Fat quantification using whole-body magnetic resonance imaging (WB-MRI) in malignant bone disease

点击翻译按钮获取中文摘要

MRI is increasingly a data-driven modality, as can be appreciated by the success of numerous quantitative imaging biomarkers. I believe MRI has the potential to be a powerful metrological tool for disease management in oncology. This PhD studentship would, beyond developing my abilities as a scientist, give me the scientific grounding and opportunity to be an active and future participant in this transformative change happening within healthcare.The Royal Marsden Hospital (RMH) / Institute of Cancer Research (ICR), where I work as an MRI clinical scientist, is a major centre for whole-body MRI (WB-MRI) with a strong track-record in development of qualitative and quantitative methods for WB-MRI. National and international guidelines recommend WB-MRI in patients with myeloma [1] and metastatic prostate cancer [2]. WB-MRI includes diffusion-weighted MRI (DW-MRI), which has high sensitivity for detection of bone lesions as well as enabling quantification of abnormal cellularity via the apparent diffusion coefficient (ADC), and fat-water imaging, which provides information about the displacement of normal marrow fat by tumours and return of fat in treated disease.WB-MRI exams usually include an approximate assessment of fat content using the relative fat fraction (FFr), quantified using the signal in magnitude images from two-point T1-weighted fat-water imaging. FFr has demonstrated clinical utility in assessment of myeloma and metastatic prostate and breast cancers, and changes in FFr may precede changes in ADC in patients with myeloma who respond to treatment [3]. There are no agreed thresholds in FFr to identify active, treated, or responding lesions. Fully quantitative assessment of fat content (proton density fat fraction, or PDFF) is not currently achieved using FFr, which includes sources of bias (T1 and T2* relaxation times, noise, choice of spectral model) [4]. Clinical applications of fat quantitation in bone marrow have also been demonstrated in non-malignant diseases [5]. Previous work in liver MRI has led to the development of successful clinical tools for PDFF estimation in liver, which remove these biases and are available on MRI scanners from all main manufacturers. The potential for improving fat quantification in malignant bone disease has not been assessed. There is, therefore, a clinical need to develop tools to provide PDFF estimation in WB-MRI for malignant bone disease, and to assess performance in assessment of bone lesions and response to treatment.Currently, I'm working on the final stages of a technical assessment of accuracy and repeatability of various Dixon techniques in the pelvis for patients with metastatic prostate cancer. This project's findings and limitations serve as a springboard for the work that would be a part of this studentship:Main Proposal Aims1. Determine the 1H spectral model in treated and untreated lesions and normal-appearing marrow. 2. Assess the sensitivity of fat quantification techniques to pre- and post-treatment measurements. 3. Determine a threshold in PDFF that distinguishes responding from non-responding lesions