The Feasibility of Assessing the Cortical Bone Trajectory Screw Placement Accuracy Using a Traditional Pedicle Screw Insertion Evaluation System.

Abstract

This was a retrospective observational study. We aimed to characterize the feasibility of assessing the accuracy of cortical bone trajectory (CBT) screw placement in midline lumbar interbody fusion using a traditional pedicle screw insertion accuracy evaluation system based on computed tomography (CT). Since Santoni and colleagues proposed CBT as an alternative approach for the treatment of lumbar degenerative disease, CBT has been biomechanically and clinically investigated in detail. The reported misplacement rate was 0%-12.5%. Therefore, these cortical screws may result in severe complications, such as nerve root, vascular, and spinal cord injuries. However, to the best of our knowledge, the accuracy of the current assessment system of cortical bone screw placement has not been described clearly. Overall, 342 cortical screws of 69 consecutive patients with lumbar degenerative disease who underwent midline lumbar interbody fusion surgery in one surgeon's initial phase were examined retrospectively. A comprehensive and detailed pedicle screw accuracy classification and grading system was introduced in our study, including 5 types of misplacement: (1) medial and (2) lateral cortical bone perforation (MCP and LCP) of the corresponding pedicle, (3) anterior cortical bone perforation of the vertebral body, (4) endplate perforation, and (5) foraminal perforation (FP). The degree of interobserver and intraobserver agreement with regard to the screw positions based on CT were used as indicators of the reliability of the modified classification system. All patients were retrospectively assessed for screw placement-related complications throughout the entire treatment course to evaluate the relationship between the procedure adequacy and neurological symptoms. The interobserver and intraobserver agreements were substantial-to-almost perfect (κ=0.78 and 0.88, respectively) in distinguishing the acceptable-placed pedicle screws from those with partial or complete cortical perforation. In the MCP and LCP-the most common types of misplacement-the interobserver agreement was substantial (κ=0.70 and 0.76, respectively), and the intraobserver agreement was almost perfect (κ=0.85 and 0.89, respectively). In total, there are 7 (2.05%) MCP and 65 (19.01%) LCP screws. The screw placement-related complication rate is significantly higher in the MCP and FP groups than that in the LCP group. Our study demonstrated that using a pedicle screw classification and grading system based on CT to assess the accuracy of CBT screw placement is feasible and practical. MCP and FP screws are more likely to cause neurological deficits with statistical significance, especially grade 2 MCP. We recommend inexperienced surgeons choose a lateral trajectory rather than a medial one if they cannot ensure accurate screw insertion. Level III.