Role of Thoracolumbar Fascia Injury Classification in Predicting Percutaneous Vertebroplasty Outcomes for Osteoporotic Vertebral Compression Fractures

Abstract

Objective

Thoracolumbar fascia injury is one of the important causes of suboptimal early analgesic outcomes following percutaneous vertebroplasty (PVP). Previous studies have predominantly treated thoracolumbar fascia injury as a binary variable, overlooking the heterogeneity in the extent of injury. This study aimed to establish an MRI-based classification system for thoracolumbar fascia injury and to evaluate the impact of different classification types on early postoperative pain relief and time to ambulation after PVP.

Methods

A retrospective analysis was conducted on the medical records of patients with osteoporotic vertebral compression fractures (OVCFs) who underwent PVP at our institution between January 2023 and January 2025. A total of 56 patients with concomitant thoracolumbar fascia injury were enrolled. Based on the longitudinal extent of fascial edema on preoperative MRI T2-weighted short tau inversion recovery (T2-STIR) sequences, patients were classified into the localized type (n=11), segmental continuous type (n=21), and extensive multisegmental type (n=24). The reproducibility of the classification was evaluated using interobserver and intraobserver reliability analyses based on Cohen’s kappa coefficients. The primary patient-reported outcome measures (PROMs) were the visual analogue scale (VAS) scores during turning over and standing. The Oswestry Disability Index (ODI) served as the secondary PROM. Additionally, the achievement rates of the minimal clinically important difference (MCID) and patient acceptable symptom state (PASS) for VAS and ODI at the final follow-up were evaluated across groups.

Results

No statistically significant differences in baseline PROMs were observed among the three groups (P >0.05). The interobserver and intraobserver reliability for the classification were substantial to almost perfect, with Cohen’s kappa coefficients of 0.85 (95% CI, 0.74–0.96) and 0.88 (95% CI, 0.78–0.98), respectively. Within the first postoperative month, the standing VAS scores and ODI in the extensive multisegmental type group were significantly higher than those in the segmental continuous type and localized type groups (P <0.05), and the segmental continuous type group was also significantly higher than the localized type group (P <0.05). At 3 months postoperatively and at the final follow-up, no statistically significant differences were observed among the three groups (P >0.05). At the final follow-up, the achievement rates of MCID and PASS did not differ significantly among the three groups (P >0.05). On postoperative days 1 and 3, the proportion of patients meeting ambulation criteria in the localized type group was significantly higher than that in the segmental continuous type and extensive multisegmental type groups (P <0.05). Following comprehensive treatment, all patients met the ambulation criteria within 7 days postoperatively.

Conclusion

The findings of this preliminary study suggest that the classification of thoracolumbar fascia injury is closely associated with the degree of early postoperative pain relief and time to ambulation after PVP. The more extensive the injury, the more severe the early residual pain and the more delayed the ambulation. Preoperative MRI-based classification of thoracolumbar fascia injury should be emphasized to more accurately predict postoperative outcomes following PVP, formulate individualized perioperative management plans, such as proactive multimodal analgesia and appropriately delayed mobilization for patients with extensive multisegmental injury, and provide patients with more targeted prognostic counseling.

Introduction

Osteoporotic vertebral compression fracture (OVCF) is the most common type of fragility fracture in the elderly population, caused by minimal force (such as a ground-level fall) or even occurring in the absence of obvious trauma (such as coughing).^1–3 Percutaneous vertebroplasty (PVP), owing to its advantages of minimal invasiveness, rapid efficacy, and short hospital stay, has become one of the preferred minimally invasive procedures for the treatment of OVCFs.^2,4,5 However, a proportion of patients still experience mild to moderate residual pain after PVP, with reported incidence rates ranging from 4.6% to 27.9%.^3,6–10 Previous studies have predominantly focused on the influence of technical factors such as cement distribution, cement leakage, and the degree of vertebral height restoration on outcomes, while relatively insufficient attention has been paid to soft tissue injuries beyond the vertebral fracture itself—particularly thoracolumbar fascia injury.¹¹

