DOI: https://doi.org/10.21203/rs.3.rs-2297814/v1
To propose a novel free-hand pedicle screw placement technique in the lumbar spine and investigate its accuracy and safety in clinical practice.
110 patients with herniation of a single lumbar disc were recruited for our study. All patients were consistently treated with posterior discectomy, interbody fusion, and transpedicular internal fixation. For 54 patients in the observation group, the pedicle screws were placed with our technique, which is located at 4 mm below the superior edge of the transverse process in line with the lateral margin of the superior articular process. For 56 patients in the control group, pedicle screws were placed according to the traditional crista lambdoidalis method. Comparisons were made in terms of the operation time, blood loss, time for exposure, the accuracy of placement, and postoperative complications. In addition, we prospectively applied our method to 64 patients with indistinguishable crista lambdoidalis and evaluated the accuracy of screw placement and clinical outcomes.
There was no significant difference in intraoperative bleeding, accuracy of placement, and postoperative complications between our technique and the traditional crista lambdoidalis method (P > 0.05). However, the exposure time before screw placement (12.8 ± 0.3 vs 17.4 ± 0.3, p = 0.001) and the total surgery time (97.2 ± 1.9 vs 102.3 ± 0.9, p = 0.020) were significantly shortened with our method. Additionally, in cases of indistinguishable crista lambdoidalis, our technique showed satisfying accuracy, with 97.6% screws placed in appropriate trajectory on the first attempt and all screws eventually positioned in the safe zone according to the Gertzbein Robbins grading. All patients experienced steady improvement after surgery, as assessed with the visual analog pain scale and the Japanese Orthopaedic Association (JOA) score.
This study highlights a safe and effective technique for pedicle screw placement in the lumbar spine, which is particularly useful in cases of vertebral fracture and severe hyperplasia or degeneration of the facet joint.
Pedicle screw implantation is the most commonly used approach to achieve three-column stability for the majority of spinal surgery procedures[1]. Indications for transpedicular fixation include congenital disease, deformities such as scoliosis and hyperkyphosis, vertebral fracture, spinal tumor, infection, and degenerative disease[2]. Given the vicinity of vital neurovascular structures, malposition of pedicle screws can cause serious complications and internal fixation failure[3]. Advances have been made in intraoperative navigation systems over the past decades. Computer tomography(CT)-based, fluoroscopy-assisted, and robot-assisted systems are reported to increase the accuracy and decrease the comorbidities of screw implantations[4]. Nevertheless, the navigation system can be cost-consuming with limited application in hospitals of small scales. So far, free-hand screwing is still the most commonly used technique, especially in emergent and limited situations.
The establishment of a proper insertion point is the first and crucial step for accurate pedicle screw implantation[5]. Various methods have been advocated, including the Roy-Camille method, Magerl’s technique, Weinstein’s method, and Crista lambdoidalis technique etc.[6–9]. The Crista lambdoidalis method describes the entry point as the apex where the isthmus crest converses the accessory process crest, which is also known as the “∧” shape crest[9, 10]. However, tactile feedback and experience-based judgment are of utmost importance in present techniques. The misplacement rate in the lumbar spine was reported to range from 5 to 41%[2, 11]. Additionally, in cases of lumbar spondylolysis, vertebral fracture, and severe hyperplasia or degeneration of the facet joint, the crista lambdoidalis can be difficult to distinguish. In this study, we highlight a safe and quantified technique for pedicle screw placement in the lumbar spine, by locating the entry point at 4mm below the superior edge of the transverse process in line with the lateral margin of the superior articular process.
The aims of this study include: (a) to propose a novel technique of free-hand pedicle screw placement in the lumbar spine, (b) to evaluate the accuracy and safety of this technique as compared to the traditional crista lambdoidalis method, and (c) to apply our technique in patients with indistinguishable crista lambdoidalis and assess its safety and validity.
Inclusion criteria were as follows: 1) patients with typical symptoms and radiological manifestations indicating disc herniation of a single segment, 2) patients with no alleviation of symptoms after 3 months’ conservative treatment, 3) patients who underwent posterior discectomy, interbody fusion and transpedicular fixation by the same surgical team in our hospital; and 4) patients who were followed for at least 1 year after surgery. Exclusion criteria were as follows: 1) patients who could not tolerate surgery, 2) patients with vertebral infection or fracture involving articular process and/or transverse process, and 3) patients with history of lumbar spine surgery. This research was approved by the ethics committee of our hospital, and written informed consent was obtained from all patients or their legal guardians.
