Pedicle screw insertion angle and pullout strength: comparison of 2 proposed strategies

Object. Minimally invasive pedicle screws inserted vertically (that is, dorsoventrally) through the pedicle, as opposed to the more common coaxial technique, offer potential advantages by minimizing soft-tissue stripping during screw placement. The screws are designed for insertion through a medial starting point with vertical trajectory through the pedicle and into the vertebral body. As such, no lateral dissection beyond the insertion point is necessary. However, the effects of this insertion technique on the screw biomechanical performance over a short- and long-term are unknown. The authors investigated the pullout strength and stiffness of these screws, with or without fatigue cycling, compared with comparably sized, traditional screws placed by coaxial technique. Methods. Twenty-one lumbar vertebrae (L-3, L-4, and L-5) were tested. Each pedicle of each vertebra was instrumented with either a traditional, coaxial pedicle screw (Group A), placed through a standard starting point, or a vertically oriented, alternative-design screw (Group B), with a medial starting point and vertical trajectory. The specimens were divided into 2 groups for testing. One group was tested for direct pullout (10 specimens) while the other was subject to pullout after tangential (toggle) cyclic loading (11 specimens). The screws were cycled in displacement control (+/- 5 mm producing similar to 4-Nm moment) at a rate of 3 Hz for 5000 cycles. Pullout tests were performed at a rate of 1 mm/minute. Results. Two-way ANOVA showed that Group B screws with a medial starting point (2541 +/- 1090 N for cycled vs 2135 +/- 1323 N for noncycled) had significantly higher pullout loads than Group A screws with a standard entry point (1585 +/- 766 N for cycled vs 1417 +/- 812 N noncycled) (p = 0.001). There was no significant effect of cycling or screw insertion type on pullout stiffness. Tangential stiffness of the Group B screws was significantly less than that of the Group A screws (p = 0.001). The stiffness of both screws in the toe region was significantly affected by cycling (p = 0.001). Conclusions. The use of Group B screws inserted through a medial starting point showed greater pullout load than a Group A screw inserted through a standard starting point. The greater pullout strength in Group B screws may be due to screw thread design and increased cortical bone purchase at the medial starting point. Nevertheless, anatomical considerations of the medial starting point, that is, pedicle or lateral vertebral body cortex breach, may limit its application. The medial starting point of the Group B screw was frequently in the facet at the L-3 and L-4 pedicle entry points, which may have clinical importance. (DOI: 10.3171/2010.11.SPINE09886)