Effects of thoracolumbar fracture combined with muscle injury on long-term outcome and prognosis: a retrospective study

Abstract

Background

To explore the relationship between muscle injury and severity, stability, and clinical outcomes in thoracolumbar fracture patients and identify prognostic factors.

Methods

Clinical data from 118 thoracolumbar fracture patients (January 2017–December 2019) were retrospectively analyzed. Detailed information on patients’ clinical conditions, muscle injury degree, and spinal injuries was collected. Evaluation included TLICS, LSC, PLC, and IVD injury classifications, Denis grading, Sander grading, and preoperative load sharing score. Multifactor and correlation analyses identified factors influencing lumbar spine function and severe muscle injury.

Results

Muscle injury incidence post-thoracolumbar fracture was 83.05%, with moderate and severe injuries at 50.85%. At follow-up, kyphotic correction loss averaged 3.26 ± 1.83°, and ODI score after internal fixation was 14.78 ± 6.17. Muscle injury correlated positively with spinal fracture severity and stability. Significant associations were found between kyphotic correction loss and muscle injury in lumbar fractures. Multifactor analysis identified TLICS grading, PLC grading, and LSC score as key factors influencing ODI and severe muscle injury.

Conclusions

This study addresses a knowledge gap on spinal fractures and associated muscle injuries, revealing a correlation between higher-grade spinal fractures and more severe muscle injury. Significantly, severe muscle injury was associated with greater kyphotic correction loss and worse mid- to long-term lumbar functional recovery.

Introduction

Thoracolumbar fracture is a common spinal injury, commonly seen in traffic accidents, falls and sports injuries [1, 2]. Patients with fractures are often accompanied by varying degrees of muscle injury [3], which may have an important impact on long-term efficacy and prognosis. At present, a hot topic of research is the influence of lumbar muscle lipomatosis [4, 5] on back pain and iatrogenic muscle injury caused by surgery [6, 7]. However, at present, the research on thoracolumbar fracture complicated with muscle injury is relatively limited, and the in-depth understanding of its impact needs to be further explored. The literature on thoracolumbar fracture has involved different treatment methods [8–10], surgical techniques and functional rehabilitation [11]. Several studies have focused on the classification of fractures [12], treatment strategies [13, 14], and timing of surgery [15]. There have also been studies on thoracic blunt aortic injury caused by thoracic vertebra fracture [16], but there is still a gap in knowledge about combined muscle injury, especially its impact on long-term recovery and quality of life of patients. Therefore, a retrospective study of thoracolumbar fractures associated with muscle injury will provide insight into the characteristics, treatment difficulties, and possible prognostic factors of this condition. Through the systematic collection and analysis of clinical and imaging data of patients, we can explore the relationship between muscle injury and fracture healing and functional rehabilitation, so as to provide scientific basis for formulating more effective treatment programs.

Data and methods

General information

The cases of thoracolumbar fracture treated in our hospital from January 2017 to December 2019 were retrospectively analyzed. Inclusion criteria: (1) thoracolumbar fracture patients aged 18–65 years with posterior pedicle screw reduction and fixation; (2) Complete image data: preoperative plain X-ray, MR And CT, complete follow-up X-ray (all data were taken and stored in the PACS system of our imaging center); (3) Complete the clinical follow-up of this study. Exclusion criteria: (1) leaping multisegmental fracture. History of osteoporosis and other spine and spinal cord; (2) The internal fixation is loose or broken; (3) anterior fixation fusion cases; (4) Cases with nerve injury. The study was approved by our hospital. Written informed consent was obtained from all individuals included in this study.

