Low‐Profile Posterior Lumbar Fusion Combined With Proximal Interlaminar Coflex Dynamic Fixation for Lumbar Degenerative Diseases: A Comparative Study

ABSTRACT

Objective

Lumbar fusion combined with topping‐off dynamic stabilization has been used to reduce adjacent‐segment stress in lumbar degenerative diseases. However, traditional pedicle screw‐based topping‐off surgery may require relatively extensive posterior exposure, which can increase soft‐tissue injury and delay postoperative recovery. This study aimed to evaluate the feasibility and effectiveness of low‐profile lumbar fusion using cortical bone trajectory screws (CBTs) and pedicle screw (PS) internal fixation combined with Coflex dynamic stabilization, compared to traditional PS‐based topping‐off.

Methods

A retrospective cohort study was conducted using data from patients with lumbar degenerative diseases treated between December 2021 and January 2023. Patients who underwent either the low‐profile topping‐off procedure or traditional topping‐off surgery were enrolled. Demographic data, perioperative parameters, radiographic outcomes, complications, and clinical outcomes were collected and compared between the two groups. Statistical analyses included Student's t‐test, chi‐square test, two‐way analysis of variance, multivariable linear regression, and binary logistic regression.

Results

A total of 229 patients undergoing single‐level L4/L5 fusion (104 in the low‐profile group and 125 in the traditional group) were followed for 35.69 ± 4.76 months (range, 24–40 months). Baseline characteristics, including age, sex, bone mineral density (BMD), diagnosis distribution, and follow‐up duration, were comparable between groups (p > 0.05). Compared with the traditional group, the low‐profile group demonstrated improved perioperative outcomes, with less intraoperative blood loss, shorter incision length, lower drainage volume, and shorter postoperative hospital stay, while operative time was similar. Radiographically, the low‐profile group showed a higher fusion success rate at final follow‐up (Bridwell Grade I + II: 91.3% vs. 80.8%). Adjacent‐segment analyses indicated better preservation of cranial adjacent‐segment radiographic parameters (L3/4) at 24 months in the low‐profile group, whereas the overall adjacent segment degeneration (ASD) incidence did not differ significantly between groups. Both groups achieved significant improvements in Visual Analog Scale (VAS), Oswestry Disability Index (ODI), and Japanese Orthopedic Association Score (JOA); although postoperative back pain VAS and ODI were statistically lower in the low‐profile group, leg pain VAS, JOA, and overall complication rates were comparable. Multivariable regression adjusting for age, sex, BMD, diagnosis type, and baseline ODI showed that surgical approach remained independently associated with final ODI, while no significant association with overall complications was observed.

Conclusion

Low‐profile topping‐off surgery is as effective as traditional topping‐off for lumbar degenerative diseases, which has the strength of being less invasive and enabling more rapid recovery.

Low‐profile topping‐off surgery, combining cortical bone trajectory screws with pedicle screws and Coflex stabilization, achieved comparable clinical outcomes to traditional topping‐off, with reduced incision length, blood loss, and drainage. This technique offers a less invasive approach and faster recovery for patients with lumbar degenerative diseases.

Abbreviations

ASD

adjacent segment degeneration

BMD

bone mineral density

CBTs

cortical bone trajectory screw

CI

confidence interval

CT

computed tomograph

IDH

intervertebral disc height

IFA

intervertebral foramen area

JOA

Japanese Orthopedic Association

LDD

lumbar degenerative disease

LDH

lumbar disc herniation

LSS

lumbar spinal stenosis

MRI

magnetic resonance imaging

ODI

Oswestry Disability Index

OR

odds ratio

PLIF

posterior lumbar interbody fusion

PS

pedicle screw

R‐ASD

radiographic adjacent segment degeneration

ROM

range of motion

S‐ASD

symptomatic adjacent segment degeneration

VAS

visual analogue scale

1. Introduction

The rising prevalence of lumbar degenerative diseases (LDD) due to aging and unhealthy lifestyles places a heavy economic burden on patients and healthcare systems [1, 2]. Posterior lumbar fusion fixation remains the primary treatment [3], but long‐term studies report complications such as pseudarthrosis, adjacent segment degeneration (ASD), loss of mobility, graft fractures, and nerve or vascular injuries [4, 5, 6]. ASD is particularly common due to loss of motion at the fused segment, leading to increased load, compensatory movement, accelerated disc and facet joint degeneration, and higher ASD risk [7, 8, 9].

Coflex is a dynamic interspinous non‐fusion fixation system made of titanium alloy, featuring a distinctive “U” shape that provides elastic expansion and excellent biomechanical properties [10]. This design maintains the lumbar segment in a mildly flexed state, limiting extension and enhancing the height of the intervertebral space and foramina [11, 12, 13, 14]. By increasing the distance between the spinous process and the posterior vertebral edge, Coflex reduces disc pressure, thereby protecting the intervertebral disc and facet joints from collapse and degeneration [15]. Its elasticity serves as a buffer between fused and adjacent vertebrae, preventing stress concentration, and promoting dynamic spinal stability [16]. Additionally, Coflex may mitigate compensatory hypermobility and help preserve adjacent‐segment biomechanics following surgery [15, 17, 18].

In previous clinical and basic research, the Coflex non‐fusion dynamic internal fixation system has been proven to be an effective treatment for LDD [17]. With the widespread application of non‐fusion technology in clinical practice, the Coflex system is used in the upper segment of PLIF surgery and, when combined with fusion technology, leads to the development of topping‐off surgery. In topping‐off, Coflex acts as a “transition zone” between fused and non‐fused segments, preserving partial range of motion (ROM) and preventing excessive vertebral movement [19, 20]. This approach may help reduce adjacent‐segment stress and stabilize decompression‐induced vertebral instability without compromising efficacy [15, 18, 21]. It offers a safe and effective treatment for lumbar degeneration.

Posterior lumbar fusion with PS fixation is commonly used for LDD, providing strong three‐column fixation, high stability, effective correction, and a high fusion rate [22, 23, 24]. However, traditional PS fixation requires extensive muscle and soft tissue dissection, increasing structural damage and delaying recovery [25]. In osteoporotic patients, reduced bone density weakens the bone‐screw interface, raising the risk of screw loosening, fixation failure, and revision surgery [26, 27]. To address these issues, Santoni et al. introduced the CBTs technique in 2009, which improves screw contact with cortical bone by modifying the trajectory, enhancing biomechanical stability [28, 29, 30]. CBTs reduce paravertebral tissue damage, preserve adjacent facet joints, and expedite recovery by utilizing a midline entry point [31, 32]. This technique is widely applied for degenerative diseases, trauma, and tumors, offering superior holding power, lower complication rates, less tissue damage, and faster rehabilitation compared to PS fixation [33, 34].

