Comparison of the Therapeutic Effects of Short-Segment Posterior Fixation With Monoaxial Pedicle Screws or Polyaxial Pedicle Screws via Injured Vertebra on Thoracolumbar Fracture

Purpose: To evaluate the therapeutic effects of short-segment posterior xation with monoaxial pedicle screw or polyaxial pedicle screw via injured vertebra on thoracolumbar fracture. Methods: All patients who underwent short-segment posterior xation with monoaxial pedicle screws or polyaxial pedicle screws in the injured vertebra of a thoracolumbar fracture (T12-L2). The clinical and radiological data such as the correction of deformity, sagittal proe and record of the perioperative morbidity of the patients were analyzed. Results: There were 63 patients (21 males and 42 females) with an average age of 44.7 years and were categorised into two groups: monoaxial pedicle screws group (group A) and polyaxial pedicle screws group (group B). There were no signicant differences in age, gender, fracture site, TLISS Score, ASIA Score, AO Classication, hospital stay, Injury-to-operation intereval, and associated injury between the two groups (P>0.05). Howere, compared with group B, the injury vertebral endplate centre ratio signicantly increased postoperatively and at the nal follow-up (P<0.05) in group A. Conclusion: Short-segment posterior xation with monoaxial or polyaxial pedicle screws via the fracture level for thoracolumbar fracture can achieve kyphosis correction, reduce sagittal alignment correction failure, and maintain anterior vertebral height. The insertion of monoaxial pedicle screws at the fracture level after thoracolumbar vertebral fracture has a ick up ffect on the central vertebral body of the injured vertebrae, which is benecial to the recovery of the vertebral endplate.


Introduction
Spine fractures are common, accounting for about 14% of all fractures [1], and the thoracolumbar spine is reportedly the most common site of spinal fractures due to trauma, accounting for 60-70% of spine fractures [2,3]. Thoracolumbar fractures can cause an unstable spine leading to pain, motion disability, and even full paralysis [4]. Posterior pedicle screw xation, to maintain the stability of the spine and avoid further nerve damage, has been popular globally [5]. Posterior pedicle screws can be inserted in either the fractured vertebra itself or the adjacent vertebrae. Many orthopaedic specialists believe that it is better to insert pedicle screws at the level of the fracture because the reinforcement of xation can help to improve kyphosis correction and biomechanical stability. Inserting monoaxial or polyaxial pedicle screws at the level of vertebral fractures, as well as segmental xation of burst fractures, improves biomechanical stability. [6,7] In addition, several studies have shown that enhanced fracture-grade screws can improve the biomechanical stability of the construct [8][9][10][11], and some authors [7,12,13] have also suggested that pedicle screw xation combined with an intermediate screw in the fractured vertebra improves biomechanical stability with better reduction, less loss of correction, fewer instrument failures, and comparable or better clinical outcomes. However, it has not yet been ascertained whether inserting a monoaxial or polyaxial pedicle screw at a vertebral fracture yields better clinical results. Our study, therefore, compared the e cacy of inserting a monoaxial or polyaxial pedicle screw at a vertebral fracture, aiming to provide a reference for the clinical options for thoracolumbar fractures.

General information
This retrospective study included patients who underwent short-segment posterior xation with monoaxial or polyaxial pedicle screws in the injured vertebra of a thoracolumbar fracture and removal of the internal xation in our hospital between June 2012 and December 2018. The enrolment criteria were as follows: (1) thoracolumbar single segmental fracture; (2) short-segment posterior xation with monoaxial or polyaxial pedicle screw via the injured vertebra; (3) patient information is complete, imaging data is complete and fracture healing, pedicle screw removal. Patients with signi cant osteoporosis, endocrine system disease, vertebral tumour, tuberculosis, ankylosing spondylitis, and other destructive vertebral structural diseases were excluded from the study.Of these patients, 42 were male and 21 were female, with an average age of 44.7 years. The patients' thoracolumbar injury classi cation and severity scale (TLICS) [14] scores were 5-7 points and their American Spinal Injury Association (ASIA) [15] scores were C-E on rst admission. According to the type of pedicle screw insertion at the vertebral fracture, the patients were categorised into two groups: group A had monoaxial pedicle screws, and group B, polyaxial pedicle screws. The fractures were classi ed according to Arbeitsgemeinschaft für Osteosynthesefragen (AO) [16] Spine Classi cation.All the patients included in the study had a minimum of 1 year of follow-up after surgery. Record of the perioperative parameters: injury-to-operation interval (days), associated injury, and hospital stay (days) on rst admission. Radiographic data: thoracolumbar anterior-posterior (AP), were taken preoperatively, three days postoperatively, and on removal of internal xation. Measurements taken include the height of the leading edge of the vertebral body adjacent to the injured vertebra and that of the injured vertebra, the height of the centre of the injured endplate and injured vertebra, the Cobb angle, and the vertebral changes for the two internal xation methods between three days postoperatively and when the internal xation device was removed. We also assessed whether the internal xation fractured or loosened.
Measurement and calculation of prevertebral height ratio ( Figure 1) The sum of the heights of the leading edges of the vertebral bodies above and below the injured vertebra was compared with twice the height of the leading edge of the injured vertebra.
Measurement and calculation of injured vertebral endplate centre ratio ( Figure 2) The sum of the heights of the upper and lower endplates of the injured vertebrae was compared with twice the height of the centre of the injured endplate.
Measurement and calculation of kyphosis of the injured vertebrae ( Figure 3) The angle between the upper endplate of the injured vertebra and the extension of the lower endplate of the injured vertebra.

