Analysis on the related factors of misplacement of freehand pedicle screws via posterior approach in degenerative scoliosis

Abstract

Background

To study the risk factors associated with misplacement of freehand pedicle screws through a posterior approach for degenerative scoliosis.

Methods

A total of 204 patients who underwent posterior pedicle screw-rod system surgery for degenerative scoliosis in our hospital from December 2020 to December 2023 were retrospectively analyzed. Patient demographics, radiographic accuracy, and surgery-related information were recorded.

Results

A total of 204 patients were included. A total of 2496 screws were placed. 2373 (95.07%) were in good position. Misplacement screws were 123 (4.93%). None of the patients had postoperative spinal nerve symptoms due to screw malposition. The misplacement rate of thoracic (T10-T12) pedicle screws was 11.11% (60/540). Misplacement of pedicle screws in the lumbar spine (L1-L5) was 3.22% (63/1956). Age, gender, surgeon, and operation time had no significant effect on misplacement of pedicle screws (P>0.05). Body mass index, Hu value, number of screw segments, Cobb angle, vertebral rotation, and spinal canal morphology had some correlation with pedicle screw misplacement. Among them, BMI, Hu value, number of screw segments, Cobb angle, and vertebral rotation grade were independent risk factors for PS misplacement (P<0.05). The height of the posterior superior iliac spine had a significant effect on pedicle screw misplacement in the lower lumbar spine (L4/5) (P<0.05).

Conclusion

BMI, Hu value, number of screw levels, Cobb angle, and vertebral rotation grade were independent risk factors for pedicle screw misplacement in patients with degenerative scoliosis. Posterior superior iliac spine height has a large impact on PS placement in the lower lumbar spine. Patients with degenerative scoliosis should be preoperatively planned for the size and direction of the placed screws by X-ray and CT three-dimensional, to reduce the misplacement rate of pedicle screws.

Introduction

With the continuous advancement of the aging population in China, the number of patients with degenerative spinal diseases is increasing. Patients with degenerative scoliosis are also increasing. Surgery became the treatment of choice for most patients [1]. Pedicle screws were first proposed by Boucher [2] as an optimal means of spinal fixation. Compared with previous fixation systems, including spinal hook-rod systems and sublaminar wires, pedicle screws have high tensile strength and significant biomechanical advantages [3]. Pedicle screw fixation is currently the most commonly used surgical method for posterior spinal surgery. However, due to the peculiar pedicle anatomy, screw placement carries a potential risk of misplacement. Misplacement of pedicle screws occurs frequently. Pedicle screw placement with the free-hand technique has been reported to have a 5–15% risk of misplacement [4]. Misplacement of pedicle screws will not only lead to the decline of biomechanical stability of the pedicle internal fixation system, affecting the fixation effect but also cause complications including nerve, blood vessels, and organ injury in severe cases. In this study, we mainly analyzed the risk factors associated with pedicle screw misplacement in degenerative scoliosis.

Subjects and methods

Research object

Patients who underwent posterior pedicle screw system internal fixation for degenerative scoliosis in our hospital from December 2020 to December 2023 were selected as the study subjects. Degenerative scoliosis is defined as spinal deformity with a Cobb angle > 10° in a skeletally mature population.

Inclusion criteria:

1. Patients with degenerative scoliosis were diagnosed based on their history, signs, and imaging studies.

2. No significant surgical contraindications.

3. Posterior spinal pedicle screw system internal fixation was performed in our hospital.

4. Patients with complete preoperative and postoperative imaging data.

Exclusion criteria:

1. Patients with previous thoracolumbar surgery.

2. Pedicle cement screw used intraoperatively.

3. Combined with spinal infectious diseases, tuberculosis, tumors, etc.

4. Patients with incomplete clinical and imaging data.

Operations

The operation was performed in a standard prone position, and all operations were performed separately by 2 groups of surgeons with extensive experience in pedicle screw instrumentation. Various pedicle screw instrumentation systems on the market were used. Thoracic screws are 3.5–5.5 mm in diameter and 35–45 mm in length and lumbar screws are 6.0–6.5 mm in diameter and 40–45 mm in length. The use of specific screws was left to the surgeon’s discretion. Determination of fusion level: selection of fusion range for degenerative scoliosis, i.e. proximal vertebra and distal vertebra. In this study, the upper and lower fusion vertebral segments were individualized according to the extent of nerve decompression. Try to avoid excessive fusion levels.

