The Role of Pre-soaked Bone Cement Gelatin Sponge Strips in Sealing Anterior Vertebral Wall Defects During Percutaneous Kyphoplasty for Osteoporotic Thoracolumbar Compression Fractures in the Elderly: A Retrospective Study

Abstract

Study Design

Retrospective cohort study.

Objectives

Unilateral percutaneous kyphoplasty (PKP) is widely used to treat osteoporotic vertebral compression fractures (OVCF) in elderly patients. Cement leakage is the most common complication and may cause serious consequences. Although various techniques have been proposed to reduce leakage, few studies have addressed cases with preexisting anterior vertebral wall defects. This study aims to evaluate whether intraoperative sealing of these defects with gelatin sponge strips pre-soaked in bone cement can reduce leakage and improve surgical outcomes.

Methods

A retrospective analysis was conducted on 86 patients who underwent unilateral single-level PKP for thoracolumbar OVCF from December 2021 to October 2024. All patients had preoperative cortical defects in the anterior vertebral wall and were divided into two groups: Group A (n = 43) received conventional PKP, while Group B (n = 43) underwent PKP with defect sealing using pre-soaked gelatin sponge strips. Clinical and radiological assessments were performed preoperatively, immediately postoperatively, at 3 months, and at 12 months.

Results

Both groups achieved favorable outcomes without severe complications. Cement leakage at the anterior vertebral wall was significantly lower in Group B than in Group A (P = 0.007). Group B also demonstrated better cement diffusion (P = 0.013) and a higher cement injection volume (P = 0.022).

Conclusions

In elderly patients with thoracolumbar OVCF and anterior vertebral cortical defects, unilateral puncture PKP combined with bone cement–soaked gelatin sponge strips effectively reduces anterior vertebral wall cement leakage and improves cement distribution, which may contribute to better procedural safety and stability.

Introduction

OVCF is a common type of fracture that poses a significant risk to the health of the elderly. Approximately 1.7 million cases of OVCF occur annually in the United States and Europe. In China, the prevalence of osteoporosis among individuals over 50 years old is about 34.65%, with nearly one-third of these cases leading to fractures. ¹ Thoracic or low back pain is the primary symptom in patients with OVCF. OVCF can lead to flattening or displacement of spinal fractures, and insufficient vertebral height may result in secondary spinal kyphosis, decreased cardiopulmonary function, depression, and reduced quality of life.^2,3 The primary goals of treating patients with OVCF are to restore mobility, alleviate pain, and prevent new fractures. Traditional conservative treatment methods include bed rest, opioid analgesics, and external bracing to reduce pain and stabilize the vertebrae. However, prolonged bed rest often leads to various complications, such as pneumonia, pressure ulcers, and deep vein thrombosis. ⁴ Additionally, OVCF patients undergoing conservative treatment face risks of prolonged pain and increased bone loss, which elevate the likelihood of further progression or recurrence of vertebral fractures. ⁵ Therefore, surgical treatment is often chosen for OVCF patients. Currently, minimally invasive procedures such as percutaneous vertebroplasty (PVP) and PKP are effective surgical methods for treating OVCF. Both procedures effectively alleviate pain and restore vertebral height in OVCF patients. Pain relief and vertebral height restoration are reliable indicators for evaluating the efficacy of bone cement. ⁶ Currently, compared to PVP, PKP allows for the injection of a larger volume of bone cement and the use of balloon expansion to restore vertebral height, thereby rapidly relieving patients’ pain. PKP is generally considered safe and effective and is therefore recommended as the preferred surgical method for OVCF.^7,8

