions, such as nerve root compression, disc degeneration, modic change, facet joint osteoarthritis, and spinal stenosis are all associated with LBP [2][3][4]. However, the main source and related mechanism of LBP are currently unknown. Paravertebral muscle (PVM) degeneration has attracted more attention from researchers. A number of studies have shown the relationship between LBP and PVM degeneration. In the process of PVM degeneration, normal muscle ber morphology changes and is replaced by adipose tissue resulting in a decrease in the stability of the vertebral column [5]. This may be one of the important reasons behind LBP. Additionally, Parkkola et al. [6] found increased fatty in ltration (FI) and reduced cross-sectional area (CSA) of PVM in patients with LBP compared with the control group. Storheim et al. [7] found that patients with less fatty in ltration of PVM had better results after the treatment of LBP. However, the pathological process and speci c mechanism of PVM from normal to degeneration are still unclear.
At present, the measurement of PVM morphology has been an effective method for re ecting FI and has been used for investigations into the etiology of LBP [8]. Magnetic resonance imaging (MRI) as a reliable measurement method, can clearly identify the characteristics of different groups of muscle and the difference in signal intensity between fat and muscle. Previous studies were mainly carried out on the T2-weighted image (T2WI). However, fat and liquid have a relatively longer T2-relaxation time and higher signal intensity on T2WI than other soft tissue such as normal skeletal muscle. Therefore, there is a possibility that fatty degeneration of PVM may be combined with in ammatory edema. In this study, we suspected that there might be different types of PVM pathological changes and that the high signal of PVM on T2WI might not only be fatty degeneration, but also in ammation edema. Studying PVM degeneration on both T2WI and FSI could easily distinguish the signal intensity between fat and edema.
A biopsy can help us understand the microstructure of PVM degeneration in patients with LBP. Bahar Shahidi et al. took multi dus muscle biopsy on 22 patients undergoing surgical treatment for the degenerative lumbar disease. They found high levels of muscle degeneration and in ammation and decreased vascularity [9]. Subsequently, Bahar Shahidi et al. found increased brogenic gene expression in the multi dus muscle of patients with chronic compared to acute LBP [10]. However, these studies employed a simple histological observation and genetic test but did not combine the histological changes in multi dus with imaging. Furthermore, although there have been previous studies about the imaging and histological changes of damaged PVM after lumbar surgery [11], studies combining imaging and histological changes of non-surgical degenerative PVM are still rare. Therefore, in the present study, we mainly studied PVM degeneration in the natural process at a microscopic and macroscopic level using MRI and histological analyses.
This study found three types of pathology in the process of PVM degeneration, and each corresponded to different imaging. Our study provides a basis for the personalized treatment of PVM-induced LBP.

Patients
A total of 70 patients with LBP who consecutively came to the Second A liated Hospital of Wenzhou Medical University were included in the study. Only patients with LBP, no prior spine surgery, no systemic in ammatory disease, no acute trauma, neoplasm, or infection were included in this study. The following data were recorded for each LBP patient: sex, age, body mass index (BMI) and symptoms duration, etc. This study were approved by the Second A liated Hospital of Wenzhou Medical University Ethics Committee and followed the guidelines of the Helsinki Declaration. All participants signed written informed consent before the experiments.

Imaging Evaluation
All patients underwent MRI examinations. The MRI system was a 1.5 Tesla Imaging System (GE Health care Milwaukee, USA). Images were analyzed with Image J (U.S. National Institutes of Health) and stored on the computer. Two spine surgeons with more than 10 years of work experience completed the assessment. The signal of the multi dus was quanti ed using the following steps on T2WI and FSI. The scale pixel was rst set, and each image converted to a grayscale 8-bit image. Image J was then used to outline CSA of the multi dus. The signal intensity was then quanti ed using a threshold technique. The high signal area in the 8-bit image was colored in red using the threshold tool of the program (Fig. 1).

Pathological Examination
A total of 25 underwent surgery for degenerative lumbar spine pathology. During their operation, multi dus about 1cm by 1cm were cut from diseased segment. Multi dus was xed, dehydrated, impregnated, embedded in para n to form wax block and sectioned for haematoxylin and eosin (HE) staining. Histological analysis was made by one experienced pathologist, blinded to the MRI nding and objective of this study.

Statistical Analysis
Statistical analysis was performed using SPSS version 24.0 (SPSS Inc, Chicago, USA). All continuous data were described as mean ± standard deviation. An analysis of variance was carried out to detect differences in age and BMI among the different types of PVM degeneration. The Kruskal-Wallis test was performed to detect differences in disease duration among the different types. The gender in different types was analyzed using the Chi-square test. P values less than 0.05 were considered statistically signi cant.

