A pedicle screw construct gives an enhanced posterior correction of adolescent idiopathic scoliosis when compared with other constructs: myth or reality

Abstract

Study design: Tricenter retrospective cohort study of 72 patients who underwent posterior correction of Lenke 1 adolescent idiopathic scoliosis (AIS). Each center represented a single surgeon using only one type of construct.

Objective: Compare the initial postoperative and 2-year follow-up correction of Lenke 1 AIS curves, after accounting for the preoperative flexibility of the curves.

Summary of background data: There are multiple reports in literature of the enhanced posterior corrective ability of the pedicle screw in the treatment of AIS. Unfortunately, none of these reports took into account the preoperative flexibility of the curve. It stands to reason that rigid curves will not correct as much as flexible curves irrespective of the nature of the construct.

Methods: Groups were as follows: Group 1 (proximal and distal hooks and segmental intraspinous collar button wires), 24 patients; Group 2 (proximal hooks, distal screws, and apical sublaminar wires), 23 patients; and Group 3 (pedicle screws only), 25 patients. The postoperative correction percentage was expressed as a ratio of the preoperative flexibility and was termed Cincinnati correction index (CCI). Mathematically speaking the CCI equals (postoperative correction/preoperative erect Cobb angle) divided by (supine bending preoperative correction/preoperative erect Cobb angle). The postoperative sagittal correction was also measured.

Results: CCI 2 (at 2-year follow-up) for Group 1 was 1.71, for Group 2 was 1.34, and for Group 3 was 1.41. The differences were not statistically significant. Within Group 1, however, there was a statistically significant difference between CCI (1.95) and CCI 2 (1.71), indicating a statistically significant loss of correction over 2 years. However, in terms of absolute values, there was only a 4 degree (average) difference between the initial and the 2-year postoperative Cobb measurement, rendering the loss off correction clinically insignificant. No such statistically or clinically significant differences were noted within Groups 2 and 3. Group 1 and Group 3 constructs further lordosed the curve by 8 degrees and 11 degrees, respectively, whereas the Group 2 construct retained or marginally increased the preoperative kyphosis.

Conclusion: The Group 3 (pedicle screw only) construct did not give an enhanced correction of Lenke 1 AIS, when the preoperative flexibility of the curve was considered. Also, contrary to popular belief, the pedicle screw construct has a lordosing effect on the thoracic spine. Therefore, we think that there is no significant advantage in using a relatively expensive pedicle screw construct in the correction of Lenke 1 AIS.