Radiomic Analysis and Data Insights for Enhancing Hepatocellular Carcinoma Immunotherapy Outcomes

ongoing

Hepatocellular carcinoma (HCC) ranks the sixth most common cancer and third leading cause of cancer-related death worldwide. Most patients are diagnosed at an advanced stage. For advanced HCC immunotherapy has been approved as the standard-of-care first-line treatment. However, only about one third of patients respond to immunotherapy. How to predict immunotherapy efficacy and early identify HCC patients who are likely to benefit from it is an urgent clinical problem. RADIANT-HOPE aims to address this problem by developing and externally validating a radiomics model based on clinical routine computed tomography (CT) images. To achieve this, a novel comprehensive framework will be proposed to automatically segment the tumor, expand the tumor margin, and extract radiomics features from intra- and peritumor regions at pretreatment triphasic CT images. These features will be unsupervised clustered into clinically meaningful sub-phenotypes, and patients are stratified into different risk groups of immunotherapy response. Multi-omic data (e.g. mRNA) will be profiled and compared between the sub-phenotypes. Radiogenomic analysis will be performed to evaluate their genomic foundation. The model will be validated in independent international cohorts and developed as software for facilitating clinical utility. This project adopts highly interdisciplinary and intersectoral approaches, integrating oncology, immunology, radiology, artificial intelligence, and bioinformatics. The proposed radiomics model has the potential to provide a non-invasive, accurate, wide accessible imaging biomarker for predicting immunotherapy efficacy and patient stratification. The model implementation in clinic will contribute to HCC personalized treatment, optimize healthcare resources allocation and significantly reduce healthcare costs. As a hepatobiliary surgeon and clinic-oriented researcher, my extensive experience in HCC and radiomics & AI makes me an ideal candidate to conduct this project.