The thoracolumbar fascia (TLF) is a critical fascial structure overlying the lumbar and dorsal musculature, playing a key role in maintaining spinal stability, transmitting muscular forces, and coordinating the movement of the lumbar and dorsal muscle groups.^12–17 During the occurrence of OVCF, vertebral body compression and collapse can subject the surrounding soft tissues to traction, tearing, or compression, and thoracolumbar fascia injury is not uncommon in patients with OVCFs.^6–10 Studies have reported that the incidence of thoracolumbar fascia injury in patients with acute OVCFs ranges from 7.4% to 45%.^10–12 Furthermore, nociceptive free nerve endings are present within the thoracolumbar fascia, and the application of noxious stimuli to the thoracolumbar fascia can elicit pain, suggesting that thoracolumbar fascia injury may be an important source of low back pain.^13–16 Therefore, in OVCF patients with concomitant thoracolumbar fascia injury, pain arising from the fascial injury is superimposed upon the pain caused by the vertebral fracture itself.¹⁶ Previous studies have identified a potential association between preoperative thoracolumbar fascia injury and residual pain after PVP.^10–12 Yang et al,¹⁰ through a retrospective analysis of 1,316 OVCF patients, confirmed that thoracolumbar fascia injury is an independent risk factor for residual pain after PVP (OR=3.805, P=0.002). However, previous studies have predominantly analyzed thoracolumbar fascia injury as a binary variable (present/absent), neglecting the inherent heterogeneity in the extent and severity of injury. The present study established an MRI-based classification system for thoracolumbar fascia injury and evaluated the impact of different injury types on early postoperative pain relief and time to ambulation after PVP, with the aim of providing more precise evidence for preoperative prognostic assessment and perioperative management.

Materials and Methods

Patient Selection

A retrospective analysis was conducted on the medical records of OVCF patients who underwent PVP at our institution between January 2023 and January 2025. This study was conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Ethics Review Committee of State Grid Beijing Electric Power Hospital (No. 2025041530205).

Inclusion Criteria

(1) Diagnosis of osteoporosis with a bone mineral density T-score ≤ −2.5 SD; (2) Single-level fresh thoracolumbar (T6–L5) vertebral compression fracture, with MRI demonstrating low signal intensity on T1-weighted imaging (T1WI) and high signal intensity on T2-weighted fat-suppressed sequences (T2WI) and T2-weighted short tau inversion recovery (T2-STIR); (3) Age ≥ 60 years; (4) Pain VAS score ≥ 4; (5) Preoperative MRI confirmation of concomitant thoracolumbar fascia injury; (6) Willing to undergo PVP surgery.

Exclusion Criteria

(1) Pathological fractures (tumor, infection, etc).; (2) Concomitant neurological injury requiring decompressive surgery; (3) MRI demonstrating posterior vertebral wall incompleteness or bony fragment retropulsion into the spinal canal; (4) History of prior spinal surgery; (5) Concomitant Alzheimer’s disease, Parkinson’s disease, or other psychiatric or psychological disorders; (6) Inability to ambulate prior to the index fracture; (7) Incomplete clinical or imaging data; (8) Local skin infection in the thoracolumbar dorsal region; (9) Coagulation dysfunction; (10) Patients who refused PVP surgery.

A total of 56 OVCF patients with concomitant thoracolumbar fascia injury were ultimately enrolled and classified into the localized type (n=11), segmental continuous type (n=21), and extensive multisegmental type (n=24) according to the thoracolumbar fascia injury classification criteria (see below for details).

MRI Diagnostic Criteria and Classification Method for Thoracolumbar Fascia Injury

Diagnostic Criteria

The diagnosis of thoracolumbar fascia injury was based on a combination of MRI findings and physical examination. On MRI, fascial edema manifested as areas of high signal intensity on T2WI and T2-STIR images and low signal intensity on T1WI. In general, thoracolumbar fascia injury appears as elongated or sheet-like areas of high signal on MRI, potentially involving multiple segments. Diagnosis based solely on T1WI and T2WI is prone to missed detection; T2-STIR images are the most sensitive for identifying fascial edema signals. Physical examination required the presence of localized tenderness at the segments corresponding to the fascial edema. Both criteria had to be simultaneously satisfied for diagnosis.

