PERCUTANEOUS PEDICLE CANNULATION: A CADAVERIC STUDY ASSESSING THE ACCURACY OF CONVENTIONAL C-ARM FLUOROSCOPY AND TWO DIMENSIONAL SPINAL NAVIGATION UTILIZED BY TWO LEVELS OF EXPERTISE

Abstract

Spinal procedures relying on percutaneous pedicle cannulation (PPC) are becoming increasingly common. The accuracy of PPC using currently available two-dimensional intraoperative imaging such as conventional C-arm fluoroscopy (CF) or computer-assisted fluoroscopy (2D_Nav) has not been evaluated. Following PPC of cadaveric spines (T4-S1) using CF and 2D_Nav, by a novice and clinical expert, the number and degree of pedicle breaches was assessed by CT. Accuracy using CF or 2D_Nav was equivalent and comparable to published reports for open pedicle cannulation. However, clinical expertise was the significant determinant of improved accuracy rather than technological factors.

To assess the accuracy of percutaneous pedicle cannulation(PPC) using currently available two-dimensional intraoperative imaging (C-arm fluoroscopy (CF) or computer-assisted fluoroscopy (2D_Nav)) for two levels of clinical expertise.

Accuracy using CF or 2D_Nav was equivalent and comparable to published reports for open pedicle cannulation. Main determinant of PPC accuracy is clinical experience, rather than technological factors.

Current technology cannot replace the need for rigorous training required to gain skill in percutaneous pedicle procedures.

Using an eleven-gauge bone biopsy needle, sixty randomized pedicles(two cadavers, T4–S1) were cannulated using CF or 2D-Nav by a staff spine surgeon or a third year orthopaedic resident. Pedicles for each vertebra were paired as internal controls for technique. After insertion of the biopsy needle, a 1.5mm aluminum tube was inserted through the needle as a marker. Using fine cut CT scans the position of each tube was assessed using a predefined grading system based on tube location relative to pedicle, direction of breech, trajectory, and position in vertebral body. Minimum score was three(ideal) and maximum was fourteen(gross misplacement). There were significantly (p< 0.05) more pedicle breaches for the resident (four CF, four 2D-Nav) compared to staff (one 2D-Nav). All breaches were thoracic with no statistical difference between number of breaches using CF and 2D-Nav for either skill group. Grade of breaches for CF (8.8) and 2D-Nav (9.4) were statistically similar. Main sources of error included poor image quality, image misinterpretation and biopsy needle flexibility causing navigational maltracking.