To improve outcomes for patients with LDD, especially those with reduced bone quality, we developed the low‐profile topping‐off technique. This approach aims to improve fixation quality, minimize surgical trauma, and preserve adjacent‐segment function by combining CBTs, PS fixation, and Coflex dynamic stabilization. The specific aims of this study were: (i) to compare perioperative outcomes and clinical efficacy between low‐profile and traditional PS‐based topping‐off surgery; (ii) to evaluate fusion status, complications, ASD incidence, and adjacent‐segment radiographic changes during follow‐up; and (iii) to explore whether the effects of the low‐profile technique varied according to BMD. These findings may provide preliminary evidence for the clinical value and future prospective evaluation of this technique.

2. Materials and Methods

2.1. Study Design and Patient Selection

A retrospective cohort study was conducted to identify consecutive patients with LDD who underwent either low‐profile topping‐off or traditional topping‐off surgery at our institution between December 2021 and January 2023. This study was performed in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Beijing Chao‐Yang Hospital, Capital Medical University (IRB No. 2024‐Ke‐429). The inclusion criteria were as follows: (1) diagnosis of lumbar disc herniation (LDH) or lumbar spinal stenosis (LSS) confirmed by clinical and radiological examinations; (2) posterior lumbar interbody fusion (PLIF) with internal fixation at the L4/L5 segment; (3) postoperative follow‐up of at least 24 months; and (4) signed written informed consent. The exclusion criteria were as follows: (1) scoliosis, spinal tumors, spinal infections, or trauma; (2) previous lumbar spine surgery; (3) secondary osteoporosis; (4) psychiatric disorders or withdrawal during follow‐up; and (5) incomplete follow‐up data or missing clinical records.

2.2. Surgical Treatment and Technique

2.2.1. Surgical Planning

In this study, “low‐profile” refers to a posterior fixation configuration in which cortical bone trajectory screws (CBTs) are used in the upper instrumented vertebra and traditional pedicle screws are used in the lower instrumented vertebra, producing a smaller arc‐shaped skin‐level notch after screw placement. This configuration creates a more medial and compact posterior fixation profile than the traditional pedicle screw‐based construct, thereby reducing the lateral extent of exposure and paraspinal muscle dissection. Thus, “low‐profile” in this study represents a technique‐specific reduction in the exposure and implant‐trajectory footprint, rather than a synonym for conventional minimally invasive spine surgery.

All operations were planned by an experienced spine surgery team based on each patient's clinical presentation and preoperative imaging. Both groups underwent posterior lumbar decompression, L4/L5 interbody fusion, and proximal Coflex dynamic stabilization according to a standardized protocol. The key difference between the two groups was the posterior fixation construct: the low‐profile group received CBT screws at L4 combined with PS fixation at L5, whereas the traditional group received conventional PS fixation at both L4 and L5. The detailed surgical steps are described below.

2.2.2. Positioning and Anesthesia

During the surgical procedure, patients were placed in the prone position on the operating table. Cushions were positioned under the thoracic spine and iliac crests to restore the natural lumbar lordosis, while the abdomen was elevated to reduce intra‐abdominal pressure. The height and orientation of the operating table were adjusted to facilitate intraoperative fluoroscopic guidance using a C‐arm. Intravenous access was established, and appropriate monitoring devices were connected. General anesthesia was administered following standard protocols. Patients first received 3–5 min of pure oxygen inhalation via a face mask, followed by endotracheal intubation through the vocal cords. Anesthetic agents were delivered through both intravenous infusion and the endotracheal tube, inducing temporary central nervous system suppression and achieving a state of general anesthesia.

2.2.3. Incision and Exposure

After routine disinfection and draping of the surgical field, a longitudinal midline incision of approximately 8 cm was made on the posterior aspect of the lumbar spine. The skin, subcutaneous tissue, and deep fascia were incised layer by layer to ensure the appropriate depth and extent of exposure. During dissection, blunt separation was performed under the periosteum along the lateral borders of the spinous processes to fully expose the laminae of L4 and L5, as well as the articular and spinous processes of L3 and L4. Throughout the procedure, care was taken to preserve the paravertebral muscles and surrounding soft tissues in order to minimize surgical trauma and reduce the risk of postoperative complications.

2.2.4. CBTs Fixation

Once adequate exposure had been achieved, the surgical field was gently expanded using a spreader. The cortical bone at the planned insertion point was meticulously removed with a burr drill. A trephine was then used to gradually advance from the cortical bone into the cancellous bone of the vertebral body. The insertion point for the CBTs was located approximately 2–3 mm below the L4 transverse process and 2–3 mm medial to the edge of the L4 isthmus. Under C‐arm fluoroscopic guidance, the screw channel was carefully prepared using a tap, creating a pathway approximately one‐third to one‐half the diameter of the screw. The integrity of the four walls of the pedicle was confirmed using a ball‐tipped pedicle probe. The Mark screw was then accurately placed into the pedicle, and its position was verified to meet the planned trajectory specifications. The typical screw trajectory showed a caudal inclination of 30°–40° and a medial inclination of 15°–20°. The final CBT trajectory was individualized according to preoperative CT‐based pedicle morphology (e.g., pedicle width, transverse diameter, and the cortical channel orientation) and confirmed intraoperatively using anteroposterior and lateral fluoroscopy to maximize cortical purchase and avoid pedicle breach. After completing decompression and interbody fusion, the appropriate CBT implant was selected and inserted.

2.2.5. PS Fixation

After exposure was completed, a spreader was used to fully reveal the bony cortex at the junction of the L5 transverse process and the pedicle. A trephine was then used to gradually penetrate into the cancellous bone of the vertebral body. The integrity of the pedicle's four walls was carefully confirmed using a ball‐tipped pedicle probe. Once proper positioning was verified, a Mark screw was inserted, and its placement was assessed under fluoroscopic guidance. After confirming the trajectory, an appropriately sized conventional pedicle screw was selected and implanted.