Classification Criteria

Two orthopedic surgeons with more than 5 years of experience in spinal MRI interpretation independently reviewed the images, and discrepancies were resolved by consensus. Based on the longitudinal extent of fascial edema (number of vertebral segments involved) on T2-STIR images, and referring to previous literature on segmental classification of spinal pathologies as well as our preliminary clinical observation that symptom severity tends to escalate when edema extends beyond 2 and 4 segments, thoracolumbar fascia injury was classified into the following three types: Localized type (Type I): Fascial edema signal on MRI T2-STIR images involving ≤ 2 vertebral segments longitudinally (Figure 1A). Segmental continuous type (Type II): Fascial edema signal on MRI T2-STIR images involving 3 to 4 vertebral segments longitudinally, presenting as a continuous band-like area of high signal (Figure 1B). Extensive multisegmental type (Type III): Fascial edema signal on MRI T2-STIR images involving ≥ 5 vertebral segments longitudinally, or presenting as multiple discontinuous large sheet-like areas of high signal (Figure 1C).

Figure 1.

(A) Magnetic resonance imaging (MRI) with T2-weighted short tau inversion recovery (T2-STIR) sequence shows an OVCF at T12 with thoracolumbar fascia injury involving a single vertebral segment (red arrow), consistent with the Localized type (Type I). (B) T2-STIR MRI demonstrates an OVCF at L2 with thoracolumbar fascia injury extending across three vertebral segments (red arrow), consistent with the Segmental continuous type (Type II). (C) T2-STIR MRI reveals an OVCF at L3 with thoracolumbar fascia injury spanning five vertebral segments (red arrow), consistent with the Extensive multisegmental type (Type III).

To evaluate the reproducibility of the proposed classification, two orthopedic surgeons independently classified the MRI scans of all 56 patients. After a 2-week interval, the same observers repeated the classification independently. Interobserver and intraobserver reliability were assessed using Cohen’s kappa coefficients.

Preoperative data recorded included sex, age, body mass index (BMI), lumbar spine bone mineral density T-score, fracture duration, fractured vertebral segment, and injury type. Fracture duration was defined as the time interval from fracture occurrence to PVP surgery. Vertebral fracture location was recorded as thoracic, thoracolumbar junction, or lumbar fracture. Injury type was categorized as overt trauma, occult trauma, or no apparent trauma. Overt trauma referred to a definite history of trauma, such as a ground-level fall; occult trauma referred to fractures caused by lifting heavy objects, lumbar torsion, or vigorous coughing. Baseline characteristics of the three groups are presented in Table 1.

Table 1.

Baseline Characteristics of Patients in the Localized, Segmental Continuous, and Extensive Multisegmental Groups

Characteristic	Localized Type (n=11)	Segmental Continuous Type (n=21)	Extensive Multisegmental Type (n=24)	P-value
Female, n (%)	8 (72.7%)	16 (76.2%)	19 (79.2%)	0.871
Age (years)	71.45±6.83	72.38±7.21	73.17±7.54	0.743
Bone mineral density (T-score)	−3.05±0.43	−3.12±0.49	−3.18±0.52	0.681
BMI (kg/m2)	24.83±2.87	24.51±3.02	24.29±3.18	0.812
Fracture duration (days)	3.36±1.12	3.65±1.48	3.81±1.67	0.716
Fracture location, n (%)				0.843
Thoracic spine	1 (9.1%)	2 (9.5%)	2 (8.3%)
Thoracolumbar junction	8 (72.7%)	15 (71.4%)	17 (70.8%)
Lumbar spine	2 (18.2%)	4 (19.0%)	5 (20.8%)
Injury type, n (%)				0.762
Overt trauma	4 (36.4%)	7 (33.3%)	7 (29.2%)
Occult trauma	3 (27.3%)	6 (28.6%)	7 (29.2%)
No apparent trauma	4 (36.4%)	8 (38.1%)	10 (41.7%)

Notes: Data are presented as mean ± standard deviation (SD) or number (%), unless otherwise specified.

