Background and purposeOnline adaptive radiotherapy (oART) potentially reduces the dose to organs at risk (OARs) as the planning target volume (PTV) margins are reduced compared to a non-adaptive approach (non-ART). This study evaluates the feasibility and dosimetric impact of cone-beam computed tomography (CBCT)-guided oART of urinary bladder cancer for the first patients treated, using patient-specific margins. Materials and methodsSixteen consecutive patients with muscle-invasive bladder cancer received two or more (median = 23) fractions as oART, and remaining fractions as non-ART. The non-ART fractions were delivered with standard population-based margins, while reduced patient-specific margins based on intra-fractional variations extracted from 2-4 fractions were applied to the primary PTV (PTV-T) during the oART fractions. Target volume and coverage, and dose to OARs were compared between non-ART and oART plans, and the oART procedure time was recorded. ResultsIn total, 297/512 fractions were delivered as oART with full re-optimization to the anatomy of the day. The median (interquartile range, IQR) oART procedure time, measured from the end of CBCT generation to completion of plan review, and quality assurance was 13.9 (11.9;16.6) min. The median (IQR) volume reduction in PTV-T volume was 33.9 (24.2;45.0)%, comparing oART and non-ART plans, resulting in median (IQR) reductions in bowel bag V45Gy of 18.8 (12.7;27.9)% and rectum V50Gy of 70.7 (35.9;94.8)%. By re-optimizing the plan to the daily anatomy, full target coverage was achieved at all oART fractions. ConclusionsoART resulted in large reductions in treatment volumes and doses to OARs, compared to non-ART, while ensuring target coverage. This indicates potential reductions in gastrointestinal toxicity.