Preoperative CT-Based Planning for Unilateral Pedicle Approach PVP Reduces Cement Leakage and Improves Radiological Outcomes in Osteoporotic Vertebral Compression Fractures: A Retrospective Cohort Study

Abstract

Background

Osteoporotic vertebral compression fractures are commonly treated with percutaneous vertebroplasty (PVP), though the conventional unilateral approach faces challenges in cement distribution and safety. This study aimed to evaluate whether a preoperative CT-based unilateral pedicle approach (CT-UPA) improves the outcomes of PVP.

Methods

This was a retrospective cohort analysis of patients undergoing PVP for single-segment thoracolumbar fractures between January 2020 and June 2024. Patients were divided into a CT-UPA group and a conventional unilateral approach (Conv-UPA) group. Perioperative indicators were recorded. Imaging assessments and clinical evaluations were conducted preoperatively, at 3 days, and 1 year postoperatively, with a final follow‑up at 1 year. Bone cement leakage was assessed via postoperative CT.

Results

A total of 222 patients were analyzed (CT-UPA: n=105; Conv-UPA: n=117). The CT-UPA group had a shorter mean surgery duration (56.43 vs 58.85 minutes, P=0.004), a higher mean injected cement volume (5.26 vs 4.09 mL, P<0.001), and a higher single-puncture success rate (96.19% vs 83.76%, P=0.002). Postoperatively, the CT-UPA group demonstrated significantly better Cobb angle correction, greater anterior vertebral height restoration, and superior cement distribution patterns. Visual Analog Scale and Oswestry Disability Index scores were significantly lower in the CT-UPA group at 3 days and 1 year. The overall cement leakage rate was significantly lower in the CT-UPA group (17.14% vs 32.48%, P=0.009), primarily due to a reduction in interspace leakage (4.76% vs 19.66%, P<0.001).

Conclusion

Preoperative CT-based planning for unilateral PVP enhances surgical precision, leading to superior radiographic outcomes, improved pain relief and function, and reduced cement leakage compared to the conventional fluoroscopy-guided method.

Introduction

Osteoporotic vertebral compression fractures (OVCF) are a common and debilitating consequence of systemic osteoporosis, primarily affecting the elderly population.¹ OVCF is defined as a reduction in vertebral body height due to decreased bone density, typically presenting with acute back pain, restricted spinal mobility, and in severe cases, spinal deformities such as kyphosis, which significantly impair patients’ quality of life.^2,3 Percutaneous vertebroplasty (PVP) has become a mainstream minimally invasive treatment aimed at stabilizing fractured vertebrae and relieving pain by injecting bone cement.⁴ Landmark randomized controlled trials, such as the recent VERTOS V trial, have established the efficacy and safety of vertebral augmentation procedures compared with active control or conservative management in patients with osteoporotic vertebral compression fractures.⁵ However, the traditional unilateral pedicle approach (UPA) PVP guided by intraoperative fluoroscopy faces challenges, including suboptimal bone cement distribution, risk of bone cement leakage, and inadequate restoration of vertebral height and sagittal alignment.^6,7 These limitations may compromise long-term biomechanical stability and clinical outcomes. Therefore, this study aims to investigate whether preoperative CT-based UPA (CT-UPA) design can improve the efficacy and safety of PVP in treating OVCF.

The pathophysiology of OVCF is rooted in osteoporosis-driven bone loss and microstructural deterioration, which reduce vertebral strength and load-bearing capacity, making the spine susceptible to fracture even under minimal trauma.^8,9 Recent advances in spinal surgery emphasize the importance of precise and personalized treatment plans. Several image-guided techniques for PVP, such as intraoperative navigation, robotic assistance, and 3D‑printed guides, have been reported to improve puncture accuracy.^10,11 However, these approaches require costly equipment and additional preparation, limiting their widespread use. In contrast, our CT‑UPA method uses only routine preoperative CT to calculate two simple parameters, including skin entry distance and needle inclination, targeting the needle tip across the midline for symmetric cement distribution. This low‑cost, reproducible strategy, which requires no extra hardware, has not been systematically compared with conventional fluoroscopy‑guided PVP, and its clinical value warrants investigation.

This study aims to evaluate the clinical value of preoperative CT-designed unilateral pedicle approach for PVP. The main innovation of this study lies in the implementation of a personalized, geometrically precise preoperative puncture plan based on CT imaging, aimed at optimizing the needle trajectory. This study seeks to provide evidence for more precise surgical techniques and contribute to the broader goal of optimizing minimally invasive treatments for osteoporosis-related spinal fractures.

Materials and Methods

Study Design and Ethics Statement

This study is a single-center retrospective cohort analysis. A retrospective analysis was conducted on the clinical data of 222 patients who underwent percutaneous vertebroplasty (PVP) for single-segment osteoporotic vertebral compression fractures (OVCF) at our hospital from January 2020 to June 2024. According to the puncture method received by the patients, they were divided into two groups: the Preoperative CT-based Unilateral Pedicle Approach Group (CT-UPA Group) (n=105) and the Conventional Unilateral Pedicle Approach Group (Conv-UPA Group) (n=117). The study protocol has been approved by the Institutional Ethics Review Board of the First Affiliated Hospital of Xinxiang Medical University (Approval No. EC-LW26-104), and was conducted in accordance to the tenets of the Declaration of Helsinki. Given the retrospective nature of the study and the fact that all data were analyzed after anonymization, the ethics committee approved the waiver of patient informed consent.

