Integration of quantitative imaging biomarkers in clinical trials for MR-guided radiotherapy: Conceptual guidance for multicentre studies from the MR-Linac Consortium Imaging Biomarker Working Group

Abstract

Quantitative imaging biomarkers (QIBs) derived from MRI techniques have the potential to be used for the personalised treatment of cancer patients. However, large-scale data are missing to validate their added value in clinical practice. Integrated MRI-guided radiotherapy (MRIgRT) systems, such as hybrid MRI-linear accelerators, have the unique advantage that MR images can be acquired during every treatment session. This means that high-frequency imaging of QIBs becomes feasible with reduced patient burden, logistical challenges, and costs compared to extra scan sessions. A wealth of valuable data will be collected before and during treatment, creating new opportunities to advance QIB research at large. The aim of this paper is to present a roadmap towards the clinical use of QIBs on MRIgRT systems. The most important need is to gather and understand how the QIBs collected during MRIgRT correlate with clinical outcomes. As the integrated MRI scanner differs from traditional MRI scanners, technical validation is an important aspect of this roadmap. We propose to integrate technical validation with clinical trials by the addition of a quality assurance procedure at the start of a trial, the acquisition of in vivo test-retest data to assess the repeatability, as well as a comparison between QIBs from MRIgRT systems and diagnostic MRI systems to assess the reproducibility. These data can be collected with limited extra time for the patient. With integration of technical validation in clinical trials, the results of these trials derived on MRIgRT systems will also be applicable for measurements on other MRI systems.

Keywords: Biomarkers; Image-guided; Magnetic resonance imaging; Multicentre study; Radiotherapy.

Fig. 1.

Example of daily imaging on an MRIgRT system. T2-weighted images and ADC maps are shown for a patient with prostate cancer who received 20 fractions of 3 Gy on the Unity system. The numbers in the corners indicate the fraction numbers. The green line in the first fraction of the T2-weighed image represents the tumour delineation. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2.

QIBs can be used during all phases of radiotherapy treatment: outcome prediction based on pretreatment or early treatment data, during treatment to adapt based on biological imaging information (BIGART), and post-treatment to assess the response after treatment. Example images are shown of patients treated on the Unity system.

Fig. 3.

Timeline for clinical trials going from biomarker discovery to QIBs for personalised RT in clinical practice.

Fig. 4.

Trial design for QIB studies on MRIgRT systems with integration of technical validation. Bias can be assessed with digital or physical phantoms by adding an accreditation phase at the start of a trial. Repeatability can be assessed by performing two measurements before the start of treatment (test-retest). The burden for the patient can be limited when the QIB measurements for the retest data are performed in the MR idle time of the first fraction before the actual treatment delivery. To assess the reproducibility on multiple systems, the test data of the MRIgRT system and the pretreatment imaging on a diagnostic scanner can be used. No extra scan session is needed for this comparison if test-retest data have already been acquired. The examples in this figure are ADC maps acquired from a patient with liver metastases treated on a 1.5T MRIgRT system (Unity). The diagnostic images were acquired on a 3T system. Phantom images are from the QIBA diffusion phantom.