Risk of vertebral artery injury: comparison between C1–C2 transarticular and C2 pedicle screws

Background context To our knowledge, no large series comparing the risk of vertebral artery injury by C1–C2 transarticular screw versus C2 pedicle screw have been published. In addition, no comparative studies have been performed on those with a high-riding vertebral artery and/or a narrow pedicle who are thought to be at higher risk than those with normal anatomy. Purpose To compare the risk of vertebral artery injury by C1–C2 transarticular screw versus C2 pedicle screw in an overall patient population and subsets of patients with a high-riding vertebral artery and a narrow pedicle using computed tomography (CT) scan images and three-dimensional (3D) screw trajectory software. Study design Radiographic analysis using CT scans. Patient sample Computed tomography scans of 269 consecutive patients, for a total of 538 potential screw insertion sites for each type of screw. Outcome measures Cortical perforation into the vertebral artery groove of C2 by a screw. Methods We simulated the placement of 4.0 mm transarticular and pedicle screws using 1-mm-sliced CT scans and 3D screw trajectory software. We then compared the frequency of C2 vertebral artery groove violation by the two different fixation methods. This was done in the overall patient population, in the subset of those with a high-riding vertebral artery (defined as an isthmus height ≤5 mm or internal height ≤2 mm on sagittal images) and with a narrow pedicle (defined as a pedicle width ≤4 mm on axial images). Results There were 78 high-riding vertebral arteries (14.5%) and 51 narrow pedicles (9.5%). Most (82%) of the narrow pedicles had a concurrent high-riding vertebral artery, whereas only 54% of the high-riding vertebral arteries had a concurrent narrow pedicle. Overall, 9.5% of transarticular and 8.0% of pedicle screws violated the C2 vertebral artery groove without a significant difference between the two types of screws (p=.17). Among those with a high-riding vertebral artery, vertebral artery groove violation was significantly lower (p=.02) with pedicle (49%) than with transarticular (63%) screws. Among those with a narrow pedicle, vertebral artery groove violation was high in both groups (71% with transarticular and 76% with pedicle screws) but without a significant difference between the two groups (p=.55). Conclusions Overall, neither technique has more inherent anatomic risk of vertebral artery injury. However, in the presence of a high-riding vertebral artery, placement of a pedicle screw is significantly safer than the placement of a transarticular screw. Narrow pedicles, which might be anticipated to lead to higher risk for a pedicle screw than a transarticular screw, did not result in a significant difference because most patients (82%) with narrow pedicles had a concurrent high-riding vertebral artery that also increased the risk with a transarticular screw. Except in case of a high-riding vertebral artery, our results suggest that the surgeon can opt for either technique and expect similar anatomic risks of vertebral artery injury.