Inclusion and Exclusion Criteria

Inclusion criteria were back pain as the primary clinical symptom, without radicular radiation or spinal cord compression symptoms. Lumbar bone mineral density T-score ≤ −2.5 as measured by dual-energy X-ray absorptiometry (DXA), meeting the diagnostic criteria for osteoporosis.¹² Preoperative MRI fat-suppression sequences or bone scans showed high signal changes in the responsible vertebral body, confirming acute or subacute fractures. The fractured vertebrae were located in the thoracolumbar region (T10-L4). Exclusion criteria included pathological fractures caused by tumors, infections, etc. Patients with severe systemic diseases of the heart, lungs, etc, who cannot tolerate surgery. Patients with incomplete clinical or imaging data. Fractured vertebrae located above T10 or below L4. And patients with two or more segments of OVCF.

Surgical Methods for the Two Groups

The choice between CT‑UPA and Conv‑UPA was made through a shared decision‑making process between the surgeon and the patient. Before surgery, the attending surgeon explained the advantages and potential limitations of both techniques to each patient. Patients were then given the opportunity to ask questions and express their preferences. The final decision was made collaboratively, taking into account the patient’s informed choice. All procedures were performed by a team of three senior spine surgeons, each with over ten years of experience in spinal surgery and more than 100 PVP procedures performed prior to the study. Before introducing the CT‑UPA technique, all participating surgeons completed a standardized training program consisting of preoperative CT planning exercises, with at least 10 mock punctures per surgeon to familiarize themselves with the measurement method and trajectory planning. This training ensured that all surgeons were proficient in the CT‑UPA technique before performing it on patients.

Patients were placed in a prone position and received percutaneous vertebroplasty (PVP) treatment under local anesthesia in the operating room. Intraoperative monitoring of vital signs, including pulse, non-invasive blood pressure, electrocardiogram, and oxygen saturation, was continuously performed. The surgical procedure utilized a vertebroplasty instrument kit provided by Shandong Guanlong Medical Supplies Co., Ltd., and polymethylmethacrylate (PMMA) bone cement provided by Heraeus Medical GmbH.

The procedure in the Conv-UPA group was performed under C-arm fluoroscopy guidance. The target vertebral body was identified on the anteroposterior (AP) fluoroscopic image, and the upper lateral quadrant of the target pedicle was marked as the skin entry point. A scalpel was used to make a skin incision of about 4 mm, followed by blunt dissection of the deep fascia. A 4.2 mm puncture needle was gradually advanced through the pedicle into the vertebral body. The puncture trajectory was confirmed alternately using AP and lateral fluoroscopy: in the AP view, the needle tip should always remain within the medial cortical margin of the pedicle; in the lateral view, the needle tip should be advanced along the posterior margin of the vertebral body. The needle tip was ultimately positioned at the anterior one-third of the vertebral body in the lateral view. Subsequently, under real-time lateral fluoroscopic monitoring, the well-mixed PMMA bone cement was slowly injected to allow early detection and termination of leakage into the epidural space, intervertebral space, or venous plexus. Injection continued until satisfactory filling of the bone cement or until it approached the posterior wall of the vertebral body.

The key aspect of the CT-UPA group’s surgery was the execution of individualized puncture planning based on preoperative CT imaging. First, on the preoperative axial CT images of the responsible vertebral body, the distance from the midline to the skin entry point and the required inclination angle of the puncture needle were precisely calculated (Figure 1). During the procedure, after identifying the target pedicle under AP fluoroscopy, the entry point was accurately marked on the skin according to the preoperatively calculated distance from the midline. Anesthesia and incision steps were identical to those in the Conv-UPA group. During puncture, the needle tip was initially placed approximately 2 mm deep into the bone surface, and AP fluoroscopy confirmed that the needle tip was at the lateral cortical margin of the pedicle. Subsequently, following the preoperatively planned inclination angle, the puncture needle was slowly rotated and advanced, moving the needle tip medially in the AP view until it reached the medial cortical margin of the pedicle, and laterally to the posterior margin of the vertebral body. The needle was further advanced to the anterior one-third of the vertebral body in the lateral view, with the needle tip located or crossing the midline in the AP view (Figure 2). The steps for bone cement injection and monitoring standards were the same as those in the Conv-UPA group.

Figure 1.

Preoperative Puncture Path Planning. Point a is the intersection at the anterior one-third of the vertebral body, point d is the outer margin of the facet joint, point b is the intersection of the line connecting points a and d with the skin, and the line connecting points b and c represents the distance from the midline to the skin entry point.

Figure 2.

(a) Preoperative MRI localization of the responsible segment; (b and c) Localization of the preoperative CT puncture axial plane; (d) Design of the distance from the midline and the puncture angle; (e and f) Intraoperative skin localization and puncture angle; (g-i) Localization of the outer margin of the pedicle, gradually advancing the needle to the inner margin of the pedicle, with the needle tip positioned at the posterior margin of the vertebral body in the lateral view; (j and k) Continuing to advance the needle to the anterior one-third of the vertebral body in the lateral view, and to the midline of the vertebral body in the AP view; (l and m) Postoperative distribution of injected bone cement.

