Evaluation of crankshaft phenomenon after posterior fusion for early onset scoliosis using an inverse trigonometric function: a multicenter retrospective cohort study.

ABSTRACT

PURPOSE: The crankshaft phenomenon (CSP) is a corrective loss after posterior surgery for early onset scoliosis (EOS). However, an accurate method for CSP evaluation has yet to be developed. In this study, we evaluated pedicle screw (PS) length and rotation angle using an inverse trigonometric function and investigated the prevalence of the CSP. METHODS: Fifty patients from nine institutions (mean age 10.6 years, male/female ratio 446) who underwent early definitive fusion surgery at ≤ 11 years of age were included. The rotation angle was calculated as arctan (lateral/frontal PS length) using radiography. Measurements were taken at the apex and lower instrumented vertebra (LIV) immediate, 2-, and 5-year postoperatively. CSP was defined as a rotation angle progression ≥ 5°. We divided patients into CSP and non-CSP groups and measured the demographic parameters, Risser grade, state of the triradiate cartilage, major coronal Cobb angle, T1-T12 length, T1-S1 length, and presence of distal adding-on (DAO). We compared these variables between groups and investigated the correlation between the measured variables and vertebral rotation. Logistic regression analysis investigated factors associated with CSP. RESULTS: The rotation angle progressed by 2.4 and 1.3° over 5 years for the apex and LIV, respectively. CSP occurred in 15 cases (30%), DAO in 11 cases (22%), and CSP and DAO overlapped in 4 cases (8%). In the CSP group, the T1-T12 length was low immediate postoperatively. The rotation angle was negatively correlated with preoperative height (r = - 0.33), T1-T12 length (r = - 0.35), and T1-S1 length (r = - 0.30). A lower preoperative T1-T12 length was associated with CSP (odds ratio 0.996, p = 0.048). CONCLUSIONS: CSP occurred in 30% of patients with EOS who underwent definitive fusion. The presence of CSP was associated with a lower preoperative T1-T12 length. LEVEL OF EVIDENCE Diagnosis, level IV.