Statistical methods
The statistical software SPSS 20 was used to analyse the data of the monoaxial and polyaxial pedicle screw groups preoperatively, three days postoperatively, and at the removal of internal xation. The measurement data were expressed as mean ± standard deviation (x ± s). Data between groups were compared using an independent sample t-test. P<0.05 was considered statistically signi cant.

Results
General information: There were 33 patients in group A (24 males and nine females), and 30 patients in group B (18 males and 12 females). X 2 test analysis (P>0.05) revealed no signi cant intergroup differences in terms of. In addition, there were no signi cant differences in hospital stay, injury-tooperation interval, and associated injuries between the two groups (P>0.05) ( Table 1). AO Classi cation and Fracture level: Of the 33 patients in group A, there were ten cases of A1 type, two cases of A2 type, and 12 cases of A3 type. There were six cases of B1 type and three cases of B2 type.
The fracture sites were T12 in two cases, L1 in 17 cases and L2 in 14 cases. Of the 30 cases in group B, the fractures were classi ed into six cases of A1 type, four cases of A2 type, nine cases of A3 type, six cases of B1 type, and ve cases of B2 type. The fracture sites were zero cases of T12, 21 cases were L1, and nine cases were L2. According to the X 2 test analysis, p>0.05; therefore, there were no statistically signi cant differences (Table 1).
TLISS Score: In the Thoracolumbar Injury Severity Score (TLISS) group, in group B, 17 people scored ve points, nine people scored six points, and seven people scored seven points. In group A, 16 people scored ve points, 11 people scored six points, and three people scored seven points. According to the X 2 test analysis, p>0.05; there were no statistically signi cant differences (Table 1).
ASIA Score: In the ASIA score group, there were seven people evaluated as C, 17 evaluated as D, and nine evaluated as E in group A; and there were three people evaluated as C, 16 evaluated as D, and 11 evaluated as E in group B. According to the X 2 test analysis, p>0.05; therefore, there were no statistically signi cant differences (Table 1) Radiographic data: The radiographic data from surgery shows that the prevertebral height ratio and the injured vertebra Cobb angle between the two groups were not signi cantly different in preoperative, postoperative (three days after surgery), and nal follow-up (removal of internal xation) evaluated by ttest analysis (P>0.05). In addition, the injured vertebral endplate centre ratios between the two groups were not signi cantly different preoperatively (P>0.05). However, the injured vertebral endplate centre ratio between the two groups were signi cantly different in the postoperative and nal follow-up checks (P<0.05) ( Table 2).