2.2.6. Coflex Insertion

Following internal fixation of L4 and L5, a Coflex interspinous implant was inserted between the L3 and L4 spinous processes. The supraspinous ligament was gently dissected and retracted to one side. With preservation of the bilateral facet joints, surrounding bone attachments and interspinous ligaments were carefully removed using a pointed rongeur and lamina clamp. The superior and inferior margins of the spinous processes were trimmed to avoid dural injury. An appropriately sized Coflex device was then selected and inserted between the spinous processes, ensuring a 3–5 mm clearance from the dura mater, and subsequently secured in position. Coflex size was selected using intraoperative trial implants based on the measured interspinous distance in a neutral position. The appropriate size was defined as one that achieved stable seating between the spinous processes without excessive distraction of the interspinous space. Final positioning and implant stability were verified by intraoperative fluoroscopy. Finally, the supraspinous ligament was sutured using absorbable sutures. Representative intraoperative photographs and fluoroscopic images are shown in Figures 1 and 2. Schematic illustrations of the low‐profile and traditional topping‐off surgical techniques are presented in Figure 3.

FIGURE 1.

Intraoperative photograph of the low‐profile topping‐off procedure.

FIGURE 2.

Intraoperative fluoroscopy of the low‐profile topping‐off procedure.

FIGURE 3.

Schematic comparison of the low‐profile and traditional topping‐off techniques. (a) Schematic sagittal representation of the low‐profile topping‐off surgical technique; (b) Schematic sagittal representation of the traditional topping‐off surgical technique; (c) Schematic coronal representation of the low‐profile topping‐off surgical technique; (d) Schematic coronal representation of the traditional topping‐off surgical technique.

2.2.7. Wound Closure and Drainage

A drainage tube was accurately placed beneath the paraspinal muscles, and the surgical incision was closed meticulously in layers, ensuring proper approximation of soft tissue and skin for optimal wound healing.

2.2.8. Postoperative Treatment

Both groups received the same postoperative care and rehabilitation protocol. Vital signs and changes in motor and sensory function of both lower limbs were closely monitored after surgery. Standard postoperative treatment included anti‐inflammatory therapy, dehydration management, glucocorticoids, and neurotrophic support. The drainage tube was removed once the daily output was less than 80 mL. After tube removal, patients were instructed to begin ambulation while wearing a lumbar brace and to perform targeted lumbar and back muscle strengthening exercises. In addition, anti‐osteoporosis therapy was reinforced, and all postoperative complications were promptly documented.

2.3. Outcome Measures

To ensure comprehensive documentation, fundamental patient information—including age, gender, diagnosis, and BMD—was recorded. In addition, surgical indicators were meticulously documented, including operation time, intraoperative blood loss, postoperative blood loss, drainage volume, and length of postoperative hospital stay. The quality of lumbar interbody fusion was assessed at the final follow‐up based on plain radiographs and CT scans using the Bridwell fusion grading system [35]. Fusion outcomes were classified into four grades: Grade I (definite fusion with trabecular bridging across the disc space); Grade II (probable fusion with graft intact and no lucency, but incomplete remodeling); Grade III (unlikely fusion with graft intact but definite lucency at the superior or inferior endplate); and Grade IV (non‐union with graft resorption or collapse). Successful fusion was defined as Grade I or Grade II. Assessment of Adjacent Segment Degeneration (ASD): Radiographic ASD (R‐ASD) was evaluated at the cranial (L3/4) and caudal (L5/S1) adjacent segments using the criteria proposed by Hilibrand et al. R‐ASD was defined as the presence of new or worsened findings compared to preoperative images, including: (1) disc height narrowing > 20%; (2) progression of osteophytes > 3 mm; (3) spondylolisthesis > 3 mm; or (4) calcification of the anterior longitudinal ligament. Symptomatic ASD (S‐ASD) was defined as newly developed or worsened clinical symptoms attributable to the adjacent segment(s) during follow‐up, in conjunction with radiographic evidence of adjacent‐segment degeneration. Symptoms included new or recurrent low back pain, radicular pain, neurogenic claudication, or neurological deficits consistent with the cranial (L3/L4) or caudal (L5/S1) adjacent level. S‐ASD events were identified through outpatient/inpatient follow‐up records and required both (1) clinical correlation and (2) imaging confirmation at the adjacent segment. Additionally, quantitative measurements of the adjacent segments were performed on lateral radiographs and CT scans. The intervertebral disc height (IDH) was measured as the average of the anterior, middle, and posterior disc heights. The intervertebral foramen area (IFA) was calculated using the sagittal reconstruction images. These parameters were recorded preoperatively and at 6, 12, and 24 months postoperatively. To evaluate patients' symptoms and functional status, standardized scoring systems were employed: the Visual Analog Scale (VAS), the Oswestry Disability Index (ODI), and the Japanese Orthopedic Association Score (JOA). These assessments were performed preoperatively, postoperatively, at 6 months, and at the final follow‐up. Relevant radiological imaging data were also collected at the same time points to allow for comprehensive analysis.

The primary outcomes were postoperative clinical outcomes, including VAS, ODI, and JOA scores. Secondary outcomes included perioperative indicators, radiological outcomes, postoperative complications, and ASD‐related clinical and radiological findings during follow‐up.

2.4. Statistical Analysis

Continuous variables are presented as mean ± standard deviation and were compared using Student's t‐test. Categorical variables are presented as counts and percentages and were compared using the chi‐square test. Statistical analyses were performed using SPSS Statistics version 25.0 (IBM Corp., Armonk, NY, USA). A two‐sided p < 0.05 was considered statistically significant.

To clarify whether the benefits of the low‐profile technique vary by bone density, a subgroup analysis was performed. Patients were stratified into three groups based on T‐scores: Normal BMD (T ≥ −1.0), Osteopenia (−2.5 < T < −1.0), and Osteoporosis (T ≤ −2.5). Interaction tests were conducted using two‐way ANOVA to evaluate the heterogeneity of treatment effects across these subgroups.

Multivariable regression analysis was performed to adjust for potential confounders. A multiple linear regression model was fitted with final ODI as the dependent variable. A binary logistic regression model was fitted with overall complications as the dependent variable. Covariates included age, sex, BMD (T‐score), diagnosis type (LSS vs. LDH), and preoperative ODI. Results are reported as regression coefficients (B) with 95% confidence intervals for linear regression and odds ratios (OR) with 95% confidence intervals for logistic regression. A two‐sided p < 0.05 was considered statistically significant. Overall complications were defined as the occurrence of any recorded perioperative or postoperative complication.

