Risk Factors for Postoperative Loss of Correction in Thoracolumbar Injuries Caused by High-Energy Trauma Treated via Percutaneous Posterior Stabilization without Bone Fusion

Background and Objectives: Percutaneous pedicle screws were first introduced in 2001, soon becoming the cornerstone of minimally invasive spinal stabilization. Use of the procedure allowed adequate reduction and stabilization of spinal injuries, even in severely injured patients. This decreased bleeding and shortened surgical time, thereby optimizing outcomes; however, postoperative correction loss and kyphosis still occurred in some cases. Thus, we investigated cases of percutaneous posterior fixation for thoracolumbar injury and examined the factors affecting the loss of correction. Materials and Methods: Sixty-seven patients who had undergone percutaneous posterior fixation for thoracolumbar injury (AO classifications A3, A4, B, and C) between 2009 and 2016 were included. Patients with a local kyphosis angle difference >= 10 degrees on computed tomography at the postoperative follow-up (over 12 months after surgery) or those requiring additional surgery for interbody fusion were included in the correction loss group (n = 23); the no-loss group (n = 44) served as the control. The degree of injury (injury level, AO classification, load-sharing score, local kyphosis angle, cuneiform deformity angle, and cranial and caudal disc injury) and surgical content (number of fixed intervertebral vertebrae, type of screw used, presence/absence of screw insertion into the injured vertebrae, and presence/absence of vertebral formation) were evaluated as factors of correctional loss and compared between the two groups. Results: Comparison between each group revealed that differences in the wedge-shaped deformation angle, load-sharing score, degree of cranial disc damage, AO classification at the time of injury, and use of polyaxial screws were statistically significant. Logistic regression analysis showed that the differences in wedge-shaped deformation angle, AO classification, and cranial disc injury were statistically significant; no other factors with statistically significant differences were found. Conclusion: Correction loss was seen in cases with damage to the cranial intervertebral disc as well as the vertebral body.