SU-F-T-651: Dosimetric Impact of Patient Setup Rotational Errors for Multi-Staged Large AVM Gammaknife Radiosurgery.

PURPOSE: Multi-staged large AVM (>10 cm3 or diameter > 3 cm) Gammaknife radiosurgery requires setting up the patient based on a pre-treatment plan over several fractions. Residual rotational errors cannot be compensated by translational repositioning of the patient and may adversely impact the dose distribution on the total AVM target. This study evaluates the effect of residual rotational setup error.

METHODS: The total AVM target was contoured and split into several small sub-targets. A Gammaknife plan was completed for the whole AVM target to meet the prescription goals. Shots were split and grouped according to the location of each sub-target to be delivered in separate staged sup-plans. Rotational errors of ±1-5° about the three rotational axes along with 5mm random translational errors were used to simulate head frame setup errors on different treatment days. Software was developed to compensate for translational errors to bring the shots back to the position of the original plan. Dose distributions were compared between the original planned dose and that resulting from residual rotational errors.

RESULTS: The volume of the target having dose differences within 3% is decreased with larger residual rotational errors (98.28%, 84.22% and 68.08% for ±1°, ±3°, and ±5° rotation, respectively). The volume of the target with dose differences within 5% (99.89%, 94.86% and 83.01% for ±1°, ±3°, and ±5° rotation, respectively) indicate that rotation less than ±3° is acceptable for treatment.

CONCLUSION: Dosimetric errors increase as rotational errors increase. The dosimetric impact was shown acceptable when rotational errors are less than 3°.