3. Results

3.1. Patient Characteristics and Perioperative Outcomes

Between 2021 and 2023, a total of 229 patients were enrolled in this study, including 102 males and 127 females. Among them, 104 patients underwent the low‐profile topping‐off procedure, with 63 diagnosed with lumbar spinal stenosis and 41 with lumbar disc herniation. The remaining 125 patients received the traditional topping‐off procedure (control group), including 81 cases of spinal stenosis and 44 cases of disc herniation. The average age was 65.44 ± 10.41 years in the low‐profile group and 65.78 ± 7.90 years in the traditional group. All procedures were performed by the same experienced spine surgery team, following a standardized surgical protocol. Perioperative parameters, pain severity (VAS score), functional status (ODI and JOA scores), and surgery‐related complications were meticulously recorded throughout the study. Baseline demographic characteristics and perioperative parameters are summarized in Table 1.

TABLE 1.

Patient demographics and perioperative parameters.

Variables	Low‐profile group (n = 104)	Traditional group (n = 125)	p
Demographics
Age (years)	65.44 ± 10.41	65.78 ± 7.90	0.783
Sex, n (%)			0.326
Male	50 (48.08)	52 (41.60)
Female	54 (51.92)	73 (58.40)
Diagnosis, n (%)			0.510
LSS	63 (60.58)	81 (64.80)
LDH	41 (39.42)	44 (35.20)
Follow‐up (months)	35.28 ± 4.52	36.04 ± 4.93	0.229
BMD (T score)	−0.75 ± 1.78	−0.46 ± 1.42	0.184
Perioperative parameters
Average surgery time (minutes)	180.75 ± 35.81	182.14 ± 41.79	0.789
Intraoperative blood loss (mL)	216.89 ± 86.63	239.94 ± 86.50	0.046 *
Postoperative drainage (mL)	189.47 ± 84.12	278.20 ± 137.10	< 0.001 *
Length of incision (mm)	88.87 ± 9.86	98.70 ± 14.32	< 0.001 *
Postoperative hospitalization (day)	7.08 ± 1.69	8.46 ± 2.47	< 0.001 *
Complications
Screw misplacement	2/104	3/125	0.586
Screw loosening	7/104	14/125	0.243
Coflex misplacement	0/104	0/125	1.000
Surgical site infection	0/104	1/125	0.546
Nerve‐root injury	0/104	0/125	1.000
Cage migration	1/104	1/125	0.703

Note: The bold values indicate statistically significant results. Statistical significance was defined as p < 0.05.

Abbreviations: BMD, bone mineral density; LDH, lumbar disc herniation; LSS, lumbar spinal stenosis.

^* p < 0.05.

Intraoperatively, the mean surgical time was 180.75 ± 35.81 min in the low‐profile group and 182.14 ± 41.79 min in the traditional group, with no significant difference (p = 0.789). The mean blood loss was significantly lower in the low‐profile group (216.89 ± 86.63 mL) compared to the traditional group (239.94 ± 86.50 mL, p < 0.05), and no intraoperative blood transfusions were required in any patient. The postoperative drainage volume was also significantly lower in the low‐profile group (189.47 ± 84.12 mL) than in the traditional group (278.20 ± 137.10 mL, p < 0.001). Additionally, the incision length was significantly shorter in the low‐profile group (88.87 ± 9.86 mm) compared to the traditional group (98.70 ± 14.32 mm, p < 0.001), and the postoperative hospital stay was shorter as well (7.08 ± 1.69 days vs. 8.46 ± 2.47 days, p < 0.001). Figure 4 illustrates the differences in incision length and surgical exposure area between the low‐profile and traditional topping‐off procedures (Table 2).

FIGURE 4.

Comparison of incision length and surgical exposure area. (a) Low‐profile topping‐off: Incision length (label A) and surgical exposure area (green shaded region). (b) Traditional topping‐off: Incision length (label B) and surgical exposure area (pink shaded region).

TABLE 2.

Comparison of clinical outcomes and radiographic fusion status between the two groups.

Variables	Low‐profile group (n = 104)	Traditional group (n = 125)	p
Bridwell fusion grade			0.044*
Grade I	85 (81.7%)	89 (71.2%)
Grade II	10 (9.6%)	12 (9.6%)
Grade III	7 (6.7%)	18 (14.4%)
Grade IV	2 (1.9%)	6 (4.8%)
VAS (leg pain)
Pre‐operation	6.52 ± 0.84	6.44 ± 0.77	0.456
Post‐operation	2.62 ± 0.73	2.76 ± 0.78	0.179
6 months	1.18 ± 0.52	1.19 ± 0.59	0.900
Final follow‐up	0.85 ± 0.55	0.93 ± 0.57	0.274
p a	0.00	0.00
VAS (back pain)
Pre‐operation	6.34 ± 1.24	6.46 ± 1.00	0.420
Post‐operation	2.57 ± 0.87	2.84 ± 0.74	0.012 *
6 months	0.78 ± 0.56	1.10 ± 0.57	< 0.001 *
Final follow‐up	0.76 ± 0.68	0.94 ± 0.62	0.041 *
p a	0.00	0.00
Oswestry Disability Index score (ODI)
Pre‐operation	64.06 ± 6.81	64.82 ± 6.20	0.379
Post‐operation	25.01 ± 3.43	26.90 ± 3.24	< 0.001 *
6 months	20.10 ± 2.52	21.94 ± 2.67	< 0.001 *
Final follow‐up	18.79 ± 2.49	19.59 ± 2.31	0.012 *
p a	0.00	0.00
Japanese Orthopedic Association score (JOA)
Pre‐operation	12.71 ± 2.51	12.44 ± 2.20	0.384
Post‐operation	21.24 ± 1.94	21.55 ± 1.61	0.193
6 months	23.55 ± 2.07	23.71 ± 2.35	0.580
Final follow‐up	26.45 ± 2.19	25.95 ± 2.74	0.127
p a	0.00	0.00

Note: The bold values indicate statistically significant results. Statistical significance was defined as p < 0.05.

Abbreviations: JOA, Japanese Orthopedic Association Score; ODI, Oswestry Disability Index; VAS, Visual Analogue Scale.

^a Pre‐operation versus final follow‐up.

^* p < 0.05.

3.2. Radiographic Outcomes and Complications

The incidence of perioperative complications (e.g., screw malposition/loosening, nerve root injury, surgical site infection) was comparable between the two groups. Postoperative CT scans showed screw malposition in two patients (1.92%) in the low‐profile group and three patients (2.40%) in the traditional group (p = 0.586). During follow‐up, screw loosening occurred in 7 patients (6.73%) in the low‐profile group and 14 patients (11.2%) in the traditional group, with no significant difference (p = 0.243). No cases of Coflex malposition or nerve root injury were observed in either group. No surgical site infections occurred in the low‐profile group, while one case was reported in the traditional group (p = 0.546). At the 24‐month follow‐up, one case of cage migration was observed in each group, with no statistically significant difference (p = 0.703).

