Developing whole-body MRI methods for assessing heterogeneity and response to immunotherapy in patients with metastatic melanoma

Abstract

Conventional response criteria are failing to accurately reflect which patients with metastatic melanoma (MM) are benefiting from treatment with immune checkpoint inhibitors (ICPI). The aim of this work was to develop techniques for whole-body MRI (WB-MRI) acquisition and analysis, to provide clinicians with tools to improve evaluation of disease distribution and response to treatment in MM patients treated with ICPI. A WB-MRI protocol was optimised and applied in 22 MM patients receiving ICPI prior to treatment, and in follow-up imaging 12 weeks and 6 months later. A total of 54 WB-MRI scans were acquired. A novel linearisable intravoxel incoherent motion model was used to estimate of the perfusion fraction from diffusion-weighted imaging (DWI). This allowed for faster computation time and uncertainty mapping of parameters. The model was compared to bi-exponential fitting in a prior prospective imaging study and tested in WB-MRI data acquired in patients with MM with mixed results. Noise-corrected, exponentially-weighted diffusion-weighted MRI (niceDWI) was previously suggested for the standardisation of inter-scanner and intra-scanner signal. Axial niceDWI and maximum intensity projections were clinically validated, however results showed no advantage to using niceDWI. Using the WB-MRI acquired in MM patients, imaging biomarkers including apparent diffusion coefficient (ADC), fat fraction (FF), enhancement fraction (EF) and transverse relaxation rate (R2) were extracted from sites of disease and normal-appearing tissues to explore treatment effects. There is preliminary evidence to suggest that changes in ADC of normal-appearing bone marrow, R2 of the spleen, and EF and FF of the thyroid correlate with response, alongside changes in the FF of soft-tissue lesions. Furthermore, the ADC and EF of the thyroid may be correlated with thyroid toxicity. In summary, this PhD has implemented and evaluated methods for data acquisition and analysis of WB-MRI, providing new quantitative insights into treatment effects in patients with MM receiving ICPI.