Developing and Optimising the Use of Magnetic Resonance Imaging and Spectroscopy in Radiotherapy (MRinRT) Pathways: Investigating Avenues to Improve Outcomes for Patients and the Assessment of Treatment Response.

Background Advances in radiotherapy technology have allowed treatments to become more conformal with steeper dose gradients, meaning target volumes are smaller and dose to surrounding normal tissues reduced. This is possible, in part, due to improvements in imaging used in radiotherapy planning and in the development of Intensity Modulated Radiotherapy. The current standard of care for most tumour sites is for target volumes to be delineated on a planning CT scan, with additional information from diagnostic imaging such as MRI or PET which can be fused with the planning CT scan. MRI is known to give clearer images of soft tissues and tissue planes allowing for more accurate target volume delivery. The fusion of MRI to the planning CT scan to aid target volume delineation is a standard of care for certain tumour sites, e.g. brain, but not in others such as oesophagogastric cancers.

Optimisation of MRI protocols for Novel Tumour sites Our group undertook a scoping review into the use of MRI for radiotherapy target volume delineation for gastric cancers and has found that although there is evidence of improved contrast resolution with MRI in comparison to CT, clinical utility for radiotherapy planning is yet to be demonstrated. The research group has also performed a pilot study assessing the addition MRI to CT planning scans for radiotherapy planning for gastric cancers which demonstrated potential for clinical benefit. The limitations of this study were small patient numbers and that MRIs used were diagnostic scans optimised for liver imaging rather than gastric cancers. An issue identified was poor anatomical correlation of the MRI to CT planning scan due to differences in stomach filling. Our study will build on this work to optimise the MRI sequences for improved clarity of imaging and to identify and standardise optimal pre-examination preparation.

It is possible to acquire a range of MRI sequences with each giving different information. Furthermore, contrast agents may be used during image acquisition to improve visualisation of the target of interest. The choice of contrast agent and sequences depends on whether the aim is to assess the size/location of the tumour or to assess a functional aspect of the tumour allowing the potential to act as a predictive or prognostic imaging biomarker. This study aims to optimise the choice of MRI sequences for target volume delineation in various anatomical regions and to assess the potential impact of MR imaging on radiotherapy planning in anatomical regions where it is not currently routine practice. In the UK there is an ongoing similar study that aims to optimise the MRI sequences for use on MRI-Linear Accelerator (MR-Linac) machines. Our study will follow similar principles in order to optimise MRI sequences for radiotherapy planning from stand-alone MRI scanners. The investigators postulate that our findings will be more generalisable compared to the recruiting MR-Linac study, as the vast majority of UK Oncology Centres do not have access to MR-Linac machines.

The potential for an MRI-Adaptive workflow Radical (curative-intent) or neoadjuvant radiotherapy treatment is often delivered as multiple treatment fractions given over several weeks. The current standard of care is to generate a radiotherapy plan based on a pre-treatment CT planning scan with guidance from additional diagnostic investigations depending on tumour site. Unless an issue is detected during treatment, e.g. detected by the on-treatment cone-beam CT(CBCT) set-up verification process, the patient would complete their whole course of treatment as per the radiotherapy plan based on the pre-treatment CT planning scan. This does not account for changes in the position, size or shape of the tumour that occur over the whole radiotherapy treatment course. Clinicians therefore add margins around the target volume to account for these potential uncertainties, leading to a larger target volumes and increased dose to surrounding normal structures, which can lead to increased toxicity. A potential solution to this issue could be to re-image and re-plan during the treatment course, termed 'adaptive radiotherapy'. MRI is already utilised for adapted radiotherapy workflow, particularly when an MR-Linac is available. This study will explore whether a diagnostic MRI may be used for RT plan adaptation. Of note, similar strategies have been assessed using CBCT with a 'plan of the day' approach in bladder cancer and there is also ongoing work assessing the benefit of adapting RT based on a mid-treatment PET scan in the PEARL study for oropharyngeal cancers.

Exploring Imaging Biomarkers MRI can also be used to determine quantitative information about functional biological processes, via the use of quantitative imaging biomarkers. These are defined as 'an objective characteristic derived from an in vivo image measured on a ratio or interval scale as an indicator of normal biological processes, pathogenic processes, or a response to a therapeutic intervention'. Current protocols include sequences which can be used to estimate functional characteristics of a tumour and surrounding tissues which could act as a biomarker to predict response to treatment or likelihood of a patient developing toxicity from treatment. However, there is limited validation of these biomarkers in clinical settings, in particular there is a lack of data assessing the repeatability and reproducibility.