Comparison of the One-Time Accuracy of Simulated Freehand and Navigation Simulated Pedicle Screw Insertion

Comparison of the One-Time Accuracy of Simulated Freehand and Navigation Simulated Pedicle Screw Insertion

Original Article Comparison of the One-Time Accuracy of Simulated Freehand and Navigation Simulated Pedicle Screw Insertion Yun-Feng Xu, Qi Zhang, Xi...

2MB Sizes 0 Downloads 49 Views

Original Article

Comparison of the One-Time Accuracy of Simulated Freehand and Navigation Simulated Pedicle Screw Insertion Yun-Feng Xu, Qi Zhang, Xiao-Feng Le, Bo Liu, Da He, Yu-Qin Sun, Ya-Jun Liu, Qiang Yuan, Zhao Lang, Xiao-Guang Han, Wei Tian

OBJECTIVE: To compare one-time accuracy rate between simulated freehand (SFH) and navigation simulated (NS) pedicle screw insertion, assuming no second chance to correct screws.

-

METHODS: A simulated, comparative, cross-sectional study was conducted on 69 patients undergoing lumbar spine surgery. An intraoperative registration system captured the planned point of entry and trajectory of pedicle screws for both SFH under direct visualization and NS under navigation-aided visualization. Pedicle screw insertion was simulated for each captured image (370 screws) using Surgimap. Rajasekaran’s method helped evaluate the point of entry accuracy and trajectory.

-

RESULTS: Accuracy rate was better for the NS method (97.8%) than for the SFH method (63.8%). Of 370 screws in the SFH group, 134 penetrated the cortex, with 31 resulting in >4 mm penetration. Of 370 screws in the NS group, 8 penetrated the cortex, <4 mm penetration. Of 134 misplaced screws in the SFH group, 64 were due to error in the point of entry, 63 were due to error in the trajectory angle, and 7 were due to both errors. Of 8 errors in the NS group, 7 were due to the point of entry.

-

CONCLUSIONS: Intraoperative navigation had significantly better one-time accuracy of pedicle screw insertion than freehand insertion and should be used to avoid injury to the pedicle and surrounding tissue from screw reinsertion.

-

Key words Intraoperative screw adjustment - Navigation simulated - One-time accuracy - Pedicle screw placement - Simulated freehand -

Abbreviations and Acronyms 3D: Three-dimensional NS: Navigation simulated SFH: Simulated freehand

WORLD NEUROSURGERY 128: e347-e354, AUGUST 2019

INTRODUCTION

P

edicle screw insertion is a key procedure for the treatment of lumbar degenerative diseases that is safe and effective1,2 and provides robust fixation with a satisfactory long-term clinical outcome. However, because of the unique anatomic structure of the pedicle, misplacement of a screw may lead to serious complications.3-5 Moreover, reinsertion of the screw can destroy the bony structure of pedicle and surrounding structures, which lowers the pullout strength of the screw, increasing the risk for loss of stabilization and recurrence of a segmental deformity.6,7 Therefore, positional accuracy of screw insertion on the first try is clinically significant. The aim of this study was to compare the one-time accuracy of simulated freehand (SFH) and navigated simulated (NS) pedicle screw insertion. MATERIALS AND METHODS Participants The study was approved by our institutional review board, and all participants provided informed consent. Patients were selected using the following inclusion criteria: lumbar degenerative disease (L1-S1) resulting in radiculopathy, 18e70 years of age at the time of surgery, and ineffective results with conservative treatment for a minimum of 6 months. Exclusion criteria were presence of scoliosis >30 , previously diagnosed osteomalacia or severe osteoporosis, history of spinal tumors or tuberculosis, and spinal infection. Based on the inclusion and exclusion criteria, 69 patients (28 men and 41 women; mean age 57.4  10.7 years) who underwent traditional open lumbar spine surgery for degenerative disease between January 2013 and February 2014 were included in the study. Indications for lumbar spine surgery included lumbar spinal stenosis (28 cases; 40.6%), spondylolisthesis (16 cases;

Department of Spine Surgery, Beijing Jishuitan Hospital and Peking University Fourth School of Clinical Medicine, Beijing, China To whom correspondence should be addressed: Wei Tian, M.D. [E-mail: [email protected]] Yun-Feng Xu and Qi Zhang are coefirst authors. Citation: World Neurosurg. (2019) 128:e347-e354. https://doi.org/10.1016/j.wneu.2019.04.151 Journal homepage: www.journals.elsevier.com/world-neurosurgery Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2019 Elsevier Inc. All rights reserved.

www.journals.elsevier.com/world-neurosurgery

e347

ORIGINAL ARTICLE YUN-FENG XU ET AL.

