Effect of the In Situ Screw Implantation Region and Angle on the Stability of Lateral Lumbar Interbody Fusion: A Finite Element Study

Objective To investigate the effect of the in situ screw implantation region and angle on the stability of lateral lumbar interbody fusion (LLIF) from a biomechanical perspective. Methods A validated L2‐4 finite element (FE) model was modified for simulation. The L3‐4 fused segment undergoing LLIF surgery was modeled. The area between the superior and inferior edges and the anterior and posterior edges of the vertebral body (VB) is divided into four zones by three parallel lines in coronal and horizontal planes. In situ screw implantation methods with different angles based on the three parallel lines in coronal plane were applied in Models A, B, and C (A: parallel to inferior line; B: from inferior line to midline; C: from inferior line to superior line). In addition, four implantation methods with different regions based on the three parallel lines in horizontal plane were simulated as types 1–2, 1–3, 2–2, and 2–3 (1–2: from anterior line to midline; 1–3: from anterior line to posterior line; 2–2: parallel to midline; 2–3: from midline to posterior line). L3‐4 ROM, interbody cage stress, screw‐bone interface stress, and L4 superior endplate stress were tracked and calculated for comparisons among these models. Results The L3‐4 ROM of Models A, B, and C decreased with the extent ranging from 47.9% (flexion‐extension) to 62.4% (lateral bending) with no significant differences under any loading condition. Types 2–2 and 2–3 had 45% restriction, while types 1–2 and 1–3 had 51% restriction in ROM under flexion‐extension conditions. Under lateral bending, types 2–2 and 2–3 had 70.6% restriction, while types 1–2 and 1–3 had 61.2% restriction in ROM. Under axial rotation, types 2–2 and 2–3 had 65.2% restriction, while types 1–2 and 1–3 had 59.3% restriction in ROM. The stress of the cage in types 2–2 and 2–3 was approximately 20% lower than that in types 1–2 and 1–3 under all loading conditions in all models. The peak stresses at the screw‐bone interface in types 2–2 and 2–3 were much lower (approximately 35%) than those in types 1–2 and 1–3 under lateral bending, while no significant differences were observed under flexion‐extension and axial rotation. The peak stress on the L4 superior endplate was approximately 30 MPa and was not significantly different in all models under any loading condition. Conclusions Different regions of entry‐exit screws induced multiple screw trajectories and influenced the stability and mechanical responses. However, different implantation angles did not. Considering the difficulty of implantation, the ipsilateral‐contralateral trajectory in the lateral middle region of the VB can be optimal for in situ screw implantation in LLIF surgery.