Part III | Chapter 11 214 perform the entire workflow from start to end, including image acquisition, image registration, contouring, and treatment plan approval. This requires a whole range of skills, including knowledge on basic MR physics and safety, MR image acquisition and interpretation, treatment plan evaluation, and plan QA.1,2 As our department hosts multiple MRI scanners, there was already a substantial amount of MRI knowledge among technicians and therapists with respect to MRI simulation and planning. This is of course not naturally the case in all radiotherapy departments. For these departments, a lack in MRI knowledge might complicate the transition to an RTT-led workflow, let alone the introduction of MRI-guided radiotherapy in the first place. To be able to implement MRIguided radiotherapy, a certain level of MRI-knowledge is needed. MRI is increasingly used for treatment simulation and planning purposes.3–5 Given that the levels of MRI education and expertise vary across institutions, the results of studies that discuss the implementation of RTT-led workflows are not (directly) applicable to all. Nonetheless, the central role of RTTs in MRI-guided radiotherapy demands qualified personnel which calls for standardised MRI education and training.2,6,7 Besides the need for MRI knowledge and training with respect to contouring, the transition to an RTT-led workflow should be backed by availability (on-call) of a radiation oncologist and physicist whenever this is required. As discussed, an RTT-led workflow is not yet implemented in all departments providing MRI-guided radiotherapy on an MR-Linac.8 This is not only caused by a lack of (MRI) expertise, but also by different opinions with respect to the distinct roles of physicians, physicists, and RTTs, and local codes of practice. Furthermore, feasibility of introducing an RTT-led workflow depends on the complexity of the treated area and chosen treatment. For example for upper gastrointestinal tumours, which is a relatively new treatment indication, the complex anatomy does not yet allow for a completely RTT-led workflow.9 Treatment times and intrafraction motion: improving speed and accuracy With MRI-guided radiotherapy workflows in which daily adaptation, including manual contour editing, is employed, treatment times can become extensive.10,11 Especially when compared to treatment times on conventional radiotherapy systems (approximately 10-15 min) the treatment times on MR-Linac systems are relatively long. Currently, treatment times for prostate cancer range between 45 and 60 min when applying full plan re-optimisation based on the anatomy of the day.12,13 This not only imposes a problem with respect to logistics (e.g. fewer patients can be treated on an MR-Linac), but also creates new potential problems. Due to the high-quality MR imaging available on MR-Linacs and the possibility of continuous imaging during dose delivery (cine-MRI), we have become even more aware of the substantial amount of intrafraction anatomical changes that can occur during the preparation and delivery of radiotherapy.14 With treatment times becoming substantially longer with increased fractional doses, there is also more room for intrafraction motion to occur. Not only during the actual beam-on, but also in the time period between daily image acquisition and treatment delivery. This in turn could negate the potential benefits from daily (interfraction) adaptive radiotherapy and might lead to target underdosage.15 To ensure adequate target coverage, relatively large Planning Target Volume (PTV) margins are required. The effect of intrafraction motion becomes even more apparent and crucial if the goal is extremely hypofractionated radiotherapy in one or two fractions, with beam-on times well beyond 10min. This underpins the need for improvements in adaptive workflow speeds to be able to reduce
RkJQdWJsaXNoZXIy MTk4NDMw