Surgical method

All patients received posterior pedicle screw system internal fixation. Specific operation: Assisted the patient in supine position, placed the soft pillow on the chest and abdomen to keep the abdomen suspended, gave the patient trachea intubation under general anesthesia, disinfected the skin of the operative area with alcohol, and positioned the vertebral fracture under the guidance of C-arm machine [XHX500 model No., Shandong Food and Drug Administration No. 2014 2300674]. A longitudinal incision was made in the posterior median with this as the center, and the subcutaneous tissue was separated obtuse, and the spinous muscle was removed layer by layer along the spinous process, so that the lamina and articular process at the injured vertebra were fully exposed. The fracture site was explored, and the needle entry point was determined according to the residual joint space between the transverse process and the upper and lower articular processes. All patients underwent short-segment posterior pedicle screw instrumentation, with bilateral pedicle screws placed in the vertebrae immediately superior and inferior to the fractured vertebra (total of four screws: two in the superior vertebra and two in the inferior vertebra). No pedicle screws were placed at the fractured vertebra level (index-level fixation was not performed). Fixation rods were then connected to the screw tails. Reduction was confirmed under C-arm fluoroscopy before final tightening. All surgical procedures were performed via a posterior-only approach with short-segment pedicle screw instrumentation. No anterior procedures, discectomy, interbody fusion, or other direct interventions for intervertebral disc injury were undertaken in any case.

After the reset is satisfied under the C-arm machine, tighten the nut. The wound was rinsed with biological saline, the drainage tube was indwelled, and the incision was sutured.

Postoperative management

Prophylactic antibiotics were used in the first 3 days after surgery, the drainage tube was removed when the incision drainage volume was less than 40 mL/d, and the thoracolumbar brace was used to promote early postoperative activity.

Observation index

1. Imaging evaluation: The local kyphotic Angle (the Angle between the upper endplate of the injured vertebra and the extension line of the lower endplate of the injured vertebra on the lateral film, LKA) of the injured vertebra was measured and recorded before surgery, immediately after surgery (3 days after surgery) and at the last follow-up X-ray. LKA loss = (LKA at last follow-up - LKA immediately after surgery). The final follow-up was uniformly set as the last X-ray taken before removal of internal fixation

2. Preoperative evaluation of related injuries: preoperative thoracolumbra injury classification score (TLICS) [17], load sharing classification (LSC) score [18, 19], posterior ligament complex (PLC) injury grade [20, 21], intervertebral disc (IVD) injury grade [22] were recorded.

3. To observe the degree of injury of thoracolumbar and lumbar muscles: mainly to observe the multifidus muscle, erector spinalis muscle and psoas major muscle. According to the muscle injury scale of British athletes and the lumbar fatty lesion scale system [23, 24], the muscle injury associated with thoracolumbar fracture was divided into four grades: grade 0 without injury (Fig. 1); Grade 1 fascia injury, or intramuscular scattered cloud high signal (Fig. 2); Grade 2 Broad continuous muscle high signal, but not more than 50% cross section or 10 cm longitudinal length (Fig. 3); Grade 3 multi-muscle broad continuous high signal, large range, more than 50% in cross section or more than 10 cm in longitudinal length, or cross section showing significant enlargement of muscle profile (Fig. 4).

4. Evaluation of lumbar function after removal of internal fixation Oswestry Disability Index (ODI) [25] was used to evaluate the degree of low back pain of patients at the last follow-up visit. The ODI scale consisted of 10 items with scores ranging from 0 to 5, and the total score was 50. Score method: ODI = actual score/50 × 100%, higher score indicates more serious dysfunction.

5. To evaluate the correlation between the degree of muscle injury and fracture-related indicators and long-term efficacy.

Fig. 1.

Thoracolumbar fracture with grade 0 muscle injury

Fig. 2.

Thoracolumbar fracture with grade 1 muscle injury

Fig. 3.

Thoracolumbar fracture with grade 2 muscle injury

Fig. 4.

Thoracolumbar fracture with grade 3 muscle injury

Statistical analysis

Data analysis was performed using SPSS 26.0 statistical software. The data of normal distribution is expressed as Mean ± SD, and the measurement data is expressed as rate (%). Pearson correlation analysis was used to evaluate the correlation of each index, and multi-factor regression analysis was used to detect the risk factors. P < 0.05 indicated that the difference was statistically significant.

