Predicting Intraoperative Vertebral Rotation in Patients With Scoliosis Using Posterior Elements as Anatomical Landmarks

Abstract

A retrospective study. To predict intraoperative vertebral rotation in patients with scoliosis using posterior elements as anatomic landmarks. In patients with scoliosis, accurate intraoperative vertebral rotation measurements are needed to avoid spinal cord injury caused by pedicle screw misplacement. Generally, we predict vertebral rotation by anatomic landmarks, that is, posterior elements that can be seen during surgery, such as spinous processes, transverse processes, or laminae. However, correlation between vertebral rotation and these anatomic landmarks is unclear. Seventy-six vertebrae (T4-T12) of 17 patients with thoracic scoliosis were measured. Patients who had severe scoliosis (>90 degrees Cobb angle) were excluded from this study. We assessed apex and adjacent vertebrae using axial computed tomography scans taken with patients in a supine position. We measured the angle between the vertebral reference line (Ve), defined as the midline of the vertebral body, and various lines based on the following anatomic landmarks of posterior elements of the vertebrae: (a) The Spinous q process line (Sp), defined as the midline of the spinous process; (b) A bisector (Bi) of the bilateral lines that pass the depressions in the laminae to the medial apex of the transverse processes; (c) a line (Tr) perpendicular to a line which passes the bilateral transverse processes; and (d) a line (La) perpendicular to the line which passes the bilateral depressions in the laminae. The average Cobb angle in A-P radiographs was 68 degrees. The average angles between Sp, Bi, Tr, and La and the vertebral reference line were 13.6 degrees (range 0 degrees -29.2 degrees), 3.0 degrees (range 0 degrees -9.0 degrees), 2.5 degrees (range 0 degrees -8.1 degrees), and 4.4 degrees (range 0 degrees -11.9 degrees), respectively. The line Tr, (runs perpendicular to the line which passes the bilateral transverse processes) had the greatest correlation to the rotation of a vertebra. On the other hand, the line Sp (defined as the midline of the spinous process) was not sufficient for predicting rotation of a vertebra and can be disregarded during the intraoperative analysis and screw placement.