Nine bladder cancer patients treated in a phase II trial of plan selection underwent 6–7 weekly repeat MRI series, each with volumetric scans acquired over a 10 min period. Adaptive re-planning on the 0 min MRI scans was performed using density override, simulating a hypo-fractionated schedule. Target coverage was evaluated on the 10 min scan to quantify the impact of intra-fractional motion.
MRIGARTanIso reduced the course-averaged PTV by median 304 cc compared to plan selection. Bladder shifts affected target coverage in individual fractions for all strategies. Two patients had a v95% of the bladder below 98% for MRIGARTiso. MRIGARTiso decreased the bowel V25 with 15–46 cc compared to MRIGARTpop.
Online re-optimised ART has a considerable normal tissue sparing potential. MRIGART with online corrections for target shift during a treatment fraction should be considered in ART for bladder cancer.