PEDICLE SCREW SIMULATED INSERTION

23.2%), lumbar disc herniation (20 cases; 29.0%), and lumbar revision surgery (5 cases; 7.4%). All surgeries were performed under general anesthesia with patients in the prone position. The operations were performed using the traditional open technique and intraoperative three-dimensional (3D) computer-assisted navigation8 by 12 board-certified specialist spine surgeons, all of whom have at least 10 years of experience in spine surgery. Each simulation was performed by only 1 surgeon. Data Capture Data on the planned point of entry and trajectory of the pedicle screw were obtained using an intraoperative 3D navigation system (Stryker Spine Navigation System, Stryker, Kalamazoo, Michigan, USA). After the spine was exposed, a calibration marker was clamped to the spinous process of the superior operated level. Intraoperative 3D images were acquired by C-arm (Arcadis Orbic 3D; Siemens Medical Solutions, Erlangen, Germany) and transferred to the navigation workstation for registration. For the SFH method, surgeons selected the point of entry and trajectory of the pedicle screw based on preoperative planning of radiographic images, fluoroscopy, and anatomic landmarks, without access to the visual display of the navigation system. Surgeons systematically planned the trajectory on preoperative radiographic images, recorded each measurement, and correlated the trajectory with known landmarks. Surgeons simulated freehand pedicle screw placement by holding a navigated probe against where they

thought the screw trajectory should be. The entry point in all lumbar levels was selected according to Magerl’s method.9 The point of entry and the trajectory (sagittal angle and transverse angle) of the pedicle screw were recorded using the screen capture tool of the navigation system (Figure 1). The process was repeated for all the screws to be used during surgery. Afterward, the same surgeon simulated the navigated screws under real-time 3D navigation, without considering the previous SFH simulation. The trajectories were saved for each screw again. To avoid possible bias, real-time screw display under navigation was not available. Additionally, the same researchers used the same method to record the screw placement plan (Figure 2).

Simulated Placements of Screws Captured data about the point of entry and trajectory of insertion of the pedicle screw for the SFH and NS methods were transferred to Surgimap Version 2.1.8 software (Nemaris, Inc., New York, New York, USA), and screw placement was simulated by a spine surgeon blinded to the used placement method, using the captured data as inputs. For each simulation, the appropriate pedicle screw size was selected based on the height and width of the pedicle, per usual procedures. For each screw placement evaluation, the same pedicle screw was used for the SFH and NS simulations (Figure 3).

Figure 1. Capture of the planning of pedicle screw insertion using the freehand method.

e348

www.SCIENCEDIRECT.com

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2019.04.151

ORIGINAL ARTICLE YUN-FENG XU ET AL.

PEDICLE SCREW SIMULATED INSERTION

Figure 2. Capture of the planning of pedicle screw insertion using the navigation method.

Accuracy Assessment (Rajasekaran’s Method) The accuracy of the SFH and NS methods was evaluated using Rajasekaran’s method,10 with the degree of screw penetration into the cortex measured using the ruler tool in Surgimap. Each virtual screw placement was evaluated by 2 independent reviewers, a spine surgeon and a radiologist, who were blinded to the placement method used. In case of disagreement, a third senior surgeon reviewed the images for adjudication. The position of the screw was classified into 4 groups for analysis: grade 0, completely within the pedicle, with no violation of the cortex; grade 1, <2 mm penetration of the cortex; grade 2, 2e4 mm penetration of the cortex; grade 3, >4 mm penetration of the cortex. Analysis of Inaccurate Screw Placement Inaccurate screw placement included error in the point of entry and/or error in the trajectory of insertion. These errors were further characterized by the trajectory of the error (medial, lateral, superior, or inferior) (Figures 4 and 5). The rate and type of errors were compared between the SFH and NS methods. The lumbosacral levels were used as explanatory factors in the analysis. Statistical Analysis Statistical analysis was performed using SPSS Version 17 for Windows (SPSS, Inc., Chicago, Illinois, USA). c2 test or Fisher