To assess the reproducibility of the preoperative CT measurement protocol, two independent observers, a spine surgeon and a radiologist, each measured the skin entry distance and needle inclination angle on a randomly selected subset of 30 preoperative CT scans. Inter‑observer reliability was evaluated using the intraclass correlation coefficient (ICC, two‑way mixed model, absolute agreement). The ICC values for both parameters exceeded 0.90, indicating excellent reproducibility.

Both groups of patients received routine wound care, pain relief, and systemic anti-osteoporosis drug treatment postoperatively. Patients were encouraged to start ambulating with lumbar support on the second day after surgery.

Data Collection

Baseline Data

All data in this retrospective study were collected through the review of our hospital’s electronic medical record system. Demographic information, smoking history, current medical history and comorbidities, trauma causes, and history of anti-osteoporosis treatment were directly extracted from patients’ admission records, progress notes, and discharge summaries.

Imaging assessment data were obtained by systematically reviewing patients’ archived imaging materials and related reports. The confirmation of the responsible fracture segment was based on a comprehensive judgment of preoperative spinal AP and lateral X-rays, computed tomography (CT), and magnetic resonance imaging (MRI) reports. Bone mineral density T-scores were sourced from preoperative DXA reports. The severity of vertebral fractures was assessed using the Genant semi-quantitative classification method, which categorizes fractures as mild (20–25% loss), moderate (25–40% loss), and severe (>40% loss) based on the percentage height loss of the anterior, middle, and posterior margins of the vertebral body on lateral X-ray images.¹³ To ensure consistency and objectivity of the classification, all imaging materials were independently evaluated by a senior radiologist or spine surgeon who was blinded to the group allocation.

Perioperative Indicators and Surgical Procedure Information

Data on perioperative and postoperative rehabilitation indicators were obtained from anesthesia records, nursing documentation, and departmental registration materials. The surgery duration was defined as the time from skin incision to wound dressing completion. The volume of injected bone cement was recorded in the surgical notes. The number of intraoperative fluoroscopy (X-ray) exposures was counted by reviewing radiology operation logs or operating room radiation exposure records. The single puncture success rate was determined based on descriptions of the puncture process in the surgical notes, with the criteria being that the puncture needle reached the anterior-middle one-third target area of the vertebral body on the first attempt according to the planned trajectory, and no mention of needle path adjustment or re-puncture during the procedure. The time to first ambulation after surgery was determined by the time from surgery to the first assisted walking, as documented by the responsible nurse in the rehabilitation records. The postoperative hospital stay was calculated based on the discharge date and the surgery date.

Postoperative Imaging Assessment

Cobb angles and anterior vertebral body heights were measured on standard lateral spine X-rays taken preoperatively, 3 days postoperatively, and 1 year postoperatively. All X-rays were taken using a digital radiography system (DigitalDiagnost C90, Philips, the Netherlands). Measurements were independently performed by a senior spine surgeon using the measurement tools in the hospital’s picture archiving and communication system (PACS).

Assessment of bone cement distribution was based on spinal CT scans performed within 3 days after surgery. The CT scans were completed using a 64-slice spiral CT scanner (SOMATOM Perspective, Siemens Healthineers, Germany). The Kim classification was divided according to the relationship between the bone cement and the upper and lower endplates on the sagittal view: Type I (cement-endplate non-contact type) and Type II (cement-endplate contact type).¹⁴ The Lin classification assessed the three-dimensional diffusion pattern of bone cement within the vertebral body: Type I (poor distribution) refers to bone cement mainly aggregating on the puncture side without crossing the midline of the vertebral body; Type II (moderate distribution) indicates that the bone cement crosses the midline but the contralateral filling is still insufficient or uneven; Type III (good distribution) indicates that the bone cement diffuses uniformly and symmetrically on both sides of the vertebral body with adequate filling.¹⁵

Representative postoperative CT images of cement distribution patterns for both groups are shown in Figure 3.

Figure 3.

Post‑operative CT images of four representative patients. (A) Axial CT of Conv‑UPA group, Lin Type I: bone cement does not cross the vertebral midline. (B) Axial CT of Conv‑UPA group, Lin Type II: cement crosses midline with insufficient contralateral dispersion. (C) Axial CT of CT‑UPA group, Lin Type III: bilateral symmetric cement dispersion. (D) CT‑UPA group, Lin Type III combined with Kim Type II. (D1) Axial view shows bilateral symmetric cement distribution; (D2) Sagittal reconstruction demonstrates cement contacts both upper and lower vertebral endplates.

Postoperative Clinical Efficacy Assessment

Pain intensity was quantitatively assessed using the Visual Analog Scale (VAS).¹⁶ This scale consists of a 10-cm long line, with one end marked as “no pain” (0 points) and the other end marked as “most severe pain” (10 points). Patients mark the point on the line corresponding to their pain level, and the distance from the “no pain” end to the marked point is measured as the score. Higher scores indicate more severe pain. This scoring was completed preoperatively, 3 days postoperatively, and 1 year postoperatively.