Observation index

1. Demographic data: including gender, age, nature of work, MBI, bone quality, etc.

2. Operation-related data: including the number of fixed segments, the number of screws, the surgeon, and the operation duration.

3. Imaging data: CT examination was performed by the radiology department of our hospital before and after the operation. The measured imaging parameters were as follows:

Hu value of CT (mean value of CT value of screw implanted vertebral body (Fig. 1A);

Fig. 1.

a is the measurement mode of vertebral body CT value. b is posterior superior iliac spine height ratio

Spinal canal morphology (divided into the normal spinal canal, stenotic spinal canal, and cloverleaf spinal canal);

The posterior superior iliac spine height ratio: refers to the ratio of the distance from the vertical line of the midpoint of the upper endplate of the S1 vertebral body to the line of the highest point of the iliac spine on both sides (OA) to the vertical line of the midpoint of the upper endplate of the S1 vertebral body to the central point of the upper endplate of the L4 vertebral body (OB). Iliac height can eliminate the influence of height and vertebral body height on iliac spine height measurement (Fig. 1B);

Vertebral rotation (Nash-Moe method is divided into 5 levels) and Cobb angle.

Radiological evaluation

Postoperatively, all patients underwent CT examination by our radiology department to assess the pedicle screw position. Screw accuracy was assessed on CT according to Gertzbein-Robbins grading [5]. (A: No cortical penetration; B: Cortical penetration<2 mm; C: 2 mm ≤ Cortical penetration<4 mm; D:4 mm ≤ Cortical penetration<6 mm; F: Cortical penetration ≥ 6 mm) (Fig. 2). Where Grade A and B are defined as accurate localization and Grade C, D, and F are defined as malposition. The main direction (medial, lateral) of PS breaking through the pedicle cortex was recorded. Each PS was measured and recorded separately.

Fig. 2.

a: Grade A screw without cortical breach. b: Class B screws. c: Invasion into medial pedicle wall, grade C screws. d, e, f: Invasion of lateral pedicle wall, grade C screws

Statistical analysis

Data were analyzed using SPSS 26.0 statistical software. Measurement data were expressed in the form of (𝒙̅±s), and enumeration data were expressed in frequency and percentage. Enumeration data were compared using the chi-square test. Differences between groups were compared using independent samples t-test. Logistic regression analysis was performed for factors that were statistically significant in univariate analysis. P<0.05 was considered statistically significant and P<0.01 was considered highly significant.

Results

Incidence of pedicle screw misplacement in DS

208 patients passed the inclusion criteria, of which 2 patients were excluded because of intraoperative use of bone cement screws, 1 patient was excluded because of combined spinal tumors, and 1 patient was excluded because of incomplete imaging data. A total of 204 patients were included. Including 118 males and 86 females. Patients ranged in age from 58 to 72 years, with an average of 64.8 ± 3.3 years. Body mass index was 15.90–33.30 kg/m², with a mean of 23.90 ± 3.01 kg/m². There were 2496 screws. There were 2086 grade A screws, 287 grade B screws, 94 grade C screws, and 29 grade D screws. 2373 (95.07%) were in a good position; 123 (4.93%) were misplaced screws (Table 1). None of the patients had postoperative spinal nerve symptoms due to screw malposition.

Table 1.

