The Cervical Vertebral Bone Quality Score Is a Novel Reliable Index Reflecting the Condition of Paraspinal Muscles and Predicting Loss of Cervical Lordosis After Open‐Door Laminoplasty

ABSTRACT

Objective

Recently, the MRI‐based cervical vertebral bone quality (C‐VBQ) scoring system has demonstrated accuracy in reflecting cervical bone quality and predicting postoperative complications from cervical spine surgery. Studies have shown that cervical bone quality is closely linked to loss of cervical lordosis (LCL) after open‐door laminoplasty. Additionally, research on lumbar VBQ indicates a strong correlation between lumbar VBQ scores and lumbar paraspinal muscle quality. However, the relationship of C‐VBQ score to cervical paraspinal muscles and LCL remains unclear. Therefore, this study aimed to explore the relationship between C‐VBQ score and cervical paraspinal muscle‐related parameters as well as postoperative LCL, in addition to exploring the risk factors associated with LCL.

Methods

A total of 101 patients who underwent standard C3–C7 open‐door laminoplasty at our institution from 2012 to 2022 were included in this study. The LCL group was defined as loss of cervical lordosis > 5° at 1‐year postoperative follow‐up. Cervical X‐rays were obtained to measure the C2–7 Cobb angle, C2–7 sagittal vertical axis (SVA), T1 slope, and cervical range of motion (ROM). The relative cross‐sectional area (RCSA) and degree of fat infiltration (DFF) of the deep cervical extensors, flexors, and all muscles were measured using image J software. Cervical CT Hounsfield unit (HU) and C‐VBQ values were measured on preoperative CT and MRI T1‐weighted mid‐sagittal images, respectively. Then, demographics, cervical sagittal parameters, ROM, paraspinal muscle‐related parameters, CT‐HU and C‐VBQ values were assessed for their correlation with LCL, and multivariate linear analysis was used to determine the risk factors associated with LCL. Finally, the relationship between C‐VBQ scores and cervical paraspinal muscle‐related parameters was evaluated.

Results

A total of 55 (54.45%) patients were included in the LCL group due to loss of cervical lordosis > 5° at 1‐year follow‐up. LCL was positively correlated to the preoperative T1 slope, Flexion ROM, C2–7 ROM, Flexion/Extension ROM, Flexion muscles DFF, Extension muscles DFF, Average DFF, and C‐VBQ scores, while it was negatively correlated to Extension ROM, Extension muscles RCSA, Total RCSA, and CT‐HU values. Furthermore, Flexion/Extension ROM, Total RCSA, Average DFF, CT‐HU, and C‐VBQ values were independent risk factors for LCL. In addition, C‐VBQ scores were significantly correlated with RCSA and DFF of Flexion and Extension muscles.

Conclusions

This study is the first to find a significant correlation between C‐VBQ scores and cervical paraspinal muscle quality. The C‐VBQ score is a comprehensive indicator that reflects the quality of the cervical bone and paravertebral muscles, and it is a novel predictor of LCL after open‐door laminoplasty.

There is a close correlation between C‐VBQ scores and the cervical deep extensors, deep flexors, and overall paraspinal muscles relative cross‐sectional area (RCSA) and degree of fat infiltration (DFF). The C‐VBQ score is a comprehensive parameter reflecting the quality of cervical bone and paraspinal muscles. In addition, there is a significant correlation between the C‐VBQ score and LCL, and the C‐VBQ score was a novel independent risk factor for LCL after open‐door laminoplasty.

1. Introduction

Cervical open‐door laminoplasty is widely used in the treatment of cervical spondylotic myelopathy (CSM). As a non‐fusion technique, open‐door laminoplasty not only achieves adequate indirect decompression of the spinal cord, but also well maintains the cervical range of motion (ROM) [1, 2]. However, postoperative loss of cervical lordosis (LCL) is one of the most common surgical complications, with a reported incidence of LCL up to 80% within a 1‐year follow‐up [3, 4]. Maintaining cervical lordosis after open‐door laminoplasty results in ideal decompression and is an important factor in ensuring the long‐term efficacy of the surgery. Therefore, it is particularly important to look for risk factors associated with LCL. Several previous studies have investigated risk factors for LCL. T1 Slope, cervical ROM, the relative cross‐sectional area (RCSA) as well as the degree of fat infiltration (DFF) of the deep muscles, and the cervical vertebral computerized tomography (CT) Hounsfield unit (HU) have been suggested to be associated with LCL [5, 6, 7]. However, more research is needed to explore and validate the risk factors associated with LCL, particularly the relationship between cervical bone and deep flexor muscle quality and LCL.