Functional status was assessed using the Oswestry Disability Index (ODI).¹⁷ This questionnaire includes 10 items closely related to daily life: pain intensity, personal care, lifting, walking, sitting, standing, sleeping, social life, travel, and occupation/housework. Each item has six statements corresponding to scores of 0 to 5, with the total score converted through a formula to a range of 0–100 points. Higher scores indicate more severe functional impairment due to back pain. Scores are categorized as follows: 0–20 for mild disability, 21–40 for moderate disability, 41–60 for severe disability, 61–80 for disability, and 81–100 for bedridden or symptom exaggeration. The ODI questionnaire was also independently completed by patients preoperatively, 3 days postoperatively, and 1 year postoperatively.

Surgical Safety Assessment

Bone cement leakage was identified and recorded through early postoperative imaging examinations. All patients underwent spinal CT scans within 3 days after surgery to precisely assess the occurrence, location, and classification of bone cement leakage. The determination and classification of leakage were independently performed by a radiologist who was blinded to the group allocation, using axial, sagittal, and coronal reconstructed CT images. Leakage was classified into paravertebral leakage (cement leakage into the paravertebral soft tissues or venous plexus), interspace leakage (cement leakage into the adjacent intervertebral disc), and intraspinal leakage (cement leakage into the epidural space of the spinal canal). The final leakage determination results were adjudicated by one evaluator.

Statistical Analysis Methods

All statistical analyses were conducted using R software (version 4.3.1, R Foundation for Statistical Computing, Austria). Continuous variable data were tested for normality using the Shapiro–Wilk test. Normally distributed data were expressed as mean ± standard deviation (Mean ± SD), and between-group comparisons were performed using independent samples t-tests. Non-normally distributed data were expressed as median (interquartile range) [M (IQR)], and between-group comparisons were performed using the Mann–Whitney U-test. Categorical variables were expressed as frequency (percentage) [n (%)], and between-group comparisons were performed using chi-square tests or Fisher’s exact tests (when expected frequencies < 5). All statistical tests were two-sided, with a P-value < 0.05 indicating statistically significant differences.

Post-hoc power analyses were conducted for the primary and secondary outcomes using the observed effect sizes with a two-sided alpha of 0.05. Given the sample sizes of 105 in the fluoroscopy-guided group and 117 in the CT-UPA group, the achieved power exceeded 80% for the primary outcome measure (bone cement distribution grade). Based on the observed difference in cement distribution grade between the two groups, the post-hoc power for the primary endpoint was calculated as 89.4%. For secondary outcomes including vertebral height restoration and VAS pain score improvement, the achieved powers were 86.2% and 72.5%, respectively. These calculations confirmed that the study had adequate statistical power to detect clinically meaningful differences between the two groups for the primary outcome, despite the unequal group sizes.

All reported outcomes were prespecified, with bone cement distribution grade (Lin classification) and overall cement leakage rate designated as primary endpoints. Other outcomes, including VAS, ODI, Cobb angle, vertebral height restoration, and leakage subtypes, were considered exploratory. All P values are reported as nominal values, and secondary findings should be interpreted accordingly.

For repeatedly measured indicators, two-way repeated measures analysis of variance (RM-ANOVA) was performed to evaluate the effects of group (CT-UPA vs Conv-UPA), time (preoperative, 3 days postoperative, and 1 year postoperative), and the group-by-time interaction. Mauchly’s test of sphericity was used to assess the sphericity assumption; when the assumption was violated (Mauchly’s test P < 0.05), the Greenhouse-Geisser correction was applied to adjust the degrees of freedom. Partial eta squared (η²p) was calculated as a measure of effect size for each significant effect. For significant main effects or interactions, post-hoc pairwise comparisons were performed using Bonferroni correction to adjust for multiple comparisons. For between-group comparisons at each individual time point, independent samples t-tests were used as supplementary analyses, with Bonferroni correction applied to control the family-wise error rate across the three time points. All statistical tests were two-sided, with a P-value < 0.05 indicating statistically significant differences.

Results

Patient Baseline Data

Table 1 shows the baseline characteristics of patients in the Conv-UPA group and the CT-UPA group. There were no significant differences observed between the two groups in terms of age, gender, BMI, smoking history, surgical segment distribution, bone mineral density T-score, hypertension, diabetes, low-energy falls, fracture Genant classification, and history of anti-osteoporosis treatment (all P values were greater than 0.05).

Table 1.

Comparison of Baseline Characteristics of Patients

Index	Conv-UPA Group (n=117)	CT-UPA Group (n=105)	t/χ2	P
Age	57.51 ± 7.69	58.36 ± 7.72	0.814	0.416
Gender (male/female)	30 (25.64%)/87 (74.36%)	18 (17.14%)/87 (82.86%)	2.358	0.125
BMI	23.16 ± 2.14	23.43 ± 2.39	0.895	0.372
Smoking history	21 (17.95%)	17 (16.19%)	0.121	0.728
Surgical segments			10.049	0.123
-T10	3 (2.56%)	6 (5.71%)
-T11	21 (17.95%)	13 (12.38%)
-T12	36 (30.77%)	18 (17.14%)
-L1	24 (20.51%)	27 (25.71%)
-L2	18 (15.38%)	23 (21.90%)
-L3	3 (2.56%)	6 (5.71%)
-L4	12 (10.26%)	12 (11.43%)
Bone mineral density T-score	3.34 ± 0.51	3.28 ± 0.49	0.860	0.391
Hypertension	58 (49.57%)	52 (49.52%)	0.000	0.994
Diabetes mellitus	26 (22.22%)	20 (19.05%)	0.339	0.560
Low-energy fall	110 (94.02%)	100 (95.24%)	0.161	0.688
Fracture Genant classification (n, %)			0.602	0.740
-mild	15 (12.82%)	12 (11.43%)
-moderate	78 (66.67%)	75 (71.43%)
-severe	24 (20.51%)	18 (17.14%)
History of anti-osteoporosis treatment	82 (70.09%)	85 (80.95%)	3.506	0.061

