Impact of bone quality on the performance of integrated fixation cage screws.

Abstract

Commercially available lumbar integrated fixation cages (IFCs) have variable designs. For example, screw-based designs have up to four screws inserted at different locations across the vertebral end plate as well as at different angles in the sagittal and transverse planes. This is important as end plate and trabecular bone quality may vary across the vertebra and may affect the screw's fixation ability, particularly if bone purchase at the bone-screw interface is poor. This study aimed to evaluate whether variations in local bone quality surrounding IFC screws inserted at different locations in the vertebrae would affect their mechanical performance. This study is an in vitro human cadaveric biomechanical analysis. Fourteen lumbar (L3 and L4) vertebrae from 10 cadavers (age: 76±10 years, bone mineral density: 0.89±0.17 g/cm²) were used for this study. Pilot holes (3.5-mm diameter×15-mm length) representing three different IFC screw orientations (lateral to medial [LM], midsagittal [MS], and medial to lateral [ML]) were created in vertebrae using an IFC guide and bone awl. The screw locations and trajectories chosen are representative of commercially available IFC designs. These pilot holes were then imaged with high-resolution microcomputed tomography to obtain a three-dimensional structure of the bone surrounding the pilot hole. Local bone morphology was then quantified by evaluating a 3-mm-thick circumferential volume surrounding the pilot hole. Integrated fixation screws were implanted into pilot holes while recording maximum screw insertional torques. Screws were toggled in the cranial direction from 10 to 50 N for first 10,000 cycles, and the maximum load was increased by 25 N for every 5,000 cycles for a total of 25,000 cycles. Total bone volume (BV) and trabecular bone volume fraction surrounding ML screws were significantly greater (p<.03) compared with those around MS screws and LM screws. The maximum insertional torque for ML screws were greater (p=.06) than LM and significantly greater (p<.02) than MS screws. The number of cycles to failure for the ML screw was significantly greater (p<.04) than that for the LM and the MS screws. Total BV (R²≤46.2%, p<.03) and the maximum insertional torque (R²≤59.6%, p<.03) provided better correlations to screw loosening compared with all the other bone quality parameters. Our findings indicate that bone quality in the vertebral body varied spatially depending on the orientation and the insertion location of the IFC screw. These alterations in local bone quality significantly affected the screw's ability to fixate to bone. These variations in bone quality may be assessed intraoperatively using screw insertional torque measurements. By understanding available bone purchase at the bone-implant interface, the appropriate implant design can be selected to maximize the fixation strength.