Sarcopenia‐osteoporosis is a syndrome in which decreased muscle mass and strength coexist with osteoporosis [8]. In recent years, an increasing number of studies have focused on the correlation between bone and paraspinal muscle quality; it has been demonstrated that osteoporosis and sarcopenia have common pathophysiological mechanisms and similar clinical features [9, 10, 11]. In addition, osteoporosis has many of the same risk factors as sarcopenia, including age, gender, and habits. Therefore, bone and muscle are gradually being considered as a whole in studies, which is conducive to providing new ideas for the study of the musculoskeletal system. However, there is still a lack of an indicator that reflects both the quality of the cervical vertebral bone and the paraspinal muscles.

The novel MRI‐based cervical vertebral bone quality (C‐VBQ) scoring system evaluates bone quality conditions by quantitatively measuring fat content in the cervical vertebral body, which has the advantages of no radiation, no additional cost, and easy accessibility [12]. Previous studies have demonstrated that the C‐VBQ score effectively reflects cervical bone quality and predicts a variety of postoperative cervical spine complications [13, 14]. Interestingly, a recent study by Li et al. [15] found that the lumbar VBQ score is a comprehensive index reflecting the quality of lumbar vertebral bone and paraspinal muscles. However, the relationship between C‐VBQ scores and the quality conditions of cervical deep flexor and extensor muscles remains unclear, and their relationship with LCL needs to be further investigated.

Therefore, the aims of this study are to (i) investigate the relationship between preoperative parameters, including C‐VBQ scores, RCSA, and DFF of cervical paraspinal muscles, and LCL after open‐door laminoplasty; (ii) determine whether the C‐VBQ score can reflect the quality of cervical paraspinal muscles.

2. Methods

2.1. Patient Population

After receiving approval from the ethics committee of West China Hospital, Sichuan University (No. 2023252), we identified a consecutive retrospective group of patients who underwent standard C3–C7 open‐door laminoplasty at a single institution. All procedures for these patients were performed between January 2012 and January 2022. Because this was a retrospective study, signing the informed consent was waived. The inclusion criteria were: (1) patients who underwent standard C3–C7 open‐door laminoplasty for cervical spondylotic myelopathy caused by ossification of the posterior longitudinal ligament or other reasons; (2) patients who had cervical spine X‐rays, CT, and MRI scans within2 weeks prior to surgery; and (3) patients who had at least 1 year of follow‐up information and cervical spine X‐rays from 1 year after the surgery. The exclusion criteria were as follows: (1) diagnoses of tumors, fractures, infections or spine deformity; (2) previous history of spine surgery; (3) Patients who have incomplete or poor‐quality follow‐up or imaging data. Finally, 101 cases completed the follow‐up. Two senior spine surgeons performed all the operations. All patients received miniature titanium plates (Medtronic Inc. USA) fixation from C3 to C7, wore a cervical gear for 3 weeks postoperatively, and received short‐term NSAID painkillers; none of them received any specialized physiotherapy after surgery and during follow‐up.

2.2. Radiographic Assessment

All patients underwent standard lateral, forward flexion, and posterior extension radiographs of the cervical spine preoperatively and lateral cervical spine radiographs at the final follow‐up. CT scans (Kv: 140, mean mAs: 182) and MRIs (TR: 561 ms, TE: 10 ms, Slice: 3 mm, FOV read: 230 mm, Average: 3) of the cervical spine were performed 1 week before surgery. Imaging parameters were measured using a picture archiving and communication system (PACS) (Siemens, syngo MultiModality Workplace, Germany).

Cervical sagittal parameters were measured as follows (Figure 1): (1) Cervical lordosis (CL), defined as the angle formed by the parallel lines of the lower end plates of C2 and C7; (2) LCL group, defined as a decrease in CL > 5° at the last follow‐up [16]; (3) C2–C7 sagittal vertical axis (SVA), defined as the horizontal distance between the vertical line from the center of the C2 vertebra and the posterior‐superior corner of the C7 vertebra; (4) T1 slope, defined as the angle between the parallel lines of the endplates on the T1 vertebrae and the horizontal line; (5) Flexion ROM (F ROM), defined as the difference between the C2–C7 Cobb angle in the neutral and flexed cervical positions; (6) Extension ROM (E ROM), defined as the difference between the C2–C7 Cobb angle in the cervical extension and neutral positions; (7) Cervical ROM, defined as the difference between the C2–C7 Cobb angle in cervical extension and flexion; (8) The ratio of F ROM/E ROM was used to indicate asymmetry in cervical spine mobility.

FIGURE 1.

