Synthetic polyurethane foams are frequently used in biomechanical testing of spinal medical devices. However, it is unclear what types of foam are most representative of human vertebral trabecular bone behavior, particularly for testing the bone-implant interface. Therefore, a study was conducted to compare polyurethane foam microstructure and screw pullout properties to human vertebrae. Cadaveric thoracolumbar vertebrae underwent microcomputed tomography to assess trabecular bone microstructure. Spine plate screws were implanted into the vertebral body and pullout testing was performed. The same procedure was followed for eight different densities (grades 5-30) of commercially available closed cell (CCF) and open cell foams (OCF). The results indicated that foam microstructural parameters such as volume fraction, strut thickness, strut spacing, and material density rarely matched that of trabecular bone. However, certain foams provided mechanical properties that were comparable to the cadavers tested. Pullout force and work to pullout for screws implanted into CCF grade 5 were similar to osteoporotic female cadavers. In addition, screw pullout forces and work to pullout in CCF grade 8, grade 10, and OCF grade 30 were similar to osteopenic male cadavers. All other OCF and CCF foams possessed pullout properties that were either significantly lower or higher than the cadavers tested. This study elucidated the types and densities of polyurethane foams that can represent screw pullout strength in human vertebral bone. Synthetic bone surrogates used for biomechanical testing should be selected based on bone quantity and quality of patients who may undergo device implantation.

Several approaches (anterior, posterior, lateral, and transforaminal) are used in lumbar fusion surgery. However, it is unclear whether one of these approaches has the greatest subsidence risk as published clinical rates of cage subsidence vary widely (7–70%). Specifically, there is limited data on how a patient's endplate morphometry and trabecular bone quality influences cage subsidence risk. Therefore, this study compared subsidence (stiffness, maximum force, and work) between anterior (ALIF), lateral (LLIF), posterior (PLIF), and transforaminal (TLIF) lumbar interbody fusion cage designs to understand the impact of endplate and trabecular bone quality on subsidence. Forty-eight lumbar vertebrae were imaged with micro-ct to assess trabecular microarchitecture. micro-ct images of each vertebra were then imported into image processing software to measure endplate thickness (ET) and maximum endplate concavity depth (ECD). Generic ALIF, LLIF, PLIF, and TLIF cages made of polyether ether ketone were implanted on the superior endplates of all vertebrae and subsidence testing was performed. The results indicated that TLIF cages had significantly lower (p < 0.01) subsidence stiffness and maximum subsidence force compared to ALIF and LLIF cages. For all cage groups, trabecular bone volume fraction was better correlated with maximum subsidence force compared to ET and concavity depth. These findings highlight the importance of cage design (e.g., surface area), placement on the endplate, and trabecular bone quality on subsidence. These results may help surgeons during cage selection for lumbar fusion procedures to mitigate adverse events such as cage subsidence.

In the thoracolumbar region, between 7% and 30% of spinal fusion failures are at risk for pseudarthrosis. From a biomechanical perspective, the nonconformity of the intervertebral graft to the endplate surface could contribute to pseudarthrosis, given suboptimal stress distributions. The objective of this study was to quantify the effect of endplate-graft conformation on endplate stress distribution, maximum Von Mises stress development, and stability. The study design used an experimentally validated finite element (FE) model of the L4-L5 functional spinal unit to simulate two types of interbody grafts (cortical bone and polycaprolactone (PCL)-hydroxyapatite (HA) graft), with and without endplate-conformed surfaces. Two case studies were completed. In Case Study I, the endplate-conformed grafts and nonconformed grafts were compared under without posterior instrumentation condition, while in Case Study II, the endplate-conformed and nonconformed grafts were compared with posterior instrumentation. In both case studies, the results suggested that the increased endplate-graft conformity reduced the maximum stress on the endplate, created uniform stress distribution on endplate surfaces, and reduced the range of motion of L4-L5 segments by increasing the contact surface area between the graft and the endplate. The stress distributions in the endplate suggest that the load sharing is greater with the endplate-conformed PCL-HA graft, which might reduce the graft subsidence possibility.