New means in spinal pedicle hook fixation

Summary Pedicle hooks which are used as an anchorage for posterior spinal instrumentation may be subjected to considerable three-dimensional forces. In order to achieve stronger attachment to the implantation site, hooks using screws for additional fixation have been developed. The failure loads and mechanisms of three such devices have been experimentally determined on human thoracic vertebrae: the Universal Spine System (USS) pedicle hook with one screw, a prototype pedicle hook with two screws and the Cotrel-Dubousset (CD) pedicle hook with screw. The USS hooks use 3.2-mm self-tapping fixation screws which pass into the pedicle, whereas the CD hook is stabilised with a 3-mm set screw pressing against the superior part of the facet joint. A clinically established 5-mm pedicle screw was tested for comparison. A matched pair experimental design was implemented to evauluate these implants in constrained (series I) and rotationally unconstrained (series II) posterior pull-out tests. In the constrained tests the pedicle screw was the strongest implant, with an average pull-out force of 1650 N (SD 623 N). The prototype hook was comparable, with an average failure load of 1530 N (SD 414 N). The average pull-out force of the USS hook with one screw was 910 N (SD 243 N), not significantly different to the CD hook's average failure load of 740 N (SD 189 N). The result of the unconstrained tests were similar, with the prototype hook being the strongest device (average 1617 N, SD 652 N). However, in this series the difference in failure load between the USS hook with one screw and the CD hook was significant. Average failure loads of 792 N (SD 184 N) for the USS hook and 464 N (SD 279 N) for the CD hook were measured. A pedicular fracture in the plane of the fixation screw was the most common failure mode for USS hooks. The hooks usually did not move from their site of implantation, suggesting that they may be well-suited for the socalled segmental spinal correction technique as used in scoliosis surgery. In contrast, the CD hook disengaged by translating caudally from its site of implantation in all cases, suggesting a mechanical instability. The differences in observed hook failure modes may be a function of the type and number of additional fixation screws used. These results suggest that additional screw fixation allows stable attachment of pedicle hooks to their implantation site. Hooks using additional fixation screws passing obliquely into the pedicle apparently provide the most rigid attachment. The second fixation screw of the prototype hook almost doubles the fixation strength. Thus, the prototype hook might be considered as an alternative to the pepdicle screw, especially in the upper thoracic region.