Cervical spine X‐ray radiographic measurements. (A) flexion, (B) lateral, and (C) extension radiographs, (a) C2–C7 Cobb angle, (b) C2–C7 sagittal vertical axis (SVA), and (c) T1 slope.

CT‐HU values were measured using the standard method published by Schreiber et al. [17]. Three levels of scanning were selected for each vertebral body of C3–C6 in the CT axial plane: the middle of the vertebral body, 2 mm below the upper endplate, and 2 mm above the lower endplate. Regions of interest (ROIs) were placed in the cancellous bone regions at each level (excluding thecortical bone and including as much of the cancellous bone as possible), and the mean HU values were calculated for each vertebra (Figure 2). Cervical HU values were defined as the mean of C3–C6 vertebral HU values.

FIGURE 2.

Examples of the Cervical CT‐HU and VBQ values measurements. (A) CT‐HU values measurements, U, 2 mm below the superior endplate, M, in the middle of the vertebral body, L, 2 mm above the inferior endplate, (B) C‐VBQ values measurements, non‐contrast enhanced T1‐weighted MRI images showing measurements of C3–C6 SI and CSF SI. CSF, cerebrospinal fluid; SI, signal intensity.

The C‐VBQ values were assessed on mid‐sagittal cut of the patient's preoperative cervical MRI T1‐weighted, non‐contrast imaging. ROIs were placed in the C3–C6 vertebrae, which should contain as much cancellous bone region as possible after excluding cortical bone, abnormal subchondral alterations, focal lesions (e.g., haemangiomas) and posterior venous plexus. Two additional ROIs are placed in the cerebrospinal fluid (CSF) anterior and posterior to the spinal cord in the C2 vertebral plane (Figure 2). When obstruction exists in the CSF at the C2 level, the ROIs are placed in the CSF at the C1 vertebral plane. Then, the median signal intensity (SI) of C3–C6 is divided by the average of the pre‐ and post‐spinal cord CSF SIs (Formula 1).

$$\mathrm{C}-\mathrm{VBQ}=\frac{{\mathrm{SI}}_{\left(\mathrm{C}3-\mathrm{C}6\right)}}{{\mathrm{SI}}_{\left(\mathrm{C}2\mathrm{CSF}\right)}}$$ (1)

Cervical paraspinal muscles cross‐sectional area (CSA) and DFF were measured using ImageJ software (version 1.52α, National Institutes of Health, Bethesda, MD, USA). All data were collected at the C5/6 level according to the methods described in previous studies [18, 19, 20]. The neck flexion muscles include (1) the longus colli and longus capitis (Lco + Lca); (2) the sternocleidomastoid muscles (SCM); extension muscles include: (3) multifidus (Mult); (4) semispinalis cervicis (SeCe); (5) semispinalis capitis (SeCa); (6) splenius cervicis and splenius capitis (SpCe + SpCa); (7) levator scapulae (LSc). The CSA of the C5 vertebra was used as a reference. The RCSA of the paravertebral muscles/vertebral body was calculated. The RCSA of all muscles was summed to represent the total RCSA (Figure 3). The percentage of fat content of each paraspinal muscle was determined using a pseudocoloring technique, in which the high signal of the fatty tissue was coated red, and the percentage of the red area of each muscle was calculated (Figure 3). The DFF of each muscle was summed and divided by 7 to represent the average DFF. The paravertebral muscle‐related parameters were the RCSA and DFF averaged on both sides.

FIGURE 3.

Measurement of the cross‐sectional area (CSA) and degree of fat infiltration (DFF) of the paraspinal muscles and vertebral body at the C5/6 level by ImageJ software. (A)—(1) Longus colli and longus capitis; (2) Sternocleidomastoid; (3) Multifidus; (4) Semispinalis cervicis; (5) Semispinalis capitis; (6) Splenius cervicis and splenius capitis; (7) Levator scapulae; V, vertebral body. (B) Image J measured the degree of fat infiltration.

The radiographic parameters were measured by two independent and experienced observers, and the average values were taken into consideration.

2.3. Clinical Function Assessment

The Japanese Orthopedic Association (JOA) score was used to assess the neurological function of the patients preoperatively and at the final follow‐up. The scale includes four dimensions: upper limb function, lower limb function, sensory function, and bladder function. The scores ranged from 0 to 17.

2.4. Statistical Analyses

Statistical analyses were performed using SPSS 29 (SPSS Inc., Chicago, IL). The intraclass correlation coefficient (ICC) was used to test intra‐ and inter‐rater reliability, and an ICC of ≥ 0.75 was considered to be of good reliability. The independent samples t test was employed to compare continuous variables between two independent groups. Categorical data were analyzed using either the χ ² test or Fisher's exact test, depending on the sample size and expected frequencies. Correlations between continuous variables were assessed using Pearson's correlation coefficient, while Spearman's rank correlation was used for ordinal or non‐normally distributed data. Multiple linear regressions and the receiver operating characteristic curve (ROC) were used to investigate the risk factors for predicting LCL. A value of p < 0.05 was considered statistically significant.

