Fixed Lumbar Apical Vertebral Rotation Predicts Spinal Decompensation in Lenke III C Type Adolescent Idiopathic Scoliosis After Posterior Thoracic Correction and Fusion

Purpose:The aim of our study is identify preoperative radiographic parameters which could predict postoperative spinal decompensation in the frontal plane in Lenke III C type adolescent idiopathic scoliosis after posterior thoracic correction and fusion with third generation instrumentation systems. There is still controversy whether combined thoracic and lumbar curves require instrumentation of both curves or may be best treated with selective thoracic fusion and instrumentation. This review was performed because we noticed spinal decompensation in some patients with Lenke III C type curves treated with selective thoracic fusion and instrumentation.

Patients and Methods: In a retrospective manner we reviewed the radiographs of 36 patients with Lenke III C type AIS who had had posterior thoracic correction and fusion. In 6 patients the sagittal plane modifier was hypo-, in 28 normo- and in 2 patients hyperkyphotic. The patients were treated either with the Cotrel-Dubousset instrumentation or the Universal Spine System. Radiographs were evaluated in terms of frontal and sagittal plane balance, curve flexibility, and curve correction with a minimum follow up of two years. Postoperative spinal decompensation in the frontal plane was investigated with respect to preoperative radiographic parameters on standing upright AP, thoracic and lumbar supine side-bending as well as lateral standing radiographs. Spinal decompensation in the frontal plane was defined as plumbline deviation of C7 of more than 2 cm with respect to the centre sacral vertical line within two years postoperatively. Two groups of patients were analyzed.

Results:26 patients (72%) showed satisfactory frontal plane alignment by means of C7 plumb line deviation (group A, 1.2 cm to the left), whereas 10 patients (28%) showed spinal decompensation (group B: 2.7 cm to the left). Group differences were significant (p=0003). The two groups were found statistically equivalent in terms of preoperative C7 plumb line deviation (p=0.112, A: 0.8 cm, B: 0.7 cm to the left), thoracic Cobb angles (p=0.093, A: 56º, B: 62º), lumbar Cobb angles (p=0.115, A: 42º, B: 47º), lumbar curve flexibility (p=0.153, A: 78%, B: 67%); thoracic kyphosis (p=0.153) and lumbar lordosis (p=0.534) and age at operation (p=0.195), Significant group differences, however could be revealed for thoracic curve flexibility (p=0.03, A: 43%, B: 25%), lumbar apical vertebral rotation (LAVR) measured according to Perdriolle (p=0.02, A: 16º, B: 22º) and the percentage of derotation of lumbar apical vertebrae in lumbar supine side-bending films in comparison to AP upright standing radiographs (p=0.002, A: 49%, B: 27%). Average thoracic curve correction was 51% in group A and 41% in group B. Group differences were significant (p=0.05). Average lumbar curve correction was 34% in group A and 23% in group B (p=0.09). No group differences could be revealed for postoperative thoracic kyphosis and lumbar lordosis measurements. Logistic regression analysis with C7 plumb line deviation of more than 2 cm postoperatively as the dependent variable yielded the amount of derotation of lumbar apical vertebrae in lumbar supine side-bending films as the only risk-factor (p=0.04). Lumbar apical vertebral rotation >40% was shown to have a positive predictive value of 81% to determine satisfactory postoperative spinal balance. Lumbar apical vertebral rotation <40 % revealed a negative predictive value of 90% to determine postoperative spinal decompensation. The sensitivity our test is 90%.

Conclusion: Fixed lumbar apical vertebral rotation, measured in terms of the percentage of derotation of lumbar apical vertebrae on lumbar supine side-bending films in comparison to AP upright standing radiographs, provided the radiographic prediction of spinal decompensation in the frontal plane after posterior thoracic correction and fusion of Lenke III C type curves.