Pedicle screw Grade

Segment	Number of PS	Grade A	Grade B	Grade C	Grade D	perforated the outer wall	perforated the inner wall	Misplacement Rate(%)
Thoracic	540	408	72	46	14	27	33	60(11.11%)
Lumbar	1956	1678	215	48	15	42	21	63(3.22%)
Total	2496	2086	287	94	29	69	54	123(4.93%)

The misplacement rate of pedicle screws was 11.11% (60/540) in the thoracic spine (T10-T12) and 3.22% (63/1956) in the lumbar spine (L1-L5). There was a statistically significant difference in the overall mean between the two groups (X² = 56.232, P<0.001) (Table 2).

Table 2.

Misplacement rates of thoracic and lumbar pedicle screws

Segment	Number of PS	Normal PS(%)	Misplaced PS(%)	X2-test
				X2	P
Thoracic	540	480(88.89)	60(11.11)
Lumbar	1956	1893(96.78)	63(3.22)	56.232	<0.001

Of the 123 screws that perforated the pedicle cortex, 69 perforated the outer wall, and 54 perforated the inner wall. Among them, 27 thoracic pedicle screws perforated the outer wall and 33 perforated the inner wall. Lumbar pedicle screws perforated 42 outer walls and 21 inner walls. The difference was statistically significant (P = 0.016) (Table 3).

Table 3.

Thoracic vs. lumbar pedicle screw perforated the outer wall and inner wall

Segment	Number of misplaced PS		perforated the outer wall (%)	perforated the inner wall (%)	X2-test
					X2	P
Thoracic	60		27(45.0)	33(55.0)	5.86	0.016
Lumbar	63		42(66.7)	21(33.3)

Univariate analysis of risk factors for Misplacement of pedicle screw in DS

The patients were divided into the normal screw group and the misplaced screw group. The analysis showed that age, gender, surgeon, and operation time had no significant effect on misplacement of pedicle screw (P>0.05). BMI, Hu value, number of screw segments, Cobb angle, vertebral rotation, and spinal canal morphology had a significant effect on pedicle screw misplacement (P<0.05) (Table 4).

Table 4.

Univariate analysis of risk factors for Misplacement of pedicle screw

Risk factors		Misplaced PS	Normal PS	Correlation coefficient	P
Age(yr)		64.9 ± 3.4	64.7 ± 3.3	-0.405	0.686
Sex	Female	24(27.9)	62(72.1)	0.434	0.510
	Male	38(32.2)	80(67.8)
BMI (kg/m2)		24.40 ± 2.65	22.80 ± 2.73	-3.886	<0.001
Hu value		122.76 ± 27.87	131.41 ± 25.08	2.190	0.030
Operation time		208.31 ± 22.62	202.46 ± 20.31	-1.824	0.070
Surgeon	Team A Team B	29(29.6) 33(31.1)	69(70.4) 73(68.9)	0.057	0.811
Number of screw segments		6(6, 6)	7(6, 7)	-5.479	<0.001
Cobb angle		32.92 ± 3.30	28.55 ± 3.83	-7.822	<0.001
vertebral rotation (Nash-Moe Grade)	0 1	11(21.6) 17(23.3)	40(78.4) 56(76.7)	14.189	0.003
	2	22(36.1)	39(65.0)
	3	12(63.2)	7(35.0)
	4	0(0)	0(0)
Spinal canal morphology	normal spinal canal	27(28.4)	68(71.6)	6.393	0.041
	stenotic spinal canal	22(26.2)	62(73.8)
	cloverleaf spinal canal	13(52.0)	12(48.0)

Multivariate analysis of risk factors for misplacement of pedicle screw in DS

Factors that were significant (P<0.05) in the univariate analysis were included in the binary logistic regression equation. The results showed that BMI, Hu value, number of screw segments, Cobb angle, and vertebral rotation grade were statistically different for pedicle screw misplacement (P<0.05) (Table 5). Forest plots were drawn to visualize the results of logistic regression analysis (Fig. 3).

Table 5.