WORLD NEUROSURGERY 128: e347-e354, AUGUST 2019

exact test was used to evaluate between-group differences, as appropriate for the data type. In all statistical analyses, a P value <0.05 was considered statistically significant. RESULTS Among the 69 cases included in our study cohort, a 1-level procedure was performed in 33 cases (47.8%), a 2-level procedure was performed in 27 cases (39.1%), a 3-level procedure was performed in 7 cases (10.1%), and a 4-level procedure was performed in 2 cases (2.9%). With regard to the segments involved, surgeries were performed at the level of L1-L2 in 1 case (1.5%), L1-L5 in 1 case (1.5%), L2-L5 in 4 cases (5.8%), L2-S1 in 1 case (1.5%), L3-L5 in 7 cases (10.1%), L3-S1 in 3 cases (4.4%), L4-L5 in 20 cases (29.0%), L4-S1 in 20 cases (29.0%), and L5-S1 in 12 cases (17.4%). In total, 370 pedicle screw placements were simulated in each group, as follows: L1, 4 screws (1.1%); L2, 14 screws (3.8%); L3, 32 screws (8.7%); L4, 114 screws (30.8%); L5, 136 screws (36.8%); S1, 70 screws (18.9%). Accuracy The SFH method had an accuracy of 63.8%, with 134 screws penetrating the cortex with the following grade distribution: grade 1, 16.4% (22 of 134); grade 2, 60.4% (81 of 134); grade 3, 23.1% (31 of 134). By comparison, the accuracy of the NS method was 97.8%, with 8 screws penetrating the cortex: grade 1, 87.5%

www.journals.elsevier.com/world-neurosurgery

e349

ORIGINAL ARTICLE YUN-FENG XU ET AL.

PEDICLE SCREW SIMULATED INSERTION

Figure 3. Virtual screw placement using Surgimap software with inputs from the freehand method (A) and the navigation method (B).

(7 of 8); grade 2, 12.5% (1 of 8). There was no occurrence of a grade 3 penetration in the NS group. The rate of accuracy was significantly better for NS than for SFH screw placement (P < 0.05) (Table 1). According to postoperative computed tomography measurements, the actual accuracy of these pedicle screws was 98.10%.

Analysis of Error for SFH Method Of the 134 incorrectly placed screws in the SFH group, 64 were due to entry point errors (47.8%), 63 were due to an incorrect penetration angle (47.0%), and 7 were due to both types of error (5.2%). The distribution of errors in location of the entry point (64 screws) was as follows: medial, 59.4% (38 of 64); lateral, 6.3% (4 of 64); superior, 23.4% (15 of 64); inferior, 10.9% (7 of 64). With regard to errors in trajectory of insertion (63 screws), the sagittal angle was too large in 39.7% of cases (25 of 63) and too small in 4.8% (3 of 63), whereas the transverse angle was too large in 27.0% of cases (17 of 63) and too small in 28.6% (18 of 63). The

e350

www.SCIENCEDIRECT.com

distribution of levels in which errors occurred was as follows (Table 2): L2, 1.5% (2 screws); L3, 6.7% (9 screws); L4, 35.1% (47 screws); L5, 39.6% (53 screws); S1, 17.2% (23 screws). Analysis of Error for NS Method Among the 8 errors in the NS group, 7 were due to entry point errors (87.5%), and 1 was due to an angle error (12.5%). The trajectory of error in the entry point was as follows: medial, 71.4% (5 screws); superior, 14.3% (1 screw); inferior, 14.3% (1 screw). In the 1 case of error in the trajectory of insertion, the transverse angle was too large, leading to penetration of the medial cortex. The distribution of levels in which errors occurred was as follows (Table 3): L4, 37.5% (3 screws); L5, 25.0% (2 screws); S1, 37.5% (3 screws). DISCUSSION This study demonstrated improved one-time accuracy on using real-time 3D navigation compared with freehand insertion. The

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2019.04.151

ORIGINAL ARTICLE YUN-FENG XU ET AL.

