Proximal junctional failure after surgical instrumentation in adult spinal deformity: biomechanical assessment of proximal instrumentation stiffness

Study design Assessment of different proximal instrumentation stiffness features to minimize the mechanical proximal junctional failure-related risks through computer-based biomechanical models. Objective To biomechanically assess variations of proximal instrumentation and loads acting on the spine and construct to minimize proximal junctional failure (PJF) risks. Summary of background data The use of less-stiff fixation such as hooks or tensioned bands, compared to pedicle screws, at the proximal instrumentation level are considered to allow for a gradual transition in stiffness with the adjacent levels, but the impact of such flexible fixation on the loads balance and complications such as PJF remain uncertain. Methods Six patients with adult spine deformity who underwent posterior spinal instrumentation were used to numerically model and simulate the surgical steps, erected posture, and flexion functional loading in patient-specific multibody analyses. Three types of upper-level fixation (pedicle screws (PS), supralaminar hooks (SH), and sublaminar bands (SB) with tensions of 50, 250, and 350 N) and rod stiffness (CoCr/6 mm, CoCr/5.5 mm, Ti/5.5 mm) were simulated. The loads acting on the spine and implants of the 90 simulated configurations were analyzed using Kruskal-Wallis statistical tests. Results Simulated high-tensioned bands decreased the sagittal moment at the adjacent level proximal to the instrumentation (1.3 Nm at 250 N; 2.5 Nm at 350 N) compared to screws alone (PS) (15.6 Nm). At one level above, the high-tensioned SB increased the sagittal moment (17.7 Nm-SB vs. 15.5 Nm-PS) and bending moment on the rods (5.4 Nm and 5.7 Nm vs. 0.6 Nm) (p < 0.05). SB with 50 N tension yielded smaller changes in load transition compared to higher tension, with moments of 8.1 Nm and 16.8 Nm one and two levels above the instrumentation. The sagittal moment at the upper implant-vertebra connection decreased with the rod stiffness (1.0 Nm for CoCr/6 mm vs. 0.7 Nm for Ti/5.5 mm; p < 0.05). Conclusion Simulated sublaminar bands with lower tension produced smaller changes in the load transition across proximal junctional levels. Decreasing the rod stiffness further modified these changes, with a decrease in loads associated with bone failure, however, lower stiffness did increase the rod breakage risk