MRI Evaluation of Multi dus
The demographic data of the 70 patients are presented in Table 1. Three different types of PVM high signals were identi ed from the MRI examination. A total of 36 patients (51.42%, 57.33 ± 7.24 years) showed high signal on T2WI and low signal on FSI, which indicated fatty degeneration (Type 1,

Histological Evaluation Of Multi dus
The demographic information of the 25 surgical patients is listed in Table 2. From the histological analyses, adipocytes in ltration was observed in 14 patients (56.00%, 60.14 ± 6.27 years). Adipocytes could be seen in muscle tissue by HE staining and the muscle cell nuclei were squeezed into periphery without in ammatory cells in ltration (Fig. 3A). A total of 3 patients (12.00%, 27.00 ± 4.36 years) showed in ammatory cells in ltration. HE staining showed in ammatory cells (mainly neutrophils) in muscle tissue without adipocytes (Fig. 3B). Finally, 8 patients (32.00%, 45.88 ± 6.15 years) showed both adipocytes and in ammatory cells in ltration. The HE staining showed both adipocytes and in ammatory cells (mainly lymphocytes) in muscle tissue, light staining of in ammation edema area, and tissue gap narrowed or disappeared (Fig. 3C). Degenerative lumbar disease of 25 surgical patients are shown in Fig. 4.
Histological results of these 25 patients were consistent with their imaging. The 14 patients with adipocytes in ltration showed type 1 on imaging; the 3 patients with in ammatory cells in ltration showed type 2 and the 8 patients with both adipocytes and in ammatory cells in ltration showed type 3 on imaging.

Demographic Data Evaluation
For patients of the different types of MRI evaluation, age (p < 0.001) and BMI (p < 0.001) indicated signi cant difference, while gender (p = 0.424) indicated no statistical difference (Table. 1). For patients of the different types of histological evaluation, age (p < 0.001) and BMI (p = 0.025) indicated signi cant difference (Table. 2). In addition, there was a statistical difference in the disease duration (p < 0.001, Table. 3). The duration of type 1 patients (58.33%) mainly were longer than 12 weeks, all type 2 patients (100%) lasted for less than 4 weeks and type 3 patients (52.00%) were mostly between 4 to 12 weeks.