However, existing studies indicate that PKP treatment has certain limitations, including cement leakage, unsatisfactory vertebral height restoration, and postoperative complications such as adjacent vertebral fractures and refractures of the treated vertebra. ⁹ In cases of insufficient cement filling, the improvement of symptoms in the injured vertebra is suboptimal, significantly increasing the likelihood of refracture (Figure 1). Conversely, overfilling with bone cement can result in excessive stiffening of the treated vertebra, thereby increasing the risk of cement leakage and fractures in adjacent vertebrae. ¹⁰ Moreover, the distribution of bone cement within the vertebral body significantly influences its mechanical support. Poor cement distribution may lead to inadequate structural stability and consequently increase the risk of vertebral refracture. ¹¹ Cement leakage can occur through various pathways, including the basivertebral venous plexus, vertebral endplates, and cortical defects. Leakage into the spinal canal or neural foramina may lead to serious neurological complications such as nerve compression or paralysis. While cement infiltration into the intervertebral disc may accelerate disc degeneration and disrupt normal load transmission, increasing stress on adjacent levels and causing subsequent fractures. ¹² Among these pathways, leakage through pre-existing anterior vertebral cortical defects deserves particular attention. Cement extravasation from these defects can directly reach major abdominal vessels, potentially causing life-threatening pulmonary embolism, or result in injury to adjacent organs such as the aorta or bowel.^13,14 Despite routine intraoperative strategies—such as optimal needle placement, adjustment of cement viscosity, low-pressure injection, real-time fluoroscopic monitoring, and limiting cement volume—to reduce leakage risks, ¹⁵ patients with anterior cortical defects remain at high risk for anterior cement leakage due to the absence of an effective mechanical barrier. To address this issue, our medical centers have adopted a novel strategy specifically for patients with substantial preoperative defects in the anterior vertebral wall. Before cement injection, bone cement–soaked gelatin sponge strips are placed into the anterior cortical defect via the working cannula, with the goal of sealing the defect and mitigating anterior cement leakage, while still allowing adequate cement filling for stabilization.

Figure 1.

A 79-Year-old Female Patient With Osteoporosis was Diagnosed With a Compression Fracture of the L2 Vertebral Body, Accompanied by Lumbar and Back pain. On April 19, 2022, the Patient Underwent a Unilateral Approach L2 Vertebral PKP, With the Bone Cement Confined to the Side of the Puncture and Not Diffused to the Opposite side. Later, Further Compression Occurred on the Side of the L2 Vertebral Body where the Bone Cement was Not Distributed. On November 14, 2023, the Patient Underwent a Second PKP Procedure for the L2 Vertebra, as Well as for the L3 Vertebra. A, B: Preoperative X-ray Images From the First Surgery (A: anteroposterior view, B: Lateral View), Showing a Mild Compression Fracture of the L2 Vertebral Body, With Prior Vertebroplasty Performed on the T12 and L4 Vertebrae (Bone Cement Visible Within the Vertebrae). C, D: Preoperative CT Images From the First Surgery (C: coronal view, D: Sagittal View), Showing Findings Consistent With the X-ray. E: Preoperative Sagittal T2-Weighted Fat-Suppressed MRI From the First Procedure Showed a High Signal Near the Inferior Endplate of L2. F, G: Postoperative X-ray Images From the First Surgery (F: anteroposterior view, G: Lateral View), Showing that the Bone Cement was Distributed Only on One Side of the Vertebral body. H, I: Preoperative X-ray Images From the Second Surgery (H: anteroposterior view, I: Lateral View), Showing Further Compression and Flattening of the Side of the L2 Vertebral Body where the Bone Cement was Not Distributed. J, K, L: Preoperative CT Images (J: coronal view, K: Sagittal View of the Side With Bone Cement distribution, L: Sagittal View of the Side Without Bone Cement Distribution) From the Second Surgery, Showing that the Side of the L2 Vertebral Body With Bone Cement Maintained Its Height Relatively Well, while the Side Without Bone Cement Showed Significant Further Compression. M, N: Preoperative Sagittal T2-Weighted Fat-Suppressed MRI Images (M: side With Bone cement, N: Side Without Bone Cement) From the Second Surgery, Showing that the Side of the L2 Vertebral Body With Bone Cement Maintained Its Height Well, With No Obvious High Signal, while the Side Without Bone Cement Showed Decreased Vertebral Height, Accompanied by High Signal Within the Vertebrae. O, P: Postoperative X-ray Images From the Second Surgery (O: anteroposterior view, P: Lateral View), Showing Partial Recovery of the Height on the Side of the L2 Vertebral Body With Further Compression

This study aims to compare the clinical efficacy, bone cement injection volume, cement dispersion, vertebral height restoration, and incidence of anterior vertebral wall cement leakage between two treatment approaches for OVCF patients with significant preoperative anterior vertebral wall defects: conventional unilateral PKP and PKP with the placement of pre-soaked bone cement gelatin sponge strips. The goal is to evaluate whether this novel technique can reduce cement leakage and thereby contribute to improved clinical outcomes, including pain relief and vertebral stability.