Result

A total of 118 patients were selected in this study, including 69 males and 49 females, with an average age of 47.66 ± 10.97 years (age range 18–63 years). Internal fixation was removed 10–24 months after surgery, with a mean of 13.50 ± 7.82 months, and telephone ODI was completed 15–36 months after surgery, with a mean of 20.47 ± 6.54 months. No iatrogenic or secondary nerve damage was found in all the selected cases, including 21 cases of T11-12, 64 cases of L1-2, and 33 cases of L3-5. There were 20 (16.95%), 38 (32.20%), 17 (14.41%) and 43 (36.44%) cases of muscle injury in grade 0, 1, 2 and 3, respectively. TLICS classification was used to evaluate the severity of thoracolumbra and lumbar spine injuries. There were 43 cases (36.44%), 24 cases (20.34%) and 51 cases (43.22%) of grade 1, grade 2 and grade 3, respectively. Denis classification system was used to evaluate the damage of dorsal ligament structure (PLC). There were 41 cases (34.75%), 20 cases (16.95%), 17 cases (14.41%) and 40 cases (33.90%) in grade 0, 1, 2 and 3, respectively. According to the Sander grading system, there were 36 (30.51%), 30 (25.42%), 23 (19.49%), and 29 (24.58%) patients with grade 0, grade 1, grade 2, and grade 3 injuries. The severity and stability of thoracolumbar fractures were assessed by the preoperative load sharing score (LSC), which was 7.57 ± 2.63. Of the 118 patients, 72 (61.0%) had an LSC score ≥ 7, indicating potentially compromised anterior load-sharing. In addition, the LKA before, after and at the last follow-up were 20.88 ± 3.89°, 2.17 ± 0.57° and 5.43 ± 1.77°, respectively, with the difference of LKA being 3.26 ± 1.83°. The ODI score is 14.78 ± 6.17, as shown in Table 1.

Table 1.

The basic clinical conditions and long-term curative effects of the patients

Index	N = 118, Mean ± SD or n(%)
Age (years)	47.66 ± 10.97
Sex
Male	69(58.47%)
Female	49(41.53%)
Fracture type
T11−12	21(17.80%)
L1−2	64(54.24%)
L3−5	33(27.97%)
TLICS classification
Grade 1	43(36.44%)
Grade 2	24(20.34%)
Grade 3	51(43.22%)
Muscle injury grading
Grade 0	20(16.95%)
Grade 1	38(32.20%)
Grade 2	17(14.41%)
Grade 3	43(36.44%)
PLC classification
Grade 0	41(34.75%)
Grade 1	20(16.95%)
Grade 2	17(14.41%)
Grade 3	40(33.90%)
IVD classification
Grade 0	36(30.51%)
Grade 1	30(25.42%)
Grade 2	23(19.49%)
Grade 3	29(24.58%)
LSC score	7.57 ± 2.63
LKA (°)
Before operation	20.88 ± 3.89
Postoperation	2.17 ± 0.57
Last follow-up	5.43 ± 1.77
LKA loss (°)	3.26 ± 1.83
ODI score	14.78 ± 6.17

The correlation between the degree of muscle injury and other relevant indicators

As shown in Table 2, in addition to gender, The degree of muscle injury was correlated with age (r = 0.730), fracture type (r = 0.769), TLICS classification (r = 0.863), PLC classification (r = 0.840), and IVD There were significant positive correlations among grading (r = 0.430), LSC score (r = 0.852), LKA difference (r = 0.803) and ODI score (r = 0.847) (p < 0.01), the degree of muscle injury was most closely related to TLICS grade, followed by LSC score. At the same time, LKA difference and ODI score were significantly positively correlated with muscle injury degree, age, fracture type, TLICS grade, PLC grade, IVD grade and LSC score (p < 0.01).

Table 2.

Correlation of various indicators of patients

Items	Muscle injury grading	Sex	Age	Fracture type	TLICS classification	PLC classification	IVD classification	LSC score	LKA loss	ODI score
Muscle injury grading	1
Sex	0.053	1
Age	0.730**	0.036	1
Fracture type	0.769**	0.102	0.511**	1
TLICS classification	0.863**	0.084	0.627**	0.615**	1
PLC classification	0.840**	0.057	0.601**	0.660**	0.749**	1
IVD classification	0.430**	0.04	0.260**	0.279**	0.296**	0.470**	1
LSC score	0.852**	0.069	0.605**	0.662**	0.719**	0.859**	0.411**	1
LKA loss	0.803**	0.087	0.583**	0.490**	0.792**	0.709**	0.349**	0.736**	1
ODI score	0.847**	0.069	0.686**	0.656**	0.776**	0.713**	0.324**	0.787**	0.714**	1

* p < 0.05 ** p < 0.01

Multi-factor analysis of ODI

In addition, by multivariate analysis, fracture type (OR: 6.409, 95%CI: 1.926–18.376, P = 0.001), TLICS grade (OR: 4.814, 95%CI: 1.777–7.501, P = 0.013), IVD grade (OR: 0.141, 95%CI: 0.022 ~ 0.900, P = 0.038), LSC score (OR: 2.635, 95%CI: 1.182 ~ 5.873, P = 0.018) and LKA difference (OR: 3.006, 95%CI: 1.230 ~ 7.349, P = 0.016) were the risk factors affecting ODI (Table 3).

