Abstract
INTRODUCTION: High rates of postoperative rod fracture at the lumbosacral junction have been reported after long spinopelvic fixation. In the prevention of rod fractures, supplemental accessory rods (ARs) and lateral interbody fusion are commonly used and reportedly effective. However, the optimal AR placement to mitigate rod stress at the lumbosacral junction is unclear. We therefore used a synthetic bone model and a finite element model concurrently to address their respective shortcomings. METHODS: Both models included the lumbar spine (L1-L5) and the pelvis, and were instrumented with a screw and rod system and lateral interbody fusion cages to closely resemble actual surgical procedures. The four different constructs were: two primary rods (PRs) without ARs, PRs+contoured long ARs, PRs+short ARs, and PRs+straight long ARs. In our synthetic model, we applied vertical load to the constructs and measured rod strain at L5-S1 using strain gauges. We calculated a mean value of the five rods in each construct. In our finite element model, we measured maximum principal stresses at L5-S1 after the application of flexion/extension, lateral bending, and axial rotation loads. RESULTS: In our synthetic bone model, there was significant reduction of rod strain by 52% in PRs+straight long ARs compared with PRs without ARs (p=0.023). A reduction of average principal stress in the finite element model was observed in PRs+straight long ARs by up to 44.2% (highest against flexion load) compared with PRs without ARs. CONCLUSIONS: We conducted concurrent biomechanical analyses using a synthetic bone model and a finite element model. We recommend straight long ARs to prevent rod fracture at the lumbosacral junction in long spinopelvic fixation.