From Aug 2018 to Aug 2020, 110 patients who met the inclusion and exclusion criteria were retrospectively enrolled in our study. Patient demographics, including patients’ age, gender, duration of symptoms and VAS score were evaluated to determine the baseline difference. Surgical information was also recorded, including the operation time, blood loss, time for exposure, the accuracy of screw placement, and postoperative complications (including surgical site infection, neurovascular injury, and internal fixation failure). Postoperative CT scans were assessed according to the Gertzbein Robbins grading system[12]. Screws completely contained within the pedicles (Grade A) and breaches within 2mm (Grade B) were considered in the safe zone with a low risk of neural damage[12].
Afterwards, we prospectively applied our technique to 64 patients with indistinguishable crista lambdoidalis, where the traditional crista lambdoidalis technique is unfeasible. The accuracy of screw placement and times of intra-operative fluoroscopy were recorded. The clinical efficacy was evaluated by the Japanese Orthopaedic Association (JOA) score and VAS score before operation and 1 month, 3 months, and 6 months after surgery[13, 14]. The Bridwell grading was adopted for assessment of interbody fusion at the last follow-up, where ‘fused with remodeling and trabeculae’ was defined as Grade I, while ‘intact graft without fully remodeling and incorporation, but no present lucency’ was defined as Grade II[15].
Patients were placed in the prone position after anesthesia. A mid-line incision was made, followed by blunt separation of the subcutaneous tissues and subperiosteal dissection of the muscles until the superior articular process and, in case of the control group, bilateral lamina were exposed. In the observation group, an awl was used to detect the superior edge of the transverse process, afterwards, a nerve raspatory was used to locate the entry point at 4mm below the superior edge of the transverse process in line with the lateral margin of the superior articular process (Fig. 1). While in the control group, screws were placed at the apex of the crista lambdoidalis[10] (Fig. 1). After identification of the entry point, a blunt opener was used to enter the vertebral pedicle. Afterward, a blunt probe was applied to ensure the intactness of the trajectory before a screw of appropriate diameter and length was implanted. Once all screws were placed, the C-arm fluoroscopy was used to examine the position of all screws. If the pedicle was breached, the trajectory was re-built and fluoroscopy was performed afterward until satisfactory implantation was achieved. After screw fixation, the posterior discectomy, and interbody fusion were subsequently performed. Finally, screws were connected with two longitudinal link rods before the wound was washed and sutured.
The normal distribution of quantitative data was determined by the Shapiro–Wilk test, and the homogeneity of variance was tested by Levene’s test. Student t-test, Mann-Whitney U test, or chi-squared test was performed to compare the mean value or data distribution among different groups. p values less than 0.05 are considered statistically significant. All statistical analyses were conducted using SPSS version 26.0 (SPSS Inc., Chicago, IL, USA).
Baseline information among the two groups was compared and listed in Table 1. There were 31 male patients and 23 female patients in the observation group, with an average age of 53.6 years. In the meanwhile, 27 male patients and 29 female patients were included in the observation group, with an average age of 52.9 years. On average, patients from both groups suffered from herniation-related symptoms for an average of 12 months before surgery. In the meanwhile, symptoms evaluated by the VAS was comparable between the two groups.