Figure 4. Examples of inaccurate entry point selection. (A) Medial location in the axial view. (B) Superior

improved accuracy with navigation significantly reduced the error rate, including >4 mm penetration of the screw into the cortex of the pedicle. Therefore, compared with freehand insertion, intraoperative navigation can significantly improve the one-time accuracy of insertion of pedicle screws and should be used to avoid injury to the pedicle and surrounding structures from screw reinsertion. Pedicle screws provide rigid fixation for correction of segmental deformity and fusion,11,12 being widely used for the treatment of spinal trauma, degeneration, deformity, and tumors. However, the pedicle is a very narrow structure, with the width of the isthmus, from L1 to L5, in adults ranging from 6.4  1.6 mm to 17.7  2.7 mm, with a height of 12.7  2.1 mm to 13.6  1.4 mm.13 Moreover, the pedicle is in close proximity to the dural sac (at an average distance of 1.5 mm), the superior nerve root (5.3 mm), and the inferior nerve root (1.5 mm).14 Therefore, inaccurate placement of pedicle screws may damage vital tissues, leading to serious surgical complications, including pedicle fracture or penetration, nerve root or spinal cord injury, vascular injury, dural rupture, and epidural hematoma.3-5,9,15-17 Penetration of the cortex also reduces the pullout strength of the pedicle screw, compromising the fixation strength.18 Therefore, precise placement of a pedicle screw on the first try is necessary to guarantee the safety of surgery, requiring surgeons to appropriately select the entry point, the sagittal and transverse angles, the path length of the screw, and the size of the screw relative to the diameter of the pedicle.12 For lumbar pedicle screw, freehand insertion under fluoroscopy guidance is normally used, according to the Roy-Camile,1 Magerl,9 Krag,19 and Weinstein20 methods. The entry point and angle of insertion are selected based on anatomic landmarks and preoperative radiography as well as surgeon’s experience. The screw position can be adjusted intraoperatively, using marking

WORLD NEUROSURGERY 128: e347-e354, AUGUST 2019

PEDICLE SCREW SIMULATED INSERTION

location in the sagittal view. (C) Inferior location in the sagittal view.

needles under two-dimensional fluoroscopy, until a satisfactory position is achieved. However, such repeated adjustment can damage the bone, which lowers the maximal insertion torque and pullout strength of the screw,6,7 and result in serious complication due to injury to surrounding structures. Of note was our finding that the freehand technique led to 64 entry point errors, of which 16 were grade 3 errors. Moreover, the most common error was a medial position of the entry point (38 screws), which increases the risk of injury to the dura, nerve roots, and other important tissues. Additionally, reliance on anatomic parameters is difficult in cases of severe spinal degeneration, particularly in the presence of segmental rotational deformities that can produce significant alterations in the transverse angle and increase the risk of penetration of the cortex and possibly of the spinal canal, resulting in damage to important tissue and internal fixation failure. Moreover, in cases of significant lumbar spine degeneration, changes in the position of the vertebrae are possible between preoperative radiographs and the time of surgery. These factors increase the difficulty in one-time accurate placement of pedicle screws using the freehand approach. In our study sample, 63 screws were inserted along an incorrect path in the SFH group, with 35 errors due to an incorrect choice of transverse angle and 28 errors due to an incorrect sagittal angle. Intraoperatively, surgeons adjust the trajectory of the screw by detecting the difference in resistance between cancellous and cortical bone, this being based on a surgeon’s experience. In this process, the cortex may be inadvertently damaged, which may lead to surrounding tissue damage. Moreover, distinguishing the resistance between cortical and cancellous bone may be difficult, particularly in patients with severe osteoporosis in whom this difference in resistance is likely to be very small. Therefore, reliance on this subjective detecting may be risky.

www.journals.elsevier.com/world-neurosurgery

e351

ORIGINAL ARTICLE YUN-FENG XU ET AL.