Table 3.

Multi-factor analysis of ODI

Items	Regression coefficient	OR value	95% CI	p value
Sex	−0.208	0.812	0.158 ~ 4.170	0.803
Fracture type	4.569	6.409	1.926 ~ 18.376	0.001
TLICS classification	2.696	4.814	1.777 ~ 7.501	0.013
Muscle injury grading	0.642	1.901	0.148 ~ 14.440	0.622
PLC classification	1.619	0.198	0.027 ~ 1.444	0.11
IVD classification	1.957	0.141	0.022 ~ 0.900	0.038
LSS score	0.969	2.635	1.182 ~ 5.873	0.018
LKA loss	1.101	3.006	1.230 ~ 7.349	0.016

P < 0.05 indicates significant

Multifactorial analysis of severe muscle injury

At the same time, TLICS classification (OR: 0.145, 95%CI: 0.068 ~ 0.644, P = 0.040), PLC classification (OR: 0.132, 95%CI: 0.310 ~ 1.251, P = 0.007), LSC score (OR: 0.294, 95%CI: 0.056 ~ 0.358, P = 0.019) were also risk factors for severe muscle injury (Table 4).

Table 4.

Analysis of multiple factors affecting muscle injury

Items	Regression coefficient	OR value	95% CI	p value
Sex	1.382	1.982	0.000 ~ 0.754	0.197
Age	−1.003	0.367	0.021 ~ 0.855	0.134
TLICS classification	1.694	0.145	0.068 ~ 0.644	0.040
PLC classification	1.135	0.132	0.310 ~ 1.251	0.007
IVD classification	0.083	0.893	0.147 ~ 1.301	0.075
LSS score	1.013	0.294	0.056 ~ 0.358	0.019
Fracture type	−0.246	1.033	0.745 ~ 1.825	0.115

P < 0.05 indicates significant

Discussion

According to the literature survey, there are relatively few studies on spinal fracture combined with muscle injury, and our study fills this knowledge gap. The muscles in the thoracolumbar section of the spine are mainly composed of the multifidus muscle, the vertical spinal muscle and the psoas major muscle, which play a key role in the all direction movement of the spine, including flexion and extension, rotation and lateral flexion, etc [26]. These muscles cooperate with intervertebral disc, ligament and other structures to maintain the stability and balance of the spine [27, 28], while muscle dysfunction will disrupt the normal function of other spinal components. And may lead to spinal disorders [29]. Gao et al. [5] showed that posterior fascia injury, paravertebral muscle steatosis and facet joint invasion were identified as independent risk factors for residual back pain after percutaneous vertebral augmentation, in which paravertebral muscle steatosis played an important role.

Given the lack of a standardized classification for acute trauma-induced paraspinal muscle injury in thoracolumbar fractures, we adapted the British Athletics Muscle Injury Classification [23, 24], which grades lower-limb injuries based on MRI T2 signal changes (fascial involvement, edema extent, and muscle fiber disruption). This system was chosen because acute paraspinal muscle trauma manifests analogous MRI findings—scattered or broad T2 hyperintensity reflecting edema/hemorrhage—and provides an objective, reproducible framework for severity assessment. Elements from Kader’s paraspinal muscle fatty infiltration grading [30] were incorporated to better account for cross-sectional involvement. This novel adaptation allowed consistent grading across our cohort despite the absence of spine-specific tools.