|
Variables |
Observation Group |
Control Group |
Effect Size (t/χ2/Z value) |
p value |
|---|---|---|---|---|
|
Cases (No.) |
54 |
56 |
||
|
Age (Mean ± SD, years) |
53.6 ± 2.0 |
52.9 ± 1.6 |
Z=-0.446 |
0.656 |
|
Gender (Male/Female) |
31/23 |
27/29 |
χ = 0.932 |
0.334 |
|
Duration of symptoms (months) |
12.1 ± 0.5 |
12.1 ± 0.9 |
Z=-1.001 |
0.317 |
|
Visual analog scale |
||||
|
Pre-operation |
7.3 ± 0.1 |
7.3 ± 0.1 |
Z=-0.397 |
0.691 |
|
Post-operation |
2.0 ± 0.2 |
2.0 ± 0.1 |
Z=-0.217 |
0.828 |
|
Operation time (min) |
97.2 ± 1.9 |
102.3 ± 0.9 |
t=-2.382 |
0.020 |
|
Blood loss (ml) |
307.4 ± 7.8 |
309.5 ± 10.5 |
t=-0.157 |
0.876 |
|
Time for exposure (min) |
12.8 ± 0.3 |
17.4 ± 0.3 |
Z=-7.633 |
0.001 |
|
Accuracy of screw implantation (No., %) |
||||
|
Success on the first attempt |
204 (94.5%) |
208 (92.9%) |
χ = 0.465 |
0.495 |
|
Re-implantation |
12 (5.5%) |
16 (7.1%) |
||
|
The Gertzbein Robbins grading |
||||
|
Grade A & B |
213 (98.6%) |
217 (96.9%) |
χ = 1.492 |
0.222 |
|
Grade C & D &E |
3 (1.4%) |
7 (3.1%) |
||
In the observation group, a total of 216 screws were placed, 204 (94.5%) of them were placed in appropriate trajectory on the first attempt based on the intraoperative fluoroscopy, while 12 (5.5%) of them were considered inappropriate and re-implantation was performed. By comparison, 224 screws were placed in the control group and the success rate on the first attempt was 92.9%. Although the success rate on the first attempt was higher in the observation group, no statistical significance was reached (p = 0.495). In the meanwhile, the accuracy of screw placement was also evaluated on postoperative CT images according to the Gertzbein Robbins grading. 184 in the observation group were rated as grade A, 29 as grade B, and 3 as grade C. The accuracy was calculated as 98.6% (213/216). In comparison, of the 224 screws in the control group, 189 were rated as grade A, 28 as grade B, and 7 as grade C. The overall accuracy was 96.8%, showing no significant difference from the 98.6% in the observation group (p = 0.222). Of note, it took shorter time to expose relevant bony landmarks in the observation group to prepare patients for screw implantation (12.8 ± 0.3 min vs 17.4 ± 0.3, p = 0.001), and the whole operation time was also significantly shortened with our technique (97.2 ± 1.9 min vs 102.3 ± 0.9 min, p = 0.020). During the follow-up, patients in both groups showed significant pain alleviation and function restoration, with no significant difference noted as evaluated with the VAS and muscle strength grading.
Furthermore, 35 male patients and 29 female patients with indistinguishable crista lambdoidalis were prospectively identified (Table 2). Pedicle screws were placed at 4 mm below the superior edge of the transverse process in line with the lateral margin of the superior articular process according to our technique. The average number of intraoperative fluoroscopies was 3.9 ± 0.7, and 279 out of 286 (97.6%) screws were placed accurately on the first attempt. All screws were placed in the safe zone, as evaluated with the Gertzbein Robbins grading on postoperative CT scans. After an average follow-up of 11.2 months, no complications were noticed in all patients. The Bridwell grading showed satisfied bone fusion, while the JOA and VAS grading indicated steady improvement of neural function in all patients (Table 3).
|
Parameter |
Value |
|---|---|
|
Age (Mean ± SD, years) |
49.9 ± 16.2 |
|
Gender (Male/Female) |
35/29 |
|
Follow-up (months) |
11.2 ± 4.7 |
|
Intraoperative fluoroscopy (times) |
3.9 ± 0.7 |
|
Accuracy of screw implantation (No., %) |
|
|
Success on the first attempt |
279/286(97.6%) |
|
Re-implantation |
7/286 (2.4%) |
|
The Gertzbein Robbins grading |
|
|
Grade A |
204 (79.7%) |
|
Grade B |
52 (20.3%) |
|
Grade C & D &E |
0 |
|
The Bridwell grading |
|
|
I |
60/64 (93.8%) |
|
II |
4/64 (6.2%) |
|
III |
0 |
|
Time |
Visual analog scale |
JOA |
|---|---|---|
|
Preoperative |
6.859 ± 1.193 |
4.984 ± 1.339 |
|
Post-operation at 1 month |
2.016 ± 1.016① |
20.094 ± 1.455① |
|
Post-operation at 3 months |
1.031 ± 0.755①② |
24.469 ± 1.112①② |
|
Post-operation at 6 months |
0.250 ± 0.471①② |
25.203 ± 1.535①② |
| Notice: ①comparison with preoperative, p＜0.05; ②Comparison with the previous review, p＜0.05 | ||
The implantation of pedicle screw is a fundamental technique in spinal surgery. The accuracy of pedicle screw placement strongly affects the outcome of spinal surgery and mainly relies on the surgeons’ experience. Inaccurate placement may incur devastating neural injuries and internal fixation failure[16]. Therefore, finding an accurate and quantified method of locating the entry point is of vital importance, especially for young surgeons with less experience.