All patients were followed for at least 24 months after surgery, with a mean follow‐up duration of 35.69 ± 4.76 months (range: 24–40 months). Radiographic evaluation at the final follow‐up revealed a significant difference in the distribution of Bridwell fusion grades between the two groups (p = 0.044). Specifically, the Low‐profile group achieved a significantly higher fusion success rate (Grade I + II) compared to the Traditional group (91.3% vs. 80.8%). The radiographic outcomes of the adjacent segments are summarized in Table 3. Preoperatively, there were no significant differences in the IDH and IFA at the cranial (L3/L4) and caudal (L5/S1) adjacent levels between the two groups (p > 0.05). At the final 24‐month follow‐up, the Low‐profile group maintained a significantly greater IDH (10.30 ± 1.24 mm vs. 9.80 ± 1.42 mm, p = 0.005) and larger IFA (115.28 ± 18.44 mm² vs. 110.31 ± 18.20 mm², p = 0.042) at the cranial adjacent segment (L3/L4) compared to the Traditional group, suggesting a potential protective effect on the cranial adjacent segment structure. However, no significant differences were observed in the radiographic parameters of the caudal adjacent segment (L5/S1) at any follow‐up time point. Regarding the incidence of ASD, based on the Hilibrand criteria, a total of 11 patients (10.6%) in the Low‐profile group and 17 patients (13.6%) in the Traditional group developed ASD at the final follow‐up. Specifically, ASD at the cranial segment (L3/L4) was observed in 9 (8.7%) and 13 (10.4%) patients, respectively, while ASD at the caudal segment (L5/S1) was found in 2 (1.9%) and 4 (3.2%) patients, respectively. Although the overall incidence of ASD was lower in the Low‐profile group, the difference did not reach statistical significance (p = 0.622). A representative case demonstrating postoperative interbody fusion and maintained adjacent‐segment height after low‐profile topping‐off surgery is shown in Figure 5.

TABLE 3.

Comparison of radiographic parameters of adjacent segments and ASD incidence between the two groups.

Variables	Low‐profile group (n = 104)	Traditional group (n = 125)	p
L3/L4 disc height (mm)
Pre‐op	11.25 ± 1.16	11.33 ± 1.33	0.642
6 months	10.91 ± 1.22	10.90 ± 1.35	0.921
12 months	10.70 ± 1.20	10.40 ± 1.37	0.073
24 months	10.30 ± 1.24	9.80 ± 1.42	0.005 *
L3/4 foramen area (mm 2 )
Pre‐op	122.64 ± 17.70	121.68 ± 18.06	0.687
6 months	120.01 ± 17.97	116.98 ± 18.06	0.206
12 months	118.40 ± 18.42	114.98 ± 17.98	0.159
24 months	115.28 ± 18.44	110.31 ± 18.20	0.042 *
L5/S1 disc height (mm)
Pre‐op	10.77 ± 1.35	11.00 ± 1.63	0.245
6 months	10.62 ± 1.35	10.84 ± 1.63	0.264
12 months	10.56 ± 1.35	10.75 ± 1.64	0.361
24 months	10.47 ± 1.36	10.61 ± 1.65	0.498
L5/S1 foramen area (mm 2 )
Pre‐op	117.09 ± 19.31	117.58 ± 19.99	0.852
6 months	115.66 ± 19.38	115.92 ± 19.91	0.921
12 months	115.42 ± 19.44	115.01 ± 20.16	0.878
24 months	114.10 ± 19.54	113.79 ± 20.04	0.907
Incidence of ASD (24 months), n (%)
Cranial segment (L3/L4)	9 (8.7%)	13 (10.4%)	0.825
Caudal segment (L5/S1)	2 (1.9%)	4 (3.2%)	0.691
Total patients	11 (10.6%)	17 (13.6%)	0.622

Note: The bold values indicate statistically significant results. Statistical significance was defined as p < 0.05.

Abbreviation: ASD, adjacent segment degeneration.

^* p < 0.05; Values are presented as mean ± standard deviation or number (percentage).

FIGURE 5.

Radiological findings of a 69‐year‐old male patient with lumbar spinal stenosis who underwent low‐profile topping‐off surgery. (a) Preoperative sagittal lumbar radiograph; (b) preoperative sagittal lumbar CT image; (c) preoperative sagittal MRI of the L4/L5 disc; (d) preoperative axial MRI of the L4/L5 disc; (e, f) immediate postoperative anteroposterior and lateral lumbar radiographs; (g) sagittal reconstructed lumbar CT image obtained 6 months postoperatively, showing radiographic signs of interbody fusion at L4/L5; (h) sagittal lumbar CT image obtained 24 months postoperatively, demonstrating maintained intervertebral height with CT‐based radiographic evidence of fusion and no obvious decrease in adjacent segment height. This representative case mainly demonstrates CT‐based fusion evidence and does not include 24‐month postoperative MRI assessment of adjacent‐segment disc degeneration. Systematic MRI follow‐up with Pfirrmann grading will be performed in future prospective studies.

3.3. Clinical Outcomes

The VAS was used to evaluate the relief of low back pain and lower limb discomfort. Statistical analysis revealed no significant difference between the two groups in terms of postoperative improvement in VAS scores for leg pain, whereas a statistically significant difference was observed between the groups in VAS scores for back pain. In both groups, pain levels were significantly reduced compared to preoperative values. In the low‐profile topping‐off group, the mean preoperative VAS score for leg pain was 6.52 ± 0.84, which decreased to 1.18 ± 0.52 at 6 months postoperatively and further declined to 0.85 ± 0.55 at the final follow‐up. In the traditional topping‐off group, the corresponding scores were 6.44 ± 0.77 preoperatively, 1.19 ± 0.59 at 6 months, and 0.93 ± 0.57 at the final follow‐up. For back pain, the low‐profile group showed a reduction in VAS scores from 6.34 ± 1.24 preoperatively to 0.78 ± 0.56 at 6 months, and further to 0.76 ± 0.68 at the final follow‐up. In the traditional group, back pain VAS scores decreased from 6.46 ± 1.00 preoperatively to 1.10 ± 0.57 at 6 months, and to 0.94 ± 0.62 at the final follow‐up. Regarding the ODI, the low‐profile group improved significantly from 64.06% ± 6.81% preoperatively to 18.79% ± 2.49% at the final follow‐up (p < 0.001), while the traditional group improved from 64.82% ± 6.20% to 19.59% ± 2.31% (p < 0.001). The extent of postoperative ODI improvement was statistically greater in the low‐profile group (p < 0.05). In terms of the Japanese Orthopedic Association (JOA) score, the mean score in the low‐profile group increased from 12.71 ± 2.51 preoperatively to 26.45 ± 2.19 at the final follow‐up (p < 0.001), while the traditional group improved from 12.44 ± 2.20 to 25.95 ± 2.74 (p < 0.001). Although both groups showed significant improvements in pain relief and functional recovery as measured by VAS, ODI, and JOA scores, not all intergroup differences in improvement reached statistical significance. Detailed clinical outcomes and Bridwell fusion grades are presented in Table 2.