Multivariate analysis of risk factors for Misplacement of pedicle screw in DS

Risk factors	β	SE	Wald	OR	95% CI for Exp (B)		P
					Lower part	Upper part
BMI	0.278	0.080	11.976	1.321	1.128	1.546	0.001
Hu value	-0.015	0.008	4.054	0.985	0.970	1.000	0.044
Number of screw levels	0.949	0.388	5.987	2.582	1.208	5.521	0.014
Cobb angle	0.400	0.080	25.061	1.492	1.275	1.744	<0.001
Vertebral rotation	-0.631	0.277	5.179	0.532	0.309	0.916	0.023
Spinal canal morphology	0.413	0.278	2.203	1.512	0.876	2.609	0.138

Fig. 3.

Multivariate regression analysis visual forest plot

Multivariate regression analysis yielded significant factors including “BMI, Hu value, number of screw segments, Cobb angle, and vertebral rotation grade” that were integrated. Logistic regression equation was established: P = 1/(1 + e-Y), e was the base of the natural logarithm, Y = (0.278 × BMI) + (− 0.015 × Hu value) + (0.949 × number of screw segments) + (0.400 × Cobb angle) + (− 0.631 × vertebral rotation grade) − 23.169. ROC curves were plotted to evaluate the discriminatory power of the model. The results showed AUROC = 0.873, 95% CI: 0.823–0.924. Hosmer-Lemeshow test P = 0.278 > 0.05. It shows that there is no statistical difference between the current model and the ideal perfect model (Fig. 4).

Fig. 4.

ROC Curve Analysis for Pedicle Screw Misplacement

Effect of the height of the posterior superior iliac spine on pedicle screw placement in the lower lumbar spine

204 patients were included, including 192 patients who underwent lower lumbar (L4 or L5) screw placement and 10 patients who had screw misplacement. A total of 740 screws were placed, including 11 misplaced screws. The misplacement rate was 1.48%. The results showed that the height of the posterior superior iliac spine had a significant effect on the misplacement of pedicle screws in the lower lumbar spine (L4/5) (P<0.05) (Table 6).

Table 6.

Relationship between iliac spine height and lower lumbar spine screw placement

Group	Mean ± SD	Difference with 95% CI	t	P
Normal group	0.65 ± 0.12	-0.09	-2.780	0.006
Misplacement group	0.74 ± 0.06	(-0.16～-0.27)

Discussion

Incidence of pedicle screw misplacement

Misplacement rates of 5–15% have been reported when PS is placed freehand [4]. Orief T [6] found varying degrees of pedicle penetration in 14.3% of percutaneous pedicle screws in patients with thoracolumbar fractures. Aigner [7] included 1068 pedicle screws and found 4.8–7.0% misplacement screws by two different classifications. At present, there are many studies on PS in thoracolumbar fractures, but there are relatively few studies on PS misplacement in degenerative scoliosis. In the free-hand technique, Laudato [8] reported a thoracic to lumbar PS accuracy of 93.6%. Kim [9] found that the accuracy of PS in the thoracic spine in adult patients with degenerative scoliosis and deformity was 93.8%. A total of 204 patients with 2496 screws were included in this study. According to Gertzbein-Robbins classification [5], 2373 (95.07%) screws were in a good position and 123 screws were misplaced, with a misplacement rate of 4.93%. This study showed high accuracy in surgery for patients with degenerative scoliosis, which is similar to previous findings. In this study, 69 PS perforated the outer pedicle wall, and 54 perforated the inner pedicle wall. Because of the peculiar pedicle anatomy, screw placement carries a potential risk of malposition. Accuracy of pedicle screw placement impacts biomechanical stability. When the pedicle screw is misplaced, it will lead to its stability decreased, and in severe cases, it will lead to screw withdrawal [10]. Korkmaz M [11] found that the pullout strength of pedicle screws decreased by approximately 71% when the lateral pedicle wall was perforated. Pedicle screw misplacement is associated with complications. Such as vascular injury, and cerebrospinal fluid leakage. Severe cases can even lead to esophageal perforation, and delayed spinal cord nerve injury [12–15]. None of the patients in this study had postoperative spinal nerve symptoms due to screw malposition. Previous studies have shown that there is a learning curve for the surgeon when placing pedicle screws, and the number of screw misplacements is significantly reduced after approximately 80 PS are placed [16]. In this study, all surgeons had placed far more than 80 screws in the past, there was no significant difference in the misplacement rate of pedicle screws, and the procedure achieved homogenization.