Perioperative Indicators and Surgical Procedure Information

There were significant differences in perioperative indicators between the two groups, with the CT-UPA group having a shorter surgery time (P=0.004), a higher volume of injected bone cement (P<0.001), and a higher single puncture success rate (P=0.002; Table 2). Although there was a slight difference in the number of X-ray exposures in the CT-UPA group, it did not reach statistical significance (P=0.081). These results suggest that the CT-UPA method may have potential advantages in improving surgical efficiency.

Table 2.

Perioperative Indicators

Index	Conv-UPA Group (n=117)	CT-UPA Group (n=105)	95% CI	t/χ2	P
Surgery duration (minutes)	58.85 ± 6.08	56.43 ± 6.26	(−4.06, −0.78)	2.916	0.004
Bone cement injection volume (mL)	4.09 ± 0.93	5.26 ± 0.94	(0.92, 1.42)	9.336	< 0.001
Number of X-ray exposures	19.23 ± 2.17	18.73 ± 2.08	(−1.06, 0.06)	1.756	0.081
Single puncture success rate	98 (83.76%)/19 (16.24%)	101 (96.19%)/4 (3.81%)	(4.76%, 20.10%)	9.206	0.002

In terms of early postoperative rehabilitation and hospital stay, the time to first ambulation was significantly higher in the Conv-UPA group compared to the CT-UPA group (P=0.011), indicating that patients in the CT-UPA group were able to start moving more quickly (Table 3). Regarding postoperative hospital stay, the Conv-UPA group was also significantly higher than the CT-UPA group (P=0.016), suggesting that patients in the CT-UPA group were able to be discharged earlier. These results suggest that the CT-UPA method may help accelerate the postoperative recovery process, thereby improving the treatment experience for patients.

Table 3.

Comparison of Early Postoperative Rehabilitation and Hospital Stay Between the Two Groups

Index	Conv-UPA Group (n=117)	CT-UPA Group (n=105)	95% CI	t	P
Time to first ambulation (hours)	32.12 ± 8.11	29.65 ± 6.28	(−4.37, −0.57)	2.555	0.011
Postoperative hospital stay (days)	5.10 ± 1.34	4.69 ± 1.15	(−0.74, −0.08)	2.436	0.016

Postoperative Imaging Assessment

The imaging indicators of the two groups before and after surgery showed that there was no significant difference in preoperative Cobb angles between the two groups (P=0.532; Table 4). However, at 3 days postoperatively (P=0.003) and 1 year postoperatively (P=0.007), the Cobb angle in the CT-UPA group was significantly lower than that in the Conv-UPA group, indicating that the CT-UPA group had better results in correcting spinal deformities. There was also no significant difference in anterior vertebral body height between the two groups preoperatively (P=0.891). At 3 days postoperatively (P=0.030) and 1 year postoperatively (P=0.004), the anterior vertebral body height in the CT-UPA group was significantly higher than that in the Conv-UPA group, showing that the CT-UPA group performed better in restoring vertebral body height.

Table 4.

Comparison of Imaging Indicators Before and After Surgery Between the Two Groups

Index	Conv-UPA Group (n=117)	CT-UPA Group (n=105)	95% CI	t	P
Cobb angle
Preoperative	11.86 ± 1.68	11.71 ± 1.75	(−0.62, 0.32)	0.625	0.532
Postoperative 3 days	10.43 ± 1.22	10.03 ± 0.74	(−0.66, −0.14)	3.007	0.003
Postoperative 1 year	9.69 ± 1.19	9.24 ± 1.24	(−0.77, −0.13)	2.744	0.007
Anterior vertebral body height
Preoperative	20.71 ± 1.71	20.74 ± 1.45	(−0.39, 0.45)	0.138	0.891
Postoperative 3 days	22.06 ± 1.08	22.36 ± 1.01	(0.03, 0.57)	2.180	0.030
Postoperative 1 year	22.63 ± 0.84	22.94 ± 0.70	(0.10, 0.52)	2.919	0.004

The distribution of bone cement 3 days postoperatively according to the Kim classification showed that in the non-contact type between bone cement and endplate, the Conv-UPA group was significantly higher than the CT-UPA group (risk difference: −14.58%; 95% CI: −24.77% to −4.39%; P=0.011), indicating that the CT-UPA group performed better in the distribution of the bone cement-endplate contact type, with more patients showing bone cement-endplate contact (Figure 4). These results suggest that the CT-UPA method may help improve the distribution of bone cement.

Figure 4.

Comparison of Bone Cement Distribution in the Two Groups at 3 Days Postoperatively (Kim Classification).

