Wednesday, September 26, 2018 7:35 AM–9:00 AM ePosters: P34. Axial rotation correction in adolescent idiopathic scoliosis with pedicle screw construct

Abstract

BACKGROUND CONTEXT Adolescent idiopathic scoliosis (AIS) is a three-dimensional deformity characterized by tilt and rotation of the spine. As rib prominence is related to the vertebral rotation, derotation of the spine to reduce deformity is desirable. At our institution, posterior spinal fusions for AIS are performed with the use of intraoperative 3D fluoroscopic imaging with navigation. Scans of the entire operative curve are taken before and after instrumentation, resulting in tomographs in which the accuracy of pedicle screw placement and the change in axial rotation can be measured. METHODS Posterior spinal fusion with pedicle screw and rod instrumentation with the use of intraoperative 3D fluoroscopic imaging and navigation was performed. Pedicle screws were placed using navigation, following standard posterior approach. The most cephalad level of instrumentation was secured with hooks. After placement of pedicle screws and a convex rod, derotation was performed using uniaxial screws with tubes over screwdrivers. A second scan was performed to ensure the appropriate position of all screws, allowing measurement of derotation intraoperative. Data collection: Institutional review board approval was obtained for retrospective review of patient charts and images, and for the creation of a patient database. All patients who received posterior spinal fusion for AIS between 2008 and 2010 were included in the database. Inclusion criteria for this review were limited to patients with AIS and with greater than 2years follow-up. This resulted in a total of 54 patients for inclusion. Imaging taken before and after instrumentation and rod de-rotation maneuver was reviewed. For each patient, apical and neutral vertebrae were identified on the standing PA radiograph. The angle of rotation (RA sag) was then measure on the axial 3D images for the apical and neutral vertebrae, by the method described by Aaro and Dahlborn. The difference between the RA sag of these two vertebrae was defined as the axial rotation measured on the pre- and post-instrumented axial images. PA standing scoliosis view was also taken on fourth postoperative day. Cobb angles were measured from each of these radiographs and submitted for analysis. Statistical analysis: Mixed linear models were used to predict axial rotation correction and calculate LSMEANS with 95% confidence intervals. Covariates included gender and continuous age. Logistic regression was employed to estimate the odds of non-normal status postsurgery when compared to normal with presurgery standing Cobb as the independent variable. The one patient with a Lenke classification of five was removed for this analysis. All analyses were conducted using the statistical package SAS, version 9.3. (SAS Institute Inc., Cary, North Carolina) RESULTS Axial rotation correction overall was a 28.7% improvement postoperatively. This change meets statistical significance (p=.0202) when adjusting for age and gender. The change in axial rotation is significantly negatively correlated with age, indicating the change in axial rotation tends to decrease as age increases. Discussion: Our study was able to show a significant improvement in axial rotation in our group of patients. The manner in which the images were taken provides, by our estimation, a standardized method of image-gathering that is most reproducible and has the least amount of variability. General anesthesia and positioning alone have shown to decrease the magnitude of measured curves. Lee et al. showed a decrease of as much as 74% in Cobb angles and 25.7% rotational measurement (using Pedriolle torismeter) by placing a patient under anesthesia and on the operative table. Our study has the unique advantage of having all images taken after dissection and preparation of the spinal elements has occurred, removing an additional layer of variability. CONCLUSIONS Derotation maneuver using convex pedicle screws significantly reduced axial rotation of the spine. This study validates one technique of derotation. Vertebral derotation resulting from correction maneuver can be quantified, making comparisons of different techniques and instrumentation possible. Future studies using different instrumentations and derotation techniques should make comparisons possible.