In this study, the misplacement rate of pedicle screws was 11.11% (60/540) in the thoracic spine (T10-T12) and 3.22% (63/1956) in the lumbar spine (L1-L5). The misplacement rate of lumbar pedicle screws was significantly lower than that of thoracic screws. We consider that because thoracic pedicle width is usually thinner than lumbar vertebrae, the rate of misplacement is higher than lumbar vertebrae. In this study, the misplacement rate of the medial wall of thoracic pedicle screws was slightly higher than that of the lateral wall. We analyzed the reason that thoracic vertebrae are mostly apical vertebrae of scoliosis, and it is generally necessary to give PS a sufficiently large abduction angle to enhance the holding force of PS. This easily gives too much abduction angle which causes PS to perforate the medial wall. However, lumbar screws perforated more lateral walls than medial walls of the pedicle. In our analysis, lumbar paravertebral muscles were more developed, and when muscle relaxants were underdosed during surgery, they may result in an inability to give adequate abduction angles to lumbar PS. When placing PS in the lower lumbar spine, obstruction of the iliac crest can also lead to failure to provide adequate abduction angles of the PS. This can cause the PS to perforate the lateral wall. Previous scholars have found that the width of thoracic pedicles is generally small in Chinese. When the diameter of the screw is too large, there is a risk of damage to nerves, blood vessels, etc. However, too small a screw size can lead to reduced strength of surgical internal fixation and easy screw breakage [17–19]. It is also one of the factors that the misplacement rate of thoracic pedicle screws is higher than that of lumbar spine. Some scholars prefer to choose relatively thick size screws, and deliberately offside the screws during screw placement to slightly perforate the outer wall of the pedicle. At this time, although the pullout strength of the screw is only 75% of that of the pedicle screw in place, the risk of injuring the thoracic spinal cord is low [20]. The perfect fit between screw diameter and pedicle width is one of the key factors for successful pedicle screw fixation surgery. Surgeons should plan screws preoperatively based on CT three-dimensional advance and select appropriate pedicle screws, which can reduce the occurrence of pedicle screw placement failure [21, 22].

Analysis of risk factors of misplacement of pedicle screws

In this study, multivariate regression analysis showed that BMI, Hu value, number of screw segments, Cobb angle, and vertebral rotation grade were independent risk factors for pedicle screw misplacement (P<0.05).

Previous studies have shown that patients undergoing surgery for degenerative spine disease are generally obese. There was a linear association between increasing body mass index and postoperative complications after fusion surgery [23–25]. Obesity has been reported to increase the risk of misplacement of PS [26]. It has also been shown that obese patients do not have an increased risk of misplacement of percutaneous pedicle screws [27]. In this study, the OR for BMI was 1.321 with a 95% confidence interval of 1.128–1.546, P = 0.001. Compared to patients with normal body mass index, overweight patients had an approximately 0.32-fold increased risk of misplacement of PS. Pedicle screw placement in obese patients often requires dissection of thick fat and muscle tissue. This may result in a deeper and obscured operative field and the inability to give the appropriate abduction angle for the screw due to obstruction by thick fat and muscle. So pedicle screws may be more easily misplaced in obese patients than in normal patients.

