Maintenance of Derotation in Adolescent Idiopathic Scoliosis: a Novel Technique Measuring Postoperative Vertebral Rotation by Pedicle Screw Position.

BACKGROUND: Current methods of measuring vertebral rotation by plain radiographs rely on anatomic landmarks that are not present in the postoperative spine or require advanced imaging. Furthermore, there are few studies on the incidence of crankshaft with modern pedicle instrumentation.

QUESTIONS/PURPOSES: We sought to (1) describe and validate a method of vertebral rotation measurement using plain radiographs and (2) measure postoperative rotation in a series of patients treated for adolescent idiopathic scoliosis.

METHODS: Patients with adolescent idiopathic scoliosis treated with surgery over a 6-year period were reviewed. Patients with computed tomography (CT) scans and radiographs within 60 days of another were included. Vertebral rotation was calculated by radiographic measurements and measured directly by CT scan. As an internal control, patients with two apical pedicle screws on all radiographs were analyzed. Rotation was measured for all patients with at least 1 year of radiographic follow-up.

RESULTS: Three thousand five hundred fifty-two instrumented spinal levels in 308 consecutive patients were reviewed. Ten patients with 93 screws were analyzed by CT and radiographs. The average discrepancy between computed tomography (CT) and radiographs was 3.3 ± 1.9°, with 81.7% (76/93) within 5°. Intra- and inter-rater reliabilities for measured axial rotation were excellent (intra-class correlation coefficient (ICC) = 0.879 and 0.900, respectively). One hundred swventy-eight patients were eligible with an average follow-up of 2.3 ± 1.2 years; 84.8% (151/178) had screw(s) visible on all images at the major curve apex. The average postsurgical rotation was 3.5 ± 2.9°; 19.2% (29/151) were measured to have a rotation over 5°, and 4.0% (6/151) demonstrated a rotation over 10°. Only 4.6% (7/151) of patients demonstrated a postoperative Cobb angle change over 10°.

CONCLUSIONS: Most major curves have apical pedicle screw instrumentation that can be followed by radiographs alone to measure rotation. Vertebral rotation measurement requires only plain radiographs and is a more sensitive determination for subtle postoperative crankshaft than change in Cobb angle.