Abbreviation: BMI, body mass index.

Surgical Procedure

A unilateral transpedicular or extrapedicular approach was employed. Under C-arm fluoroscopic guidance, the trocar was advanced to the anterior-middle third of the vertebral body. Polymethylmethacrylate (PMMA) bone cement was prepared to the late filament-drawing phase (toothpaste-like consistency) and slowly injected under continuous lateral fluoroscopic monitoring. Injection was immediately terminated when the cement reached the posterior vertebral wall or when a leakage tendency was observed. Operative time, cement volume, and leakage events were recorded. Postoperative systemic anti-osteoporosis therapy was administered.

Patient-Reported Outcome Measures (PROMs)

PROMs were used to evaluate treatment outcomes in this study. The VAS (0–10 points) was used to assess thoracolumbar pain during turning over and standing at preoperative and multiple postoperative time points (day 1, day 3, 1 week, 1 month, 3 months, and final follow-up). The ODI (0–100%) was used to evaluate functional disability preoperatively and at 1 week, 1 month, 3 months, and final follow-up postoperatively. At the final follow-up, change scores were calculated (Δ = final follow-up score − preoperative score). The MCID thresholds were defined as: ΔODI of −9.5, and ΔVAS for turning over and standing of −2.2.¹⁸ The PASS thresholds were defined as: VAS ≤ 3 and ODI ≤ 22.¹⁸ The ambulation criterion was defined as a standing VAS < 5; patients with VAS ≥ 5 continued bed rest with lumbar paraspinal muscle exercises.¹⁸ The turning over and standing VAS scores and the number of patients meeting ambulation criteria on postoperative days 1, 3, and 7 were compared among the three groups to analyze the impact of thoracolumbar fascia injury classification on early outcomes after PVP. Nonsteroidal anti-inflammatory drugs (NSAIDs) were not routinely administered postoperatively; they were given on an as-needed basis only when VAS ≥ 4, and the pre-medication VAS score was recorded.

Statistical Analysis

Statistical analysis was performed using SPSS version 22.0. Categorical data were compared among groups using the chi-square test and expressed as frequencies (percentages). Continuous data conforming to a normal distribution were analyzed using one-way analysis of variance (ANOVA) and expressed as mean ± standard deviation (), with pairwise comparisons performed using the least significant difference (LSD) t-test. VAS scores (turning over and standing) and ODI at different time points were compared among groups using two-way repeated measures ANOVA. The significance level was set at α = 0.05 (two-sided).

Results

General Characteristics

All patients underwent surgery successfully without serious complications. No significant differences were observed among the three groups in operative time, cement volume, cement leakage, follow-up duration, or incidence of subsequent adjacent vertebral fractures (Table 2).

Table 2.

Comparison of VAS Scores (Turning Over and Standing) and ODI Among the Three Groups at Different Time Points

