Part III | Chapter 11 218 evaluated in terms of (acute) toxicity, these margin reductions are very promising and a first step towards high-precision SBRT on 1.5 T MR-Linac systems. Currently, data on toxicity (physician- and patient-reported) and oncological outcomes is prospectively gathered in patients treated with the sub-fractionation workflow in combination with smaller PTV margins. This will allow evaluation of the clinical effects of this advancement. Nevertheless, the sub-fractionation workflow is limited by its simplicity, as it cannot act upon sudden anatomical changes during beam-on (e.g. gas in the rectum). Also, there is a limit to the number of sub-fractions per fraction with respect to clinical feasibility. This depends on the overall beam-on time and the time needed for imaging and replanning (steps 3-4 and steps 6-7 of Figure 1 in chapter 5), which typically takes around 5-7 min for a prostate cancer case with currently available technology. The optimal number of sub-fractions also depends on the tumour site and the expected intrafraction motion. With beam-on times of several minutes for the two sub-fractions schedule presented in chapter 5, there is still room for significant motion that could affect the delivered dose. Therefore, further dose escalation or margin reduction (e.g. to 1 mm margins) is currently not safely possible. To counteract all systematic and sudden intrafraction motion on-the-fly, we ultimately need continuous, real-time intrafraction adaptive methods. These methods could be based on real-time accumulation of the dose delivered to the target(s) and OARs. To be able to employ intrafraction plan adaption based on the delivered dose, methods to accurately reconstruct and accumulate the delivered dose during treatment are warranted. Dose reconstruction methods have already been extensively described in literature, for example by Skouboe et al.42, who employed dose reconstruction using fiducial markers and kilovolt (kV) imaging during radiotherapy. Menten et al.33 were one of the first to describe a dose reconstruction method for MRI-guided radiotherapy on a 1.5 T MR-Linac using the machine log files and 2D cine-MR imaging obtained during irradiation. Kontaxis et al.43 adopted a similar set-up for dose reconstruction, but as input 3D cine-MR imaging was used and additionally rotations of the prostate were considered. Regardless of the exact method, it is important that specific characteristics of MR-Linac treatment, such as the electron return effect, are also taken into account when reconstructing the delivered dose.27 The output of 3D intrafraction motion tracking algorithms, such as described in chapter 4, could be used as input for such dose reconstruction pipelines. The question remains whether or not there is a need for real-time, online dose accumulation for the treatment of prostate cancer. Other intrafraction adaptive methods, such as multileaf collimator (MLC)-tracking or target tracking with exceptional gating (stop irradiation when the target crosses a pre-defined boundary and restart when the target is again within the boundary), are not based on online dose accumulation.44,45 Nevertheless, these methods could also nicely counteract intrafraction motion and improve treatment delivery accuracy. Especially MLC-tracking would allow for fast and accurate treatments, since the treatment does not have to be interrupted as the MLC positions and shape of the beam are continuously adapted to the moving target.44,45 The method that best suits the target area should be chosen. Towards improved outcomes with MRI-guided radiotherapy All the aforementioned technological developments in the end have to serve a clinical purpose: reduced toxicity rates and improved oncological outcomes for our patients. There is no need for technological advancements that do not improve patient care. At this point in time, MR-Linac
RkJQdWJsaXNoZXIy MTk4NDMw