The Role of Radiomics in the Era of Immune Checkpoint Inhibitors: A New Protagonist in the Jungle of Response Criteria.
Angelo CastelloMassimo CastellaniLuigia FlorimonteLuca UrsoLuigi MansiEgesta LopciPublished in: Journal of clinical medicine (2022)
Immune checkpoint inhibitors (ICI) have demonstrated encouraging results in terms of durable clinical benefit and survival in several malignancies. Nevertheless, the search to identify an "ideal" biomarker for predicting response to ICI is still far from over. Radiomics is a new translational field of study aiming to extract, by dedicated software, several features from a given medical image, ranging from intensity distribution and spatial heterogeneity to higher-order statistical parameters. Based on these premises, our review aims to summarize the current status of radiomics as a potential predictor of clinical response following immunotherapy treatment. A comprehensive search of PubMed results was conducted. All studies published in English up to and including December 2021 were selected, comprising those that explored computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET) for radiomic analyses in the setting of ICI. Several studies have demonstrated the potential applicability of radiomic features in the monitoring of the therapeutic response beyond the traditional morphologic and metabolic criteria, as well as in the prediction of survival or non-invasive assessment of the tumor microenvironment. Nevertheless, important limitations emerge from our review in terms of standardization in feature selection, data sharing, and methods, as well as in external validation. Additionally, there is still need for prospective clinical trials to confirm the potential significant role of radiomics during immunotherapy.
Keyphrases
- computed tomography
- contrast enhanced
- positron emission tomography
- magnetic resonance imaging
- clinical trial
- lymph node metastasis
- dual energy
- diffusion weighted imaging
- image quality
- deep learning
- current status
- magnetic resonance
- pet ct
- pet imaging
- high intensity
- healthcare
- systematic review
- human health
- social media
- health information
- big data
- squamous cell carcinoma
- climate change
- artificial intelligence