Time Point	Localized Type (n=11)	Segmental Continuous Type (n=21)	Extensive Multisegmental Type (n=24)	P-value
Turning over VAS score
Preoperative	7.45±0.52	7.57±0.49	7.63±0.55	0.612
Postoperative day 1	3.82±0.65Δ	4.05±0.71Δ	4.27±0.74Δ	0.231
Postoperative day 3	2.91±0.54Δ	3.12±0.58Δ	3.31±0.63Δ	0.198
Postoperative week 1	2.09±0.47Δ	2.24±0.51Δ	2.38±0.53Δ	0.342
Postoperative month 1	1.86±0.33Δ	1.97±0.37Δ	2.11±0.41Δ	0.271
Postoperative month 3	1.64±0.31Δ	1.71±0.35Δ	1.79±0.38Δ	0.583
Final follow-up	1.58±0.27Δ	1.63±0.29Δ	1.71±0.31Δ	0.491
Standing VAS score
Preoperative	7.73±0.58	7.81±0.62	7.92±0.65	0.714
Postoperative day 1	3.64±0.67Δ	4.21±0.73Δ▲	4.87±0.81Δ▲#	0.000
Postoperative day 3	2.87±0.54Δ	3.54±0.62Δ▲	4.13±0.71Δ▲#	0.000
Postoperative week 1	2.41±0.46Δ	3.07±0.53Δ▲	3.68±0.64Δ▲#	0.000
Postoperative month 1	2.18±0.38Δ	2.79±0.47Δ▲	3.52±0.56Δ▲#	0.000
Postoperative month 3	2.05±0.31Δ	2.13±0.33Δ	2.24±0.36Δ	0.374
Final follow-up	1.91±0.26Δ	1.98±0.28Δ	2.07±0.31Δ	0.421
ODI (%)
Preoperative	70.18±7.24	71.05±7.63	72.13±7.89	0.731
Postoperative week 1	22.47±5.12Δ	28.35±5.74Δ▲	36.19±6.47Δ▲#	0.000
Postoperative month 1	21.83±4.76Δ	26.94±5.31Δ▲	32.57±6.12Δ▲#	0.000
Postoperative month 3	21.25±4.13Δ	23.48±4.57Δ	25.31±4.94Δ	0.127
Final follow-up	20.87±2.54Δ	21.34±2.71Δ	21.79±2.83Δ	0.583

Notes: Data are presented as mean ± standard deviation. Δ P < 0.05 compared with preoperative values; ▲ P < 0.05 compared with the localized type group; # P < 0.05 compared with the segmental continuous type group. Bold indicates statistically significant differences.

Abbreviations: VAS, visual analogue scale; ODI, Oswestry Disability Index.

Therapeutic Effect Evaluation

No statistically significant differences in baseline PROMs were observed among the three groups (Table 3). At all postoperative time points, VAS scores for turning over, VAS scores for standing, and ODI in all three groups were significantly improved compared with preoperative values (P <0.05, Table 3). No statistically significant differences in VAS scores for turning over were observed among the three groups at any postoperative time point (P >0.05, Table 3). However, regarding standing VAS scores and ODI, statistically significant differences were observed among the three groups within the first postoperative month (postoperative day 1, day 3, 1 week, and 1 month) (P <0.05, Table 3): the extensive multisegmental type group was significantly higher than the segmental continuous type and localized type groups, and the segmental continuous type group was also significantly higher than the localized type group. At 3 months postoperatively and at the final follow-up, no statistically significant differences were observed among the three groups (P >0.05). At the final follow-up, the achievement rates of MCID and PASS for VAS scores during turning over, VAS scores during standing, and ODI did not differ significantly among the three groups (P >0.05, Table 4).

Table 3.

Comparison of Perioperative and Follow-Up Characteristics Among the Three Groups

Characteristic	Localized Type (n=11)	Segmental Continuous Type (n=21)	Extensive Multisegmental Type (n=24)	P-value
Operation time (min)	31.18±4.92	32.05±5.31	33.17±5.68	0.512
Cement volume (mL)	3.61±0.38	3.55±0.44	3.49±0.51	0.673
Cement leakage, n (%)	1 (9.1%)	2 (9.5%)	3 (12.5%)	0.891
Follow-up duration (months)	18.45±5.12	17.86±5.43	18.21±5.67	0.934
Secondary adjacent vertebral fracture, n (%)	1 (9.1%)	2 (9.5%)	4 (16.7%)	0.672

Notes: Data are presented as mean ± standard deviation or number (%). No statistically significant differences were observed among the three groups for any of the above characteristics (all P > 0.05).

Table 4.

Comparison of MCID and PASS Achievement Rates for VAS and ODI at Final Follow-Up Among the Three Groups