Methods

Patients

Following approval from the institutional review board, we conducted a retrospective analysis of data from patients with OVCF who underwent PKP at the Department of Spine Surgery between December 3, 2021 and October 24, 2024. Due to the retrospective nature of the study, written informed consent was waived. The inclusion criteria for this study were as follows: (1) age >50 years, (2) single-segment PKP performed at the T10-L3 levels, (3) ineffective conservative treatment, defined as persistent back pain after ≥2 weeks of standard non-operative management (including bed rest, analgesics, and anti-osteoporotic medication) or severe back pain intolerable to conservative treatment, as documented in the clinical records, (4) minimum follow-up period of 10 months to 1 year postoperatively, (5) preoperative magnetic resonance imaging (MRI) showing significant high T2-weighted signal and vertebral height loss in the injured vertebra, (6) preoperative bone density testing indicating osteoporosis, (7) stable vertebral compression fractures without severe complications such as spinal cord or nerve root compression; all fractures were preoperatively classified according to the AO classification system and identified as type A1 vertebral compression fractures, (8) anterior cortical bone defects were identified on preoperative three-dimensional CT scans as visually apparent discontinuities of the anterior vertebral wall cortex, as judged by experienced clinicians, (9) presence of a history of trauma leading to the vertebral compression fracture, regardless of trauma severity. The exclusion criteria for this study included: (1) severe spinal deformities, (2) incomplete preoperative and postoperative imaging data (preoperative X-rays, 3D CT, MRI, postoperative X-rays), (3) absence of clinical assessments at preoperative, immediate postoperative, 3 months, or last postoperative follow-up, (4) unstable vertebral fractures or accompanying nerve damage, (5) vertebral fractures without osteoporosis, (6) patients unable to cooperate with the procedure. According to these criteria, a total of 86 patients were included in this study. Among them, 43 patients underwent conventional unilateral PKP (Group A), while the remaining 43 patients received unilateral PKP with the insertion of pre-soaked gelatin sponge strips before bone cement injection (Group B). Patients were allocated to Group A or Group B using an alternating assignment method based on their order of presentation (i.e., the first patient was assigned to Group A, the second to Group B, and so on). All procedures were performed under local anesthesia.

Demographic and Perioperative Data Collection

This study obtained patient demographic profiles, perioperative data, and radiological records by querying the internal medical records management system of our medical center. The demographic characteristics of the included patients included age, gender, height, weight, body mass index (BMI), and medical history. Perioperative data and radiological records included preoperative symptom duration, preoperative comorbidities, severity of preoperative pain symptoms, specific vertebral segments affected by compression fractures, severity of vertebral compression, results of preoperative bone density testing, surgical duration, volume of bone cement injected intraoperatively, intraoperative complications, postoperative pain relief, distribution of cement dispersion, restoration of vertebral height, vertebral anterior wall cement leakage, changes in the wedge angle of the fractured vertebra, changes in the local kyphotic angle, postoperative complications, length of hospital stay, postoperative modified Oswestry Disability Index (ODI), as well as long-term outcomes and patient satisfaction assessed during follow-up and review visits.

Surgical Technique

Before the surgery, all patients were thoroughly informed about the procedural details, and all signed informed consent forms for the surgery. All the surgeries were performed using a unilateral approach, where the procedure was carried out through a single-sided puncture to access the vertebral body for injection of the bone cement. All procedures were performed by experienced spine surgeons (with more than eight years of practice). The injection of bone cement was standardized to occur during the late wire-drawing phase to ensure optimal viscosity and minimize leakage risk.

Group A (Conventional PKP Surgery): Patients were placed in a prone position with cushions under the chest and pelvis to maintain the natural physiological curvature of the spine. The target vertebral body was identified under fluoroscopy, and the puncture point was marked on the skin. The surgical area was then routinely disinfected and draped. Local anesthesia was administered at the puncture site. A small incision was made at the puncture site using a scalpel. Under fluoroscopic guidance, a puncture needle was percutaneously inserted into the vertebral body. A working channel was established through the puncture needle, and the channel diameter was gradually enlarged using a solid drill to facilitate the entry of subsequent instruments. A balloon dilator was inserted and advanced to the predetermined position within the vertebral body under fluoroscopic guidance. Contrast agent was slowly injected under fluoroscopic monitoring to expand the balloon, creating a cavity within the vertebral body and partially restoring the vertebral height. After expansion, the balloon was deflated and removed. Bone cement was prepared to the appropriate viscosity for injection. The bone cement was slowly injected into the cavity within the vertebral body through the working channel, and its distribution was monitored using fluoroscopy. Once the bone cement had solidified, the working channel and all instruments were removed. The incision was sutured and covered with a sterile dressing.

