Room: Davidson Ballroom A
Purpose: In single-isocenter-multiple-target (SIMT) Linac stereotactic radiosurgery (SRS) treatments, small angular errors can lead to large shifts for targets distant to isocenter. A solution is putting extra margins on off-center targets. However, margins are limited by voxel size and can lead to over- or under-compensation. This effect can be big for small targets commonly seen in SRS treatment. We propose a method to estimate potential coverage drop-off for each target based on distance to isocenter and compensate this decrease in coverage by prescribing the target to a higher dose.
Methods: SIMT brain SRS cases were planned using 5 non-coplanar partial-arcs VMAT for Varian Truebeam (6MV-FFF beam). Isocenter to target distance was 5-9cm. PTVs range from 0.28cc-18.79cc. Each target prescribed 20Gy to PTVD95. Dose was computed in Eclipse using 1mm^3 grid-size. Different GTVs were retrospectively created from PTVs by taking the erosion of [0, 1, 2, 3] mm to simulate different GTV-PTV margin design. Random shifts were applied to one case to establish relationships between the drop-off of GTV D95/D99 and target offset. Angular error of 2 degrees was introduced in remaining 3 cases (9 off-center targets) in Eclipse and the GTV coverage recorded.
Results: The GTV coverage drop-off with angular error matched well with our model. For brain metastases cases tested, with Â±2Â° error, GTV D95 drop-off can be as large as 15%, 10% and 5% and GTV D99 drop-off can be 20%, 15%, and 10% for 0, 1, 2mm margin respectively for the off-center target studied.
Conclusion: The method for correcting residual setup angular error in SIMT Linac SRS treatment was validated. The increase of prescription dose needed can be found based on target distance from rotation axis and physicianâ€™s choice of D99/D95 as the evaluation metric.