3. Results

A total of 101 patients (72 male and 29 female) were finally enrolled in this retrospective study, and all demographic, clinical outcomes, and radiological data are shown in Tables 1 and 2 Both intra‐ and inter‐rater reliability were greater than 0.75 (Intra‐rater: C‐VBQ: 0.88, Total muscles RCSA: 0.76; Inter‐rater: C‐VBQ: 0.81, Average muscles DFF: 0.75). Mean age was 56.33 ± 9.81 years, mean follow‐up time was 16.09 ± 3.21 months, and the average body mass index (BMI) was 25.22 ± 2.06 kg/m². At the last follow‐up, all patients had significantly improved JOA scores. The mean C2–C7 cobb angle decreased from 13.30° ± 10.53° preoperatively to 9.88° ± 11.33° at the last follow‐up. T1 Slope decreased from 24.86 ± 9.01 preoperatively to 21.45 ± 6.74 at the last follow‐up, while C2–C7 SVA increased from 2.54 ± 1.90 cm preoperatively to 5.49 ± 5.54 cm. This means that not only does LCL occur after open‐door laminoplasty, but an imbalance of the entire sagittal plane of the cervical spine can appear. Other relevant parameters are shown in Tables 1 and 2.

TABLE 1.

Basal characteristics and clinical outcomes between LCL and NLCL groups.

Variable	Total	LCL (n = 55)	NLCL (n = 46)	p
Sex
Male	72	38	34	0.594
Female	29	17	12
Age (years)	56.33 ± 9.81	57.07 ± 10.35	54.24 ± 8.78	0.146
BMI (kg/m2)	25.22 ± 2.06	25.01 ± 2.12	25.47 ± 1.97	0.264
Follow‐up (months)	16.09 ± 3.21	15.93 ± 2.86	16.20 ± 3.43	0.723
JOA score
Preoperation	12.23 ± 1.28	12.23 ± 1.05	12.25 ± 1.53	0.987
Follow‐up	14.14 ± 0.72	13.94 ± 0.67	14.58 ± 0.88	< 0.001

Abbreviations: BMI, body mass index; JOA score, Japanese Orthopedic Association Score.

TABLE 2.

Radiographic parameters between LCL and NLCL groups.

Variable	Total	LCL (n = 55)	NLCL (n = 46)	p
C2–C7 Cobb (°)
Preoperation	13.30 ± 10.53	14.37 ± 10.75	12.02 ± 10.22	0.266
Follow‐up	9.88 ± 11.33	7.27 ± 10.43	12.99 ± 11.70	0.011
T1 slope (°)
Preoperative	24.86 ± 9.01	27.41 ± 9.39	22.19 ± 8.60	0.004
Follow‐up	21.45 ± 6.74	21.00 ± 7.04	21.99 ± 6.40	0.465
C2–C7 SVA (cm)
Preoperative	2.54 ± 1.90	2.74 ± 2.32	2.32 ± 1.19	0.270
Follow‐up	5.49 ± 5.54	7.58 ± 6.68	3.00 ± 1.72	< 0.001
Flexion ROM (°)
Preoperative	29.42 ± 8.12	33.38 ± 8.11	24.68 ± 5.05	< 0.001
Extension ROM (°)
Preoperative	9.71 ± 2.31	8.20 ± 1.65	10.50 ± 1.60	< 0.001
C2‐7 ROM (°)
Preoperative	38.12 ± 7.62	40.58 ± 8.35	35.18 ± 5.39	< 0.001
F ROM/E ROM
Preoperative	3.80 ± 2.04	4.96 ± 2.09	2.40 ± 0.60	< 0.001
Muscles
Preoperative
Lco + Lca
RCSA	0.41 ± 0.15	0.45 ± 0.13	0.38 ± 0.16	0.017
DFF	18.87 ± 5.77	21.25 ± 4.97	16.03 ± 5.39	< 0.001
SCM
RCSA	1.30 ± 0.17	1.28 ± 0.18	1.32 ± 0.16	0.245
DFF	18.11 ± 5.94	21.33 ± 5.20	14.26 ± 4.27	< 0.001
Mult
RCSA	0.81 ± 0.21	0.38 ± 0.19	0.44 ± 0.15	0.086
DFF	22.48 ± 6.68	25.53 ± 6.28	18.82 ± 5.16	< 0.001
SeCe
RCSA	0.53 ± 0.22	0.49 ± 0.18	0.56 ± 0.17	0.049
DFF	22.99 ± 7.64	26.89 ± 6.55	18.34 ± 6.13	< 0.001
SeCa
RCSA	0.71 ± 0.11	0.68 ± 0.08	0.73 ± 0.09	0.004
DFF	17.04 ± 7.39	18.68 ± 6.28	16.68 ± 6.19	0.118
Spca +Spce
RCSA	0.80 ± 0.13	0.77 ± 0.12	0.85 ± 0.05	< 0.001
DFF	21.23 ± 6.51	19.38 ± 5.58	17.48 ± 5.50	0.089
Lsc
RCSA	1.16 ± 0.08	1.11 ± 0.07	1.21 ± 0.06	< 0.001
DFF	20.77 ± 6.01	23.54 ± 5.21	17.44 ± 5.19	< 0.001
Flexion muscles
RCSA	1.71 ± 0.18	1.73 ± 0.15	1.70 ± 0.20	0.392
DFF	36.98 ± 11.03	42.59 ± 9.16	30.28 ± 9.25	< 0.001
Extension muscles
RCSA	3.60 ± 0.45	3.43 ± 0.35	3.79 ± 0.28	< 0.001
DFF	101.51 ± 27.93	114.47 ± 21.85	88.76 ± 20.68	< 0.001
Total RCSA
Preoperative	5.33 ± 0.53	5.16 ± 0.43	5.49 ± 0.44	< 0.001
Average DFF (%)
Preoperative	20.21 ± 5.35	20.37 ± 4.14	18.01 ± 4.03	0.005
CT‐HU values
Preoperative	356.93 ± 94.74	325.83 ± 92.97	394.12 ± 83.53	0.002
C‐VBQ values
Preoperative	2.60 ± 1.06	3.15 ± 1.03	1.94 ± 0.62	0.001