Group B (Modified PKP Surgery): Preoperative three-dimensional CT scans were used to precisely evaluate the specific location, orientation, and extent of the anterior vertebral wall defect. Patients were placed in a prone position with cushions placed under the chest and pelvis to maintain the natural physiological curvature of the spine. The target vertebral body was identified under fluoroscopy, and the puncture point was marked on the skin. The surgical area was routinely disinfected and draped. Local anesthesia was administered at the puncture site, followed by a small incision made with a scalpel. The lateral and cranio-caudal angulations of the puncture needle were planned according to the defect site. Under C-arm fluoroscopic guidance, the needle trajectory was adjusted in real time to ensure the needle tip accurately reached the predetermined target point. A working channel was established through the puncture needle, and its diameter was gradually enlarged using a solid drill to facilitate the insertion of subsequent instruments. A balloon dilator was then inserted and advanced to the predetermined position within the vertebral body under fluoroscopic guidance. Contrast agent was slowly injected under fluoroscopic monitoring to expand the balloon, thereby creating a cavity within the vertebral body and partially restoring vertebral height. After expansion, the balloon was deflated and removed. During the procedure, gelatin sponge strips were prepared by first cutting them into rectangular blocks measuring 4 × 3 × 5 mm. These blocks were then rolled into slender cylindrical shapes and impregnated with low-viscosity bone cement. The sponge strips were promptly withdrawn to ensure uniform coating with bone cement, facilitating their smooth passage through the working channel without obstructing the needle pathway, and guaranteeing unobstructed subsequent bone cement injection into the vertebral body. The prepared bone cement-impregnated gelatin sponge strips were then placed into the anterior portion of the cavity within the vertebral body. Their position and stability in effectively sealing the anterior vertebral wall defect were confirmed under anteroposterior and lateral X-ray fluoroscopy, thereby reducing the risk of bone cement extravasation. Subsequently, bone cement was injected to fill the vertebral cavity, with its distribution continuously monitored under fluoroscopy. After the bone cement solidified, the working channel and all instruments were removed, and the incision was sutured and covered with a sterile dressing (Figure 2). Postoperative CT scans confirmed that the bone cement-impregnated gelatin sponge strips were accurately placed at the anterior vertebral wall defect site identified on preoperative CT (Figure 3).

Figure 2.

A 62-Year-old Female Patient With Osteoporosis was Diagnosed With a Vertebral Compression Fracture at the L1 Level, Presenting With Low Back pain. The Patient Underwent PKP via a Unilateral Approach. During the Procedure, Pre-soaked Bone Cement-Impregnated Gelatin Sponge Strips Were Used to Seal the Cortical Defect at the Anterior Edge of the L1 Vertebral Body, Followed by the Injection of Bone Cement. A, B: Preoperative X-ray Findings (A: anteroposterior view, B: Lateral view). C: Preoperative Sagittal CT Scan Showing a Significant Cortical Defect at the Anterior Edge of the L1 Vertebral body. D: Preoperative Sagittal T1-Weighted MRI Showing a Low Signal Within the L1 Vertebral body. E: Preoperative Sagittal T2-Weighted Fat-Suppressed MRI Showing a High Signal Within the L1 Vertebral body. F, G: Postoperative X-ray Images (F: anteroposterior view, G: Lateral View) Demonstrating the Diffusion of Bone Cement to the Contralateral Side of the L1 Vertebral body. H: Intraoperative Localization and puncture. I: Balloon Expansion During surgery. J: Placement of Pre-soaked Bone Cement-Impregnated Gelatin Sponge Strips to Seal the Cortical Defect at the Anterior Edge of the Vertebral body. K: Injection of Bone Cement Into the Vertebral Body

Figure 3.

Sealing Effect of the Anterior Vertebral Wall defect. A: Preoperative Sagittal 3D CT Image, With the White Arrow Indicating the Location of the Anterior Vertebral Wall defect. B: Postoperative Sagittal 3D CT Image, With the White Arrow Indicating the Position of the Bone Cement Impregnated Gelatin Sponge Used for sealing. C: Preoperative Axial 3D CT Image, With the White Arrow Indicating the Location of the Anterior Vertebral Wall defect. D: Postoperative Axial 3D CT Image, With the White Arrow Indicating the Position of the Bone Cement Impregnated Gelatin Sponge Used for Sealing

Clinical and Radiographic Evaluation

In evaluating clinical outcomes, the ODI was utilized to assess the degree of patient disability. ODI scores were measured preoperatively, immediately postoperatively, at 3 months postoperatively, at the last postoperative follow-up. Additionally, patients’ thoracic or lumbar Visual Analog Scale (VAS) scores were assessed at the following time points: preoperatively, immediately postoperatively, at 3 months postoperatively, and at the last postoperative follow-up.

