A mathematical model for estimating the intraoperative lowest instrumented vertebra (LIV) tilt angle using preoperative supine left side-bending (LSB) radiographs in adolescent idiopathic scoliosis (AIS) patients with lenke type 1 and 2 non-AR curves

Purpose: To devise a mathematical model for estimating the intraoperative lowest instrumented vertebra (LIV) tilt angle using preoperative supine left side-bending (LSB) radiographs in adolescent idiopathic scoliosis (AIS) patients with Lenke type 1 and 2 (non-AR curves), and to review its clinical and radiological outcomes.

Methods: The mathematical model for the adjusted LSB LIV tilt angle (α) measured preoperatively, was expressed as the sum of preoperative LSB LIV tilt angle (x) and LIV displacement angle (y) (α = x + y). This model was validated through inter-rater and intra-rater analysis in Part I of the study. The α angle derived was applied to estimate the intraoperative LIV tilt angle. In part II of the study, clinical and radiological outcomes of 50 Lenke type 1 and 2 (non-AR curves) AIS patients operated using the α angle were reviewed. The difference between the intraoperative LIV tilt angle achieved (β) and the preoperative α angle was determined (∆LIV tilt angle = β-α).

Results: The α angle had excellent inter-rater and intra-rater intraclass correlation coefficients (0.982; 0.907). 42 patients had positive ∆LIV tilt angles whereas 8 patients had negative ∆LIV tilt angles. The overall incidence of distal adding-on (AO) was 10.0% (n = 5/50). Patients with negative ∆LIV tilt angles had a higher incidence of distal AO (n = 4/8, 50.0%) than patients with positive ∆LIV tilt angles (n = 1/42, 2.4%) (p = 0.001).

Conclusion: Achieving an intraoperative LIV tilt angle (β) greater than or equal to the preoperative α angle derived (β ≥ α) may help avoid the distal AO phenomenon.