The distribution of bone cement 3 days postoperatively according to the Lin classification showed that the CT-UPA group had significantly fewer cases of poor bone cement distribution (Type I) compared to the Conv-UPA group (P<0.001), while the CT-UPA group had a significantly higher number of cases with good bone cement distribution (Type III) (P<0.001; Table 5). This indicates that the bone cement distribution in the CT-UPA group was more diffuse and complete (P<0.001), whereas in the Conv-UPA group, it was more concentrated and insufficient (P<0.001). There was no significant difference between the two groups in the moderate distribution Type II category (P=0.687).

Table 5.

Comparison of Bone Cement Distribution in the Two Groups at 3 Days Postoperatively (Lin Classification)

Index	Conv-UPA Group (n=117)	CT-UPA Group (n=105)	95% CI	χ2	P
Type I	45 (38.46%)	9 (8.57%)	(−39.96%, −19.82%)	26.858	< 0.001
Type II	60 (51.28%)	51 (48.57%)	(−16.27%, 10.85%)	0.163	0.687
Type III	12 (10.26%)	45 (42.86%)	(21.99%, 43.21%)	30.819	< 0.001

Postoperative Clinical Efficacy Assessment

There was no significant difference in preoperative VAS pain scores between the two groups (95% CI: −0.37 to 0.05; P=0.128; Figure 5). At 3 days postoperatively (95% CI: −0.40 to −0.04; P=0.015) and 1 year postoperatively (95% CI: −0.09 to −0.01; P=0.018), the VAS pain scores in the CT-UPA group were significantly lower than those in the Conv-UPA group, indicating that the CT-UPA group had better results in reducing patients’ short-term and long-term postoperative pain. This suggests that the CT-UPA method can more quickly relieve postoperative pain and maintain good analgesic effects over the long term, which is important for improving patients’ quality of life and postoperative recovery.

Figure 5.

Comparison of VAS Pain Scores Before and After Surgery Between the Two Groups.

Table 6 presents the repeated measures ANOVA results for radiological and clinical outcomes. For Cobb angle, there was a significant main effect of time (F = 148.27, P < 0.001) and a significant main effect of group (F = 6.84, P = 0.010), but the group-by-time interaction was not significant (F = 0.86, P = 0.402). For anterior vertebral body height, there was a significant main effect of time (F = 267.52, P < 0.001) and a significant main effect of group (F = 4.92, P = 0.028), with no significant group-by-time interaction (F = 0.74, P = 0.460). For VAS scores, there were significant main effects of time (F = 452.38, P < 0.001) and group (F = 6.21, P = 0.014), as well as a significant group-by-time interaction (F = 4.82, P = 0.015). For ODI scores, there were significant main effects of time (F = 1687.52, P < 0.001) and group (F = 5.63, P = 0.019), along with a significant group-by-time interaction (F = 4.13, P = 0.023). Post-hoc pairwise comparisons with Bonferroni correction demonstrated that the CT-UPA group had significantly better outcomes at both 3 days and 1 year postoperatively for all four indicators (all P < 0.05), consistent with the findings presented in Table 5. The significant group-by-time interactions for VAS and ODI indicate that the treatment benefits in pain relief and functional recovery became more pronounced over time in the CT-UPA group.

Table 6.

Repeated Measures ANOVA Results for Radiological and Clinical Outcomes

Index	Effect	F	df	P
Cobb angle	Time	148.27	(1.52, 334.40)	< 0.001
	Group	6.84	(1, 220)	0.010
	Group × Time	0.86	(1.52, 334.40)	0.402
Anterior vertebral body height	Time	267.52	(1.68, 369.60)	< 0.001
	Group	4.92	(1, 220)	0.028
	Group × Time	0.74	(1.68, 369.60)	0.460
VAS	Time	452.38	(1.54, 338.80)	< 0.001
	Group	6.21	(1, 220)	0.014
	Group × Time	4.82	(1.54, 338.80)	0.015
ODI	Time	1687.52	(1.61, 354.20)	< 0.001
	Group	5.63	(1, 220)	0.019
	Group × Time	4.13	(1.61, 354.20)	0.023

There was no significant difference in preoperative ODI scores between the two groups (P=0.325; Table 7). At 3 days postoperatively (P=0.005) and 1 year postoperatively (P=0.029), the ODI scores in the CT-UPA group were significantly lower than those in the Conv-UPA group, indicating that the CT-UPA group had better results in improving patients’ short-term and long-term functional impairment. This is important for enhancing patients’ quality of life and postoperative recovery.

Table 7.

Comparison of Oswestry Disability Index Before and After Surgery Between the Two Groups

Index	Conv-UPA Group (n=117)	CT-UPA Group (n=105)	95% CI	t	P
Preoperative	79.39 ± 10.42	78.04 ± 9.81	(−4.04, 1.34)	0.987	0.325
Postoperative 3 days	37.26 ± 5.53	35.21 ± 5.28	(−3.48, −0.62)	2.816	0.005
Postoperative 1 year	18.47 ± 6.89	16.45 ± 6.84	(−3.83, −0.21)	2.191	0.029

Surgical Safety Assessment

In terms of overall leakage rates, the Conv-UPA group was significantly higher than the CT-UPA group (P=0.009; Table 8). There were no significant differences between the two groups in paravertebral leakage (P=0.779) and intraspinal leakage (P=0.692). However, in interspace leakage, the Conv-UPA group was significantly higher than the CT-UPA group (P<0.001). These results suggest that the CT-UPA method may have advantages in reducing bone cement leakage, particularly interspace leakage, which can help lower surgical risks and improve postoperative recovery for patients.

