Posterior Short-Segment Instrumentation with Intermediate Screw versus Long-segment Instrumentation in Thoracolumbar Fracture Treatment

Background Data: Application of either longor short-segment instrumentation with intermediate screw to correct thoracolumbar spine fracture is still controversial. Both surgical techniques have their advantages and disadvantages. Even though early clinical results of these surgeries are usually satisfactory, a high failure rate and progressive kyphosis remain a concern. Purpose: To assess the efficiency and safety of short-segment pedicle screw fixation in thoracolumbar junction fracture with an intermediate screw placement compared to the more popular long-segment technique. Study Design: A retrospective comparative study Patients and Methods: A total of 61 patients were reported in this study and divided into two groups: Group I (33 patients) in which patients were surgically treated by short-segment instrumentation with intermediate screw placement at the fractured level and Group II in which all patients were surgically treated using long-segment instrumentation excluding the fracture level (28 patients). Outcome parameters included Visual Analogue Score (VAS) score of back pain and American Spinal Injury Association (ASIA) score of neurological status. Results: Although the VAS of back pain rapidly improved in Group I compared to Group II during the one-year follow-up, both had the same end results with no significant difference between both groups (P > 0.05). In Group I, the preoperative Cobb’s angle was 19.3 ± 3.7, which was corrected in the immediate postoperative X-ray to 6.8 ± 2.6 and maintained on the final follow-up at one year at 7.97 ± 1.67. In Group II, the mean preoperative Cobb’s angle was 18.7 ± 3.8, which was corrected in the immediate postoperative X-ray to 5.8 ± 1.6 and maintained on the final follow-up at one year at 7.89 ± 1.67. However, there was no statistically significant difference between both groups regarding angle correction (P > 0.05). In perioperative data, Group I showed less operative time (137.73 ± 16.96) than Group II (153.57 ± 19.525) with a highly significant statistical difference (P < 0.001). In addition,


INTRODUCTION
Surgical treatment of thoracolumbar fractures in neurologically intact patients has shown good outcomes. Those fractures are usually unstable and require surgical stabilization to maintain reduction and kyphosis correction and promote bony fusion and early mobilization. 20,21 The surgical approaches include the posterior approach, anterior approach, or a combination of both. The posterior approach is utilized more frequently because of its familiarity and smaller incision areas. 12,15 The application of either long-or short-segment construct in the management of thoracolumbar spine fracture is still controversial 15 . The longsegment construct is stiffer and provides good stabilization, thus minimizing the risk of posttraumatic kyphosis and screw/rods breakage. However, the main disadvantage of this technique is the loss of vertebral motion 13 . On the other hand, the short-segment construct has the advantages of a smaller surgical field, less blood loss, and reduced surgical time and hospital cost while preserving more motion segments. However, it increases the possibility of kyphosis and subsequently chronic pain and a higher incidence of instrumentation failure. 6,8,15 In order to avoid such insufficiency with the short-segment construct, procedures supporting the anterior column have been introduced such as vertebroplasty or kyphoplasty and screws placement at the fractured vertebra. 4,16 Many biomechanical studies 11,12 h ave demonstrated that the placement of intermediate screws at the fractured vertebra can significantly increase the stiffness and stability of the construct and reduce the stress on pedicel screws, hence preventing kyphosis progression and reducing the chance of implant failure. A recent systematic review 9 showed that the addition of intermediate screws results in significantly less postoperative pain, better radiological outcomes, and less implant failure. However, most differences in the outcome were small and the overall quality of the evidence was moderate to low. More studies on this topic are needed and should be directed more at patientreported outcomes and at which specific fracture types benefit the most from intermediate screws.
In this study, we aimed to compare the clinical and radiological outcomes after short-segment instrumentation with intermediate screw versus long-segment instrumentation in thoracolumbar fractures.

