Rate and Factors Associated with Misplacement of Percutaneous Pedicle Screws in the Thoracic Spine.

Introduction: Thoracic percutaneous pedicle screw (PPS) fixation is technically challenging because of the complexity of the spinal anatomy involved. Furthermore, owing to the proximity of critical neurovascular structures, serious complications have been reported because of misplaced thoracic pedicle screws. Therefore, it is important to know the factors associated with the misplacement of thoracic PPS, but there have been few reports to date. Methods: The present study included 663 PPSs inserted from T4 to T12 in 127 patients. The accuracy of pedicle screw placement was assessed using computed tomography (CT) scans conducted within two weeks postoperatively. We compared the screws in the misplaced group (Group M) and the optimal placed group (Group O) for sex, age, body mass index, the consecutive surgery numbers, type of disease, instrumented level, laterality, the pedicle diameter, the inclination angle of the transverse process, and Hounsfield units (HU) at the base of the transverse processes of the instrumented vertebrae. Results: Screw misplacement was observed in 28 (4%) of 663 screws on CTs conducted within two weeks postoperatively. In univariate analysis, there was a statistically significant difference between Group M (n=25) and Group O (n=638) for insertion level, the pedicle diameter, and the HU value of the transverse process. In multivariate logistic regression analysis, T4-6 level (T4-6; odds ratio [OR]=12.083, 95% confidence interval [CI]: 3.219-45.355) and greater HU value at the transverse process (OR=1.009, 95% CI: 1.004-1.014) were identified as independent factors associated with the misplacement of thoracic PPS. Conclusions: The misplacement of thoracic PPS was observed in 28 (4%) of 663 screws. The vertebral level (T4-6) and greater HU values at the base of the transverse process were identified as independent factors associated with the misplacement of thoracic PPS.