Effect of image uncertainty on the dosimetry of trigeminal neuralgia irradiation.

OBJECTIVE: Our objective was to quantify the uncertainty in localization of the trigeminal nerve (TGN) with magnetic resonance imaging (MRI) and computed tomography (CT) and to determine the effect of this uncertainty on gamma-knife dose delivery.
METHODS: An MR/CT test phantom with 9, 0.6-mm diameter, copper rings was devised. The absolute ring positions in stereotactic space were determined by the angiographic module of the LGP software. The standard deviation, sigma, in the difference between the absolute and MR-measured or CT-measured coordinates of the rings was determined. The trigeminal nerve in 52 previously treated patients was contoured and expanded by 1sigma and 2sigma margins to model the uncertainty in the location of the nerve. For gamma-knife treatment, a single isocenter was used and was located at the distal cisternal portion of the trigeminal nerve root. Irradiation methods included a 4-mm collimator, 90 Gy to isocenter and a 4&8-mm collimator, 70 Gy to isocenter. A patient outcome survey that sampled pain relief and morbidity was done.
RESULTS: The MR coordinate sigma was 0.7 mm left-right, 0.8 mm anterior-posterior, and 0.6 mm superior-inferior, and the CT coordinate sigma was 0.4 mm left-right, 0.2 mm anterior-posterior, and 0.2 mm superior-inferior. A 45% higher dose line covered the TGN with the 4&8-mm method. No significant increase in pain reduction or morbidity occurred.
CONCLUSIONS: The uncertainty of target location by MRI is more than twice that found in CT imaging. The 4&8-mm collimator method covers the trigeminal root cross section with a higher isodose line than does the 4-mm method. This higher dose did not significantly reduce pain or increase morbidity.