Table 8.

Comparison of Bone Cement Leakage Between the Two Groups

Index	Conv-UPA Group (n=117)	CT-UPA Group (n=105)	95% CI	χ2	P
Overall leakage	38 (32.48%)	18 (17.14%)	(−26.47%, −4.21%)	6.900	0.009
Paravertebral leakage	12 (10.26%)	12 (11.43%)	(−7.02%, 9.36%)	0.079	0.779
Interspace leakage	23 (19.66%)	5 (4.76%)	(−23.17%, −6.63%)	11.141	< 0.001
Intraspinal leakage	3 (2.56%)	1 (0.95%)	(−5.02%, 1.80%)	0.157	0.692

Discussion

This study provides compelling evidence that preoperative CT-UPA for PVP has clinical advantages over Conv-UPA in the treatment of OVCF. By integrating precise anatomical measurements from preoperative CT imaging into the puncture planning process, the CT-UPA technique not only improves the technical efficiency of the surgery but also achieves superior postoperative outcomes in spinal alignment, vertebral height restoration, pain relief, functional recovery, and surgical safety. These findings highlight the potential of personalized image-guided interventions to address the inherent limitations of traditional fluoroscopy-guided PVP, such as suboptimal cement distribution and higher leakage risk.

The observed reduction in surgical time and increase in single puncture success rate with the CT-UPA technique logically stem from preoperative rehearsal of the surgery. By predefining the distance of the skin entry point from the midline and the precise needle tilt angle, the intraoperative phase transitions from a dynamic, fluoroscopy-intensive search for the optimal path to confidently executing a pre-validated plan.¹⁸ This minimizes adjustments and repetitions often required in traditional methods when the initial trajectory is found to be suboptimal, directly saving time. This finding aligns with broader principles in minimally invasive spine surgery that preoperative navigation or planning can reduce variability in surgery.¹⁹ While some studies on navigated PVP have reported similar improvements in puncture accuracy and time, our approach relies on standard CT rather than complex intraoperative navigation systems, making its benefits potentially more accessible.²⁰ This distinction is important because previous image‑guided protocols often emphasized technology rather than simplicity, whereas our method prioritizes a quantifiable geometric rule that any surgeon can readily adopt.

The most significant improvements with the CT-UPA technique lie in its ability to better correct spinal deformities and more effectively restore vertebral body height. Traditional unilateral approaches often result in the needle tip being laterally positioned within the vertebral body. Starting from this position, bone cement tends to aggregate asymmetrically on the same side, leading to uneven vertebral reinforcement and limited potential for height restoration through indirect mechanisms.²¹ In contrast, the CT-UPA plan intentionally positions the needle endpoint at or across the midline of the vertebral body. This central or contralateral positioning creates a strategic reservoir where bone cement can spread more symmetrically under lower pressure, filling the anterior portion of the vertebral body and more reliably contacting both endplates, as evidenced by favorable Kim and Lin classification results.²² The biomechanical principle is robust: symmetric, endplate-to-endplate cement bridging forms a more stable internal cast that can better resist axial loads and bending moments, thereby maintaining the height restoration achieved during the cement injection. Previous biomechanical studies have linked symmetric cement distribution with improved vertebral stiffness and strength recovery.²³ Benek et al recently reported that a larger cement volume in unilateral kyphoplasty for thoracolumbar fractures significantly improved anterior vertebral height restoration.²⁴ Their findings support the importance of adequate cement volume and distribution, consistent with our observation that CT‑UPA enabled greater cement volume and superior height restoration compared with the conventional approach. Our clinical outcomes confirm this, showing that the enhanced distribution in the CT-UPA group is associated with better maintenance of anterior height and kyphosis correction over one year, indicating sustained biomechanical efficacy.

In addition to mechanical benefits, the CT-UPA technique has also shown clear advantages in postoperative pain management and functional outcomes. The effective pain relief from PVP is attributed to the stabilization of micro-movements at the fracture site. More symmetrical and complete bone cement filling, especially in contact with the endplates, may provide more comprehensive stabilization of the fractured trabeculae throughout the vertebral body, thereby achieving more direct and sustained pain relief.²⁵ Furthermore, better sagittal alignment (reduced Cobb angle) can reduce abnormal biomechanical stress on adjacent spinal segments and paravertebral muscles, which might otherwise lead to persistent pain and functional limitations.²⁶ The earlier ambulation and shorter hospital stays observed in the CT-UPA group further reflect this enhanced early stability and pain control, enabling faster functional recovery. This observation is supported by literature indicating that inadequate cement filling due to insufficient mechanical stability can lead to persistent pain.²⁷ In contrast, the Conv-UPA approach is limited by intraoperative fluoroscopic visualization, which may restrict the volume of cement that can be safely injected, resulting in suboptimal analgesic effects. The VAS and ODI differences between the two groups, though statistically significant, fell below established MCID thresholds. Nevertheless, these modest patient-reported differences should be interpreted alongside the clinically meaningful improvements in radiographic alignment, vertebral height restoration, cement distribution, and, most notably, a fourfold reduction in interspace leakage, which together support the overall benefit of the CT-UPA technique.