PEDICLE SCREW SIMULATED INSERTION

Figure 5. Examples of inaccurate direction selection. (A) Transverse angle is too large in the axial view (medial slant). (B) Sagittal angle is too large in the

This underlines the fact that average values of recorded parameters from anatomic studies may not be sufficient to achieve accurate patient-specific pedicle screw placement. Intraoperative two-dimensional fluoroscopy can ameliorate these difficulties to some degree, but often does not provide sufficient clarity of the

sagittal view (superior slant). (C) Sagittal angle is too small in the sagittal view (inferior slant). (D) Transverse angle is too small in the axial view (lateral slant).

Table 2. Analysis of Errors with Simulated Freehand Pedicle Screw Placement

Entry point

Table 1. Comparison of One-Time Accuracy of Freehand and Navigation Methods Rajasekaran’s Grade

Freehand

Navigation

P Value Direction

Grade 0

236

362

Grade 1

21

7

Grade 2

83

1

Grade 3

30

0

Total

370

370

63.78%

97.84%

Accuracy

e352

www.SCIENCEDIRECT.com

<0.05

Deviation

Grade 1

Grade 2

Grade 3

Total

Medial

7

24

7

38

Lateral

1

2

1

4

Superior

2

6

7

15

Inferior

0

6

1

7

Medial

3

13

1

17

Lateral

4

12

2

18

Superior

3

13

9

25

Inferior

1

2

0

3

Both (entry point and direction)

1

3

3

7

Total

22

81

31

134

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2019.04.151

ORIGINAL ARTICLE YUN-FENG XU ET AL.

PEDICLE SCREW SIMULATED INSERTION

Table 3. Analysis of Errors with Navigation Simulated Pedicle Screw Placement Deviation Entry point

Direction

Case

Medial

5

Superior

1

Inferior

1

Medial

1

Total

8

local anatomy, requiring the surgeon to rely on innate knowledge of the local 3D anatomy. Moreover, an assistant with fluoroscopy expertise is needed to manipulate the fluoroscopy system intraoperatively, which makes it difficult to determine whether images are presented to the surgeons in a standard anatomic position (i.e., without angulation). This nonstandardization can be a significant factor in reducing the accuracy of placement of the pedicle screw. Weinstein et al.20 reported a high rate of both false-positive and false-negative alignments of two-dimensional fluoroscopy images with the local anatomy to guide screw placement, even when using C-arm fluoroscopy. In an evaluation of the accuracy of 131 lumbar pedicle screws placed under fluoroscopy guidance in 31 patients, using postoperative computed tomography, Castro et al.21 reported a penetration of the medial wall of the pedicle in 29.0% of screws. In a meta-analysis, Kosmopoulos and Schizas22 reported an accuracy of 87.3% for lumbar pedicle screw insertion without navigation. An additional concern is the exposure of surgeons and patients to radiation during intraoperative fluoroscopy for screw adjustment and position confirmation, with the exposure level being 10e12 times higher for spinal surgery than for other types of orthopedic surgeries.23 Longterm exposure to radiation can result in serious health risk to medical personnel.23-25 Therefore, there are many drawbacks to freehand pedicle screw placement under fluoroscopic guidance. Advancement in intraoperative real-time navigation has led to obtaining accurate 3D anatomic information and tracking of the location of surgical tools, allowing surgeons to confirm the optimal position of pedicle screws, which is very important for surgeons to improve one-time accuracy rate. Using real-time 3D navigation for the placement of 220 pedicle screws for minimally invasive spinal surgery in 46 patients, Villavicencio et al.26 reported an accuracy of 98.5%. Verma et al.27 performed a metaanalysis on the accuracy of pedicle screw placement; for

REFERENCES 1. Roy-Camille R, Saillant G, Mazel C. Internal fixation of the lumbar spine with pedicle screw plating. Clin Orthop Relat Res. 1986;203:7-17. 2. Gaines RW Jr. The use of pedicle-screw internal fixation for the operative treatment of spinal disorders. J Bone Joint Surg Am. 2000;82A:1458-1476. 3. Pihlajämaki H, Myllynen P, Böstman O. Complications of transpedicular lumbosacral fixation for