In the muscle injury classification of British athletes [24], the lower limb muscle injury was divided into four grades according to T2 imaging on MR: fascia and intramuscular edema, edema of different degrees with high signal and even fissure changes in muscles. Kader DF et al. [30] observed the high signal area in the paravertebral muscle on the T2 axial image as fat replacement, and divided the paravertebral muscle fat into 4 grades according to the ratio of the cross-sectional area of fat. With comprehensive reference to the above classification methods, this study divided the muscle injury associated with thoracolumbar fracture into four grades to more accurately reflect the degree of injury. Among them, grade 0 had no injury; Grade 1 fascia injury, or intramuscular scattered cloudy high signal; Grade 2 broad continuous muscle high signal, but not more than 50% cross section or longitudinal length not more than 10 cm; Grade 3 multi-muscle broad continuous high signal, large range, more than 50% cross section or more than 10 cm longitudinal length, or cross section showing significant enlargement of muscle profile.

The results showed that the incidence of muscle injury after thoracolumbar fracture was higher by 83.05% (98/118), of which the moderate and severe injury accounted for more than 50.85% (60/118). Further correlation analysis found significant positive associations between the degree of muscle injury and the severity of the fracture (TLICS, PLC, IVD) and spinal stability assessment (LSC score) (p < 0.01). The results of multivariate regression analysis showed that TLICS grading, PLC grading, and LSC score were risk factors for severe muscle injury, further emphasizing that higher grade spinal fractures may be associated with more severe muscle injury. In general, thoracolumbra and lumbar spine fractures are mostly high-energy injuries, and extreme flexion of the spine at the moment of injury causes tensile injury of muscles. Some injuries may also be combined with direct muscle injury, such as direct impact injury in traffic accidents [1]. These factors lead to destruction of muscle fibers and denervation damage [31, 32], and the damage worsens with the increase of injury energy. Therefore, the risk and management of muscle injury should also be considered in the management of spinal fractures, especially when assessing the severity and stability of the fracture.

IVD injury was assessed preoperatively using an adapted Pfirrmann-based grading system on MRI [22], and higher grades were associated with worse long-term functional outcomes in multivariate analysis. However, no patients received specific treatment directed at the injured disc (e.g., discectomy or interbody reconstruction), as the uniform surgical strategy was posterior short-segment fixation only. This posterior-only approach proved sufficient for maintaining correction and achieving good functional results in this neurologically intact cohort, even in cases with grade 3 IVD injury. These findings align with evidence that many thoracolumbar fractures with disc involvement can be effectively managed posteriorly without anterior supplementation when there is no neurological deficit. Nonetheless, severe IVD injury may contribute to residual back pain or dysfunction through persistent biomechanical instability, warranting consideration in treatment planning.

The short-term effect of posterior reduction and fixation for thoracolumbar fracture is good, but the long-term effect is not satisfactory [33]. Loss of postoperative kyphotic correction showed a clear trend in our study, with LKA of 3.26 ± 1.83°, similar to previous reports [34, 35]. Notably, we found for the first time a significant positive association between kyphotic correction loss and muscle injury associated with lumbar fracture, a finding that highlights the possibility that mild muscle injury may heal on its own, while severe muscle injury is associated with greater disability, subsequent fatty infiltration, and potential compromise of muscle stability and sagittal alignment of the spine. Some studies have also pointed out that kyphotic deformity and degeneration will further increase the tension of paravertebral muscle, leading to strain and fatization, and eventually evolve into degenerative changes of lumbar spine [36, 37].

In addition, ODI results of follow-up patients showed that most of the patients complained of mild discomfort 15–36 months after surgery, and the average ODI score was 14.78 ± 6.17, suggesting that patients with thoracolumbar fracture had various degrees of low back pain and mobility disorders. In addition, our study further showed that medium and long term ODI was closely related to muscle injury and LKA loss, and was affected by fracture type, TLICS grade, IVD grade, LSC score and other factors, which was consistent with previous studies [38].

Regarding surgical strategy, all patients in this cohort were treated with short-segment posterior instrumentation spanning one level above and one below the fractured vertebra, without intermediate screws at the index level. The mean preoperative Load Sharing Classification (LSC) score was 7.57 ± 2.63, with many patients scoring ≥ 7, a threshold at which anterior column support or longer/more robust fixation is often recommended due to increased risk of construct failure [18, 19]. Despite this, we observed satisfactory maintenance of kyphotic correction (mean loss 3.26 ± 1.83°) and good functional outcomes (mean ODI 14.78 ± 6.17) at mid- to long-term follow-up, with no implant loosening or breakage in the included cases. This suggests that, in neurologically intact patients with careful selection and rigorous postoperative management, short-segment fixation can yield acceptable results even in higher LSC cases. However, cases with implant failure were excluded per study criteria, which may introduce selection bias toward successful outcomes. The fractured vertebra itself was not instrumented in any case, consistent with traditional short-segment technique rather than index-level screw augmentation.