The safety of the CT-UPA method also represents a significant advancement over traditional techniques. The incidence of bone cement leakage, particularly interspace leakage into adjacent intervertebral discs, was reduced in the CT-UPA group. Leakage remains the most common complication of PVP, and disc leakage poses a potential risk to adjacent vertebral fractures.²⁸ The mechanisms by which CT-UPA mitigates this risk are multifactorial. Precise midline trajectories allow for slow, low-pressure injection of bone cement from a central position, as the cement can flow laterally to fill the vertebral body rather than being forcefully pushed toward the cortical walls or basal vertebral plexus from a peripheral starting point.²⁹ There is a higher possibility of achieving bilateral filling, which can reduce the temptation to over-inject cement on the puncture side in an attempt to cross the midline, a situation that often increases intradiscal pressure and leads to leakage.³⁰ The reduction in interspace leakage is particularly noteworthy, consistent with engineering principles of fluid dispersion. Lower leakage rates not only reduce the risk of potential complications such as adjacent vertebral fractures but also contribute to a smoother postoperative recovery.³¹ In a recent cohort study, Tabanli et al identified cement leakage as a risk factor for new vertebral fractures following vertebral augmentation and demonstrated that anti‑osteoporosis treatment significantly reduced this risk.³² This further highlights the clinical importance of minimizing cement leakage—a key advantage demonstrated by our CT‑UPA technique. These findings align with previous studies that have emphasized the importance of accurate needle placement in preventing cement leakage.³³ However, while some early studies reported no significant difference in leakage rates between unilateral and bilateral approaches, our study indicates that the integration of CT-based planning can further enhance safety beyond what is achievable with fluoroscopy alone.³⁴

An important consideration is the additional radiation exposure from the preoperative CT scan. While intraoperative fluoroscopy time showed a trend toward reduction with CT‑UPA, the preoperative CT does contribute to the overall radiation burden. However, the effective dose of a single spinal CT is relatively low and falls within the acceptable range for diagnostic imaging. More importantly, this modest additional exposure is justified by the tangible clinical benefits, including improved cement distribution, better vertebral height restoration, more effective pain relief, and reduced cement leakage. Given these meaningful advantages that directly impact patient recovery and safety, we consider the risk‑benefit profile of the CT‑UPA approach to be favorable.

However, this study has several limitations that must be acknowledged. The retrospective design inherently introduces selection bias and limits the ability to establish causal relationships. Most critically, because the CT‑UPA technique was progressively adopted during the study period, earlier patients predominantly received Conv‑UPA while later patients received CT‑UPA, introducing time‑dependent confounding from surgeon learning, protocol refinements, and patient selection drift; we did not perform propensity score matching or multivariable regression to adjust for these temporal covariates, which may have biased our estimates. Our conclusions should therefore be viewed as exploratory rather than definitive. As a single-center study, these findings may not be generalizable to other institutions with varying levels of surgical expertise and imaging capabilities. The non-random allocation of surgical techniques may reflect the evolution of institutional preferences over time, potentially confounding the results. While a one-year follow-up period is sufficient to evaluate early and intermediate outcomes, it is insufficient to assess long-term complications such as adjacent segment fractures, cement fragmentation, or late sequelae of asymptomatic cement leakage. The longevity of radiological and clinical benefits observed within one year needs to be validated in future long-term studies. The integration of 3D reconstruction software or computer-assisted surgical planning systems could further refine trajectory planning, especially for severely deformed vertebrae or challenging pedicle anatomy, and should be explored in future research. Additionally, although the participating surgeons were highly experienced and underwent standardized training before adopting the CT‑UPA technique, the potential influence of a learning curve cannot be completely excluded, particularly during the early implementation phase. A prospective study with formal learning curve assessment would be valuable to confirm the reproducibility of our findings across different institutional settings and surgeon experience levels. Finally, the study population was limited to single-segment thoracolumbar fractures, and the efficacy and safety of the CT-UPA protocol for multi-segment fractures or fractures in the high thoracic or lower lumbar regions warrant separate investigation.

Conclusion

This study demonstrates that preoperative CT-designed PVP unilateral pedicle approach improves the precision, efficacy, and safety of the surgery compared to traditional guidance methods. By promoting optimal needle placement, it facilitates a more symmetrical and stable bone cement structure, thereby better restoring vertebral height and correcting kyphotic deformity, achieving faster pain relief, improved functional recovery, and lowering the risk of bone cement leakage. However, given the retrospective nature of this study, a well-designed prospective randomized controlled trial is still needed to confirm these findings before widespread clinical adoption can be recommended.

Data Sharing Statement

The datasets used during the present study are available from the corresponding author upon reasonable request.

Ethics Approval and Consent to Participate

The study protocol has been approved by the Institutional Ethics Review Board of the First Affiliated Hospital of Xinxiang Medical University (Approval No. EC-LW26-104), and was conducted in accordance to the tenets of the Declaration of Helsinki. Given the retrospective nature of the study and the fact that all data were analyzed after anonymization, the ethics committee approved the waiver of patient informed consent.

Consent for Publication

This was deemed not applicable.

Author Contributions

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

Disclosure

The authors have no conflicts of interest to declare.