RELATIONSHIP BETWEEN VERTEBRAL VESSELS AND CORTICAL PATH SCREWS IN CORTICAL TRANSFIXATION

Introduction: This study aims to evaluate the safety of using the cortical path screw with transfixation of the second cortical bone in relation to the vascular structures. Methods: This retrospective observational study (level of evidence: III, study of non-consecutive patients) analyzed data from the medical records of patients who underwent computed angiotomography scans of the abdomen at Hospital Mater Dei, measuring, in millimeters, the distance between the point of the lumbar vertebra considered the anatomical reference for the transfixation of the second cortical bone and the vascular structures adjacent to the spine (abdominal aorta, inferior vena cava, iliac vessels, segmental lumbar arteries). Results: Forty-eight patients were evaluated, with a mean age of 60 years (±8 years, 41-75), of whom 52% were male and 48% female. The measurements obtained between the pre-vertebral vessels and the possible screw exit points did not demonstrate contact in any of the vertebrae studied. Conclusions: The measurements obtained suggest the safety of using the cortical path screw transfixing the second cortical bone. Knowing the position of the vessels is essential to reduce intra-and postoperative complications related to spinal instrumentation. Level of evidence III; Study of non-consecutive patients.


INTRODUCTION
] This is the gold standard technique when it comes to instrumentation for the treatment of thoracolumbar spine injuries. 106] Although experimental, the bicortical fixation techniques emerged as an alternative to reduce fixation failures.Pedicle screws used bicortically in the thoracolumbar spine run up against a risk of vascular injury. 15he main objective of this study is to evaluate the relationship in millimeters, in the bicortical technique described by Resende, between the exit point of the cortical screw and the lumbar blood vessels (abdominal aorta, inferior vena cava, iliac blood vessels, lumbar segmental arteries).The secondary objective is to evaluate the same relationship considering the sex of the patient.11.4.1.1011-Manufacturer: Carestream, Rua Pequetita, 215, Bairro Vila Olímpia, São Paulo, SP, Brazil) for image analysis.
The simulation process began with the identification of the pedicle and pars interarticularis (coronal plane) to obtain the point of entry of the screw, located at five o'clock (left pedicle) and seven o'clock (right pedicle).This was followed by the elaboration of the path, via a line inclined at 25° caudo-cranially (sagittal plane) and 10° medio-laterally (axial plane) (Figure 2).
Once the simulation of the ideal cortical screw path had been performed bilaterally for each vertebra, the exit point in the second cortical layer was obtained using the software.This point was projected in three planes: axial, sagittal, and coronal (Figure 3).
The distance between the screw exit point in the transitional region of the superior vertebral plate and the lateral wall and the vessels was measured in millimeters (mm).In the coronal plane, the distances to the lumbar segmental arteries at levels L1 to L5 on the right and left were evaluated.In the axial plane, the distance to the aorta and vena cava at levels L1 to L5, and the right and left common iliac arteries and the right and left common iliac veins, at levels L4 and L5 were measured according to the anatomical variations of each patient.
The research data were processed using statistical program R, version 3.6.3(Manufacturer: R Development Core Team, Free Software Foundation -51 Franklin Street, Fifth Floor, Boston, MA 02110 USA).Analysis of the confidence interval (CI) of 95% was used to evaluate the means of the clinical variables.The Kolmogorov-Smirnov test verified the assumption of normality of the sample distribution with a p-value greater than 5%.In the bivariate analysis, the parametric Student´s t test was used to evaluate the differences between the sex of the patients and the clinical variables.

METHODS
This is a retrospective observational study that collected and analyzed data from the medical records of patients who had undergone spiral computed angiotomography (CTA) of the abdomen (Toshiba 160-channel Aquilion PRIME model TSX 303A -Manufacturer: Toshiba Medical System Corporation 1385, Shimoishigami Otawara-shi, Tochigi Japan), at the Hospital Mater Dei (HMD) during the period from January 2019 to June 2020.The study was approved by the Institutional Review Board (Identification: 40678720.4.0000.5128),and the Informed Consent Form was waived.
Male and female patients over 18 years of age submitted to CTA of the abdomen were included.Cases with a history of trauma, lumbar spine surgery, anatomical changes, or skeletal immaturity identified during the examination were excluded.
Simulation of the cortical screw trajectory followed Santoni's original description 14 and was performed using Carestream Picture Archiving and Communication System software (PACS -version

RESULTS
Forty-eight patients were included in the study, 25 (52%) of whom were male and 23 (48%) of whom were female.The mean age was 60 years (±8 years, minimum age 41 years, maximum age 75 years).
The mean, maximum, and minimum distances and standard deviations from the transfixation point of the bicortical screw to the aorta (Table 1), vena cava (Table 2), right (Table 3) and left (Table 4) lateral segmental arteries, the right (Table 5) and left (Table 6) common iliac arteries, and the right and left common iliac veins (Table 7) are summarized below.For the vena cava, there was a difference between the means.For the right common iliac vein, the p-value (p<0.05) was statistically significant when compared between the sexes.The data are summarized together in Figure 4.
As for the correlation between the clinical variables and sex, an association was identified in the variables Vena Cava (L1-L4), Aorta (L2-L4) and the common iliac artery.The men had a greater distance between the screw and the vascular structure than the women.The respective distances in millimeters for men and women were 35.6