Radiographic assessment includes the use of preoperative X-rays, 3D CT, and MRI to confirm new vertebral compression fractures in specific spinal segments. Additionally, it involves evaluating the specific compression sites of newly developed compression fractures and assessing the severity of vertebral compression and kyphosis based on preoperative X-rays, 3D CT, and MRI. The specific compression areas of the injured vertebra are categorized into the upper third, middle third, and lower third of the vertebral body.^16-19 Vertebral geometry is represented by the wedge angle of the compressed vertebra. This angle is calculated as the angle between the lines connecting the upper and lower endplates of the injured vertebra. The degree of kyphotic deformity is assessed using the three-segment kyphotic angle, which takes into account the adjacent vertebrae above and below the injured vertebra. This angle is measured as the angle between the line connecting the upper endplate of the adjacent vertebra above and the line connecting the lower endplate of the adjacent vertebra below. Vertebral height should be recorded at the point of the most severe collapse. This measurement is an important parameter for evaluating structural improvements post-surgery. Osteoporosis is assessed using preoperative dual-energy X-ray absorptiometry (DEXA). Intraoperative X-rays are used to evaluate cement leakage, while postoperative X-rays are used to assess the dispersion of bone cement and the recovery of vertebral height. Cement leakage is defined as the diffusion of bone cement beyond the upper and lower endplates and the anterior and posterior boundaries of the vertebra. This study focuses specifically on cement leakage at the anterior aspect of the vertebra. The dispersion of bone cement is assessed through X-ray examination to observe whether the cement has spread to the opposite side of the vertebra. The recovery of vertebral height is calculated using the following formula: [ (Postoperative Vertebral Height – Preoperative Vertebral Height) / Preoperative Vertebral Height ] × 100%. Additionally, X-rays taken immediately postoperatively, at 3 months, and at the last postoperative follow-up were used to assess mid- to long-term postoperative complications, including vertebral fracture recurrence, the occurrence of new fractures in adjacent vertebrae, and abnormalities in bone cement, such as cement degradation or bone resorption around the cement. This assessment could evaluate the long-term safety and durability of the procedure, identify late-onset adverse events affecting spinal stability or function, and provide evidence to guide postoperative management and improve patient outcomes. The formula for calculating the mid- to long-term postoperative complication rate is as follows: Mid- to Long-Term Postoperative Complication Rate = (Number of Patients with Mid- to Long-Term Postoperative Complications / Total Number of Patients) × 100%.

Consistency Verification

Two spine surgeons independently evaluated all imaging parameters in a blinded manner. Interobserver consistency was assessed by comparing their measurements, and any discrepancies were resolved through discussion and consensus. Intraclass correlation coefficients (ICCs) were calculated to quantitatively evaluate the level of agreement between the two observers.

Statistical Analysis

All continuous variables are presented as mean ± standard deviation. ICCs were used to assess the consistency of measurements among different assessors, with results categorized as poor (0−0.39), fair (0.4−0.74), and excellent (0.75−1). For continuous variables, data normality was first assessed using the Shapiro-Wilk test, and homogeneity of variance was evaluated using the Levene test for comparisons between two groups of patients. Following confirmation of normal distribution and variance homogeneity, independent samples t-tests were used to compare differences between groups. For data not following a normal distribution, Mann-Whitney U tests were employed for comparisons. For data (VAS scores and ODI) obtained from the same group of patients at different follow-up time points, the Friedman test was used to assess differences among multiple variables. Pairwise comparisons following the Friedman test were conducted using the Nemenyi test. Chi-square tests were used for categorical variables. Statistical significance was set at P < 0.05. All analyses were performed using SPSS version 23.0 (IBM Corporation, Armonk, NY, USA).

Results

Perioperative-Related Data

Between December 2021 and October 2024, this study included 86 patients diagnosed with OVCF (13 males and 73 females) based on predefined inclusion and exclusion criteria. All patients underwent single-segment, unilateral PKP in the thoracolumbar region at the Department of Spine Surgery. The average age of these patients was 70.85 ± 6.02 years, with an age range of 54 to 85 years. Based on the surgical approach, the patients were divided into two groups: Group A, which included 43 patients, and Group B, which also included 43 patients. A detailed description of the surgical procedures can be found in the Methods section. In group A, fractures occurred at the T10 vertebra in one patient, T11 in three, T12 in six, L1 in nineteen, L2 in seven, and L3 in seven. In group B, fractures were at T10 in two patients, T11 in five, T12 in eleven, L1 in seventeen, L2 in four, and L3 in four. The baseline characteristics of the two groups, including age, gender, height, weight, BMI, duration of preoperative symptoms, preexisting comorbidities, surgical duration, and length of hospital stay, were similar, with no statistically significant differences. The mean volume of bone cement injected into the vertebra was higher in Group B (5.63 ± 1.10 mL) than in Group A (5.11 ± 0.99 mL), with a mean difference of 0.52 mL (95% CI: 0.0761 to 0.9704, P = 0.022, Cohen’s d = 0.50), indicating a medium effect size (Table 1).