3.4. Subgroup and Multivariable Analyses

Subgroup analysis stratified by BMD is shown in Table 4. The sample sizes of the BMD subgroups were as follows: in the low‐profile group, 56 patients had normal BMD, 25 had osteopenia, and 23 had osteoporosis; in the traditional group, the corresponding numbers were 77, 39, and 9, respectively. Overall, the low‐profile group maintained favorable perioperative outcomes across BMD strata, particularly with respect to postoperative drainage volume. A statistically significant interaction was identified for postoperative drainage (P for interaction = 0.009), indicating that the magnitude of drainage reduction differed across BMD categories. No significant interaction was observed for operative time, intraoperative blood loss, incision length, postoperative hospitalization, final VAS for back pain, final VAS for leg pain, or final ODI (all p for interaction > 0.05), suggesting that the between‐group differences in these outcomes were generally consistent across BMD strata. A significant interaction was also observed for the final JOA score (p for interaction = 0.007). However, the direction of the between‐group differences was not uniform across BMD strata, indicating heterogeneity in functional recovery by bone density and warranting cautious interpretation.

TABLE 4.

Subgroup analysis stratified by bone mineral density (BMD).

Variables/subgroups	Low‐profile group	Traditional group	p	p for Interaction
Average surgery time (min)				0.216
Normal BMD	189.21 ± 36.03	183.25 ± 44.87	0.397
Osteopenia	167.40 ± 30.27	182.46 ± 38.62	0.087
Osteoporosis	174.65 ± 36.54	171.33 ± 26.68	0.780
Intraoperative blood loss (mL)				0.083
Normal BMD	224.84 ± 81.44	230.57 ± 86.24	0.697
Osteopenia	196.84 ± 91.52	245.77 ± 83.85	0.036 *
Osteoporosis	219.35 ± 93.78	294.89 ± 86.73	0.046 *
Postoperative drainage (mL)				0.009 *
Normal BMD	206.25 ± 81.15	261.82 ± 140.82	0.005 *
Osteopenia	141.20 ± 74.67	290.64 ± 129.58	< 0.001 *
Osteoporosis	201.09 ± 84.13	364.44 ± 107.19	0.001 *
Length of incision (mm)				0.412
Normal BMD	87.45 ± 8.16	98.16 ± 14.53	< 0.001 *
Osteopenia	88.12 ± 10.97	100.05 ± 14.34	< 0.001 *
Osteoporosis	93.13 ± 11.54	97.44 ± 13.47	0.413
Postoperative hospitalization (day)				0.816
Normal BMD	7.07 ± 1.56	8.44 ± 2.50	< 0.001 *
Osteopenia	7.16 ± 1.99	8.41 ± 2.31	0.025 *
Osteoporosis	7.00 ± 1.71	8.89 ± 3.14	0.119
VAS (back pain) at final follow‐up				0.641
Normal BMD	0.71 ± 0.59	0.91 ± 0.63	0.072
Osteopenia	0.80 ± 0.76	1.03 ± 0.63	0.224
Osteoporosis	0.83 ± 0.78	0.78 ± 0.44	0.827
VAS (leg pain) at final follow‐up				0.098
Normal BMD	0.71 ± 0.56	0.87 ± 0.50	0.101
Osteopenia	1.12 ± 0.53	0.97 ± 0.71	0.350
Osteoporosis	0.87 ± 0.46	1.22 ± 0.44	0.062
ODI at final follow‐up				0.159
Normal BMD	18.73 ± 2.32	19.69 ± 2.09	0.016 *
Osteopenia	19.36 ± 3.11	19.26 ± 2.58	0.890
Osteoporosis	18.30 ± 2.10	20.22 ± 2.99	0.105
JOA at final follow‐up				0.007 *
Normal BMD	26.93 ± 2.00	25.62 ± 2.96	0.003 *
Osteopenia	26.52 ± 2.22	26.31 ± 2.37	0.718
Osteoporosis	25.22 ± 2.24	27.22 ± 1.56	0.009 *

Note: Values are presented as mean ± standard deviation. BMD groups were defined as: Normal (T‐score ≥ −1.0), Osteopenia (−2.5 < T‐score < −1.0), and Osteoporosis (T‐score ≤ −2.5). Interaction p‐value indicates whether the difference between groups varies significantly across BMD strata. The bold values indicate statistically significant results. Statistical significance was defined as p < 0.05.

Abbreviations: BMD, bone mineral density; JOA, Japanese Orthopedic Association Score; ODI, Oswestry Disability Index; VAS, Visual Analogue Scale.

^* p < 0.05.

In multivariable linear regression adjusting for age, sex, BMD, diagnosis type, and preoperative ODI, the low‐profile technique remained independently associated with a lower final ODI (B = −0.868, 95% CI: −1.497 to −0.240, p = 0.007). In binary logistic regression, surgical technique was not significantly associated with overall complications after adjustment (OR = 0.677, 95% CI: 0.284–1.537, p = 0.360) (Table 5).

TABLE 5.

Multivariate regression analysis of factors associated with final ODI and complications.

Variables	Multiple linear regression (dependent: final ODI)			Binary logistic regression (dependent: complications)
	Coefficient (B)	95% CI	p	Odds ratio (OR)	95% CI	p
Group (low‐profile vs. traditional)	−0.868	−1.497, −0.240	0.007 *	0.677	0.284, 1.537	0.360
Age (years)	0.002	−0.033, 0.036	0.915	1.010	0.966, 1.058	0.651
Sex (Female vs. Male)	−0.810	−1.438, −0.181	0.012 *	0.946	0.419, 2.177	0.895
BMD (T‐score)	0.036	−0.161, 0.233	0.721	1.035	0.798, 1.348	0.793
Diagnosis (LSS vs. LDH)	0.320	−0.330, 0.971	0.333	1.471	0.636, 3.347	0.358
Preoperative ODI	−0.013	−0.061, 0.035	0.600	1.018	0.956, 1.087	0.582

Note: The bold values indicate statistically significant results. Statistical significance was defined as p < 0.05.