placement of 5992 screws in 1228 patients, they reported an accuracy rate of 93.3% with real-time navigation compared with 84.7% without navigation. This difference was statistically significant. In our study, the one-time accuracy for pedicle screw placement was 97.8% under navigation and 63.8% for freehand insertion (P < 0.05). Of note, even with navigation, 8 screws were misplaced, with 7 grade 1 errors and 1 grade 2 error, all of which were located at L4, L5, and S1 levels. Of these errors, 7 were pointof-entry errors, of which 5 screws were too medial, which occurred at L5 and S1 levels. In our study cohort, all procedures were performed via median incision, with dissection of the paravertebral muscles. Strong paravertebral muscles may limit the choice of lateral entry point, especially at the L5-S1 segment, which is located under the attachments of strong paravertebral muscles and is blocked by the wing of the iliac bone, making retraction difficult. Therefore, open pedicle screw placement may not be errorfree, even under 3D real-time navigation. Our study has some limitations. First, our SFH screw placement does not consider the effect of tactile feedback when freehand screws are placed. Typically, although surgeons would localize their starting point using anatomic landmarks, they cannulate the pedicle with a gearshift probe and rely on tactile feedback to gauge if their probe is still within bone. In addition, we did not consider the length of the screw, which is an important intraoperative parameter, as screws that are too short will reduce the pullout force of the screw and decrease the strength of fixation. In contrast, implanting screws that are too long may cause extension beyond the vertebral body and cause damage to surrounding organs, nerves, and blood vessels.15 Software within the navigation systems could be used to assist surgeons in selecting the optimal screw length for a patient. CONCLUSIONS We provide evidence of improved one-time accuracy using realtime 3D navigation compared with freehand insertion. This improved accuracy with navigation significantly reduced the error rate, including >4 mm penetration of the screw into the cortex of the pedicle. Therefore, compared with freehand insertion, intraoperative navigation can significantly improve one-time accuracy of insertion of pedicle screws and should be used to avoid injury to the pedicle and surrounding structures from screw reinsertion. ACKNOWLEDGMENTS We thank Yu-Zhen Sun, Jing Zhang, and Teng-Fei Ma, 3 secretaries at our hospital, for their outstanding work.

non-traumatic disorders. J Bone Joint Surg Br. 1997; 79:183-189. 4. Jutte PC, Castelein RM. Complications of pedicle screws in lumbar and lumbosacral fusions in 105 consecutive primary operations. Eur Spine J. 2002; 11:594-598. 5. Coe JD, Arlet V, Donaldson W, et al. Complications in spinal fusion for adolescent idiopathic scoliosis in the new millennium. A report of the Scoliosis Research Society Morbidity and Mortality Committee. Spine (Phila Pa 1976). 2006;31:345-349.

WORLD NEUROSURGERY 128: e347-e354, AUGUST 2019

6. Matityahu A, Hurschler C, Badenhop M, et al. Reduction of pullout strength caused by reinsertion of 3.5-mm cortical screws. J Orthop Trauma. 2013;27:170-176. 7. Marmor M, Mirick G, Matityahu A. Screw stripping after repeated cortical screw insertion—can we trust the cancellous “bailout” screw? J Orthop Trauma. 2016;30:682-686. 8. Tian W, Xu Y, Liu B, et al. Lumbar spine superiorlevel facet joint violations: percutaneous versus open pedicle screw insertion using intraoperative

www.journals.elsevier.com/world-neurosurgery

e353

ORIGINAL ARTICLE YUN-FENG XU ET AL.

3-dimensional computer-assisted navigation. Chin Med J (Engl). 2014;127:3852-3856. 9. Magerl FP. Stabilization of the lower thoracic and lumbar spine with external skeletal fixation. Clin Orthop Relat Res. 1984;189:125-141. 10. Rajasekaran S, Vidyadhara S, Ramesh P, Shetty AP. Randomized clinical study to compare the accuracy of navigated and non-navigated thoracic pedicle screws in deformity correction surgeries. Spine (Phila Pa 1976). 2007;32:E56-E64. 11. Chi JH, Lee R, Mummaneni PV. Concepts of surgical correction-segmental derotation and translation techniques. Neurosurg Clin N Am. 2007;18: 325-328. 12. Iampreechakul P, Chongchokdee C, Tirakotai W. The accuracy of computer-assisted pedicle screw placement in degenerative lumbrosacral spine using single-time, paired point registration alone technique combined with the surgeon’s experience. J Med Assoc Thai. 2011;94:337-345.

