Experimental study on the loosening of pedicle screws implanted to synthetic bone vertebra models and under non-pull-out mechanical loads

Background: Pull-out tests have been widely applied to orthopaedic devices such as pedicle screws for the last decades. Despite their wide applicability, this particular pedicle screw testing method does not represent the load pattern of implanted vertebrae with the use of pedicle screw spinal constructs. Purpose: In the present study a new custom made foam model (Sawbone) of human vertebra and a corresponding load pattern for pedicle screw spinal implant construct were introduced. They allow for the investigation of loosening at the bone-screw interface and constitute an alternative test method to pullout tests. Methods: The new custom made foam model of human vertebra was developed and made out of biomechanical materials Sawbones in three different variants, including varied density in order to simulate normal, osteopenic, and osteoporotic quality of bone. Mechanical testing methods followed reproducible, non-pull-out loading protocols based originally on the ASTM F- 1717 standard which enables to introduce loosening at the bone-screw interface. Results: The results include the pattern of screw loosening at the bone-screw interface, measured loosening area at the interface, and mechanical parameters characterising bone-screw interface such as initial stiffness and residual displacements. Conclusions: The new custom made foam model of human vertebra followed by non-pull-out load protocol enabled investigations of the loosening at the bone-screw interface. During the tests a characteristic toggling profile of loosening was found with a significant widening at the tip of the screw for the osteoporotic quality of bone. It was proposed that the model and the corresponding load protocol may extend the current testing methods based on the pull-out testing, leading to the more accurate testing of pedicle screw spinal constructs focused on the bone-implant interface.