Long-term patient experience with online MR-guided radiotherapy: adaptive versus non-adaptive workflow

TL;DR

This study compares patient experiences and outcomes between adaptive and non-adaptive workflows in MR-guided radiotherapy, finding that while adaptive methods take longer and cause more short-term discomfort, they do not increase toxicity and may offer faster recovery.

Contribution

The study provides new insights into patient-reported outcomes and toxicity of adaptive versus non-adaptive MR-guided radiotherapy workflows.

Findings

Adaptive treatment sessions were significantly longer than non-adaptive ones (71 vs. 36 minutes).

Patients in adaptive treatment reported higher discomfort related to duration, immobility, and sensory effects.

Adaptive therapy showed faster return to baseline post-treatment (6–8 weeks vs. 6–12 months).

Abstract

Magnetic resonance–linear accelerator (MR-Linac) systems enable high-precision radiotherapy through real-time MR guidance and daily online adaptive treatment planning. While online adaptation offers substantial dosimetric advantages, it extends treatment session durations on an already resource-intensive platform. This study aimed to evaluate patient-reported outcome measures (PROMs) and long-term toxicity profiles associated with MR-guided radiotherapy, with a particular focus on the impact of online adaptive workflows.

This subgroup analysis of an ongoing prospective observational study comprises patients treated with the MRIdian Linac at the Department of Radiation Oncology at Heidelberg University Hospital between January 2019 and May 2021. Online plan adaptation was implemented in February 2020. A custom-designed in-house questionnaire (PRO-Q) was employed to assess patient experience with MR-guided treatment. Toxicity was classified according to the Common Terminology Criteria for Adverse Events (CTCAE v. 5.0).

A total of 231 patients were included, comprising 130 non-adaptive and 101 adaptive treatments across 286 target volumes. Baseline patient characteristics, prior systemic therapy, and median planning target volumes (36.4 mL vs. 35.3 mL) were comparable between groups. Adaptive treatment was associated with significantly prolonged session durations (median 71 minutes vs. 36 minutes; p<0.01). During adaptive treatment, patients reported significantly higher discomfort in domains related to treatment duration, immobility, and sensory perceptions (e.g., tingling) as per PRO-Q responses. No statistically significant differences in overall toxicity were observed. However, patients undergoing adaptive therapy exhibited a faster return to baseline status post-treatment (6–8 weeks vs. 6–12 months).

Online plan adaptation at the MR-Linac increased treatment times and was associated with less favorable short-term patient-reported outcomes, yet it was delivered safely without compromising toxicity or oncologic outcomes. These results support adaptive MR-guided radiotherapy as a feasible and technically promising approach, while highlighting the need for further studies with validated PROMs and cost-benefit analyses to define its clinical value.

Figures