Improve Consistency In Radiation Shot Placement Using T1-Weighted And Fiesta Mri Sequences In Trigeminal Neuralgia Gamma Knife Radiosurgery
T.J. Wang,R. Brisman,J.N. Shah,E.J. Yoshida,T. Liu
DOI: https://doi.org/10.1016/j.ijrobp.2009.07.558
2009-01-01
Abstract:T1-weighted MRI is most commonly used for treatment planning in trigeminal neuralgia Gamma Knife radiosurgery. Fast imaging employing steady state acquisition (FIESTA) MRI is a newer sequence that provides enhanced visualization of the trigeminal cistern and thus, further distinguishes trigeminal nerve anatomy. The purpose of our study was to assess whether T1-weighted MRI, in combination with FIESTA MRI, can improve consistency in target delineation and precision in radiation shot placement among physicians with varying Gamma Knife radiosurgery experience. We studied 10 trigeminal neuralgia patients who received Gamma Knife radiosurgery. T1-weighted and FIESTA MRI sequences were acquired for all patients. For each patient, three physicians with 1, 5, and 12 years of Gamma Knife experience performed target delineation of the symptomatic trigeminal nerve and brainstem, as well as placed a 4-mm radiation shot using (1) T1-weighted MRI alone, (2) FIESTA MRI alone, and (3) combined T1-weighted and FIESTA MRIs. Each physician was blind to the others' treatment plans. The radiation shot was placed at the root entry zone of the symptomatic trigeminal nerve such that the 50% isodose line touched the brainstem. The volumetric variations of the trigeminal nerve contours and the geometric variations of the shots among physicians were analyzed. Significant variation existed in trigeminal nerve delineation among the physicians, regardless of MRI sequence. Standard deviation of the trigeminal nerve volumes ranged from 2.4 mm3 (4%, standard deviation divided by the mean) to 63.0 mm3 (110%) using T1-weighted MRI alone, 3.3 mm3 (4%) to 12.7 mm3 (84%) using FIESTA MRI alone, and 11.5 mm3 (9%) to 28.5 mm3 (112%) using combined T1-weighted and FIESTA MRIs. For radiation shot placement, the mean geometric variation was 2.3 mm for all imaging methods. However, the standard deviation of shot-placement variation using combined MRI sequences was 42% lower compared to T1-weighted MRI alone, and 36% lower compared to FIESTA MRI alone. Furthermore, 50% fewer cases of shot-placement variation > 5 mm occurred with combined MRI sequences than with T1-weighted MRI alone. We found that patients with poorly visualized trigeminal nerves on T1-weighted MRI benefited most from FIESTA MRI. Variability in contouring the trigeminal nerves among physicians using T1-weighted MRI alone, FIESTA MRI alone, and combined T1-weighted and FIESTA MRIs was substantial. However, using T1-weighted MRI in combination with FIESTA MRI reduced radiation shot-placement variation among physicians with different experiences, and therefore may improve reproducibility and outcome for trigeminal neuralgia Gamma Knife radiosurgery.