John S. Kirkpatrick, MD, MS, and
Mark N. Melkerson, MS
John S. Kirkpatrick, MD, MS, is assistant professor of orthopaedic surgery, University of Alabama at Birmingham, and director, Spine Biomechanics Laboratory, Birmingham Veterans Administration Hospital.
Mark N. Melkerson, MS, is Branch Chief, Orthopaedic Devices Branch, Division of General and Restorative Devices, Office of Device Evaluation, Food and Drug Administration.
A standard for Test Methods for Spinal Implant Constructs in a Corpectomy Model (F1717) has been adopted by the American Society for Testing and Materials (ASTM).
Researchers attending a two-day international workshop reviewed experiences on how the provisional standard on spinal implant testing, which was adopted in 1994, was being used, the results and what other testing procedures were being used. The workshop was titled, "Spinal Implants: What we have learned from testing, failure analysis and device retrieval."
Sponsored by the Committee on Medical and Surgical Devices of ASTM, the workshop included a broad range of topics from testing groups in France, Germany, the United Kingdom and the United States. The researchers discussed a variety of specific test methods, including finite element simulations, synthetic, animal and cadaveric substrates, as well as long-term animal and human retrieval studies.
The goals of the workshop were to provide a means to assess the current knowledge base and current standards in use, and to encourage the development of new standards activities.
Finite element analysis was presented as a popular analytical tool for studying spine biomechanics. Increasingly, it can be used to simulate clinically-relevant injuries and treatments. The potential future areas of development included modeling the entire trunk, including muscle effects and evaluating the time-dependent changes within the disk and bone. A significant limitation of the finite element method is the lack of experimental validation which is critical to the appropriate use of this tool.
Researchers presented a variety of testing substrates, including a proposed mechanical model of the motion segment. Cadaveric animals and human cadaveric models were presented as ways to evaluate the acute mechanical changes of implants. Researchers discussed a long-term animal model to demonstrate long-term changes of bone mineral content and disk with respect to varying implant stiffness. A method to simulate muscle forces on cadaveric specimens using cables was believed to better simulate in vivo mechanics.
A synthetic corpectomy model was utilized by a number of investigators to evaluate the stress within the implants, construct stiffness and the effect of stress risers within implants. Testing of isolated interconnections was found to result in failures similar to those in construct testing.
Discussion of the experience with the provisional standard for spinal device testing focused on ensuring that the failures found in the testing were clinically-relevant failure modes, used similar constraints and load applications and whether testing should be done in air or solution. The application of testing methods to the cervical spine were demonstrated in evaluating anterior plates of several designs. Anchor screw pull-out evaluations were tested, using a porcine vertebral body model.
The workshop concluded with two clinical papers relevant to device evaluation. One was an evaluation of 186 scoliosis cases which had a reoperation rate for hardware removal of 17 percent. None of these were for mechanical device failure. Reoperations were for pseudoarthrosis and device-related pain. Results from hardware removal for pain were successful in 10-12 patients.
The other clinical paper was a histologic study, following device removal. Fibrous tissue interfaces with the implant was a consistent finding. Eleven of 36 devices also had an epithelioid cell layer adjacent to the implant. A black metallic debris was noted in nine of the specimens, typically near the interconnections of the devices.