Table 1.

Demographic and Surgical Data Among Groups

Statistical variables	Group A	Group B	P Value
Number of patients	43	43	-
Sex			0.132
Male	9	4
Female	34	39
Age (years)	71.16 ± 6.46	70.53 ± 5.61	0.631
Height (m)	1.60 ± 0.06	1.60 ± 0.06	0.574
Weight (kg)	62.06 ± 9.45	63.12 ± 8.36	0.584
BMI (kg/m2)	24.19 ± 3.40	24.66 ± 3.02	0.500
Duration of preoperative symptoms (d)	4.30 ± 10.13	10.30 ± 30.24	0.655
Preexisting comorbidities	27	32	0.245
Surgical duration (min)	29.91 ± 10.07	33.19 ± 10.70	0.147
Length of hospital stay (d)	5.37 ± 1.27	4.86 ± 1.32	0.071
Volume of bone cement injected (ml)	5.11 ± 0.99	5.63 ± 1.10	0.022 a

^aP value <0.05.

Clinical Outcomes

Both groups achieved favorable surgical outcomes. Detailed clinical results are presented in Table 2, demonstrating significant postoperative relief of back pain and functional improvements throughout the follow-up period. There were no statistically significant differences between the two groups in terms of postoperative pain relief and functional scores. In addition, within-group comparisons of the VAS and ODI scores at preoperative and various postoperative follow-up time points for both patient groups revealed significant statistical differences. Subsequent pairwise comparisons within each group were then conducted. No significant differences were found between the VAS and ODI scores at 3 months and the last postoperative follow-up in both groups. However, significant differences were observed in all other pairwise comparisons.

Table 2.

Comparison of Clinical Treatment Results Among Groups

Statistical variables	Group A	Group B	P Value
VAS for back pain
Preoperative	4.16 ± 0.75	4.35 ± 0.81	0.274
Postoperative	2.47 ± 0.80	2.33 ± 0.64	0.375
3-month follow-up	0.02 ± 0.15	0.09 ± 0.29	0.171
Last postoperative follow-up	0.00 ± 0.00	0.02 ± 0.15	0.320
ODI
Preoperative	82.21 ± 6.53	82.93 ± 6.01	0.596
Postoperative	16.53 ± 4.56	17.67 ± 4.51	0.247
3-month follow-up	9.81 ± 3.14	9.21±3.86	0.428
Last postoperative follow-up	7.81 ± 3.22	7.00±3.60	0.272

VAS: Visual analog scale; ODI: Oswestry disability index.

Radiological Findings

Seventeen cases of bone cement leakage through the anterior vertebral wall occurred in Group A, compared to six cases in Group B. Additionally, one case of intradiscal bone cement leakage was observed in Group A, and one case of bone cement leakage into small veins occurred in Group B; however, none of these leakage events were associated with any serious clinical symptoms. During the follow-up period, three cases of new vertebral compression fractures occurred at adjacent segments in Group A, corresponding to a mid- to long-term complication rate of 6.98% (3/43). In Group B, four cases of adjacent segment fractures were observed, resulting in a complication rate of 9.30% (4/43). No other clinically significant mid- to long-term complications were reported in either group. These findings indicate a low incidence of delayed complications and reinforce the safety of the modified technique. The ICC analysis indicated high consistency between the two observers in the radiographic measurements. Specifically, the ICC value for the percentage of vertebral body height restoration in Group A was 0.927, the ICC value for the change in the wedge angle of the fractured vertebra in Group A was 0.926, and the ICC value for the change in local kyphosis angle in Group A was 0.897. The ICC value for the percentage of vertebral body height restoration in Group B was 0.907, the ICC value for the change in the wedge angle of the fractured vertebra in Group B was 0.917, and the ICC value for the change in local kyphosis angle in Group B was 0.878. Radiographic data analysis showed no statistically significant differences between the two groups in terms of preoperative bone mineral density, the location of vertebral damage (upper, middle, lower regions), the severity of vertebral compression, vertebral height restoration, changes in the wedge angle of the fractured vertebra, changes in the local kyphosis angle, and the incidence of adjacent vertebral fractures. Notably, a higher number of patients in Group B had cement dispersion to the opposite side compared to Group A, with a statistically significant difference (P < 0.05). Additionally, fewer patients in Group B experienced cement leakage through the anterior vertebral wall compared to Group A, and this difference was statistically significant (P < 0.05) (Table 3).