Research limitations and future directions

Although our study reached some important conclusions, there are also some limitations. First, the sample size was relatively small and the study design was retrospective, which may lead to selection bias. Second, this study only focused on one type of treatment, and failed to explore the impact of different treatment methods on muscle injury and clinical results. Future studies can further explore the mechanism of spinal fracture combined with muscle injury by expanding the sample size and using a multicenter, prospective study design to provide a more comprehensive basis for more accurate treatment. Additionally, the uniform use of short-segment posterior fixation without index-level screws or anterior supplementation limits generalizability to patients with very high LSC scores (≥ 7–9), where alternative strategies (e.g., intermediate screws, longer constructs, or combined approaches) are frequently advocated to enhance construct stability. Furthermore, the exclusion of cases with internal fixation loosening or breakage may bias the results toward more favorable radiographic and functional outcomes, and precludes analysis of the potential role of severe muscle injury in contributing to implant failure. Nevertheless, the exclusion of patients with neurological injury, while intended to homogenize the cohort and isolate the effects of paraspinal muscle injury on pain, function, and sagittal alignment, limits the external validity of our findings. These results are primarily applicable to neurologically intact thoracolumbar fractures and may not fully reflect outcomes in the broader trauma population, where neurological deficits are common and often exert a greater influence on long-term prognosis than muscle injury alone. Furthermore, functional outcomes were evaluated exclusively using the ODI, which primarily assesses pain-related lumbar disability. While ODI is reliable and relevant for this population, it may not fully encompass broader health-related quality of life, psychological aspects, or objective functional recovery. Complementary instruments such as the SF-36 or EQ-5D (for general health status) and metrics like return-to-work rates were not available in this retrospective cohort. Future prospective studies should employ multimodal outcome measures to provide a more holistic assessment of long-term recovery. Additionally, ODI assessments were conducted at variable follow-up intervals (15–36 months post-surgery), and no statistical adjustments were made for follow-up duration in the analyses. Although all evaluations occurred after internal fixation removal and within a mid- to long-term timeframe, this variability could potentially confound results, as longer follow-up might permit progressive muscle changes or degenerative processes affecting functional outcomes. Future prospective studies with standardized follow-up timing would help mitigate this issue.

Conclusion

This study fills a gap in the knowledge of spinal fractures associated with muscle injury, highlighting that higher-grade spinal fractures may be associated with more severe muscle injury, and that severe muscle injury was associated with greater kyphotic correction loss and worse mid- to long-term lumbar functional recovery. These findings suggest that preoperative assessment of paraspinal muscle injury should be incorporated into prognostic evaluation and treatment planning for thoracolumbar fracture patients.

Abbreviations

TLICS

Thoracolumbra injury classification score

LSC

Load sharing classification

PLC

Posterior ligament complex

IVD

Intervertebral disc

ODI

Oswestry Disability Index

LSC

Load sharing score

Authors’ contributions

MSZ is resposible for the guarantor of integrity of the entire study, study concepts & design, definition of intellectual content, literature research, clinical studies, data acquisition & analysis, statistical analysis, manuscript preparation; JBX is resposible for the clinical studies, data acquisition & analysis; SQL is resposible for the clinical studies, data acquisition & analysis; MY is resposible for the clinical studies, manuscript editing & review. All authors read and approved the final manuscript.

Funding

The authors have not received any funding support.

Data availability

All data generated or analysed during this study are included in this. Further enquiries can be directed to the corresponding author.

Declarations

Ethics approval and consent to participate

This was a study undertaken in compliance with the Declaration of Helsinki.The study was approved by the First Affiliated Hospital, Yijishan Hospital, Wannan Medical College. Informed consent was waived because of the retrospective nature of the study, with all data anonymized for privacy protection.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.