Mri-only treatment planning for anal and rectal cancer radiotherapy (MANTA-RAY)

Introduction: A barrier to the widespread clinical implementation of magnetic resonance imaging (MRI)-only radiotherapy planning for pelvic sites is the limited assessment in the literature demonstrating technical achievability and benefit. The MRI-only treatment planning for anal and rectal cancer radiotherapy (MANTA-RAY) study aimed to validate synthetic-CT (computed tomography) model generation accuracy, show the viability of cone beam CT (CBCT) patient positioning using synthetic-CT and MRI as a reference image, assess the patient experience of radiotherapy MRI scans and quantify potential clinical benefits of MRI-only patient treatments for anal and rectal cancer sites. Methods: The MANTA-RAY study recruited 46 patients with anal and rectal cancers who received radiotherapy CT and MRI scans after informed consent. A deep learning synthetic-CT model was trained and validated in terms of dosimetric calculation accuracy. Differences in CBCT patient position registration accuracy were assessed when using synthetic-CT and MRI as the reference images. The patient experience of the MRI simulation was assessed. A planning study quantified the differences between MRI- and CT-only planned treatments. Results: The synthetic-CT model had excellent dosimetric accuracy (planning target volume (PTV) D95% dose difference to CT = 0.1%). MRI and synthetic-CT reference images had systematic differences to CT in all translational and rotational dimensions of <1 mm and