Stereotactic Shifts During Frame-Based Image-Guided Stereotactic Radiosurgery: Clinical Measurements

To determine the magnitude and reason for discrepancies between frame- and cone beam computed tomography (CBCT)-determined stereotactic coordinates, we reviewed frame-based Gamma Knife radiosurgery procedures in which CBCT was performed before treatment.

Clinical and treatment documentation was reviewed for 150 frame placements for which stereotactic coordinates were defined via both frame and fiducials on computed tomography imaging and CBCT. Treatment planning system-reported rotational and translational differences and standard deviations (SDs) between frame-based and CBCT-based stereotactic coordinates were recorded. Potential clinical predictors for increased differences were collected. Multiple linear regressions were performed to evaluate for associations with increased translations and rotations.

The absolute mean of the measured pitch, yaw, and roll shifts was 0.14 degrees (range -0.71-0.63 degrees, SD 0.19 degrees), 0.16 degrees (range -0.50 to 0.83 degrees, SD 0.21 degrees), and 0.12 degrees (range 0.37-0.51 degrees, SD 0.15 degrees), respectively. The absolute mean of the measured shifts in the left-right, anteroposterior, and superior-inferior direction was 0.29 mm (range -1.29 to 0.82 mm, SD 0.35 mm), 0.24 mm (range -0.59 to 0.33 mm, SD 0.19 mm), and 0.24 mm (range -0.69 to 0.91 mm, SD 0.27 mm), respectively. Three cases (2.0%) exceeded 1 mm in translational difference, all in the left-right direction (1.05, 1.13, and 1.29 mm). Lower Karnofsky Performance Scale status was associated with greater translational differences (vector magnitude, P = .023) and rotation (pitch, P = .044; yaw, P = .002). Usage of longer total pin length (sum of all 4 fixation pin lengths) was associated with increased rotation but not with translation (P < .001 and P = .56, respectively).

CBCT imaging in this cohort of frame-based cases suggests that the discrepancy in stereotactic coordinates is less than 1 mm or degree in most cases. Low Karnofsky Performance Scale status and longer total pin length correlate with larger differences between frame-defined and CBCT-defined stereotactic coordinates.