Table 3.

Comparison of Imaging Data Among Groups

	Group A	Group B	P Value
Preoperative bone mineral density	−3.07 ± 1.35	−2.99 ± 0.92	0.767
Location of vertebral damage			0.301
The upper third	14	21
The middle third	19	15
The lower third	10	7
Severity of vertebral compression			0.637
Mild	27	24
Moderate	11	15
Severe	5	4
Vertebral height restoration (%)	23.66 ± 27.41	40.46 ± 84.63	0.219
Wedge angle changes in the fractured vertebra	3.57 ± 3.63	4.06 ± 3.84	0.543
Local kyphosis angle changes	2.71 ± 3.52	3.63 ± 2.42	0.164
Adjacent vertebral fracture incidence (%)	6.98	9.30	0.693
Bone cement dispersion to the contralateral side	33	41	0.013 a
Bone cement leakage through the vertebral anterior wall	17	6	0.007 a

^aP value <0.05.

Discussion

OVCF is severe consequence of osteoporosis, primarily caused by primary osteoporosis, with an increasing prevalence annually. PKP is currently the preferred treatment for OVCF; however, controversy remains regarding the clinical efficacy and selection between unilateral and bilateral surgical approaches.^20,21 A systematic review and meta-analysis evaluating the advantages of unilateral PKP for treating osteoporotic vertebral compression fractures demonstrated that unilateral PKP was associated with significantly shorter operative time, lower bone cement volume, and reduced radiation exposure compared to bilateral PKP. No significant differences were observed between the two approaches in terms of bone cement leakage, kyphotic angle correction, pain relief (VAS score), or functional improvement (ODI score). Therefore, while both unilateral and bilateral PKP yield comparable clinical outcomes, unilateral PKP offers advantages in terms of reduced operative time, bone cement usage, and radiation exposure, ultimately lowering surgical risks and complications. ²² Optimizing unilateral PKP may provide greater clinical benefits, making it a potentially superior option for OVCF treatment.

PKP for osteoporotic thoracolumbar fractures is associated with a high risk of bone cement leakage, the most common postoperative complication with an incidence of 14%–18%, particularly in patients with significant preoperative anterior vertebral wall defects. ²³ Intraoperative leakage of bone cement through the anterior vertebral wall not only increases the incidence of cement-related complications but may also necessitate premature termination of cement injection, potentially leading to suboptimal clinical outcomes due to insufficient cement volume. ²⁴ Therefore, strategies to minimize bone cement leakage while ensuring therapeutic efficacy warrant further investigation. Gelatin sponge, widely used for surgical hemostasis, possesses favorable permeability, biocompatibility, and absorbability. Previous clinical studies have explored the use of bone cement–gelatin sponge composites in PVP to reduce cement leakage. Findings suggest that, compared to conventional PVP, this approach effectively decreases the rate of cement leakage in patients with Kümmell’s disease at stages I and II. ²⁵ Consequently, in the context of unilateral PKP for OVCF with anterior vertebral wall defects, the pre-soaked gelatin sponge strips may serve as an effective barrier, potentially reducing the risk of anterior cement leakage.

The findings of this study indicate that, compared to conventional unilateral PKP performed in Group A, the use of pre-soaked bone cement gelatin sponge strips in Group B significantly reduced bone cement leakage at the anterior vertebral wall defect. Additionally, the cement distribution in Group B was superior to that in Group A, and the cement injection volume in Group B was greater than in Group A. Although the difference in cement leakage rates was statistically significant (39.5% vs 14.0%), the relatively small sample size may still limit the robustness of this finding. A post hoc power analysis based on the observed leakage rates and sample size (n = 43 per group) indicated a statistical power of approximately 74%, suggesting that larger-scale studies are needed to further validate this result. In contrast, no statistically significant difference was observed in the incidence of adjacent vertebral fractures between the two groups (7.0% vs 9.3%). Given the low event rates and limited sample size, a post hoc power analysis revealed a statistical power of only approximately 9%, indicating that the study was markedly underpowered to detect any meaningful difference in this secondary outcome. Therefore, these findings should be interpreted with caution. In this study, pre-soaked bone cement gelatin sponge strips were used, allowing the originally radiolucent gelatin sponge to become radiopaque under X-ray fluoroscopy. This, in turn, enabled more effective sealing of the bone defect in the anterior vertebral wall. So, the possible explanation for these results is that the pre-soaked gelatin sponge strips effectively sealed the anterior vertebral defect, thereby reducing cement leakage and allowing the surgeon to inject a greater volume of bone cement into the injured vertebra. This prevented insufficient cement injection and led to increased intravertebral pressure, which in turn enhanced cement dispersion. It is important to note that, to date, no definitive evidence has established a positive correlation between the degree of postoperative pain relief and the volume of injected bone cement in PKP. ²⁶ Meanwhile, related studies have emphasized that ensuring adequate and symmetric distribution of bone cement within the fractured vertebra is crucial for improving surgical outcomes and reducing the risk of recurrent fractures.^27-29