Characteristic	Localized Type (n=11)	Segmental Continuous Type (n=21)	Extensive Multisegmental Type (n=24)	P-value
MCID VAS (turning over)				0.812
Not achieved	1 (9.1%)	3 (14.3%)	4 (16.7%)
Achieved	10 (90.9%)	18 (85.7%)	20 (83.3%)
MCID VAS (standing)				0.763
Not achieved	2 (18.2%)	4 (19.0%)	5 (20.8%)
Achieved	9 (81.8%)	17 (81.0%)	19 (79.2%)
MCID ODI				0.841
Not achieved	3 (27.3%)	5 (23.8%)	7 (29.2%)
Achieved	8 (72.7%)	16 (76.2%)	17 (70.8%)
PASS VAS (turning over)				0.791
Not achieved	2 (18.2%)	3 (14.3%)	4 (16.7%)
Achieved	9 (81.8%)	18 (85.7%)	20 (83.3%)
PASS VAS(standing)				0.874
Not achieved	2 (18.2%)	5 (23.8%)	6 (25.0%)
Achieved	9 (81.8%)	16 (76.2%)	18 (75.0%)
PASS ODI				0.823
Not achieved	3 (27.3%)	6 (28.6%)	7 (29.2%)
Achieved	8 (72.7%)	15 (71.4%)	17 (70.8%)

Notes: Data are presented as number (%). No statistically significant differences were observed among the three groups for any MCID or PASS measure (all P > 0.05).

Abbreviations: MCID, minimal clinically important difference; PASS, patient acceptable symptom state; VAS, visual analogue scale; ODI, Oswestry Disability Index.

On postoperative days 1 and 3, the ambulation achievement rates in the localized type and segmental continuous type groups were both superior to those in the extensive multisegmental type group (P<0.05), with rates of 63.6% and 90.9% in the localized type group, 47.6% and 76.2% in the segmental continuous type group, and only 20.8% and 54.2% in the extensive multisegmental type group, respectively. By postoperative day 7, the achievement rate reached 100.0% in all three groups, with no significant intergroup difference (P=0.412, Table 5).

Table 5.

Comparison of the Number of Patients Meeting Ambulation Criteria at Different Postoperative Time Points Among the Three Groups

Time Point	Localized Type (n=11)	Segmental Continuous Type (n=21)	Extensive Multisegmental Type (n=24)	P-value
Postoperative day 1, n (%)	7 (63.6%)	10 (47.6%)▲	5 (20.8%)▲#	0.021
Postoperative day 3, n (%)	10 (90.9%)	16 (76.2%)▲	13 (54.2%)▲#	0.031
Postoperative week 1, n (%)	11 (100.0%)	21 (100.0%)	24 (100.0%)	0.412

Notes: Data are presented as number (%). ▲ P < 0.05 compared with the localized type group; # P < 0.05 compared with the segmental continuous type group. Bold indicates statistically significant differences. At postoperative week 1, all patients in the three groups met the ambulation criteria, with no statistically significant difference among the groups (P = 0.412).

The interobserver and intraobserver reliability for the proposed classification were substantial to almost perfect, with Cohen’s kappa coefficients of 0.85 (95% CI, 0.74–0.96) and 0.88 (95% CI, 0.78–0.98), respectively, indicating that the MRI-based classification could be consistently applied by different observers and by the same observer over time.

Discussion

Studies have demonstrated that PVP can rapidly relieve pain and promote early mobilization in patients with osteoporotic vertebral compression fractures; however, a proportion of patients still experience mild to moderate residual pain after PVP.^6–10 Yan et al¹⁹ identified a potential association between preoperative thoracolumbar fascia injury and residual pain after PVP. Yang et al,¹⁰ through a retrospective analysis of 1,316 OVCF patients, confirmed that thoracolumbar fascia injury is an independent risk factor for residual pain after PVP (OR=3.805, P=0.002). However, previous studies have predominantly treated thoracolumbar fascia injury as a binary variable, overlooking the inherent differences in the extent of injury. Building upon prior work, the present study is the first to propose a three-type classification system for thoracolumbar fascia injury based on MRI T2-STIR images and to systematically evaluate the differential impact of different classification types on outcomes after PVP. Our results demonstrated that within the first postoperative month, standing VAS scores and ODI in the extensive multisegmental type group were significantly higher than those in the segmental continuous type and localized type groups, and the segmental continuous type group was also significantly higher than the localized type group (P <0.05); whereas at 3 months postoperatively and at the final follow-up, no statistically significant differences were observed among the three groups (P >0.05). These findings suggest that the extent of thoracolumbar fascia injury has a significant impact on early postoperative pain relief after PVP—the more extensive the injury, the more severe the early residual pain—but does not significantly affect medium- to long-term outcomes, thereby providing a more precise reference for clinical prognostic assessment.

