Background: Optimal implant densities and configurations for thoracic spine instrumentation to treat adolescent idiopathic scoliosis remain unknown. The objective was to computationally assess the biomechanical effects of implant distribution on 3D curve correction and bone-implant forces.
Methods: 3D patient-specific biomechanical spine models based on a multibody dynamic approach were created for 9 Lenke 1 patients who underwent posterior instrumentation (main thoracic Cobb: 43°-70°). For each case, a factorial design of experiments was used to generate 128 virtual implant configurations representative of existing implant patterns used in clinical practice. All instances except implant configuration were the same for each surgical scenario simulation.
Findings: Simulation of the 128 implant configurations scenarios (mean implant density=1.32, range: 0.73-2) revealed differences of 2° to 10° in Cobb angle correction, 2° to 7° in thoracic kyphosis and 2° to 7° in apical vertebral rotation. The use of more implants, at the concave side only, was associated with higher Cobb angle correction (r=-0.41 to -0.90). Increased implant density was associated with higher apical vertebral rotation correction for seven cases (r=-0.20 to -0.48). It was also associated with higher bone-screw forces (r=0.22 to 0.64), with an average difference between the least and most constrained instrumentation constructs of 107N per implant at the end of simulated instrumentation.
Interpretation: Low-density constructs, with implants mainly placed on the concave side, resulted in similar simulated curve correction as the higher-density patterns. Increasing the number of implants allows for only limited improvement of 3D correction and overconstrains the instrumentation construct, resulting in increased forces on the implants.
Keywords: Adolescent idiopathic scoliosis; Biomechanical modeling; Curve correction; Implant density; Implant distribution; Pedicle screw.
Copyright © 2016 Elsevier Ltd. All rights reserved.