In recent years, the assessment of bone quality by computed tomography (CT) has been reported, with more Hounsfield unit (Hu) measurements⁽²⁸⁾. Hu measures the density of tissue and can be used to centrally assess cancellous bone, thereby avoiding interference from the degenerative cortex. Some studies have demonstrated that the Hounsfield score is predictive of screw loosening after spinal surgery [29, 30]. However, no studies have been reported on its association with misplacement of pedicle screws. Zaidi [28] reviewed the utility of Hu in assessing bone quality, setting the threshold for osteoporosis at 110 Hu. In this study, we found that the Hu value of patients in the pedicle screw misplacement group was significantly lower than that of patients in the normal group. Pedicle screw misplacement rates are high when a bone is relatively osteoporotic. We analyzed the reason possibly because the screw easily deviated from the original open screw path during screw placement in osteoporotic patients. Before surgery, the surgeon can estimate the bone quality of the patient by measuring the Hounsfield unit, to give anti-osteoporosis and other treatments. The misplacement rate of pedicle screws can be reduced by planning the size and direction of the placed screws through CT three-dimensional before surgery.

In this study, the risk of misplacement of pedicle screws increased as the number of surgical levels increased. Because of the increased level of operation, the surgeon’s energy consumption is too high and fatigue may occur, which leads to an increased risk of misplacement of pedicle screws. Previous studies have shown that preoperative coronal Cobb angle and vertebral rotation grade are risk factors for PS misplacement in spinal surgery [31]. Zhu [32] found that patients with a Cobb angle > 80 ° had a higher rate of PS misplacement than patients with a Cobb angle < 80 °in congenital scoliosis. Harimaya [33] found a positive correlation between vertebral rotation and the occurrence of PS malposition in infants and adolescents. In this study, Cobb angle and vertebral rotation grade were found to be independent risk factors for PS misplacement in degenerative scoliosis. For patients with larger Cobb angles and higher vertebral rotation grades, clinicians should plan the direction and position of each PS preoperatively. Navigation systems can be used to place screws intraoperatively, which is very helpful for scoliosis screw placement [34].

Relationship between the height of posterior superior iliac spine and lower lumbar screw placement

In this study, we used the posterior superior iliac spine height ratio to describe the iliac spine height, which eliminates the influence of height and vertebral body height on the posterior superior iliac spine height measurement. Previous studies have analyzed the influence of posterior superior iliac spine height in relation to endoscopic transforaminal surgery. A higher iliac spine may have some effect on the endoscopic channel [35]. The results of this study found that patients with a high ratio of posterior superior iliac spine height meant that the iliac spine height was close to the superior endplate of the L4 vertebral body. Such patients are more prone to screw misplacement in the lower lumbar spine. And we found in our clinical work that PS screw placement in the lower lumbar spine is more difficult in patients with a higher posterior superior iliac spine, and due to the hindrance of a higher iliac spine, it is difficult to give an appropriate abduction angle for PS, which easily leads to screw misplacement.

This study has several limitations. First, this study was retrospective and may have generated some bias during data collection. This study is a single-center study and may be underrepresented. A multicenter prospective study will be performed for subsequent studies to validate the results of this study. At present, robotic screw placement in spinal surgery is emerging, and relevant comparative studies will be carried out subsequently.

Conclusion

Misplacement of pedicle screws occurs frequently during spinal surgery. BMI, Hu value, number of screw levels, Cobb angle, and vertebral rotation grade were independent risk factors for PS misplacement in patients with degenerative scoliosis. Posterior superior iliac spine height has a large impact on PS placement in the lower lumbar spine. Patients with degenerative scoliosis should be preoperatively planned for the size and direction of the placed screws by X-ray and CT three-dimensional, to reduce the misplacement rate of PS.

Abbreviations

DS

Degenerative scoliosis

PS

Pedicle screws

BMI

Body mass index

CT

Computed tomography

Hu

Hounsfield unit

Author contributions

GCJ: data acquisition, literature search, manuscript preparation, and was a major contributor in writing the manuscript. WRY: data acquisition, literature search, manuscript preparation. RN: revised and reviewed the manuscript for the final publication. RN and LJ analyzed and interpreted the patient’s procedure related to the surgery. LQ and WY: made charts. ZF and CLY: data curation and literature search. All authors read and approved the final manuscript.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Data availability

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

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the Yichang Central People’s Hospital. No. 2023-253-01. Informed consent to participate was waived by the Ethics Committee of the Yichang Central People’s Hospital due to the retrospective nature of the study design.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.