PATIENTS AND METHODS
This is a retrospective cohort study that included patients surgically treated for thoracolumbar spine fracture via posterior approach and transpedicular screw fixation between January 2016 and August 2020 at Ain Shams University Hospitals, Cairo, Egypt. The protocol of our study obtained approval from the research ethics committee of our institute, Faculty of Medicine at Ain Shams University (reference number: FWA 00006444). Being a retrospective study, patients´ consent for participation in the study and publication was not applicable. The study was conducted according to the WMA Declaration of Helsinki-Ethical Principles for Medical Research Involving Human Subjects.
Patients included were those with a single-level posttraumatic fracture at the thoracolumbar junction (T10-L2), with less than 50% canal compromise, local kyphosis with Cobb's angle 5 less than 25°, type A and B injuries according to AO Spine classification 19 , complete contact, clinical and radiological data, and minimum 12-month follow-up. Patients with more than one level of fracture, more than 50% canal compromise, or Cobb's angle of more than 25°, osteoporosis or pathological fractures, type C injuries according to AO Spine classification, and incomplete data or follow-up and intubated patients or those with disturbed conscious level were excluded. The included patients underwent either posterior short-segment instrumentation with intermediate screw placement at the fractured level (Group I) or long-segment instrumentation with two segments above and below the fracture vertebra, which were excluded from fixation (Group II). Allocation for either of the two techniques was according to the attending operating surgeon's preference. Preoperative Evaluation Clinical Evaluation. The full medical history of the patients was taken, and they were subjected to neurological assessment and general and local back examination. Preoperative evaluation of back pain severity was conducted according to the Visual Analogue Scale (VAS). 7 On the other hand, the neurological status was documented at the time of admission, according to American Spinal Injury Association (ASIA) impairment scale 2 to be compared with postoperative data. Radiological Evaluation. All patients were submitted for plain radiographs and multislice computed tomography (MSCT) scan of the spine. Radiographs were performed in both AP and lateral positions to assess the fracture characteristics and local kyphotic angle. The local kyphosis (Cobb) angle was calculated by measuring the angle between the upper endplate of the vertebra one level up and the lower endplate of the lower vertebra. MSCT scan assesses the type of fracture, degree of canal compromise, size of pedicles to choose the appropriate screw size and to assess the integrity of the pedicles at the fractured level for screw placement. Magnetic Resonance Image (MRI) of the spine was performed for most cases to determine the neural tissue compromise, intervertebral disc affection, and integrity of the posterior ligamentous complex. Outcome Measures: Data were collected from patients' medical records of our hospital, including the immediate postoperative period and then at 3, 6, and 12 months postoperatively. Patients' documented data included the following: assessment of pain score using VAS and neurological status using the ASIA scale. Radiological evaluation was performed by plain radiographs (AP and lateral) to assess the angle of correction and restoration of the sagittal plane, while MSCT scan to assess screws' position and decompression of the spinal canal. Surgical Technique: All patients were operated in a prone position under general anesthesia with endotracheal intubation. Standard midline posterior spine approach with bilateral subperiosteal muscle separation was performed in all cases. Under fluoroscopic guidance, transpedicular screw stabilization was performed in all patients who underwent either posterior bilateral short-segment pedicular screw instrumentation with bilateral intermediate screw placement at the fractured level (Group I) or bilateral long-segment pedicular screw instrumentation with two segments above and below the fractured vertebra with the exclusion of the fractured vertebra from fixation (Group II). Posterolateral fusion was performed using bone fragments and autologous bone graft through either decompressed bone or iliac crest graft. Single-level laminectomy was done only when there were neurological deficits due to cord contusion or with retropulsed segment occupying more than 40% of the spinal canal on CT scan. Examples and case demonstrations are demonstrated in Figures 1 and 2. Statistical Analysis: Data were analyzed using Statistical Program for Social Science (SPSS). Quantitative data The EGYPTIAN SPINE Journal were expressed as mean ± standard deviation (SD). Moreover, qualitative data were expressed as frequency and percentage. Independent sample t-test of significance, paired sample t-test of significance, and probability (P-value) were used. P-value < 0.05 was considered significant, P-value < 0.001 was considered highly significant, while P-value > 0.05 was considered insignificant.

RESULTS
This retrospective study reported 61 patients diagnosed with single-level thoracolumbar AO Spine type A or B fracture. Patients were divided into two groups: short-segment fixation (Group I) with a total number of 33 patients (54%) and longsegment fixation (Group II) of 28 patients (46%). There were 53 males (87%) and 8 females (13%) with a mean age at time presentation of 38 ± 9.2 (29-47) years, with no statistical difference between both study groups regarding demographic data (age and sex).