PEDICLE SCREW SIMULATED INSERTION

16. Samdani AF. Regarding “Delayed presentation of aortic injury by pedicle screws: report of two cases and review of the literature. J Vasc Surg. 2008;48: 1067-1068 [author reply: 1068].

24. Jones DP, Robertson PA, Lunt B, Jackson SA. Radiation exposure during fluoroscopically assisted pedicle screw insertion in the lumbar spine. Spine (Phila Pa 1976). 2000;25:1538-1541.

17. Di Silvestre M, Parisini P, Lolli F, Bakaloudis G. Complications of thoracic pedicle screws in scoliosis treatment. Spine (Phila Pa 1976). 2007;32: 1655-1661.

25. Stratakis J, Damilakis J, Hatzidakis A, Theocharopoulos N, Gourtsoyiannis N. Occupational radiation exposure from fluoroscopically guided percutaneous transhepatic biliary procedures. J Vasc Interv Radiol. 2006;17:863-871.

18. Acikbas SC, Arslan FY, Tuncer MR. The effect of transpedicular screw misplacement on late spinal stability. Acta Neurochir (Wien). 2003;145:949-954 [discussion: 954-955]. 19. Krag MH, Van Hal ME, Beynnon BD. Placement of transpedicular vertebral screws close to anterior vertebral cortex. Description of methods. Spine (Phila Pa 1976). 1989;14:879-883. 20. Weinstein JN, Spratt KF, Spengler D, Brick C, Reid S. Spinal pedicle fixation: reliability and validity of roentgenogram-based assessment and surgical factors on successful screw placement. Spine (Phila Pa 1976). 1988;13:1012-1018.

13. Lien SB, Liou NH, Wu SS. Analysis of anatomic morphometry of the pedicles and the safe zone for through-pedicle procedures in the thoracic and lumbar spine. Eur Spine J. 2007;16:1215-1222.

21. Castro WH, Halm H, Jerosch J, Malms J, Steinbeck J, Blasius S. Accuracy of pedicle screw placement in lumbar vertebrae. Spine (Phila Pa 1976). 1996;21:1320-1324.

14. Ebraheim NA, Xu R, Darwich M, Yeasting RA. Anatomic relations between the lumbar pedicle and the adjacent neural structures. Spine (Phila Pa 1976). 1997;22:2338-2341.

22. Kosmopoulos V, Schizas C. Pedicle screw placement accuracy: a meta-analysis. Spine (Phila Pa 1976). 2007;32:E111-E120.

15. Wegener B, Birkenmaier C, Fottner A, Jansson V, Durr HR. Delayed perforation of the aorta by a thoracic pedicle screw. Eur Spine J. 2008; 17(Suppl 2):S351-S354.

e354

www.SCIENCEDIRECT.com

26. Villavicencio AT, Burneikiene S, Bulsara KR, Thramann JJ. Utility of computerized isocentric fluoroscopy for minimally invasive spinal surgical techniques. J Spinal Disord Tech. 2005;18:369-375. 27. Verma R, Krishan S, Haendlmayer K, Mohsen A. Functional outcome of computer-assisted spinal pedicle screw placement: a systematic review and meta-analysis of 23 studies including 5,992 pedicle screws. Eur Spine J. 2010;19:370-375.

Conflict of interest statement: This study was funded by the Beijing Natural Science Foundation of China (Grant Nos. 7174311 and Z170001) and National Natural Science Foundation of China (Grant No. U1713221). Received 1 March 2019; accepted 17 April 2019

23. Rampersaud YR, Foley KT, Shen AC, Williams S, Solomito M. Radiation exposure to the spine surgeon during fluoroscopically assisted pedicle screw insertion. Spine (Phila Pa 1976). 2000;25: 2637-2645.

Citation: World Neurosurg. (2019) 128:e347-e354. https://doi.org/10.1016/j.wneu.2019.04.151 Journal homepage: www.journals.elsevier.com/worldneurosurgery Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2019 Elsevier Inc. All rights reserved.

WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2019.04.151