Abbreviations: C‐VBQ, cervical vertebral bone quality score; DFF, the degree of fat infiltration; E ROM, ranges of extension motion; F ROM, ranges of flexion motion; HU, Hounsfield Units; LCo + LCa, longus colli and longus capitis; LSc, levator scapulae; Mult, multifidus; RCSA, relative cross‐sectional area; ROM, range of motion; SCM, sternocleidomastoid; SeCa, semispinalis capitis; SeCe, semispinalis cervicis; SpCe +SpCa, splenius cervicis and splenius capitis; SVA, sagittal vertical axis.

3.1. Risk Factors for LCL

All parameters were compared between the LCL and NLCL groups (Tables 1 and 2), and baseline data such as gender, age, BMI, and follow‐up time were not significantly different between the two groups. For preoperative parameters, the LCL group had a higher T1 slope, greater Flexion ROM, C2–C7 ROM, a higher Flexion ROM/Extension ROM ratio, and a smaller Extension ROM. For paraspinal muscle‐related parameters, the RCSA of SeCe, SeCa, Spca + Spce, Lsc, and total Extension muscles were significantly lower in the LCL group than in the NLCL group, and Lco + Lca, SCM, Mult, SeCe, Lsc, total Flexion muscles, and total Extension muscles had significantly higher DFF than the NLCL group. In addition, CT‐HU values were significantly lower and C‐VBQ values were significantly higher in the LCL group than in the NLCL group.

The results of the correlation analysis between preoperative relevant parameters and LCL are shown in Table 3. The Extension ROM (r = −0.466, p < 0.001), Extension muscles RCSA (r = −0.527, p < 0.001), Total RCSA (r = −0.465, p = 0.001), and CT‐HU values (r = −0.239, p = 0.016) are significantly negatively correlated with LCL. However, the T1 slope (r = 0.297, p = 0.013), Flexion ROM (r = 0.364, p = 0.003), C2–C7 ROM (r = 0.247, p = 0.010), F/E ROM (r = 0.395, p < 0.001), Flexion muscles DFF (r = 0.456, p < 0.001), Extension muscles DFF (r = 0.534, p < 0.001), Average DFF (r = 0.531, p < 0.001), and C‐VBQ scores (r = 0.494, p < 0.001) are significantly positively correlated with LCL.

TABLE 3.

Correlation between various factors and LCL.

	Age	BMI	C2–C7 Cobb	T1 sl ope	C2–C7 SVA	Flexion ROM	Extension ROM	C2–C 7 ROM	F/E ROM	Flexion muscles RCSA	Flexion muscles DFF	Extension muscles RCSA	Extension muscles DFF	Total RCSA	Average DFF	HU value	Cervical VBQ value
LCL
r	0.099	−0.055	0.158	0.297	0.179	0.364	−0.466	0.247	0.395	−0.034	0.456	−0.527	0.534	−0.465	0.531	−0.239	0.494
p	0.273	0.583	0.114	0.013	0.073	0.003	< 0.001	0.010	< 0.001	0.736	< 0.001	< 0.001	< 0.001	0.001	< 0.001	0.016	< 0.001

Abbreviation: F/E ROM, Flexion ROM/Extension ROM.