Journal
The most common mode of trauma in both groups was falling from height (62.3%), followed by road traffic accidents (RTA) (37.7%). The most common level of thoracolumbar vertebrae fracture for the entire cohort was L1 (37.7%), followed by L2 (29.6%) and D12 (23%), with the least incidence in D11 (6.6%). Regarding types of fractures, they were variable, with the most common fracture being burst fracture (Type A3 and A4 fractures according to AO Spine morphological classification system) (42.6%) and compression fracture (Type A2 fractures) (28%), while the remaining cases were type B fractures (29.4%). Both groups showed no statistical difference regarding the mode of trauma and type of fracture. With regard to neurological status, most cases in both Group I (81%) and Group II (82%) were ASIA E, with no statistically significant difference between the two groups in preoperative and postoperative ASIA scores (P-value > 0.05) ( Table 1). The perioperative data, including operative time and blood loss, number of patients who had laminectomy during fixation, hospital stay, and follow-up period, is summarized in Table 2. Statistical analysis showed that there was a highly significant statistical difference (P < 0.001) between both groups regarding operative time, with less time needed in Group I (137.73 ± 16.96) than that in Group II (153.57 ± 19.525). In addition, there was a significant statistical difference (P < 0.05) in terms of intraoperative blood loss, as Group I (270.45 ± 35.363) showed less blood loss than Group II (292.5 ± 34.788). However, other parameters showed no statistically significant difference between both groups ( Table 2). The mean preoperative VAS of back pain for Group I (short-segment construct) was 5.6 ± 2.1 and improved to 2.9 ± 1.47 at six months and 1.39 ± 0.56 at the one-year final follow-up. In Group II (long-segment construct), the mean preoperative VAS of back pain was 5.8 ± 2.2 and improved to 3.96 ± 1.43 at the six-month follow-up and 1.82 ± 0.48 at the end of the first postoperative year. Although the VAS of back pain rapidly improved in Group I compared to that of Group II during the one-year follow-up, both had nearly the same end results and there was no statistically significant difference between both groups regarding pain scales by the end of the first postoperative year with P-value In Group I, the mean preoperative Cobb's angle was 19.3 ± 3.7°, which was corrected in the immediate postoperative radiographs to 6.8 ± 2.6° and maintained on the final follow-up at one year at 7.97±1.67°. In Group II, the mean preoperative Cobb's angle was 18.7 ± 3.8°, which was corrected in the immediate postoperative X-ray to 5.8±1.6° and maintained on the final follow-up at one year at 7.89±1.67°. However, there was no statistically significant difference between short-segment fixation and long-segment fixation with regard to angle correction, using an independent sample t-test with P-value > 0.05 (Table 3). No intraoperative complications were reported in all our study cases of both groups, except for 2 cases of dural tear during laminectomy (one case in each group), that were managed intraoperatively by direct suturing of the dura with an application of muscle graft, while no postoperative CSF collection or leak was detected in these two cases. Regarding postoperative complications, only mild superficial wound infection was detected in 3 cases (one case in Group I and two cases in Group II), which was managed conservatively with local antibiotics and frequent dressings.

DISCUSSION
In our study, two patient groups underwent surgical treatment of thoracolumbar junction fracture, Group I (short-segment instrumentation with intermediate screw) and Group II (long-segment instrumentation). The demographic, clinical, and radiological properties were homogenous with no statistically significant difference. The measured outcomes of both groups, especially pain scores, kyphotic angle correction, and neurological Table 1. Preoperative data of the study patients.

Parameters
Group I Group II P-value  4 showed that shortsegment fixation with intermediate screws provides radiological results similar to longsegment posterior fixation in which two levels above and two levels below the fracture are fixated. However, a short-segment 6-screw construct could decrease the size of the immobilized segment and therefore retain more spinal motion than a longsegment fixation. 10 The present study confirms that intraoperative fracture reduction and correction of kyphotic angle in patients with thoracolumbar junction fractures can be achieved by short-segment instrumentation with the placement of intermediate screws in the fractured level with good clinical and radiological outcomes comparable to long-segment instrumentation. Moreover, short-segment instrumentation with intermediate screws can save two or more segments of vertebral motion compared to long-segment construct.
There are some limitations to our study. First, this study is a retrospective study done in a single institute. Second, patients' allocation was not random but according to the surgeon's preference. Third, the sample size is small and lacks long-term follow-up. Thus, a multicenter comparative study with a long-term follow-up is recommended to establish the obtained results. Recently, AO Spine developed patient-reported outcome measures specific for trauma. This AO Spine PROST is considered a valuable tool for use in the clinical setting and research. It can contribute to the reduction of ongoing controversies in spine trauma care when used in a global setting, allowing us to compare different treatment options in a valid and reliable fashion. 18 CONCLUSION Short-segment instrumentation with intermediate pedicle screws at the fractured level is an effective method for the treatment of thoracolumbar junction fractures with clinical and radiological outcomes like long-segment instrumentation. Moreover, short-segment instrumentation with intermediate screws can save two segments of vertebral motion, with less operative time and blood loss. However, a multicenter study with long-term follow-up is highly recommended.