The cutoff values of ≤2, 3–4, and ≥5 vertebral segments were determined based on clinical experience and preliminary observation of symptom clustering, although future receiver operating characteristic (ROC) curve analysis with larger samples may further optimize these thresholds.

Recent evidence has also highlighted the importance of evaluating predictors of postoperative outcomes and new fractures following vertebral augmentation. Tabanli et al²⁰ reported the incidence of new fractures in patients treated with kyphoplasty/vertebroplasty versus conservative methods, providing contemporary data that underscores the necessity of identifying high-risk patients. In the context of our study, the extent of thoracolumbar fascia injury may not only predict early residual pain but also potentially influence the risk of subsequent adjacent segment complications due to altered spinal biomechanics, which warrants further investigation.

Notably, VAS scores for turning over did not differ significantly among the three groups at any time point (P >0.05), whereas standing VAS scores and ODI demonstrated significant intergroup differences in the early postoperative period. This phenomenon may be attributable to the fact that turning over primarily relies on trunk rotational muscles and places relatively lower tensile demands on the thoracolumbar fascia,^12,14,16 whereas standing and walking require the thoracolumbar fascia to bear greater load-transfer functions. The more extensive the injury, the more severely the load-transfer function is compromised, resulting in more pronounced pain during standing. This also explains why VAS scores for turning over showed no significant differences among the three groups, while standing VAS scores and ODI exhibited significant classification-dependent differences.

Regarding time to ambulation, the present study found that on postoperative days 1 and 3, the proportion of patients meeting ambulation criteria in the localized type group was significantly higher than that in the extensive multisegmental type group, and the segmental continuous type group was also significantly higher than the extensive multisegmental type group (P <0.05). These findings suggest that the more extensive the thoracolumbar fascia injury, the more difficult early postoperative ambulation becomes, which is consistent with the impact of injury extent on standing pain. Following comprehensive treatment (physical therapy, lumbar paraspinal muscle exercises, pharmacological analgesia, etc), all patients met the ambulation criteria within 7 days postoperatively, and the intergroup differences disappeared, suggesting that proactive comprehensive perioperative intervention can effectively narrow the gap in ambulation time among patients with different classification types.

Therefore, we recommend tailored perioperative management based on the classification: for patients with extensive multisegmental injury, proactive multimodal analgesia, including regular NSAIDs rather than as-needed administration when clinically appropriate, should be initiated early; early weight-bearing ambulation should be appropriately delayed with a focus on bedside paraspinal muscle exercises; and more intensive rehabilitation supervision is warranted during the first postoperative month.

From an anatomical perspective, the thoracolumbar fascia is a circumferential structure composed of multiple aponeurotic and fascial layers that separates the paraspinal muscles from the posterior abdominal wall musculature. Through its continuity with the latissimus dorsi and transversus abdominis muscles, it integrates the mechanical actions of the limbs, abdomen, and spine, constituting a critical component of the paraspinal stabilizing structures.^13–16 Extensive multisegmental injury involves a broader area, causing more severe disruption of the overall biomechanical function of the thoracolumbar fascia and resulting in more pronounced spinal instability, which may be an important reason for the more severe early pain and delayed ambulation in patients with the extensive multisegmental type.

The thoracolumbar fascia is richly innervated with nerve endings.^13,14 Studies have shown that the density of calcitonin gene-related peptide (CGRP)-positive fibers in the thoracolumbar fascia is three times that of spinal muscles,²¹ and the majority of CGRP and substance P (SP) immunoreactive sensory fibers are located in the outer fascial layer and subcutaneous tissue.¹⁵ Under inflammatory conditions, CGRP- and SP-positive fibers in the thoracolumbar fascia increase, which may represent the neurobiological basis of pathological fascial pain.^13,16 Extensive multisegmental injury involves more fascial territory, resulting in a broader inflammatory response and a higher degree of neural sensitization, thereby leading to more severe pain responses. Furthermore, similar to ligaments, the fascia has a far less abundant blood supply compared with muscle and heals more slowly.¹⁴ Some studies have found that complete reversal of thoracolumbar fascia abnormalities may require more than one month.¹⁸ The longer healing time required for extensive multisegmental injury also explains why early residual pain is more persistent in patients with this type.

