ESTRO 2025 - Abstract Book

S3817

Physics - Radiomics, functional and biological imaging and outcome prediction

ESTRO 2025

reflect radiotherapy-induced changes. These findings encourage future multi-centre studies linking ADC to clinical outcomes in rectal cancer.

Keywords: Diffusion-weighted MRI, MRI-Linac, rectal cancer

References: 1.

Schurink NW et al. Br J Radiol. 2019;92(1096):20180655. de Mol van Otterloo SR et al. Front Oncol. 2020;10. Bisgaard ALH et al. Phys Imaging Radiat Oncol. 2022;21:146–52. Fernandez Salamanca M et al. Front Oncol. 2024;14:1–7. Fernando N et al. Phys Imaging Radiat Oncol. 2024;30:100570.

2. 3. 4. 5.

3317

Digital Poster Quantitative MR and delta-radiomics for longitudinal monitoring of treatment response following prostate cancer radiation therapy Yu-Feng Wang 1,2 , Sirisha Tadimalla 1,2 , Niluja Thiruthaneeswaran 2 , Lois Holloway 3,4 , Sandra Turner 2 , Amy J Hayden 2 , Mark Sidhom 5,4 , Jarad Martin 6 , Annette Haworth 1,2 1 Institute of Medical Physics, University of Sydney, Sydney, Australia. 2 Sydney West Radiation Oncology Network, Western Sydney Local Health District, Westmead, Australia. 3 Ingham Institute for Applied Medical Research, Liverpool Hospital, Liverpool, Australia. 4 South Western Sydney Clinical School, University of New South Wales, Sydney, Australia. 5 Cancer Therapy Centre, Liverpool Hospital, Liverpool, Australia. 6 Department of Radiation Oncology, Calvary Mater Newcastle, Waratah, Australia Purpose/Objective: Early detection of recurrent disease following prostate radiation therapy (RT) enables salvage focal therapies. Standard post-treatment prostate specific antigen (PSA) testing lacks sensitivity, specificity, and spatial information in detecting local recurrences. Here, we assess the feasibility of using quantitative magnetic resonance imaging (qMRI) and qMRI-based radiomic features to detect early treatment response in prostate cancer following RT. Material/Methods: Longitudinal qMRI was acquired from twenty-one patients from the Sequential Imaging Biofocussed Radiotherapy trial prior to and at 6-, 12-, and 18-months post RT (ANZCTR UTN U1111-1221-9589). Maps of apparent diffusion coefficient (ADC), diffusion coefficient (D), perfusion fraction (f), hypoxia score (HS), forward transfer rate constant (K trans ), extracellular extravascular volume fraction (v e ), total transverse relaxation rate (R2*), and longitudinal relaxation time (T1) was computed from the diffusion weighted imaging (DWI), dynamic contrast enhanced (DCE) MRI and relaxation time mapping data. 413 radiomic feature maps were extracted from the T2-weighted images and qMRI parameter maps from DWI and relaxation time mapping. Analysis of treatment-related changes considered both region-of-interest (ROI) and voxel-wise approaches within the tumour at each time point. Statistical significance of treatment-related changes was assessed with rANOVA and post hoc two-tailed t-test. The repeatability coefficient, representing the minimum detectable change, estimated from a previously reported in vivo test-retest study was used to assess the potential to reliably detect treatment-related changes. Patients receiving RT in combination with androgen deprivation therapy (ADT) were analysed separately from those receiving RT-only due to the known impact on qMRI. Results: Based on the ROI analysis, with the exception of f, significant changes in all qMRI parameters within the tumour were detected by 18-months post-RT, potentially reflecting a reduction of cellularity and vascularity due to radiation damage. However, only K trans , HS, and T1 had average (ROI) cohort changes exceeding the repeatability coefficient,