DISCUSSION
This study was the first in the Brazilian literature to simulate and measure the distance from the screw tip to the vessel, using the technique described by Resende.In the measurements obtained, there was no contact between the prevertebral vessels and the possible screw exit points in any of the vertebrae studied.
The cortical trajectory screw can reach up to four points of contact with the cortical bone, namely, the entry point into the pars interarticularis, the inferomedial wall of the pedicle, the anterolateral wall of the pedicle, and the lateral cortex of the vertebral body, touching it, but not piercing it. 21he modification proposed by Resende used the insertion technique described by Santoni, combining the perforation of the second cortical bone, insertion of the screw, and transfixation.Biomechanical studies in swine vertebrae have shown a 46% increase in pullout force with the bicortical screw. 14,16Results identified according to a 95% confidence level, Kolmogorov-Smirnov normality test.SD: Standard deviation, IL: Inferior limit, SL: Superior limit.Results identified according to a 95% confidence level, Kolmogorov-Smirnov normality test.rcIv: Right Common Iliac Vein, lcIv: Left Common Iliac Vein, SD: Standard deviation, IL: Inferior limit, SL: Superior limit.The screw is directed towards the supero-lateral region of the vertebral body and shorter screws are used.Thus, with transfixation of the second cortical bone, the exit extremity region is in the cranial third and the posterior two thirds of the wall of the vertebral body wall.][23] Also corroborating the present findings, a study reviewed 664 cases of vertebral instrumentation and identified 15 (0.22%) cases of invasion of the vascular structure, reinforcing that injury to the large vessels of the thoracolumbar spine is rare and is preceded by a screw that touched or deformed the vessel. 17s for surgical technique, one study analyzed 65 CTA images of the lumbar spine to determine the best positioning of the bicortical screw in relation to the large lumbar vessels and considered that a distance of 5 or more mm between the screw and the vessel is safe. 15egarding the variability between patients, one author compared the path of the abdominal aorta in healthy individuals and in patients with degenerative lumbar scoliosis and concluded that the vessel follows its course without deviation or change in the distance from the spine.However, in patients with kyphosis, the distance may be increased by moving the spine away from the vessel.The author reported that the reduced elastic capacity of the tissues in elderly patients have a beneficial effect as the vessels remain in their anatomical position. 24TA was chosen for image analysis with the goal of increasing sensitivity in the identification of small-caliber vessels, such as the lumbar segmental arteries.They arteries originate on the posterior surface of the aorta and follow a dorsolateral course in the middle third of the vertebral body. 25The diameter of the segmental artery increases proportionally from L1 to L4, with its smallest diameter in L5, a fact that corroborates the technical difficulty encountered in identifying the arteries at this level, even using images captured after the administration of contrast and digitalized to facilitate it. 26 study of the morphology of the vena cava reported that its distance from the anterior cortex of the vertebral bodied tends to increase as it ascends through the abdomen and that the distance tends to be smaller in females, considering the degenerative changes suffered in the lumbar spine secondary to menopause. 27n our samples, aiming for the correct positioning of the cortical screw by following the Santoni technique, we observed that even after extending the screw during simulation, there was no contact with the vessels studied. 16It is important to emphasize that, despite the study cited, there is no consensus in the current literature regarding the safe distance from the screw to the vessel and several studies report that the screw in contact with the vessel can cause the formation of pseudoaneurysms. 28,29 noteworthy strong point of our study is that the measurements were taken in a systematic manner, using an objective method, easily reproducible by other research centers.Among the limitations, it should be noted that the measurements were taken by a single trained researcher and with data collection at a single center, limiting the sample analyzed.Another limitation was the exclusion of patients with any deformity or previous spine surgery, given that the bicortical screw technique is more recommended in the elderly who experience a greater number of changes and a predisposition for treatment failure.

CONCLUSIONS
In the measurements obtained through analysis of CTA exams, no contact between the prevertebral vessels and any of the possible cortical screw exit points in the transfixation of the second cortical bone was observed, corroborating the authors' expectations, and demonstrating that the modification to the technique is safe in relation to the vessel surrounding the lumbar spine.This is the first study to corroborate the safety of using this technique.Additional studies are needed to further specify safe surgical practices and the safest location for transfixation of the second cortical bone using the cortical trajectory screw technique.

Figure 1 .
Figure 1.Trajectory of the cortical screw proposed by Santoni.(1) point of entry (3 mm medially to the lateral edge of the pars), (2) in the inferomedial wall of the pedicle, (3) in the antero-lateral wall of the pedicle, and (4) in the lateral edge of the superior endplate of the instrumented vertebra.

Figure 2 .
Figure 2. (A) A line drawn at 25° caudo-cranially, (B) the pedicle located in the coronal plane and the point of entry, (C) point of entry in the sagittal plane, (D) (E) trajectory of the screw in the sagittal plane, (E)(F) point of exit in the axial plane.

Figure 3 .
Figure 3.The exit point was identified in the axial plane and the measurements for the iliac artery (D21), the vena cava (D1), and the right and left common iliac arteries were taken.The measurements for the segmental arteries were taken in the coronal plane.

Figure 4 .
Figure 4. Measurements and confidence intervals for all the vessels.

Table 1 .
Measurements in relation to the aorta in millimeters.
Results identified according to a 95% confidence level, Kolmogorov-Smirnov normality test.SD: Standard deviation, IL: Inferior limit, SL: Superior limit.

Table 2 .
Measurements in relation to the vena cava in millimeters.

Table 3 .
Measurements in relation to the right lateral segmental artery in millimeters.

Table 4 .
Measurements in relation to the left lateral segmental artery in millimeters.

Table 5 .
Measurements in relation to the right common iliac artery in millimeters.

Table 6 .
Measurements in relation to the left common iliac artery in millimeters.

Table 7 .
Measurements related to the right and left common iliac veins in millimeters.