Geometric and dosimetric evaluation of CTV contour adaptations by radiation therapists for adaptive prostate radiotherapy on a 0.35 T MR-Linac

Background & purpose: Magnetic resonance guided adaptive radiotherapy (MRgART) enables hypofractionated prostate radiotherapy by daily contour adaptation. The MRgART workflow, however, is labour intensive and in many institutes still requires the presence of the radiation oncologist (RO). Transferring the online contour adaptation task to the radiation therapist (RTT) will release the clinician from attending each treatment fraction making MRgRT more efficient and cost effective. In this study we investigate the viability of RTT-led prostate MRgART on a low-field MR-linac, by assessing the interobserver variations of RTT- and RO-generated CTV contour adaptations as well as the resulting dosimetric effects. Materials & methods: Four RTTs and four ROs performed CTV contour adaptations on first fraction data in ten patients. Delineations were compared against a gold standard contour using target volume, Dice similarity coefficient (DSC), and 95th percentile Hausdorff distance. In addition, a dosimetric evaluation was performed on all first fractions by performing plan adaptations based on all RTT contour adaptation and comparing these to the clinically delivered plan. Finally, a full-treatment simulation was performed in four patients to investigate the dosimetric effects of the RTTs’ contour adaptations throughout an entire treatment. Results: RTTs with no experience in prostate delineation prior to this study spent more time on CTV contour adaptations. The geometric and dosimetric analyses, however, showed no statistically significant differences between both groups. Conclusions: This study confirmed that RTTs perform similarly to ROs in carrying out online contour adaptations. These results indicate the feasibility of initiating a transition in contour adaptation tasks from ROs to RTTs.