The results of the multiple linear regression are shown in Table 4; F/E ROM, Total RCSA, Average DFF, Mean CT‐HU values, and C‐VBQ values are independent risk factors for LCL. The results of the ROC curve analysis are shown in Figure 4 and Table 5, with corresponding AUC values of 0.863, 0.823, 0.880, 0.736, and 0.850, respectively.

TABLE 4.

Multivariate linear regression analysis of potential contributing factors and LCL.

Variables	b	SE	p
F/E ROM	1.446	0.738	0.050
Total RCSA	−1.887	0.934	0.046
Average DFF	4.003	1.521	0.011
Mean CT‐HU values	−0.029	0.012	0.018
C‐VBQ values	1.900	0.782	0.016

Abbreviations: b, Unstandardized Coefficients; C‐VBQ, Cervical Vertebral Bone Quality Score; DFF, the degree of fat infiltration; HU, Hounsfield Units; RCSA, relative cross‐sectional area; ROM, range of motion.

FIGURE 4.

Receiver operating characteristic (ROC) curve for prediction of LCL. CT‐HU, computerized tomography Hounsfield Unit; C‐VBQ, cervical vertebral bone quality; DFF, degree of fat infiltration of the paraspinal muscles; F/E ROM, flexion/extension range of motion; RCSA, relative cross‐sectional area of the paraspinal muscles.

TABLE 5.

The AUC information for the ROC curve analysis.

Predictors	AUC	Std	p	95% CI
F/E ROM	0.863	0.038	0.001	0.787–0.938
Total RCSA	0.823	0.041	0.001	0.743–0.904
Average DFF	0.880	0.034	0.002	0.814–0.946
Mean CT‐HU values	0.736	0.050	0.013	0.638–0.834
C‐VBQ values	0.850	0.038	0.001	0.775–0.924

Abbreviations: C‐VBQ, Cervical Vertebral Bone Quality Score; DFF, the degree of fat infiltration; HU, Hounsfield Units; RCSA, relative cross‐sectional area; ROM, range of motion.

3.2. Relationship Between C‐VBQ Scores and Paravertebral Muscle‐Related Parameters

Given that both paraspinal muscle‐related parameters and C‐VBQ scores were significantly correlated with LCL, we further analyzed the relationship between C‐VBQ scores and paraspinal muscle‐related parameters. The RCSA of Cervical Flexion muscles, Extension muscles, and total muscles were correlated with the C‐VBQ score (r = −0.325, p = 0.001; r = −0.873, p < 0.001; r = −0.861, p < 0.001, respectively). Similar results were also found for the DFF (r = 0.821, p < 0.001; r = 0.741, p = 0.001; r = 0.795, p < 0.001, respectively) (Table 6 and Figure 5). Similarly, the CT‐HU values showed a significant linear relationship with the C‐VBQ values (r = −0.370, p < 0.001) (Figure 6).

TABLE 6.

Correlation between the C‐VBQ score and cervical paravertebral muscle‐related parameters.

	RCSA		DFF
	r	p	r	p
Flexion muscles	−0.325	0.001	0.821	< 0.001
Extension muscles	−0.873	< 0.001	0.741	0.001
Total/average	−0.861	< 0.001	0.795	< 0.001

FIGURE 5.

Scatter plots of the (A) C‐VBQ score and RCSA, (B) C‐VBQ score and DFF.(A) Total muscles RCSA = Extension muscles + Flexion muscles RCSA; (B) Average DFF = (Extension muscles + Flexion muscles DFF)/7. DFF, degree of fat infiltration of the paraspinal muscles; RCSA, relative cross‐sectional area of the paraspinal muscles.

FIGURE 6.

Correlation between C‐VBQ score and CT‐HU values.

4. Discussion

In this study, we demonstrated for the first time that MRI‐based C‐VBQ scores were strongly correlated with RCSA and DFF of the deep cervical paraspinal muscles, and furthermore, we also showed that C‐VBQ scores and paraspinal muscle‐related parameters were strongly correlated with LCL after open‐door laminoplasty. Multivariate linear regression analysis showed that, like cervical paraspinal muscle‐related parameters and CT‐HU values, the C‐VBQ score is a novel risk factor for LCL.