Furthermore, in this study, a comparison of preoperative and postoperative VAS and ODI scores within each group revealed significant differences, indicating that both Group A and Group B effectively alleviated pain and achieved satisfactory clinical outcomes. However, despite these findings, the precise mechanism through which PKP relieves pain remains unclear. According to prevailing research perspectives, pain relief may be attributed to enhanced vertebral stability following bone cement injection. Additionally, local necrosis of neural tissue induced by the cement may also contribute to the reduction in pain.^30,31

Certain technical considerations should be emphasized during the procedure. When performing pedicle puncture, the lateral angulation should be maximized to ensure that the expansion balloon approaches or even crosses the midline. A thorough preoperative review of imaging data is essential, and multiple fluoroscopic views in both anteroposterior and lateral projections should be performed intraoperatively. In cases where there are obvious cortical defects, an appropriate amount of gelatin sponge should be applied to minimize the risk of cement leakage. During the placement of gelatin sponge, gentle manipulation is crucial to ensure smooth passage through the working channel. Additionally, bone cement injection should ideally be performed during the “threading phase,” when its viscosity is relatively low, allowing for better flowability and improved dispersion within the vertebra.

Limitations

This study has several limitations. Firstly, it is a retrospective dual-center study with a relatively small sample size, which may potentially impact the generalizability of the results. Future research should include large-scale, prospective randomized controlled trials to provide more robust evidence. Secondly, the follow-up period of 10 months to 1 year may limit the comprehensive assessment of long-term treatment outcomes. Therefore, further studies with extended follow-up periods are needed to provide a more thorough understanding of long-term outcomes. Thirdly, anterior cortical defects were defined based on qualitative assessment of preoperative CT images, relying on the visual judgment of experienced spine surgeons. The lack of precise quantitative criteria may introduce a degree of subjectivity and potential inter-observer variability. Future studies should incorporate objective imaging-based measurements to improve the consistency and reproducibility of inclusion criteria. Fourthly, biomechanical differences between vertebral levels (e.g., T10 vs L3) were not accounted for in the analysis. These segmental differences could potentially influence load distribution and cement behavior, thereby introducing confounding bias. Future studies should consider stratified analysis or statistical adjustment based on vertebral levels to improve the robustness of the findings. Fifthly, this study used an alternating assignment method instead of strict randomization for patient allocation. While this approach balanced group sizes, it may introduce selection bias due to its predictability. Future randomized controlled trials are needed to confirm our findings. Finally, there was a significant difference in cement volume between the two groups, which may be related to technical factors or patient-specific considerations. This difference was not adjusted for in the current analysis and may introduce selection or procedural bias. Future studies should consider standardizing cement volume or incorporating it into multivariable models to better control for its potential confounding effect.

Conclusion

For patients with OVCF presenting with significant anterior vertebral wall defects preoperatively, unilateral PKP combined with pre-soaked bone cement gelatin sponge strips can significantly reduce the incidence of anterior wall cement leakage and enhance cement dispersion. While this technique achieves comparable pain relief and functional outcomes to conventional PKP, its advantages in cement leakage control and distribution suggest it may be a safer and more effective option for selected patients, warranting further clinical investigation.

ORCID iD

Chong Sun https://orcid.org/0000-0001-8068-5107

Ethical Consideration

This clinical retrospective study of two centers was in accordance with the declaration of Helsinki. The work was compliant with the STROCSS criteria. All the procedures of this study were approved by the ethics committee of the Affiliated Hospital of Qingdao University (QYFY WZLL 29866) and Zhucheng People’s Hospital (2025-EC-35). Written informed consent was waived.

Data Availability Statement

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