2237 - A Mathematical Framework for Determining the Effect of Rotational Errors on Single Iso Multi Target Stereotactic Radiosurgery

Purpose/Objective(s): To generate a mathematical model to derive the effect of rotational errors for overlap of tumor and radiation as a function of distance from isocenter and magnitude of rotation

Materials/Methods: Single Isocenter Multi-Target (SIMT) Radiosurgery enables the simultaneous treatment of multiple targets within a single plan. This technique, however, is highly sensitive to patient positioning, and small rotations can cause significant reduction in dose coverage. The dose coverage degradation is a function of the size of the target, the magnitude of rotation, as well as the distance of the target from the isocenter. We developed a mathematical model to simulate the effects of these rotational errors, while incorporating size and distance parameters. Our model approximates the shapes of target and radiation dose coverage as spheres and assumes that the dose distribution precisely encompasses the target in the absence of any setup errors (conformality index 1). The effects of these rotational errors could then be computed mathematically and the percentage of the target covered by the prescribed dose could be determined. A python code was created to generate graphs depicting target coverage vs degree of rotation for various distances from isocenter.

Results:

The risk of compromised coverage is greatest for small targets located further from isocenter and is especially significant if margin size is too small. For a 0.5 cm diameter met, a one-degree rotation leads to a coverage (V100) decrease from 95% at 1 cm from isocenter to 65% at 7 cm from isocenter, in the absence of any margins. An addition of 1 mm margin improves the V100 to 100% and 97%, respectively, for the same tumor size and rotation. Our models show that when smaller targets are further from the SIMT isocenter, it is appropriate to strongly consider a separate isocenter and plan for these targets, such that treatment efficacy is not compromised. As the target size increases, a small margin is usually sufficient for acceptable coverage at clinically relevant rotations. For a 2.0 cm met, a 0.5 mm margin is sufficient for a V100 of 97% at one-degree rotation even if the met is 7 cm from isocenter.

Conclusion:

While analysis of rotation has been performed with selected patient data, to our knowledge this is the first mathematical analysis and generated model. Our software can be used to compute the amount of degradation in dose coverage for any rotation, tumor size, and distance from isocenter. The software can be used as a tool for optimal margin selection as well as settings to consider separate isocenters based on target size and distance from isocenter for SIMT Radiosurgery.