3202 - Dosimetric Impact of Baseline Shift Corrections In MR-Linac Comprehensive Motion Management

Purpose/Objective(s): The Elekta Unity MR-Linac Comprehensive Motion Management (CMM) system enables real-time anatomical monitoring with automatic gated beam delivery. When target persistently drifts beyond the gating window, Baseline Shift (BLS) corrections utilize rapid Segment Aperture Morphing (SAM) to recalculates the plan with a simple shift of the radiation aperture, allowing for correction of small target drifts within approximately one minute. This study evaluates the dosimetric integrity of BLS plans across varying target drift magnitudes to assess target dose preservation.

Materials/Methods: Retrospective dose calculations were performed for 10 patients treated with CMM across multiple sites (lung, liver, adrenal, pancreas, and prostate) using the Elekta Monaco 6.2.2 treatment planning system with the GPUMCD algorithm, with calculation grid size of 0.3 mm and statistical uncertainty of 1%. Simulated BLS plans were generated with translational shifts of 5 mm, 7 mm, and 10 mm along six cardinal directions (left/right [LR], superior/inferior [SI], anterior/posterior [AP]), including combined multi-directional displacements. The unshifted plan served as the reference. Dosimetric parameters D_2cc, D₉₅ and D_min were compared between the BLS and the reference plan.

Results: BLS-adapted plans generally exhibited modest dose elevations relative to reference plans. In the SI direction, dosimetric deviations demonstrated an inverse relationship with shift magnitude, mean D₉₅ values for 5 mm, 7 mm, and 10 mm shifts were +1.09% ± 2.37%, -0.18% ± 0.62%, and -0.30% ± 0.86%, respectively. Conversely, in LR and AP directions, the 7 mm shift produced the greatest deviation. In LR direction the mean value of D₉₅ for 5 mm, 7 mm and 10 mm shifts are +0.79% ± 1.54%, +2.43% ± 2.05% and +2.07% ± 2.04% respectively. In AP direction the mean values of D₉₅ are +0.57% ± 1.96%, +2.34% ± 2.04% and +1.63% ± 1.46% for 5 mm, 7 mm and 10 mm plans respectively. SI group exhibited significantly different values compared to both the LR and AP groups (p < 0.0167, Mann-Whitney U test with Bonferroni correction). Multi-directional shifts increased dosimetric complexity and deviation, with mean differences of +3.47% (D_2cc), +1.89% (D₉₅), and +1.50% (D_min) relative to reference plans.

Conclusion: BLS plans provide an efficient solution for managing persistent target drift and generally maintain acceptable target coverage. Dosimetric deviations exhibited direction-dependent behavior: an inverse trend with shift magnitude in the SI direction. The increase in deviation of dose for 7 mm and 10 mm plans when compared to 5 mm plans were observed in LR and AP direction, likely attributable to the leaf width of 7.2 mm at isocenter limiting aperture conformity for intermediate shifts of the target. Multi-directional shifts further increase deviations from the reference plan.