4.1. The C‐VBQ Score

Dual‐energy X‐ray absorptiometry (DEXA) remains the gold‐standard clinical modality for bone quality assessment through lumbar spine or hip bone mineral density (BMD) measurements, albeit providing only indirect cervical spine evaluations [21]. Nevertheless, this technique has notable limitations, including its inability to provide site‐specific cervical bone assessments and susceptibility to extraneous factors such as vascular calcification and degenerative osseous changes [22]. thereby restricting its cervical spine applications. Quantitative computed tomography (QCT) emerges as a promising alternative, offering enhanced reliability through its capacity to minimize superimposition artifacts that compromise DEXA accuracy [23, 24]. However, significant radiation exposure and cost constraints have restricted its widespread clinical adoption, particularly in cervical bone quality assessment.

At present, the use of cervical vertebral HU values based on cervical three‐dimensional CT measurements to reflect C‐VBQ has been widely investigated. Lee et al. [25] discovered that the vertebral HU value provides a reliable means of assessing bone density in the cervical spine with good accuracy. Similarly, Wang et al. [26] reported that cage subsidence in single‐level ACDF is associated with lower preoperative HU values of the vertebra. Additionally, another study found a significant correlation between cervical HU values and lumbar BMD values [27]. CT‐HU values are more accurate in assessing cervical bone quality due to their site specificity; however, the additional radiation exposure to the patient is still an issue to be considered.

Recently, Ehresman et al. [28] introduced the VBQ score based on T1‐weighted MRI and confirmed the correlation of the VBQ score with the DEXA T‐score and its clinical value in predicting lumbar osteoporosis in patients. Unlike the M‐score proposed by Bandirali et al. [29], which depends on the signal‐to‐noise ratio, the VBQ score includes CSF signal intensity to adjust for baseline signal differences between devices, giving it a wider range of applicability. The application of the VBQ score is not limited to the lumbar spine. Soliman et al. [12] proposed the C‐VBQ score and demonstrated its good application in cervical surgery. Based on this, numerous studies have demonstrated the effective use of the C‐VBQ score in predicting postoperative complications associated with cervical spine surgery, such as cage subsidence following fusion [13, 30]. In recent years, there has been increasing evidence of a significant correlation between the cervical vertebral bone density and LCL after open‐door laminoplasty; however, further investigation is needed to explore the relationship between C‐VBQ score and LCL [18]. Furthermore, considering the excellent visualization of both bone and paraspinal soft tissues provided by MRI, the relationship between the C‐VBQ score and paraspinal muscle quality deserves to be revealed.

4.2. The Predictors of LCL After Open‐Door Laminoplasty

LCL after open‐door laminoplasty appears to be an inevitable postoperative complication that could potentially result in poor clinical outcomes [4, 31]. Our results showed that at 1 year postoperatively, 55 (54.45%) patients lost more than 5° of cervical lordosis, and the mean C2–C7 cobb angle was reduced from 13.30° ± 10.53° to 9.88° ± 11.33° preoperatively.

Previous studies have demonstrated many of the risk factors associated with LCL after open‐door laminoplasty [5, 6]. The results of Lee et al. [32] showed that smaller extension ROM was an independent predictor of LCL, and similarly, another study showed that extension ROM was strongly associated with LCL [33]. In recent years, an increasing number of studies have focused on cervical flexion ROM; Takashi et al. [31] demonstrated that larger preoperative flexion ROM was an independent risk factor for LCL, and Hu et al. [18] found that flexion ROM/extension ROM independently predicted LCL. In our study, flexion/extension ROM, which represents asymmetry of cervical ROM, was moderately correlated with LCL and was an independent risk factor for LCL.

The paraspinal muscles play an important role in maintaining sagittal balance in the cervical spine [34]. Kim et al. [6] showed that the CSA and DFF of the deep extensor muscles were strongly associated with LCL. However, most studies have focused only on the deep extensor muscles, with fewer studies focusing on the deep flexor muscles and the overall quality of the paraspinal muscles of the cervical spine. In our study, flexion muscles DFF was significantly and negatively correlated with LCL; meanwhile, multiple linear regression analysis showed that both total RCSA and average DFF of flexion and extension muscles were risk factors for LCL. Therefore, even though the extension muscles provide a major role in maintaining cervical lordosis, the role of the flexion muscles should not be neglected, and surgical stripping and injury of the extension muscles may accelerate the degeneration of the flexion muscles, which may lead to LCL.