Discussion
The PVM mainly comprises psoas, erector spinae, and multi dus muscle and is considered to have two functions: to stabilize and move the lumbar vertebral column. Kalichman et al. found that fatty degeneration in the PVM was common in adults and was strongly associated with LBP [5]. Therefore, studying PVM can help to discover the potential mechanism of lumbar instability and LBP. Moreover, Guo et al. found no signi cant difference in psoas or erector spinae between the patients with LBP and normal people [12]. Barker et al. reported multi dus as the largest and most medial PVM that is signi cantly stronger in normal persons than LBP patients and is sensitive to pathological changes [13]. This indicates that multi dus might be more involved in maintaining spinal stability and preventing chronic LBP.
MRI, computerized tomography (CT) and ultrasound are ideal methods for assessing the morphology of PVM. Many researchers have inclined to use MRI to analyze PVM in patients with lumbar diseases due to its clear muscular contour and high-fat resolution [14]. Hu et al. advised MRI rather than CT for the measurement of CSA and FI by comparing the intra-and inter-reliability [8]. Fortin et al. found that an increase in fat was caused by age and BMI in a 15-years longitudinal MRI study [15]. Previous studies mainly evaluated the degree of fat degeneration through signal intensity on T2WI. But since both fat and liquid show high signal on T2WI, measurement on T2WI cannot clearly distinguish between fatty degeneration and in ammatory edema of PVM. Thus, combining T2WI and FSI could provide a reasonable contrast between fat and liquid. CSA and FI are major indicators to evaluate PVM degeneration. Chon et al. found that CSA of multi dus was smaller in patients with chronic lumbar radiculopathy than in normal people [16]. Resorlu et al. observed that patients with ankylosing spondylitis had decreased CSA of multi dus, and it was related to the disease duration [17]. In a cross-sectional observation study, Barker et al. revealed that CSA of multi dus was smaller in patients with chronic LBP than in normal [13]. Moreover, Fortin et al. systematically reviewed the studies on the morphology of PVM in patients with LBP and the control group. They concluded that CSA of PVM in patients with LBP was signi cantly smaller than that in the control group [18]. Nonetheless, studies on the pathological mechanism of PVM degeneration are still rare. Therefore, in the present study, we mainly examined the pathological process for assessing PVM change.
The results of our study showed different types of signal changes in PVM. Type 1 included the patients with high signal on T2WI and low signal on FSI on MRI imaging and with adipocytes in ltration on histological analyses. Type 2 comprised of the patients with high signal on both T2WI and FSI and showed in ammatory cells (mainly neutrophils) in ltration. Type 3 included the patients with high signal on T2WI and partial signal suppression on FSI showed both adipocytes and in ammatory cells (mainly lymphocytes) in ltration. We further analyzed the age, BMI and disease duration and proportion of each type of patient. Among the 70 patients studied, type 1 had the largest proportion and mainly composed of middle-aged and elderly, the largest BMI, and chronic LBP. Type 2 had the least proportion and was mainly made up of the young, with the lowest BMI, and mainly acute LBP. The age, BMI, duration and proportion of type 3 cut across both type 1 and type 2. In ammatory cells in type 2 were mainly neutrophils, which indicated acute in ammation whereas in type 3 the in ammation cells were mainly lymphocytes indicating chronic in ammation. Bahar Shahidi et al. found that muscle cells actively degenerated as opposed to simple atrophy, a process that may be facilitated by in ammation [9]. Hatice et al. found that chronic in ammation and cytokine-mediated brosis in patients with ankylosing spondylitis contributed to fatty degeneration and atrophy in the PVM [19]. In addition, Paul et al. reported a parallel increase in the expression of pro-in ammatory cytokines during the degeneration of PVM [20]. Therefore, we reasonably speculated that PVM degeneration in patients with LBP is a gradual process, and in ammation is an early pathological manifestation that might gradually transform into fatty degeneration.
Currently, many treatment modalities are used for LBP such as bed-rest, functional training of back muscle, physiotherapy, and medication, but their effectiveness are still inconclusive. Previous treatment strategies did not take into account the different pathological types of PVM degeneration. According to the imaging and histological results of this study, we speculate that personalized treatments are more likely to be effective for patients with different pathological changes. For patients with type 1 PVM degeneration, functional training for back muscle should be encouraged, because fatty degeneration is di cult to reverse with medication treatments or physiotherapy only. Bed-rest seems to not be of much bene t and may actually be detrimental because of fat deposit and muscle atrophy in these patients [21].
Storheim et al. reported that comprehensive training had a signi cant tendency of reversing the degeneration of PVM in patients with subacute LBP [22]. Hides et al. found that stability training could improve the CSA of multi dus in young athletes with LBP and reduce the pain [23]. However, according to the results of this study, we do not recommend high-intensity back muscle training for type 2 patients due to its acute in ammation. Bed-rest and non-steroidal anti-in ammatory drugs could be the best option, which can inhibit the progress of in ammation. Although most acute LBP is considered self-limiting, previous evidence found that a high percentage of individuals will experience recurrent symptoms leading to poor functional outcomes over time [24]. Bed-rest and drug treatment can effectively prevent type 2 patients from turning acute to chronic LBP. Because type 3 patients have a combined fatty degeneration and in ammation, we recommend that they should be treated in stages: rstly bed-rest and use of nonsteroidal anti-in ammatory drugs to inhibit the progress of in ammation, then performing functional training. For many patients with acute LBP, clinicians may recommend two days of bed-rest rather than longer periods [25]. The results of this study can provide certain scienti c guidance for the speci c time of bed-rest. In individuals with persistent recurrence or chronic symptoms, muscle tissue is relatively slow to respond to traditional rehabilitation measures [26]. Therefore, understanding different types of pathological changes in PVM is crucial to determining appropriate treatment strategies.
There were some limitations in this study that need further discussion and investigations. The sample size in the study was small hence a need for further clinical research to support our personalized treatment methods for LBP and demographic data evaluation. Secondly, the study did not establish the exact interaction between PVM degeneration and LBP, hence should be further studied. Based on this study, the speci c transformation mechanism that may exist between fatty degeneration and in ammation in PVM needs further clari cation.

Conclusion
Based on the ndings of this study, there are three types of pathology in the process of PVM degeneration, which may provide a basis for the personalized treatments of PVM-induced LBP. Besides, in ammation presents as an early pathological manifestation of PVM degeneration that gradually transforms into fatty degeneration.   The duration of less than 4 weeks manifests as acute LBP, 4 to 12 weeks manifests as subacute LBP and more than 12 weeks manifests as chronic LBP *indicates signi cant difference   Degenerative lumbar pathology of 25 surgical patients. There were 2 patients with lumbar disc herniation, 4 patients with lumbar spinal stenosis, and 8 patients with lumbar disc herniation combined with lumbar spinal stenosis in type 1 (56.00%); All three patients in type 2 (12.00%) presented lumbar disc herniation; Type 3 patients (32.00%) included 4 of lumbar disc herniation, 1 of lumbar spinal stenosis, and 3 of lumbar disc herniation combined with lumbar spinal stenosis.