Developing rapid response MRI-guided palliative radiotherapy for metastatic spinal cord compression

Abstract

Purpose or Objective Implementation of a rapid response metastatic spinal cord compression (MSCC) pathway using diagnostic (dCT) imaging and adaptive MR Linac treatment has the potential to reduce waiting times, align healthcare processes, and improve the patient experience during a global pandemic. This work presents the preliminary feasibility testing of a rapid response, single appointment MSCC pathway on the Elekta Unity MR Linac (MRL) using dCT imaging, as the diagnostic MR field of view is too small. Materials and Methods Retrospective radiotherapy plan data were collected from ten patients who had received urgent/emergency palliative spinal radiotherapy on conventional linear accelerators at our institution. The most recent dCT images prior to the treatment planning scans were imported from the picture archiving communication system (PACS) to the MRL treatment planning system (Monaco V5.40, Elekta). New treatment plans were then created on these dCT images to prepare for delivery on the MRL (figure 1). In order to facilitate this target contours were required and generated by the treatment planner with support from a clinical oncologist. Departmental policy for single dose radiotherapy was used in the creation of these plans (8Gy treated in a single post field). To test implementation on the MRL, MR scans were acquired and treatment delivered to a 3D abdominal phantom (CIRS) using MR-CT registration and the adapt to shape (ATS) workflow. ATS ensures that if anatomical changes have occurred since dCT the contours can be adapted on the day to reflect these changes, allowing for online plan adaptation Results Ten plans were created on imported dCT. Treatment site ranged from upper thoracic spine to sacrum. Target volumes ranged from 156- 508 cm3, D95% ranged from 767-874cGy (figure 2). Acceptable coverage was achieved on all plans but proved more challenging on those with larger treatment volumes. Approved plans were exported to Mosaiq (V2.83, Elekta) to test delivery on a phantom on the MRL. Conclusion For MSCC we were able to import and transfer data, produce acceptable treatment plans on dCT images and preliminary testing on the MRL. Attempted delivery of these plans highlighted technical issues that need to be overcome prior to clinical implementation. These included the lack of origin and landmarking information on dCT can make patient positioning challenging due to limitations of online shifts. Additionally, bulk density overrides are required for Hounsfield unit to electron density conversion as dCT images may come from different diagnostic CT scanners for which commissioning data is not available. Further work is on-going to confirm dosimetry.