In the past decades, various methods have been proposed for the localization of entry points. Roy-Camille suggests the entry point as the intersection of two lines: the vertical line is the prolongation of the facet joint while the horizontal line passes through the middle of the insertion of the transverse process or 1mm below the joint line[6]. Magerl’s technique describes the entry point as the intersection of the lateral border of the superior facet joint and the midline of the transverse process[7]. Weinstein prefers the entry point as the lateral and inferior corner of the superior articular facet, what Weinstein calls the ‘nape of the neck’ of the superior articular facet[8]. The crista lambdoidalis method describes the entry point as the vertex where the isthmus crest converses the accessory process crest[9]. Among these techniques, both the Roy-Camille method and the Magerl’s method use the midline of the transverse process as the reference, which requires thorough exposure of the operative field, therefore longer operation time and more blood loss. Additionally, the transverse process can be asymmetrical, and using the midline of the transverse process as the reference can be misleading. In a study conducted by Ebraheim et al., fifty dry lumbar specimens were obtained to measure the distance from the projection point of the lumbar pedicle axis to the midline of the transverse process for each level of the lumbar vertebrae[17]. From L1 to L3, the projection point of the pedicle axis lays 3.9 mm, 2.8 mm and 1.4 mm above the midline of the transverse process, respectively[17]. Therefore, using the midline of the transverse process as the reference tends to increase the risk of breaching the inferior wall of the pedicle and increase the risk of causing nerve root injury. In a study carried out by Yu et al., three methods (the Roy-Camille method, the Magerl’s method, and the crista lambdoidalis method) were simulated on 3D reconstructed images to compare the distance from the entry point to the axis of the pedicle[18]. In their study, the crista lambdoidalis method proved to be superior from L1 to L4, while the Magerl’s technique showed comparable accuracy in L3-L4 and has the highest accuracy at L5[18]. However, the overall existence of the crista lambdoidalis is reported to be around 94.5% and it’s especially lower in L5[9]. Besides, crista lambdoidalis can be indistinguishable in cases of deformities, lumbar spondylolysis, vertebral fracture, or severe hyperplasia of the lumbar facet joint. Therefore, to reduce the risk of screw malposition and to shorten the learning curve for young surgeons, we proposed a valid and quantified method, by locating the entry point at 4mm below the superior edge of the transverse process in line with the lateral margin of the superior articular process.
Compared with traditional methods, the entry point located with our method tends to have a superior and lateral position, which decreases the risk of violating the spinal canal and the nerve roots. In the meanwhile, our method provides a qualified strategy to locate the entry point, which proves to be related with higher success rate and less complications. According to the Gertzbein Robbins grading, 184 screws in the observation group were completely placed within the pedicles, while 29 screws were placed with breaches less than 2mm, and 3 screws were placed with breaches between 2mm to 4mm. In total, 98.6% of the screws were considered to be contained in the safe zone with our method, compared to 96.9% in the control group. During the follow-up, all patients showed steady improvement in pain alleviation and neurologic restoration, with no serious neurovascular complications being reported. In the meanwhile, compared with traditional methods, our method requires less extensive exposure and paraspinal dissection. According to recent studies, erector spinae and multifidus muscle injury resulted from surgical dissection are associated with postoperative low back pain, functional disability, and proximal junctional kyphosis[19, 20]. With our method, both the exposure time before screw implantation and the total operation time is significantly shortened (p < 0.05), which can reduce intraoperative blood loss, improve postoperative pain and decrease the risk for developing postoperative infection. Additionally, in cases of lumbar spondylolysis, vertebral fracture, and severe hyperplasia or degeneration of the facet joint, the crista lambdoidalis can be difficult to distinguish. In these cases, using the crista lambdoidalis as the reference could be confusing. In our study, we prospectively applied our method to 64 patients with indistinguishable crista lambdoidalis. 97.6% of the screws were placed in proper trajectory on the first attempt, and all screws were placed within the ‘safe zone’ as assessed with the Gertzbein Robbins grading. Last but not the least, by proposing a quantified approach for screw implantation, the learning curve for inexperienced surgeons could be significantly shortened, meanwhile the incidence of malposition could also be decreased.
Despite the superiority shown in our study, there are several limitations to be addressed. The average follow-up was only 12 months in our study, longer follow-up is expected to examine the incidence of long-term complications, such as pedicle screw loosening, pull out, or internal fixation failure. In addition, this study demonstrates experience in our single center, multi-center validations are critical for the evaluation and amendment of our technique.
This study highlights a valid and quantified technique for pedicle screw placement in the lumbar spine, which can help inexperienced spinal surgeons to better locate the entry point of pedicle screws, especially in cases where the crista lambdoidalis is indistinguishable.
Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.