We propose iRADIOMICS, a highly innovative and potentially clinical practice changing tool, which will allow for better management of patients undergoing immunotherapy. iRADIOMICS is based on in-depth interrogation of the molecular imaging (FDG PET/CT) data, extracting "invisible" information based on physical description of the imaging information. Based on the promising preliminary results of our pilot study, we hypothesise that radiomics analyses of FDG PET/CT scans of patients treated with immunotherapy (iRADIOMICS) can better predict response to immunotherapy compared to the current standards (iRC). iRADIOMICS will be assessed in a prospective clinical study, involving 30 patients with metastatic non-small-cell lung cancer, treated with anti-programmed death-1 (anti-PD1) antibodies. Patients will undergo FDG PET/CT before the administration of anti-PD-1, at 1 month and 4 months after the administration. Afterwards, the patients will be imaged with FDG PET/CT every 6 months. Additionally, the patients will undergo diagnostic CT scan every 3 months to allow for comparison to the current standard (irRC).
Correlation of iRADIOMICS and irRC With Survival
Diagnostic Test: FDG PET/CT
We hypothesise that molecular imaging-based RADIOMICS analysis of FDG PET/CT data (termed iRADIOMICS) provides more information than standard anatomical imaging-based irRC analysis regarding the assessment of the effectiveness of immunotherapy and will have a stronger predictive power. It is widely accepted that molecular imaging (e.g. PET/CT) reflects changes in tissues much sooner than anatomical imaging (CT, MRI). Therefore, we expect that an immunotherapy assessment tool based on FDG PET/CT should outperform anatomical-imaging-based irRC also timewise. Although we do expect an initial increase in FDG PET uptake (mainly due to the metabolic activity of tumour infiltrating lymphocytes (TILs)), followed by a late decrease, we argue that the predictive power of FDG PET can be even further increased by including an in-depth analysis of additional imaging features - the aforementioned "radiomics texture features". Many studies across different types of cancer have found a correlation between the presence of TILs and patient survival. Therefore we expect that iRADIOMICS signature of responders will be different from the irRADIOMICS signature of non-responders to antiPD1 immunotherapy due to the different levels of TILs infiltration, different TILs spatial distribution within the tumour, and different composition of immunosuppressive tumour microenvironment containing different levels and spatial distribution of various immunosuppressive cells, such as myeloid-derived suppressive cells (MDSC), regulatory T cells (Treg), tumour-associated macrophages (TAM), regulatory dendritic cells (DCreg) and others 30. Thus we anticipate that it might be possible to assess the response to immunotherapy at just one imaging time-point, preferably already in the pseudo-progression phase, thus much earlier than with irRC. Based on the assumption that irRADIOMICS might be able to detect differences in tumour immunosuppressive microenvironment, we further hypothesise that it might be also possible to predict, which patients are most likely to benefit from anti-PD1 immunotherapy already before the therapy.