Case Study: Simulation-Free MRI-Guided Adaptive Prostate Radiotherapy on Elekta Unity Using Diagnostic MRI for Reference Planning

Published: 19.02.26

Download Case Study as PDF

Overview

At The Royal Marsden NHS Foundation Trust, in collaboration with The Institute of Cancer Research (London, UK), a planning study explored a simulation-free pre-treatment workflow for prostate MR-guided online adaptive radiotherapy [1]. Traditionally, treatment planning requires a dedicated simulation scan, which can add weeks of waiting before patients begin treatment. This work used diagnostic MR scans instead, mimicking an MR-only prostate online adaptive radiotherapy workflow on the Elekta Unity MR-Linac to assess suitability for reference plan generation.

Context

Site: The Royal Marsden / The Institute of Cancer Research (London, UK)
System: 1.5T MR-Linac (Unity, Elekta AB)
Patient cohort: Patients treated with 36.25 Gy in 5 fractions within the HERMES trial
Clinical setting: Online adaptive radiotherapy (oART), where the reference plan serves as a template for plan adaptation

Objective

To test whether non radiotherapy-dedicated diagnostic MRI can be used for reference prostate planning on the MR-Linac, generating adapted plans equivalent to those clinically delivered.

Method

Previous imaging was reviewed for all 5-fraction patients treated within the HERMES trial to assess eligibility. A total of 21 out of 24 patients had diagnostic imaging suitable for reference planning.

If eligible:

Diagnostic MR images were prepared to enable MR-only reference planning (interpolation, tilt removal, external extension of the image where FOV did not include entire patient).
Target and organ-at-risk reference structures were autosegmented using Contour+ without clinician input.
A synthetic CT was created using population average electron density.
MR-Linac reference plans were generated as per HERMES protocol on the generated synthetic CT images.
Online plan adaptation was simulated using existing clinical treatment images and structure sets.
The new “simulation-free” adapted plans were compared with existing clinical adapted plans.

Results

All reference images and treatment plans were successfully created.
All adapted treatment plans:
– achieved target dose and conformality
– met all mandatory clinical goals
– showed no detriment to organ-at-risk dose or plan deliverability
Synthetic CT accuracy was high, with 2mm/2% gamma results above 98.9%.

Conclusion / Why it matters

This study has shown that diagnostic MRI can be used for reference prostate planning on the MR-Linac. Diagnostic MR images generated adapted plans equivalent to those clinically delivered that originated from radiotherapy-simulation reference plans.

Using diagnostic MR for reference planning can save weeks in the radiotherapy pathway of high volume departments, possibly improving treatment efficiency and patient experience.

Next steps