There are few studies focusing on the relationship between cervical bone quality conditions and LCL. Hu et al. [18] showed that the mean HU value in the C2–C7 vertebrae was an independent risk factor for LCL; they suggested that the loss of bone trabeculae in the vertebral body increases the risk of minor fractures and wedge changes at the anterior margin of the vertebral body, which can lead to the development of LCL. Lovecchio et al. [35] demonstrated a positive correlation between cervical HU values and cervical lordosis. However, as a parameter that also reflects the condition of vertebral cancellous bone quality, there are no studies evaluating the relationship between C‐VBQ scores and LCL. In this study, the C‐VBQ values were significantly and positively correlated with LCL, and multiple linear regression analyses showed that C‐VBQ values were another independent risk factor for LCL.

Additionally, similar to previous studies, we observed that patients in the LCL group had lower JOA scores at final follow‐up compared to those in the NLCL group. Open‐door laminoplasty aims to expand the spinal canal volume and allow posterior drift of the spinal cord for indirect decompression. The presence of LCL may hinder sufficient posterior drift, potentially impacting postoperative outcomes. Nevertheless, regardless of LCL occurrence, patients' JOA scores significantly improved after surgery compared to preoperative levels, highlighting the continued effectiveness and importance of this procedure.

4.3. Correlation Between C‐VBQ Scores and Cervical Paravertebral Muscle‐Related Parameters

Muscle and bone are highly homologous in terms of tissue development, while their functional deterioration can both lead to fractures or further loss of body function [36, 37]. Furthermore, sarcopenia is often accompanied by osteoporosis; therefore, Binkley et al. [8] introduced the concept of Osteosarcopenia. A study by Locquet et al. [38] showed that patients with sarcopenia had up to four times the risk of concurrent osteoporosis compared to non‐sarcopenic patients. Therefore, studies of the musculoskeletal system should consider muscle and bone as a whole and think about the essence of muscle‐skeletal interactions. However, there is a lack of an indicator that reflects both the quality of bone and muscle. Given that MRI can adequately visualize both bone and muscle, we consider that the MRI‐based VBQ score could be a potential candidate. Interestingly, Li et al. [15] have demonstrated a significant correlation between lumbar VBQ scores and lumbar paraspinal muscle‐related parameters, proving the potential value of VBQ scores in reflecting lumbar paraspinal muscle quality conditions. However, it remains unclear whether C‐VBQ scores have the same value. Our findings showed that C‐VBQ scores were significantly associated with RCSA and DFF in the cervical deep flexors, deep extensors, and the overall cervical muscles. This is a new finding that demonstrates that the C‐VBQ score is equally a comprehensive index reflecting the quality of Cervical bone and paraspinal muscle. This not only helps to study the cervical spine bones and muscles as a whole but also extends the application of the C‐VBQ score.

4.4. Limitations and Strengths

In this study, we demonstrate for the first time that the C‐VBQ score is not only a reliable indicator of C‐VBQ but also closely associated with the RCSA and DFF of the deep cervical flexors, extensors, and paraspinal muscles. Additionally, we establish that the C‐VBQ score is significantly correlated with LCL following open‐door laminoplasty, identifying it as a novel independent risk factor for LCL.

This study still has limitations. First, this study was a single‐center retrospective study that primarily included patients who underwent open‐door laminoplasty; the sample size was limited, so the results may not be applicable to a broader population. Second, there is still no consensus on the cut‐off value of the C‐VBQ score for assessing cervical bone quality, and errors may still exist as it is a manually measured parameter. This study only preliminarily verified the strong relationship between the C‐VBQ score and the quality of cervical bone and paraspinal muscles; however, determining the ideal cut‐off value requires a larger sample size and a multicenter prospective study. Finally, due to the large number of variables included in this study, other factors such as the quality of each cervical paraspinal muscle or vertebra and anti‐osteoporotic treatment were not further analyzed. These factors may provide additional information about the risk factors associated with LCL and need to be further investigated.

5. Conclusion

In this study, we found for the first time that the C‐VBQ score not only reflects the condition of cervical bone quality, but also strongly correlates with the cervical deep flexors, deep extensors, and the overall paraspinal muscles RCSA and DFF. The C‐VBQ score is a comprehensive indicator of the quality of cervical bone and paraspinal muscles. In addition, our findings demonstrate for the first time that the C‐VBQ score was strongly associated with LCL after open‐door laminoplasty and was a novel independent risk factor for LCL.

Author Contributions

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by J.L., X.L. The first draft of the manuscript was written by J.L, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Disclosure

All authors listed meet the authorship criteria according to the latest guidelines of the International Committee of Medical Journal Editors, and all authors are in agreement with the manuscript.

Ethics Statement

This retrospective study was in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The ethics committee of West China Hospital, Sichuan University approved this study (No. 2023252).

Consent

The information relating to the patients is completely anonymized and the submission does not include images that may identify the person.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

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