Abbreviations: BMD, bone mineral density; CI, confidence interval; LDH, lumbar disc herniation; LSS, lumbar spinal stenosis; ODI, Oswestry Disability Index.

^* p < 0.05; Covariates adjusted in the models included age, sex, BMD, diagnosis type, and preoperative ODI.

4. Discussion

4.1. Main Findings

The present study compared low‐profile topping‐off surgery with traditional pedicle screw‐based topping‐off surgery for LDD. The main findings were as follows: (i) the low‐profile technique achieved comparable clinical improvement to the traditional technique, while showing less intraoperative blood loss, lower postoperative drainage volume, shorter incision length, and shorter postoperative hospital stay; (ii) the low‐profile group showed a higher fusion success rate and better preservation of cranial adjacent‐segment radiographic parameters at 24 months, although the overall ASD incidence was not significantly different between groups; and (iii) subgroup and multivariable analyses suggested that the perioperative advantages of the low‐profile technique were generally maintained across BMD strata, and surgical technique remained independently associated with final ODI after adjustment for potential confounders. These findings indicate that low‐profile topping‐off may represent a less invasive and clinically effective alternative to traditional PS‐based topping‐off, while further prospective studies are needed to confirm its long‐term value.

4.2. Technical Rationale and Perioperative Advantages of the Low‐Profile Technique

Currently, PLIF remains a widely adopted surgical procedure for lumbar degenerative disorders, including lumbar spinal stenosis and lumbar disc herniation [22]. However, conventional PLIF typically requires extensive posterior exposure with substantial paraspinal muscle dissection and prolonged retraction, which may increase operative trauma and contribute to greater intraoperative blood loss, more severe postoperative pain, and delayed functional recovery [26, 36, 37]. In addition, elimination of motion at the fused segment may result in stress concentration at the index level and compensatory hypermobility at adjacent segments, thereby increasing the risk of ASD [38, 39]. Accordingly, an important clinical objective is to optimize surgical strategies that maintain adequate decompression and stability while minimizing soft‐tissue injury, facilitating postoperative recovery, and potentially reducing biomechanical burden on adjacent segments.

When compared to the traditional topping‐off surgery, the low‐profile topping‐off procedure offers a substantial improvement in surgical outcomes by integrating CBTs with PS technology for internal fixation. The design of the CBTs entry point is situated closer to the midline, effectively minimizing the dissection extent of soft tissue and paravertebral muscles, thereby preserving the normal anatomical structure [40, 41]. Additionally, the CBTs trajectory exhibits a caudal tilt ranging from 30° to 40° in the sagittal plane. When combined with PS, the extension lines of the upper and lower vertebral fixation screws create an included angle, subsequently mitigating skin incision and soft tissue trauma [42]. Due to the innovative design of the CBTs entry point and trajectory, the low‐profile topping‐off surgery demonstrates a minimized impact on the exposure area, significantly reducing potential harm to paravertebral muscles, soft tissues, and adjacent segmental facet joints.

During the treatment of osteoporosis patients, it was noted that the decrease in bone density was more pronounced in cancellous bone compared to cortical bone [43]. The trajectory of CBTs traverses through cortical bone structures, including the inner wall, outer wall, and upper outer wall of the pedicle. Santoni et al. further confirmed via CT scanning the presence of abundant high‐density bone surrounding the CBTs [28]. When compared to PS screws, CBTs technology exhibits a larger contact area between the screw track and cortical bone, subsequently enhancing the screw's holding force with the bone [41]. This feature confers significant biomechanical advantages to CBTs in the treatment of osteoporosis or reduced bone density patients [30]. Ding et al.'s research demonstrated that in single‐segment lumbar fusion surgery, CBTs fixation demonstrates comparable clinical symptom improvement to PS‐fixed TLIF. Notably, the low‐profile topping‐off group exhibited significantly lower rates of postoperative complications and screw loosening compared to the traditional topping‐off group. Additionally, perioperative outcomes such as intraoperative blood loss, drainage volume, and incision length were superior in the low‐profile topping‐off group compared to the traditional topping‐off group. These findings provide robust evidence supporting the clinical utilization of CBTs in the management of osteoporosis patients [27].

4.3. Radiographic Outcomes and Adjacent‐Segment Degeneration

Previous biomechanical and clinical studies suggest that Coflex‐based topping‐off constructs may improve the biomechanical environment of adjacent segments by reducing stress transfer and preserving partial motion, thereby potentially lowering the risk of ASD [43, 44]. Our Table 3 findings provide partial clinical‐radiographic support for this mechanism; overall, the 24‐month total ASD incidence in both groups was at the lower end of the range reported in previous studies [44, 45]. Specifically, compared with the traditional topping‐off group, the low‐profile topping‐off group showed better preservation of disc height and IFA at the cranial adjacent segment (L3/L4) at 24 months, whereas no significant between‐group differences were observed at the caudal adjacent segment (L5/S1). However, regarding ASD incidence, we did not observe a statistically significant reduction in the low‐profile group at 24 months. Although cranial, caudal, and total ASD incidence rates were numerically lower in the low‐profile group, all between‐group comparisons were non‐significant. This suggests that, within the current follow‐up period, the two topping‐off strategies showed generally comparable performance in preventing ASD incidence. Taken together, within this 24‐month follow‐up, the advantage of the low‐profile topping‐off technique appears to be more evident in radiographic preservation (particularly at L3/L4) than in a clear statistical reduction in ASD incidence compared with the traditional technique. Longer follow‐up and larger cohorts are needed to determine whether this radiographic benefit translates into a significant reduction in ASD incidence.

4.4. Bone Mineral Density and Subgroup Findings

Further stratification by BMD (T‐score categories) provided additional insight into whether the comparative advantages of the low‐profile technique varied by bone quality. The subgroup results suggested that the perioperative advantages of the low‐profile technique were generally preserved across BMD strata, while a significant interaction was identified for postoperative drainage volume, indicating that the magnitude of this benefit differed according to bone density category. In addition, a significant interaction was observed for the final JOA score, suggesting possible effect modification by bone density. This finding should be interpreted cautiously. In the osteoporosis subgroup, although the main preoperative baseline characteristics were generally balanced between the two groups, the traditional group showed a higher final JOA score than the low‐profile group. This unexpected result may be related to several factors. First, in osteoporotic patients, the theoretical biomechanical advantage of CBTs may be attenuated because both cortical and cancellous bone quality can be compromised, which may affect screw‐bone interface strength and functional recovery. Second, postoperative recovery in this subgroup may also be influenced by fusion progression, rehabilitation tolerance, general physical reserve, and the effectiveness of anti‐osteoporosis treatment. Finally, given the relatively small sample size of the osteoporosis subgroup, this interaction may also reflect random variation. Therefore, this subgroup finding should be regarded as exploratory rather than definitive evidence that either technique is superior in osteoporotic patients.