Discussion
Research has con rmed the bene cial therapeutic effect of short-segment posterior xation with pedicle screws in the injured vertebra as treatment for thoracolumbar fractures [6][7][8][9][10][11][12][13]. Biomechanically, the screws at the fracture level function as a push point with an anterior vector, creating a "lordorizing" force that restores the anterior vertebral height and corrects the kyphosis [17]. Some clinical studies have shown that inserting monoaxial or polyaxial pedicle screws at the fracture level could achieve better kyphosis correction, less sagittal alignment correction failure, and better maintenance of anterior vertebral height [7,8,10,12,18]. Short-segment posterior xation can be achieved by the insertion of either a monoaxial or a polyaxial pedicle screw into the injured vertebra; however, no studies have compared the e cacy of the two options.
Our retrospective study shows that both options markedly improved the outcome of patients postoperatively. This was re ected in considerable improvements whether analysed according to sex, age, fracture injury classi cation, fracture site, TLISS score, ASIA Score, AO classi cation, hospital stay, injury-to-operation interval, or associated injury after treatment. However, no statistically signi cant differences were observed between the two groups p>0.05. These results support the proposition that inserting either a monoaxial or polyaxial pedicle screw in the injured vertebra is an effective surgical treatment.
In our study, we measured and calculated the prevertebral height ratio, injured vertebral endplate centre ratio, and the kyphosis of the injured vertebrae. As shown in Table 2, the prevertebral height ratio and the kyphosis of the injured vertebrae postoperatively and at the nal follow-up were greatly improved. This con rms that inserting a monoaxial or polyaxial pedicle screw in the injured vertebra can correct the deformity through vertebral endplate augmentation with its buttress effect (bending force) as with the rod-sleeve method, which was until recently commonly used in spinal instrumentation [7,[19][20][21][22]. No statistically signi cant differences were observed, however, between the two groups. The injured vertebral endplate centre ratios in the postoperative and the nal follow-up were greatly improved, and the correction effect of group A was better than that of group B. In addition, signi cant differences were observed between the two groups (p<0.05). Many reports [17] have proposed that the screw at the fracture level may provide a mass effect that prevents the vertebra from collapsing. It may also help to support the anterior column, which is vital for the stability of the construct. However, it is not clear why there is a better correction effect in group A compared to group B.
Polyaxial pedicle screw heads are vulnerable to fatigue failure; the region between the screw head and shaft has been found to fail rst in many biomechanical studies [23][24][25]. Further, the use of additional intermediate monoaxial pedicle screws may result in a stiffer construct and a reduced level of von Mises stress on the pedicle screws than on the polyaxial pedicle screw models [26]. In addition, the monoaxial screws can slap the collapsed endplate, reset the endplate fracture, maintain the reduction, reduce the degeneration of the intervertebral disc injury, and perhaps better maintain the stability of the spine; thus, reducing the incidence of back pain. Furthermore, the head of the polyaxial pedicle screw is movable and cannot support the injured vertebral body. The monoaxial pedicle screws inserted at the fracture level showed higher stability in exion and extension than the similarly placed polyaxial pedicle screws [27].
Moreover, in the operation as shown in Figure 4), we selected the midpoint of the transverse process on both sides of the injured vertebral body and the intersection of the anterior line of the superior articular process as the needle point. We then placed two locating needles, and kept them at an angle of about 30 degrees to the sagittal plane, the positioning needle, and the cross-section form an angle; the vertebral bodies above and below the injured vertebrae are in the direction of the traditional needle insertion, and the sagittal plane is formed at an angle of about 15 degrees, so that the positioning needle is parallel to the cross-section; The vertebral pedicle screw installed in the manner of needle insertion can support the sling and the role of the vertebral endplate. Thus, our results provided a reference for the treatment of thoracolumbar fractures with severe vertebral endplate damage using invasive intermediate monoaxial pedicle screw xation.
This study has many limitations. First, the number of patients included in the study was small, and this was a retrospective study. Second, a selection bias may exist because this study included patients referred to our teaching hospitals. Third, it is necessary to discuss several factors including different patient conditions, the variability in bone density, muscle forces, vertebral size, the length and diameter of pedicle screws, and the degree of joint degeneration of the body.

Conclusions
In conclusion, based on our results, short-segment posterior xation with monoaxial or polyaxial pedicle screws at the fracture level of a thoracolumbar fracture achieved kyphosis correction, reduced sagittal alignment correction failure, and maintained anterior vertebral height. We found short-segment posterior xation with monoaxial pedicle screws at the fracture level of a thoracolumbar fracture to be better than polyaxial pedicle screws in improving the height of the centre of the injured endplate. After a thoracolumbar vertebral fracture, the insertion of a monoaxial pedicle screws at the fracture level has a ick up effect on the central vertebral body of the injured vertebrae, which is bene cial to the recovery of the vertebral endplate.   Injury vertebral endplate center ratio(%)** :The sum of the heights of the upper and lower endplates of the injured vertebrae is compared with the height of the center of the injured endplate. Figure 1 The sum of the heights of the leading edges of the vertebral bodies above and below the injured vertebra was compared with twice the height of the leading edge of the injured vertebra((a+c)/2b). The angle(α) between the upper endplate of the injured vertebra and the extension of the lower endplate of the injured vertebra.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.