At the final follow-up, the achievement rates of MCID and PASS did not differ significantly among the three groups (P >0.05), suggesting that the classification of thoracolumbar fascia injury does not affect the medium- to long-term efficacy of PVP. This is consistent with the findings of Yan et al.¹⁹ For OVCF patients with concomitant thoracolumbar fascia injury, clinicians can inform them that as the fascial injury heals and edema resolves, early residual low back pain will gradually subside, and medium- to long-term outcomes will be comparable to those of patients with localized injury, thereby alleviating patients’ psychological burden.

Additionally, it is worth discussing the generalizability of our classification to patients undergoing kyphoplasty (PKP) or other vertebral augmentation techniques. Although this study focused exclusively on PVP, the biomechanical and nociceptive mechanisms of thoracolumbar fascia injury are independent of the vertebral augmentation method utilized. Therefore, it is reasonable to hypothesize that the proposed classification could similarly predict early residual pain and delayed ambulation in PKP patients. However, given the different cement distribution characteristics and potential subtle differences in vertebral height restoration between PVP and PKP, further studies are needed to validate the applicability of this classification in cohorts undergoing other augmentation techniques.

This study has the following limitations: First, this is a single-center retrospective study with a relatively small sample size, particularly with only 11 cases in the localized type group, which may limit the statistical power to detect small effect sizes and increase the risk of type II errors, especially in subgroup analyses. Although the observed intergroup differences in early VAS and ODI reached statistical significance, indicating sufficient power for the primary outcomes, these findings should still be interpreted with caution. External validation of the proposed classification in independent, multicenter cohorts is strongly recommended before its widespread clinical implementation. Second, this was a retrospective study with inherent selection bias; prospective studies are needed in the future to more accurately reflect the impact of thoracolumbar fascia injury classification on PVP outcomes. Third, this study did not perform a detailed analysis of the specific fascial layers involved (posterior, middle, or anterior layers), nor did it evaluate quantitative indicators such as signal intensity of fascial injury or the transverse extent of injury; future studies may further explore the relationship between injury layers and classification types. Fourth, residual pain after PVP is influenced by numerous factors,^3,6–10 and thoracolumbar fascia injury classification is only one of them; large-sample multivariable analyses are needed to control for confounding factors.

Conclusions

The findings of this preliminary study suggest that the classification of thoracolumbar fascia injury is closely associated with the degree of early postoperative pain relief and time to ambulation after PVP. The more extensive the injury (extensive multisegmental type), the more severe the early residual pain and the more delayed the ambulation; however, different classification types do not significantly affect the medium- to long-term efficacy of PVP. In the preoperative evaluation of OVCF patients, MRI-based classification of thoracolumbar fascia injury should be emphasized to more accurately predict early postoperative outcomes following PVP, formulate individualized perioperative management plans (eg, proactive multimodal analgesia and appropriately delayed mobilization for extensive multisegmental type), and provide patients with more targeted prognostic counseling.

Funding Statement

The authors thank the funding sources for their support. This work was funded by the Scientific Research Cultivation Fund of Capital Medical University (PYZ24154) and the 2024 Youth Project of the Open Research Fund of the State Key Laboratory of Neurology and Oncology Drug Development (SKLSIM-2024108).

Data Sharing Statement

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

Ethics Approval and Consent to Participate

Ethical approval was obtained from the Ethics Committee of State Grid Beijing Electric Power Hospital (No. 2025041530205). And this study was conducted in accordance with the Declaration of Helsinki of the World Medical Association. Informed consent was obtained from all individual participants included in the study.

Consent for Publication

Patient gave written informed consent for their personal or clinical details along with any identifying images to be published in this study.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no competing interests.