4.5. Perioperative Benefits, Clinical Efficacy, and Adjusted Analyses

This study demonstrated that the low‐profile topping‐off group had significantly reduced intraoperative blood loss (216.89 ± 86.63 mL, p = 0.046), incision length (88.87 ± 9.86 mm, p < 0.001), and postoperative drainage volume (189.47 ± 84.12 mL, p < 0.001) compared to the traditional topping‐off group. Regarding clinical efficacy, although the low‐profile group showed statistically superior improvements in back pain VAS and ODI, the inter‐group differences (0.18 for back pain VAS and 0.80 for ODI) did not exceed the Minimally Clinically Important Difference (MCID) thresholds established in previous studies (1.2 for back pain VAS and 12.8 for ODI) [46]. This indicates that the low‐profile technique provides comparable pain relief and functional recovery to the traditional method. These findings highlight the advantages of the low‐profile topping‐off technique, which innovatively combines PS and CBTs fixation. Consequently, the principal advantage of this approach lies in its more minimally invasive profile and faster postoperative recovery, while achieving long‐term clinical efficacy that is at least comparable to that of the conventional technique. Importantly, multivariable regression further supported the independence of these findings after adjustment for age, sex, BMD, diagnosis type, and preoperative ODI: the low‐profile technique remained associated with a lower final ODI, whereas the adjusted odds of overall complications were not significantly different between groups (Table 5). Notably, the magnitude of the ODI difference was small, consistent with our MCID‐based interpretation.

4.6. Limitations and Future Directions

There exist several limitations in the present study. First, the relatively small sample size and the single‐institution study design may introduce a degree of bias. To enhance the universality and precision of the study, we intend to expand the sample size and actively pursue multicenter research. In addition, it should be noted that the MCID thresholds used to interpret clinical improvement were derived from previous studies involving different patient populations and surgical contexts, and thus their direct applicability to the present cohort may be limited and should be interpreted with caution. Second, as this study employs a retrospective cohort design, potential biases and omissions in case selection and information gathering may arise. To address these concerns, our research team has initiated prospective research on this surgical procedure and aims to continually promote this clinical research work, thereby providing a more precise evaluation of surgical outcomes. In particular, selection bias cannot be fully excluded because surgeons may preferentially choose surgical approaches based on patients' baseline characteristics, and residual confounding may remain despite statistical adjustment. Thirdly, although adjacent‐segment radiographic parameters and ASD incidence were evaluated using radiographs and CT scans, systematic 24‐month MRI follow‐up was not available for all patients. Therefore, MRI‐based assessment of adjacent disc degeneration and Pfirrmann grading could not be performed. Future prospective studies should include standardized postoperative MRI follow‐up to better evaluate adjacent‐segment disc degeneration. Fourthly, postoperative outcomes were mainly evaluated using VAS, ODI, JOA, radiographic parameters, and recorded complications, while several long‐term patient‐centered endpoints were not systematically collected in the current retrospective follow‐up database. These included reoperation due to adjacent segment disease, frequency of epidural steroid injections, chronic analgesic use, and general quality‐of‐life measures such as the SF‐36 with physical and mental component summaries. Therefore, patient‐centered recovery domains, including physical function, emotional role, social participation, and long‐term treatment burden, could not be comprehensively assessed. Future prospective studies should incorporate these endpoints to more fully evaluate the durability and clinical value of the topping‐off technique.

5. Conclusion

Low‐profile topping‐off surgery achieved clinical outcomes comparable to those of traditional pedicle screw‐based topping‐off surgery, while showing advantages in reducing surgical invasiveness, including less blood loss, shorter incision length, lower drainage volume, and shorter hospitalization. Although no significant reduction in overall ASD incidence was observed during the current follow‐up period, the low‐profile technique showed favorable fusion outcomes and better preservation of cranial adjacent‐segment radiographic parameters. Further prospective multicenter studies with longer follow‐up are needed to confirm its long‐term value.

Author Contributions

Honghao Yang: methodology, software, data curation, formal analysis, investigation, writing – original draft. Kuan Li: data curation, investigation, validation, visualization, methodology, conceptualization, writing – original draft, writing – review and editing, formal analysis, supervision. Mingzheng Zhao: investigation, methodology, software, formal analysis, writing – original draft. Jun Miao: software, validation. Yong Hai: writing – review and editing, conceptualization, supervision, resources, project administration. Yuzeng Liu: conceptualization, funding acquisition, validation, resources, project administration, writing – review and editing. Yaoshen Zhang: conceptualization, methodology, supervision, resources, writing – review and editing. Jing Su: investigation, data curation, supervision, formal analysis. Yangpu Zhang: methodology, investigation, validation, formal analysis. Shixuan Guo: investigation, visualization, data curation, writing – original draft. Dongchenlei Wang: data curation.

Funding

This study was supported by the National Clinical Medical Research Center of Orthopedics and Sports Rehabilitation Innovation Fund (2021‐NCRC‐CXJJPY‐17); Clinical Research Incubation Program of Beijing Chao‐Yang Hospital (CYFH202316); Research and Translational Application of Clinical Diagnosis and Treatment Technology in Capital City (Z211100002921027); Beijing Natural Science Foundation ‐ Haidian Original Innovation Joint Fund (L252114); Clinical Research Enhancement Program of Beijing Chao‐Yang Hospital, Capital Medical University (CYTS2025C31).

Ethics Statement

The studies involving human participants were reviewed and approved by the Research Ethics Committee of Beijing Chao‐Yang Hospital (ID 2024‐Ke‐429). The patients/participants provided their written informed consent to participate in this study; No animal studies are presented in this manuscript; No potentially identifiable human images or data is presented in this study. No potentially identifiable human images or data are presented in this study. We solemnly promise that all the methodology of this study has been carried out in accordance with the relevant guidelines and provisions of the Declaration of Helsinki, among others.

Consent

Informed consent was obtained from all study participants for identifiable medical records and imaging information of participants published in this journal. All experimenters involved in this experiment consent to publish this article.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.